JAMA Ophthalmology

Duffy, Rosemary E.; Brown, Sally E.; Caldwell, Kathleen L.; Lubniewski, Anthony; Anderson, Nicole; Edelhauser, Henry; Holley, Glenn; Tess, Ara; Divan, Hozefa; Helmy, Mark; Arduino, Matthew; Jarvis, William R.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1167pmid: 10980761

BackgroundToxic endothelial cell destruction (TECD) syndrome after intraocular ophthalmic surgery is rare and can result from exposure to a variety of toxins. During January 8 to 14, 1998, 6 patients developed TECD with corneal edema associated with unreactive or dilated pupils at Hospital A.MethodsA case patient was any Hospital A patient with TECD within 24 hours after surgery during January 5 to 14, 1998 (epidemic period). A control was any hospital A ophthalmic surgery patient without TECD during the epidemic period. The medical records of hospital A ophthalmology surgery patients during the pre-epidemic (ie, September 1, 1997-January 4, 1998) and epidemic periods were reviewed. Inductively coupled plasma atomic emission spectrometry was used to detect trace inorganic elements on sterilized surgical instruments. Cannulated surgical instruments and laboratory rinsates were perfused directly to the corneal endothelium of isolated rabbit and human corneas. Corneal endothelial ultrastructure and swelling were assessed.ResultsThe rate of TECD at hospital A was higher during the epidemic than pre-epidemic period (6/12 vs 0/118, P<.001). The only change during the periods was the introduction, on November 5, 1997, of a new sterilization method, AbTox Plazlyte, for sterilization of ophthalmic surgery instruments. Findings from spectrometry revealed that copper and zinc residues were higher in instruments sterilized with Plazlyte than in those sterilized with ethylene oxide (median copper value, 7.64 mg/L vs 0.14 mg/L, respectively, P= .02; median zinc value, 5.90 mg/L vs 1.35 mg/L, respectively, P= .2). Corneal endothelial perfusion of Plazlyte sterilized–instrument rinsates or laboratory solution with copper and zinc produced irreversible damage, similar to toxic corneal endothelial destruction, to rabbit and human corneas.ConclusionA new sterilization method degraded brass to copper and zinc on cannulated surgical instruments resulting in TECD of the cornea.TOXIC ENDOTHELIAL cell destruction (TECD) syndrome is a rare complication of intraocular surgery.It is clinically manifest by unexpected profound corneal edema and opacification within 24 hours after surgery. Toxins implicated in TECD include detergent residues on ophthalmic instruments,topical antiseptic solutions,or preservatives in intraocular medications.Corneal endothelial cell toxic effects have been demonstrated experimentally with presurgical topical antiseptic solutions,intraocular irrigating solutions,high concentrations of intraocular medications,antibiotics in corneal storage media,preservatives in medications,detergent residues on instruments,hydrogen peroxide,or intraocular air.Corneal edema following ophthalmic surgery also can result from mechanical trauma, high intraocular pressure, or inflammation.Severe corneal endothelial decompensation can require corneal transplantation. Of 1.4 million cataract surgeries performed in the United States annually,it is estimated that 0.62% are complicated by corneal edema or corneal transplantation requiring rehospitalization.Because instruments routinely used in ophthalmic surgery often have small lumens, careful cleaning and sterilization are essential. Steam autoclaving and ethylene oxide (ETO) are two commonly used methods for sterilization of ophthalmic instruments. Each technique has its drawbacks. Deposition of rust on instruments and a decrease in sharpness can be seen with steam autoclaving. The National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention (CDC), Atlanta, Ga, consider ETO to be an occupational carcinogen and reproductive toxin.Because of this and the environmentally harmful effects of ETO, the Environmental Protection Agency, Washington, DC, is encouraging health care facility personnel to reduce use of this form of sterilization.Recently, a new method of sterilization using plasma gas technology such as the AbTox Plazlyte sterilization system (AbTox Inc, Mundelein, Ill) has been introduced.This method sterilizes instruments by vaporization of a mixture of peracetic acid, acetic acid, and hydrogen peroxide in combination with low temperature (40°C) and an electromagnetic field; the vapor is removed with argon, oxygen, and hydrogen gas.During January 8 to 14, 1998, 6 of 8 patients receiving elective intraocular surgery at one hospital (hospital A) developed endothelial cell destruction within 24 hours after surgery. When no source for the outbreak could be identified, CDC was invited to assist in the investigation.METHODSCASE AND CONTROL DEFINITIONWe defined a case patient as any patient undergoing intraocular surgery from January 5 to 14, 1998 (epidemic period) at hospital A who had cornea edema and opacification (TECD) within 24 hours after surgery. We defined a control patient as any hospital A patient undergoing intraocular surgery during the epidemic period who did not develop TECD. To identify case patients and determine the background rate, we reviewed medical records of patients undergoing intraocular surgery at hospital A during the pre-epidemic (September 1, 1997-January 4, 1998) and epidemic periods.ANALYTICAL EPIDEMIOLOGY AND STATISTICAL METHODSA case-control study was conducted. Control patients were all patients who had an intraocular surgical procedure on the same day as the case patients. Data were collected on standardized forms, entered into a computer, and analyzed using Epi Info version 6.04 software (CDC) or SAS for personal computers (SAS Institute Inc, Cary, NC). Data collected included race; age; sex; dates of admission, surgery, and discharge; preoperative and postoperative vision assessment (≤24 hours and at 1-year follow-up); reason for eye surgery; type of surgery performed; type and model of intraocular lens (IOL) implant; type of anesthesia; all operative personnel including surgeons, anesthesiologists, and scrub or circulating nurses; medications used before, during, or after surgery; underlying disease; and instrument sterilization methods.An ophthalmic surgical procedure (cataract removal) was observed. Infection control policies and procedures before, during, and after ophthalmic surgery were reviewed and observed, including the cleaning, rinsing, and packaging of surgical instruments.LABORATORY STUDIESTo determine whether the method of sterilization may have been associated with adverse outcomes, we randomly selected 16 ophthalmic surgery instruments (ie, scissors, forceps, cannulas, hooks) from a hospital A ophthalmic surgical kit for chemical evaluation. All 16 instruments were placed in Micropaks (Riley Medical Inc, Auburn, Me), a plastic microsurgical instrument tray, and sterilized (Figure 1). Then, 8 instruments were sterilized with the Plazlyte system and 8 with ETO at hospital A. Following sterilization, these instruments in trays were forwarded to CDC for chemical analysis. At CDC, the 16 instruments were removed from the trays and rinsed with water polished to 18 MΩ-cm resistance with a Milli-Q system (Millipore Corp, Bedford, Mass). The rinsates were then evaluated by inductively coupled argon plasma atomic emission spectrometry(JY 70 plus; Instrument SA Inc, Edison, NJ) to determine the presence of any trace elements. To determine the effect of instrument rinsing on residual trace element contaminants, 9 of 16 instruments were washed an additional 3 times with Milli-Q water and evaluated after each rinse by atomic emission spectrometry for residual trace elements.Figure 1.Assessment methods for ophthalmic instrument sterilization. ICP-AES indicates inductively coupled argon plasma atomic emission spectrometry; ETO, ethylene oxide; and BSS, balanced salt solution.Next, to determine whether the type of packaging and the method of sterilization may have been associated with the adverse outcome, we resterilized 15 instruments, 8 of the initial 16 and 7 additional instruments (Figure 1). To determine if the adverse reactions could be associated specifically with hospital A sterilization (Plazlyte or ETO) machines, we sent instruments to be sterilized to hospital B in Dallas, Tex. Eight instruments were individually packaged in a Tyvek Mylar (DuPont, Wilmington, Del) paper and plastic pouch; 4 were sterilized in Plazlyte, (2 at hospital A; 2, hospital B) and 4 in ETO (2 at hospital A; 2, hospital B). The remaining 7 instruments were sterilized at hospital A in Micropak trays—5 in Plazlyte and 2 in ETO. To determine if any trace elements were present, these 15 instruments were rinsed in balanced salt solution (BSS) (Cytosol Laboratories Inc, Braintree, Mass), and the rinsates were evaluated by atomic emission spectrometry.To detect whether residual peracetic (peroxyacetic) acid and hydrogen peroxide were present on sterilized ophthalmic instruments, 10 mL of the rinsate was collected from the instruments. Then, an analytical titration method was performed in which hydrogen peroxide was titrated with ceric sulfate and the presence of peracetic acid was back titrated using starch indicator and sodium thiosulfate.IN VITRO EXPERIMENTSAn in vitro rabbit corneal endothelial perfusion model was used to evaluate the effects of the poststerilized instrument rinsates and various test solutions.Additionally, a human corneal endothelial perfusion model using Optisol GS–stored eye bank corneas (Chiron, Claremont, Calif) was used for evaluating laboratory-prepared solutions of 10-mg/L copper and 10-mg/L zinc (n = 5).New Zealand white rabbits weighing 2 to 3 kg were anesthetized with an intramuscular mixture of ketamine hydrochloride, 0.5 mL (30 mg/kg of body weight), and xylocaine hydrochloride, 0.5 mL (4 mg/kg of body weight). The rabbits were euthanized with an intracardiac injection of pentobarbital sodium, 324 mg/mL (1.0 Euthanasia-5 solution [1 mg/2.5-kg rabbit]; Henry Schein Inc, Fort Washington, NY), and the eyes were enucleated with the conjunctiva and eyelids intact. All experiments were conducted under the Association for Research in Vision and Ophthalmology, Bethesda, Md, guidelines for animal research.The rabbit corneas were mounted in an in vitro specular microscope. The test solutions were perfused directly to the corneal endothelium (rabbit and human) after a 1-hour stabilization perfusion period with a control solution (BSS Plus; Alcon Laboratories Inc, Ft Worth, Tex). Temperature was maintained at 37°C, perfusion pressure at 15 mm Hg, and perfusion rate at 30 µL per minute. There were 31 solutions used; 9 were rinsates from Plazlyte-sterilized instruments; 5, rinsates from ETO-sterilized instruments; 6, laboratory-prepared solutions of 10-mg/L copper or 10-mg/L zinc; 4, 10-mg/L copper only; and 7, 10-mg/L zinc only. Each of 31 rabbits had 1 cornea perfused with a test solution and the other cornea perfused with a BSS Plus control solution. We measured the change in corneal thickness (micrometers) over time (hours) and calculated a swelling rate (micrometers per hour) by linear regression analysis. At the end of perfusion, the corneas were fixed in 2.5% glutaraldehyde (Electron Microscopy Sciences, Fort Washington, Penn) in 0.1 mol/L cacodylate buffer and examined by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) to determine if TECD had occurred.The tissue parameters for human corneas stored in Optisol GS were as follows: age 60.0 ± 11.4 years; time from death to enucleation, 3.2 ± 0.7 hours; and storage time, 5.5 ± 0.5 days.RESULTSDESCRIPTIVE AND ANALYTICAL EPIDEMIOLOGYSix patients met the case definition. Toxic endothelial cell destruction after intraocular ophthalmic surgery was more frequent during the epidemic than pre-epidemic period (6/12 vs 0/118; P<.001). All case patients were men and ranged in age from 43 to 85 years (median age, 67 years). All had long-term systemic diseases, such as coronary artery disease, diabetes, or hypertension (Table 1).Table 1. Comparison of Case and Control Patients, Hospital A, January 5-14, 1998*VariableCase Patients (n = 6)Control Patients (n = 6)PContinousAge, y67 (43-85)67 (47-82)NSLength of surgery, min52 (17-119)63 (38-92)NSLength of anesthesia, min120 (85-185)129 (85-163)NSAnesthesia severity score2.8 (1-4)2.8 (2-4)NSPreoperative vision, range20/40-20/20020/40-hand motions.02Postoperative vision ≤24 hours postoperatively, range20/400-hand motions20/25-20/400.02Postoperative corneal thickness ≤24 hours postoperatively3 (2+-4+)0 (clear-trace).002Categorical (No.)RaceWhite54NSNonwhite12Sex, M66NSUnderlying diseaseCataracts45NSLens dislocation1. . .NSGlaucoma1. . .NSBullous keratopathy. . .1NSType of surgical procedureExtracapsular cataract extraction or phacoemulsification with posterior chamber intraocular lens45NSLens reposition1. . .NSGlaucoma filtering procedure1. . .NSCorneal transplant. . .1NS*Values are given as median (range) except where indicated. NS indicates not significant; ellipses, not applicable.Two case patients had extracapsular cataract extraction and a posterior chamber IOL implant; 2, cataract removal by phacoemulsification and a posterior chamber IOL implant; 1, repositioning of a previously implanted anterior chamber IOL; and 1, a trabeculectomy for glaucoma. Case patients had corneal edema associated with visual loss within 24 hours of postoperative examination. At 1-year follow-up, 3 of 6 case patients had received a corneal transplant for persistent corneal edema. Of 3 remaining case patients, corneal edema has resolved in 2, and 1 has slowly resolving corneal edema (Table 2).Table 2. Case Patients, Corneal Edema, Visual Acuity, and Pupillary Examination Preoperatively, Within 24 Hours, and at 1-Year Follow-up, Hospital A, 1998 and January 1999*Case PatientSurgical ProcedureCorneal EdemaVisual AcuityPupillary Examination Findings≤24 h Postoperatively1-y Follow-upPreoperatively≤24 h Postoperatively1-y Follow-up≤24 h Postoperatively1-y Follow-up1AC IOL repositioning2+Corneal edema resolved20/40Hand motionsHand motions†Nonreactive, dilated pupilNonreactive, dilated pupil2Trabeculectomy3+Corneal transplant (performed 7 months postglaucoma filter)20/40Hand motionsCounting fingers‡Nonreactive, small pupilNonreactive, small pupil3ECCE with PC IOL3+Slowly resolving cornea edema20/200Counting fingers20/150Partially reactive, dilated pupilPartially reactive, dilated pupil4ECCE with PC IOL3+Corneal transplant (performed 9 months postcataract surgery)20/200Counting fingers20/300Reactive, dilated pupil§Nonreactive, dilated pupil5Phacoemulsification with PC IOL4+Corneal transplant (performed 4 months postcataract surgery)20/70Counting fingers20/30Nonreactive, dilated pupilNonreactive, dilated pupil6Phacoemulsification with PC IOL3+Corneal edema resolved20/6020/40020/40Nonreactive, dilated pupilNonreactive, dilated pupil*AC IOL indicates anterior chamber intraocular lens; ECCE, extracapsular cataract extraction; and PC IOL, posterior chamber intraocular lens.†Vision with the development of optic atrophy and an afferent pupillary defect.‡Advanced glaucomatous atrophy and afferent pupillary defect noted prior to filtering procedure.§Three days postoperatively.When we compared case with control patients, there were no significant differences in the following preoperative, intraoperative, or postoperative conditions: medications, including local or general anesthesia; operating room; operative personnel, including surgeons, anesthesiologists, and scrub or circulating nurses; preoperative vision; length of surgery or time under anesthesia; type of surgery; or brand of lens implant. In contrast, case patients had poorer vision and greater postoperative corneal thickness than controls (Table 1).After introduction of the AbTox Plazlyte system, some ophthalmology personnel noticed blue-green residues on some of the instruments and commented that they smelled a pungent odor similar to acetic acid when opening instrument packs. Therefore, we next reviewed intraocular surgery procedures, including instrument cleaning and reprocessing. There had been no changes in surgical procedures. In contrast, surgery instrument sterilization practices had changed with the introduction of the AbTox Plazlyte sterilization system in November 1997. Postoperatively, instruments were washed with sterile water, placed in a solution of Klenzyme (Steris Co, St Louis, Mo) and tap water for 10 minutes, rinsed with tap water, washed by hand in a solution of Klenzyme and tap water, and rinsed again with tap water. Next, they were hand-rinsed sequentially with tap water, deionized water, and distilled water. Then, all instruments were hand-dried; the smaller instruments and those with lumens were blown dry with compressed air. All ophthalmic surgery instruments to be sterilized were placed in Micropaks and sterilized in the Plazlyte machine.LABORATORY STUDIESNext, we evaluated the relationship between sterilization method and residual inorganic metals (copper and zinc), including the influences of packaging materials on metal residuals and the importance of rinsing. First, we assessed the relationship between sterilization method (Plazlyte vs ETO) and presence of residual trace elements. Eight of 16 instruments sterilized in Micropaks by using either ETO or Plazlyte had measurable amounts of copper and zinc by atomic emission spectrometry; all 8 were cannulas with brass hubs and very small lumens, which are placed directly into the eye during surgery (Figure 1). Of 8 cannulated instruments sterilized, the 4 sterilized in the Plazlyte machine had higher levels of copper and zinc than the 4 sterilized in ETO, although the latter did not reach statistical significance (median copper values: 7.64 vs 0.14 mg/L, respectively, P= .02; median zinc values: 5.90 vs 1.35 mg/L, respectively, P= .20).Second, we assessed the influence of type of packaging (Micropaks vs Tyvek Mylar) and method of sterilization (Plazlyte vs ETO) on instrument copper or zinc residuals. Also, we evaluated whether copper or zinc would be found on cannulas after sterilization in hospital B's machines (Plazlyte vs ETO). The 8 cannulas placed in Micropaks and sterilized by using the Plazlyte method had rinsates with significantly higher levels of copper and zinc than rinsates from instruments sterilized in Micropaks and ETO (Table 3). All rinsates from Plazlyte-sterilized cannulas individually packaged in Tyvek Mylar had lower levels of copper and zinc than the rinsates from cannulas sterilized in plastic Micropaks. The copper and zinc residuals were significantly lower in ETO-sterilized instruments whether in Tyvek Mylar or Micropaks. Ophthalmic surgical instruments sterilized with Plazlyte at hospital A and at hospital B had similar copper and zinc levels.Table 3. Levels of Copper and Zinc in Cannulated Intraocular Surgery Instrument Rinsate by Type of Packaging and Sterilization Method, Hospital A, December 1997 and April 1998Packaging TypeSterilization MethodNo. of CannulasCopper, mg/LMedian PZinc, mg/LMedian PMicropak*Ethylene oxide60.050.0020.83.03AbTox-Plazlyte†89.1607.16IndividualEthylene oxide4‡0.060.040.15.06AbTox-Plazlyte4‡0.2800.89MicropakAbTox-Plazlyte89.160.0067.16.006IndividualAbTox-Plazlyte4‡0.0280.89MicropakEthylene oxide60.050.900.83.10IndividualEthylene oxide4‡0.0550.15*Riley Medical Inc, Auburn, Me.†AbTox Inc, Mundelein, Ill.‡Two cannulas from hospital A and 2 from hospital B.Third, we determined if sequential rinsing of sterilized ophthalmic surgical instruments would affect the residual amount of copper or zinc. Nine instruments, 2 of which were cannulas, sterilized in the Plazlyte machine, rinsed once, were rinsed 3 additional times. Findings from atomic emission spectrometric analysis of these 2 cannula rinsates showed a rapid decrease in the trace elements in the first 5 mL wash, with very low levels of copper and zinc in the subsequent washes (Table 4). The other noncannulated instruments had little to no initial measurable amounts of copper and zinc. When we examined the Plazlyte-sterilized cannulas for residual sterilant (ie, peracetic acid or hydrogen peroxide), neither was detected.Table 4. Residual Copper and Zinc in Plazlyte-Sterilized Cannulated Intraocular Surgery Instruments After Sequential RinsingRinse No.Cannula Levels 1Cannula Levels 2Copper, mg/LZinc, mg/LCopper, mg/LZinc, mg/LFirst5.7903.46020.1008.36Second0.0972.2100.0161.00Third0.0060.0090.0050.04Fourth0.0050.0090.0020.01Rabbit ExperimentsNext, we performed in vitro rabbit corneal endothelial perfusions to assess whether the rinsates from the instruments would cause TECD. Nine rabbit corneas were perfused with rinsates from Plazlyte-sterilized cannulas (ie, BSS Plus rinsed through the sterilized cannulas); 5 were infused with rinsates from ETO-sterilized cannulas. Additionally, 6 were perfused with a laboratory-made solution of 10-mg/L copper, 10-mg/L zinc, and BSS Plus; 4 with 10-mg/L copper only; and 7 with 10-mg/L zinc only.The rabbit corneas perfused with ETO cannula rinsates or BSS Plus alone (controls) showed minimal corneal swelling (7.0 ± 1.1 µm/h vs 3.8 ± 1.7 µm/h, respectively) (Figure 2) with no significant differences between the two. Findings from SEMs and TEMs revealed no structural damage with a normal hexagonal mosaic (SEM), intact borders, and normal intracellular organization (TEM) with no evidence of endothelial cell edema (Figure 3).Figure 2.Corneal swelling rates in rabbit corneas following endothelial perfusion of ethylene oxide rinsates compared with balanced salt solution (BSS) Plus (Cytosol Laboratories Inc, Braintree, Mass) control. This graph represents 5 individual rinsates with corresponding copper (Cu) and zinc (Zn) levels indicated. The sum (mean ± SEM) represents an average of the 5 rinsate perfusions.Figure 3.A, Scanning electron micrograph (SEM) of corneal endothelium perfused with balanced salt solution (BSS) Plus (Cytosol Laboratories Inc, Braintree, Mass) for 3 hours. The endothelial cells are hexagonal with tight junctions (original magnification ×540). B, Transmission electron micrograph (TEM) of corneal endothelium perfused with BSS Plus, which shows endothelial cells with intact borders and normal intracellular organization (original magnification ×4495). C, An SEM of corneal endothelium perfused with ethylene oxide rinsate, which is very similar to control (original magnification ×540). D, A TEM of corneal endothelium perfused with ethylene oxide rinsate, which reveals normal intracellular organization (original magnification ×4495).In contrast, rabbit corneas perfused with Plazlyte cannula rinsates swelled at 35.5, 38.0, and 16.6 µm/h, vs BSS Plus, 11.2 ± 1.5 µm/h, and corneas exposed to the laboratory-made solutions of copper and zinc swelled at 34.0 ± 3.0 µm/h vs BSS Plus, 8.5 ± 1.8 µm/h, P<.001 (Figure 4and Figure 5). Rabbit corneas perfused with either 10-mg/L zinc only (24 ± 2.7 µm/h vs BSS Plus, 13.1 ± 2.0 µm/h, P<.01) or 10-mg/L copper only (21 ± 4.2 µm/h vs BSS Plus, 5.9 ± 2.9 µm/h, P<.03) showed less corneal swelling than those perfused with both (Figure 5). Electron micrographs of Plazlyte rinsate–perfused endothelial cells demonstrated marked endothelial cell destruction (SEM) with cell borders pulling apart and endothelial cell edema (TEM) compared with normal-appearing controls (Figure 6). Findings from SEM and TEM of endothelial cells perfused with laboratory-made solutions of copper and zinc showed endothelial cell damage (SEM) with intracellular vacuolization and disorganization (TEM) compared with normal-appearing controls (Figure 7).Figure 4.Swelling rates in rabbit corneas following endothelial perfusion of Plazlyte rinsates (AbTox Inc, Mundelein, Ill) compared with balanced salt solution (BSS) Plus (Cytosol Laboratories Inc, Braintree, Mass) control. This graph represents 9 individual rinsates with corresponding copper (Cu) and zinc (Zn) levels for each rinsate. The last bar graph is the mean ± SEM of the BSS Plus control perfusions. (Four corneas used for controls were damaged during the mounting procedure.)Figure 5.Corneal swelling rates (mean ± SEM) following endothelial perfusion of rabbit corneas of laboratory-prepared 10-mg/L copper (Cu) plus 10-mg/L zinc (Zn) combined and separately compared with balanced salt solution (BSS) Plus (Cytosol Laboratories Inc, Braintree, Mass) control. All concentrations are 10 mg/L for copper and zinc. The last 2 bars represent corneal swelling (mean ± SEM) of human Optisol GS–stored corneas (Chiron, Claremont, Calif) following endothelial perfusion of 10-mg/L copper plus 10-mg/L zinc compared with BSS Plus control.Figure 6.A, Scanning electron micrograph (SEM) of corneal endothelium perfused with balanced salt solution (BSS) Plus (Cytosol Laboratories Inc, Braintree, Mass) for 3 hours. The endothelial cells show a regular hexagonal pattern with tight junctions (original magnification ×540). B, Transmission electron micrograph (TEM) of corneal endothelium perfused with BSS Plus, which shows endothelial cells with intact borders and normal intracellular organization (original magnification ×4495). C, An SEM of corneal endothelium perfused with Plazlyte rinsate (AbTox Inc, Mundelein, Ill), which demonstrates marked endothelial cell destruction (original magnification ×540). D, A TEM of corneal endothelium perfused with Plazlyte rinsate, which shows cell borders pulling apart and endothelial cell edema (original magnification ×4495).Figure 7.A, Scanning electron micrograph (SEM) of corneal endothelium perfused with balanced salt solution (BSS) Plus (Cytosol Laboratories Inc, Braintree, Mass) for 3 hours. The endothelial cells show normal hexagonality with intact borders (original magnification ×540). B, Transmission electron micrograph (TEM) of corneal endothelium perfused with BSS Plus, which shows endothelial cells with tight junctions and regular intracellular organization (original magnification ×4495). C, An SEM of corneal endothelium perfused with laboratory-made solution of 10-mg/L copper plus 10-mg/L zinc, which shows endothelial cell damage (original magnification ×540). D, A TEM of corneal endothelium perfused with laboratory-made solution of 10-mg/L copper and zinc, which reveals intracellular vacuolization and disorganization (original magnification ×4495).Human ExperimentsWe performed in vitro corneal endothelial perfusions with Optisol GS–stored human corneas to assess how human corneas perfused with laboratory-prepared samples of 10-mg/L copper and 10-mg/L zinc would compare with results of the perfused rabbit corneas of 10 mg/L-copper and 10 mg/L-zinc. Similar to results seen in the rabbit, the human corneas swelled significantly (28.2 ± 0.9 µm/hr vs BSS Plus, 14.2 ± 0.9 µm/hr, P<.001). Electron micrographs reveal endothelial cell destruction (SEM) with large vacuolization and endothelial cell edema (TEM) (Figure 8).Figure 8.A, Scanning electron micrograph (SEM) of human corneal endothelium perfused with balanced salt solution (BSS) Plus (Cytosol Laboratories Inc, Braintree, Mass) for 3 hours. The endothelial cells show a normal mosaic with intact borders and no evidence of edema. B, Transmission electron micrograph (TEM) of human corneal endothelium perfused with BSS Plus, which shows normal intracellular organization (original magnification ×4495). C, An SEM of human corneal endothelium perfused with laboratory-made solution of 10-mg/L copper plus 10-mg/L zinc, which reveals endothelial cell destruction (original magnification ×540). D, A TEM of human corneal endothelium perfused with laboratory-made solution of 10-mg/L copper and zinc, which shows large vacuoles and endothelial cell edema (original magnification ×4495).COMMENTWe investigated an outbreak of profound corneal endothelial decompensation observed within 24 hours after surgery. Our initial epidemiologic investigation failed to identify any risk factors for adverse outcome. The recent changes that occurred in ophthalmic surgery instrument sterilization methods led us to evaluate the relationship between these changes and the adverse outcomes. We found that rinsates of cannulated sterilized ophthalmic surgery instruments with small lumens had measurable copper and zinc levels. Regardless of the packaging, our laboratory experiments showed that rinsates from cannulas sterilized with the Plazlyte system had elevated levels of both copper or zinc compared with instruments sterilized in ETO; these levels were elevated whether the instruments were sterilized at Hospital A or Hospital B, suggesting that the method of sterilization rather than specific hospital practices was responsible. Rinsate from cannulas sterilized with Plazlyte in surgical kits had the highest levels of copper and zinc.Since the Plazlyte sterilization machine was introduced for instrument sterilization in November 1997, it is unclear why more patients have not developed TECD. We hypothesize that cannulated instruments that were manufactured with chrome-covered brass hubs may have been oxidized by the acetic and peracetic acids from the Plazlyte sterilization machine. The chrome's degradation, through repeated reprocessing, exposed the brass; then, brass decomposed into its primary elements, copper and zinc. It also is possible that the acetic and peracetic acids, used during the Plazlyte sterilization process, did not quickly dissipate, especially in those cannulas sterilized in Micropaks. The increase in contact time of the acetic and peracetic acids on the instruments may have allowed oxidation of the brass to occur, each time exposing more and more of the brass. This process may have been repeated each time the cannulas were sterilized. Following the sterilization, copper and zinc would have remained in the cannula channel and then flushed into the patient's eye during the surgical procedure, producing TECD.Since the oxidation process would have occurred over time, the breakdown of the brass by the acetic and peracetic acids may have been related to the number of times the cannulas were sterilized in the Plazlyte machine or the age of the cannulas. Hospital personnel did not always record which kit instrument was used or how many times it was previously used or sterilized. Hospital personnel did not perform many (<5 per day) intraocular ophthalmic surgery procedures, so it is possible that some surgical kits were used rarely and not repeatedly resterilized; thus, little oxidation of the cannula brass hubs of these instruments may have occurred. These later instruments may have been used during the eye procedures of those patients who did not develop TECD. Alternatively, the copper and zinc residuals were water soluble. Sufficient rinsing before the procedure could reduce or eliminate the copper and zinc residuals, reducing or diminishing the risk of their infusion into the eye. Hospital A personnel noted that it was their policy to rinse routinely the instruments and irrigate cannulas before their use in the eye.Contamination of instruments with the trace metals copper and zinc would allow the formation of free hydroxyl and peroxyl radicals within the eye. Any of these reactive elements would indiscriminately combine with intraocular tissues, resulting in the unexpected TECD. Additionally, the smell of acetic acid on some instruments suggests that residual hydrogen peroxide was present, although we were not able to detect it.The AbTox Plazlyte machine had not been approved by the Food and Drug Administration (FDA), Washington, DC, for instrument sterilization. The FDA had approved an earlier design of the Plazlyte machine for use on stainless steel items but not instruments with small lumens or hinges, such as many intraocular surgery instruments. Furthermore, instruments sterilized in the AbTox Plazlyte machine require careful cleaning and drying before sterilization to avoid residual chemicals or water, which can react with the Plazlyte chemicals or gas to produce by-products.The results of this investigation led to a CDC Morbidity Mortality Weekly Reportand an FDA safety alert about the use of AbTox Plazlyte Sterilization System in sterilizing eye instruments.Until the FDA approves its use, cannulated instruments should not be sterilized in the Plazlyte machine.Corneal endothelial cells are a sensitive indicator of potential toxins inadvertently introduced during ophthalmic surgery. Our case patients demonstrated the potentially catastrophic effect that copper and zinc contamination can have on intraocular tissues. This investigation documented the importance of active postmarket surveillance for complications associated with newly introduced medical devices or sterilization methods. Furthermore, it emphasizes that medical personnel should ensure that the FDA has approved a medical device or sterilization method before it is used on patients or patient equipment. As new sterilization methods and devices are developed and implemented, health care personnel must be aware of and determine whether changes in practice are safe to avoid adverse patient outcomes.DBGlasserROSchultzRAHyndiukThe role of viscoelastics, cannulas, and irrigating solution additives in post-cataract surgery corneal edema: a brief review.Lens Eye Toxic Res.1992;9:351-359.DBGlasserPathophysiology of corneal endothelial dysfunction.In: Chandler JW, Sugar J, Edelhauser HF, eds. External Disease: Cornea, Conjunctiva, Sclera, Eyelids, Lacrimal System. St Louis, Mo: Mosby–Year Book Inc; 1994:8.1-8.19.ACBreebaartRNuytsEPelsHFEdelhauserFDVerbraakToxic endothelial cell destruction of the cornea after routine extracapsular cataract surgery.Arch Ophthalmol.1990;108:1121-1125.RBPhinneyBJMondinoJDHofbauerCorneal edema related to accidental Hibiclens exposure.Am J Ophthalmol.1988;106:210-215.SMMac RaeBBrownHFEdelhauserThe corneal toxicity of presurgical skin antiseptics.Am J Ophthalmol.1984;97:221-232.HFEdelhauserRAHyndiukAZeebROSchultzCorneal edema and the intraocular use of epinephrine.Am J Ophthalmol.1982;93:327-333.HFEdelhauserRGonneringDLVan HornIntraocular irrigating solutions: a comparative study of BSS Plus and lactated Ringer's solution.Arch Ophthalmol.1978;96:516-520.KKadonosonoNItoFYazamaEffects of intracameral anesthesia on the corneal endothelium.J Cataract Refract Surg.1998;24:1377-1381.FJGarcia-FerrerJSPeposePRMurraySRGlaserJHLassWRGreenAntimicrobial efficacy and corneal endothelial toxicity of DexSol corneal storage medium supplemented with vancomycin.Ophthalmology.1991;98:863-869.RAHyndiukROSchultzOverview of the corneal toxicity of surgical solutions and drugs: and clinical concepts in corneal edema.Lens Eye Toxic Res.1992;9:331-350.KGreenLCheeksDAStewardDTraskRole of toxic ingredients in silicone oils in the induction of increased corneal endothelial permeability.Lens Eye Toxic Res.1992;9:377-384.NOhguroMMatsudaSKinoshitaThe effects of denatured sodium hyaluronate on the corneal endothelium in cats.Am J Ophthalmol.1991;112:424-430.CHChangCPLinHZWangCytotoxicity of intracameral injection drugs to corneal endothelium as evaluated by corneal endothelial cell culture.Cornea.1995;14:71-76.RMNuytsHFEdelhauserEPelsACBreebaartToxic effects of detergents on the corneal endothelium.Arch Ophthalmol.1990;108:1158-1162.AArtolaJLAlioJLBellotJMRuizProtective properties of viscoelastic substances (sodium hyaluronate and 2% hydroxymethylcellouse) against experimental free radical damage to cornea endothelium.Cornea.1993;12:109.RJOlsonAir and the corneal endothelium: an in vivo specular microscopy study in cats.Arch Ophthalmol.1980;98:1283-1284.HFEdelhauserDBGlasserSurgical pharmacology intraocular solutions and drugs for cataract surgery.In: Buratto L, ed. Phacoemulsification Principles and Techniques. Thorofare, NJ: Stark Inc; 1998:275-292.Agency for Health Care Policy and ResearchCataracts in Adults: Management of Functional Impairment.Rockville, Md: US Dept of Health and Human Services, Agency for Health Care Policy and Research; 1993. AHCPR publication 93-0542.JKCannerJCJavittAMMcBeanNational outcomes of cataract extraction, III: corneal edema and transplant following inpatient surgery.Arch Ophthalmol.1992;110:1137-1142.National Institute for Occupational Safety and HealthCurrent Intelligence 52: Ethylene Oxide Sterilizers in Health Care Facilities: Engineering Controls and Work Practices.Cincinnati, Ohio: US Dept of Health and Human Services, Public Health Service, Centers for Disease Control, NIOSH; 1989. DHHS Pub No. (NIOSH) 89-115.National Institute for Occupational Safety and HealthCurrent Intelligence 35: Ethylene Oxide (ETO): Evidence of Carcinogenicity.Cincinnati, Ohio: US Dept of Health and Human Services, Public Health Service, Centers for Disease Control, NIOSH; 1981. DHHS publication (NIOSH) 81-130.MLynchGas plasma low-temperature sterilization technology.Infect Control Today.1998;2:53-56.EABryceEChiaGLogelinJASmithAn evaluation of the AbTox Plazlyte Sterilization System.Infect Control Hosp Epidemiol.1997;18:646-653.MKimberlyDPaschalScreening for selected toxic elements in urine by sequential-scanning inductively-coupled plasma atomic emission spectrometry.Anal Chim Acta.1985;174:203-210.EnvironmentalProtection AgencyMethods for the Determination of Metals in Environmental Samples.Washington, DC: Environmental Protection Agency; 1994. 600/R-94-111, PB94-184942.Not AvailableEuropean Pharmacopoeia, Vol 3-5.2nd ed. Sainte-Ruffine, France: Council of Europe; 1980. European Treaty Series No. 50.BEMcCareyHFEdelhauserDLVan HornFunctional and structural changes in the corneal endothelium during in vitro perfusion.Invest Ophthalmol.1973;12:410-417.TKimGPHolleyJHLeeGBroockerHFEdelhauserThe effects of intraocular lidocaine on the corneal endothelium.Ophthalmology.1998;105:125-130.Not AvailableCorneal decompensation after Intraocular Ophthalmic Surgery—Missouri, 1998.MMWR Morb Mortal Wkly Rep.1998;47:306-309.Food and Drug AdministrationA Nationwide Warning Against the Use of "AbTox" Plazlyte Sterilization System.Rockville, Md: US Dept of Health and Human Services, Food and Drug Administration; 1998.Food and Drug AdministrationFDA Safety Alert: Warning Regarding the Use of the AbTox Plazlyte Sterilization System.Rockville, Md: US Dept of Health and Human Services, Food and Drug Administration; April 13, 1998.Accepted for publication February 8, 2000.This study was supported in part by the National Institutes of Health, National Eye Institute, Bethesda, Md, grant EY00933 (Dr Edelhauser).Use of trade names and commercial sources is for identification purposes only and does not imply endorsement by the Public Health Service or the US Department of Health and Human Services.We thank Douglas R. Dobson, BS, and H. Denny Donnell, MD, MPH, at the Missouri Department of Health, Jefferson City, and Phillip Boston, BS, and Jeffery A. Hangartner, BS, at the Food and Drug Administration, St Louis, Mo, for their assistance in this investigation.Toxic Endothelial Cell Destruction Syndrome Investigative Team: Sandra Sides, RN, Chaille Fisher, RN, BSN, John Cochran Division, Veteran Affairs Medical Center, St Louis, Mo; and Johnny M. Khoury, MD, Department of Ophthalmology and Visual Sciences, Washington University, St Louis, and the Ophthalmology Section, Surgical Service of the St Louis Veteran Affairs Medical Center.Reprints: William R. Jarvis, MD, CDC, Hospital Infections Program, 1600 Clifton Rd NE, MS-E69, Atlanta, GA 30333.

Smolek, Michael K.; Klyce, Stephen D.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1179pmid: 10980762

BackgroundKeratoconus has long been considered to be an ectasia produced by stromal stretching. Although stretching should result in increased corneal surface area, previous observations of topography during progression of keratoconus have suggested that surface area may actually be conserved. A novel objective surface area measurement based on corneal topography was tested and applied to data from actual corneas under various conditions for comparative analysis.SettingThe LSU Eye Center clinic videokeratography archives.MethodsTMS-1 videokeratography files (Tomey Corp, Cambridge, Mass) were obtained from 6 groups of corneas: normal (n = 29), keratoconus from mild to severe states (n = 51), topographically judged keratoconus-suspect conditions (n = 10), postoperative photorefractive keratectomy for myopia (n = 39), with-the-rule corneal astigmatism (n = 17), and keratoglobus (n = 1). Additionally, 3 different spherical test surfaces were analyzed to verify the accuracy of the process. Only maps with no missing data out to ring 29 were used. The cumulative surface area from center to periphery was determined by calculating and summing the area of individual patches along consecutive annular rings. Mean surface area with respect to mean chord radius was plotted for each corneal condition, and curve fitting was used to extend each result to a 5.85-mm limbus. Means, SEs, and 95% confidence intervals were calculated at intervals for statistical comparisons among all groups. Computer-generated surfaces helped to evaluate the relationship between shape and surface area.ResultsWhen videokeratographic test targets were used, surface area error was less than 2%, which was deemed acceptable. Normal corneas had a mean ± SE surface area of 120.3 ± 2.2 mm2, whereas all keratoconus corneas combined had a mean ± SE surface area of 116.2 ± 3.4 mm2. The difference was not significant at any chord radius (analysis of variance, P<.05). The keratoglobus cornea was found to have a surface area of 129.9 mm2, which was 7.98% greater than normal. An individual with progressive keratoconus exhibited no appreciable trend toward increasing surface area during a 76-month period. The corneas in the other groups resembled normal corneas in their total surface area.ConclusionsWith the exception of the single case of keratoglobus, corneal surface area tended to be conserved near a value of 120 mm2for all groups in the study, including corneas with keratoconus. Surface area is remarkably insensitive to curvature change near the vertex. Flattening seen in the periphery of corneas with keratoconus suggests that biomechanical coupling compensates for any increase in curvature occurring in the region of the cone itself. Thus, it seems that keratoconus is not a true ectasia as is keratoglobus, but rather a specialized type of warpage, at least in mild to moderate forms of the disease.KERATOCONUS is a degenerative corneal disease characterized by a localized region of stromal thinning that is spatially associated with a cone-shaped deformation of the surface.It has been inferred that keratoconus is an ectasia resulting from stromal stretching.The term ectasiais defined as a dilation, expansion, or distension, all of which invoke the notion of an increase in surface area by a process of stretching. Despite wide acceptance that keratoconus represents a true ectasia, it has never been confirmed by objective measurements, nor have any accurate corneal surface area measurements been made from living corneas (Table 1).We hypothesized that total corneal surface area during the progression of keratoconus remains essentially identical to that of normal corneas. The reasons for this counterintuitive hypothesis are 2-fold. First, simple mathematical calculations reveal that the total surface area of a hemisphere is not appreciably increased by the addition of a region of localized steepening. Second, the corneal periphery covers a proportionately greater surface area than the central cornea, such that a minimal amount of peripheral flattening might compensate for the emergence of a cone elsewhere on the surface. Because of its subtle nature, peripheral flattening is a rarely recognized sign of keratoconus.An example of superior peripheral flattening of corneas with inferiorly located keratoconus has been shown with cinemakeratography, an animated form of videokeratography that enhances visualization of curvature change.Table 1. Previously Published Corneal Surface Area ValuesSurface Area Estimate, mm2StudyMethod Used130MauriceSphere geometry150EhlersGeneral shape estimate132KwokEllipsoid geometry136.3Hanna et alLotmaraspherical cornea104Watsky et al2-Dimensional plane projected areaThis study describes a novel method for the objective measurement of surface area using data obtained from videokeratography. During the first part of the study, a computer algorithm was designed and tested using mathematical formulas relating surface area to radius of curvature, elevation height, and location on spherical surfaces. The method was extended by an iterative, patchwork approach to measure surface area on a variety of closed surfaces representing theoretical corneal shapes, including spheres and rotationally symmetric and asymmetric ellipsoids. Next, surface area measurements using the algorithm were validated with known test surfaces imaged using videokeratography. Finally, the surface areas of normal, astigmatic, keratoconus-suspect, keratoconus, keratoglobus, and surgically altered cornea groups were calculated using the algorithm in conjunction with patient videokeratographs. Averaged results from each group were compared with results from the normal group or with modeled surfaces of similar shapes. Surface area for keratoconus progression during a 76-month period was also measured to determine if any trends toward increasing surface area were present.MATERIALS AND METHODSSURFACE AREA ALGORITHMA computer algorithm was derived to calculate the surface area of a 3-dimensional conic section surface modeling the cornea, a videokeratographically acquired test surface, or a videokeratographically acquired surface of a living cornea. Using this algorithm, the surface was divided from center to periphery into concentric rings of known elevation, with 256 patches distributed at equal meridional angles along each ring. A value of 256 patches per ring was chosen to comply with the data acquisition format of the TMS-1 videokeratoscope (Tomey Corp, Cambridge, Mass).The surface area of each patch was calculated by first determining its axial radius of curvature using the following relationship: r = y2/2s + s/2, where s is the sagitta determined for a chord radius, y (Figure 1).We define chord radiusas the distance from the axis of rotation of a conic section or the videokeratographic axis to a point on the surface that defines the location of the patch. The sagitta and chord radius values can be generated from equations describing a theoretical surface, or in the case of videokeratography, using data contained in the RAD and HIT files of the TMS-1. Note that values for radius of curvature also can be back-calculated from the DIO file. However, this approach was tried and found to generate increasing, albeit minor, inaccuracies in surface area measurements toward the periphery. These inaccuracies may be consistent with the minor peripheral curvature inaccuracies reported previously.Figure 1.Surface area geometry. Top, The chord radius (y) and sagitta (s) are used to determine a radius of curvature (r) at point P on a patch of the surface. Bottom, For a height (h) the area of a spherical segment with radius of curvature (r) can be determined and divided by 256 to estimate the effective area of a individual patch on any surface. By keeping the patch area small, the method of calculating its surface area relative to a best-fitting sphere can be applied to any surface shape.Given an elevation height (h) and radius of curvature (r) for each patch, the equation A = 2πrh was applied to determine the total surface area (A) of a segment of height h for a sphere of radius r (Figure 1).The surface area (A) of the segment was then divided by 256 to determine the effective surface area of a single patch. The process was repeated for each patch on the surface. Using this approach, the surface area of any complex surface could be determined. When modeling a theoretical surface in a patchwork fashion, h was given a constant height of 50 µm, but when deriving area from TMS-1 videokeratographic data, h was determined from the HIT file by obtaining the mire inter-ring elevation height. Only the first 29 rings were used in the analysis of videokeratographic data. Although there were more patches (the total number depending on the shape) measured for a modeled surface, compared with 7424 patches when using a videokeratograph, the modeled surface patches were proportionately smaller and this difference was found to have no effect on the total surface area measurement. Although the surface area of each patch tended to increase from center to periphery, individual patch areas were, on average, only 0.016 mm2when videokeratographic data were used.The area of each patch was added to a running total for each ring. The total surface area of each ring was then added to a cumulative total surface area out to a chord radius of at least 6 mm for modeled surfaces or out to ring 29 of the videokeratograph. Because chord radius varies for each patch and is dependent on the shape of the surface, the mean chord radius for each ring was determined in order to plot surface area as a function of chord radius.MATHEMATICAL MODEL TESTINGThe computer algorithm was tested by generating a wide variety of spherical and ellipsoidal surfaces of the following formula: x2/a2+ y2/b2+ z2/c2= 1, where x, y, and z are coordinates of surface points and a, b, and c are the axis intercepts of the surface along the x, y, and z directions. Spheres of various sizes as well as both rotationally symmetric and asymmetric ellipsoids can be modeled by choosing the appropriate axis intercept values. The surface areas of theoretical shapes (or any portion of the shapes) were then computed using a noniterative approach,and the results compared with the iterative approach of the computer algorithm to confirm the validity of the algorithm.TMS-1 VALIDATION TESTINGBefore analyzing the corneas of living eyes, we assessed the ability of the TMS-1 system to generate height and chord radius values that could provide accurate area measurements, as well as the algorithm's ability to convert those data into surface area measurements. Videokeratography maps for 3 spherical test surfaces having vertex radii with curvatures of 9.3 mm, 7.8 mm, and 6.5 mm were provided by Computed Anatomy Inc (New York, NY) and subsequently analyzed for surface area by comparing the results with models of spheres with the same radius of curvature. Maps of more complex test surfaces were not available at the time of this study.CORNEAL MAP ANALYSISTMS-1 videokeratography maps were obtained from adult patient records at the LSU Eye Center, New Orleans, La. Videokeratography had institutional review board approval at the LSU Eye Center and informed consent was obtained from all patients undergoing this procedure. Maps were obtained for normal corneas (n = 29); clinically diagnosed keratoconus corneas that included 17 mild, 24 moderate, and 10 severe cases (n = 51); topographically judged keratoconus suspects (n = 10); postoperative photorefractive keratectomy for myopia cases (n = 39); with-the-rule corneal astigmatism cases (n = 17); and a keratoglobus cornea (n = 1). It is important to note that there were twice as many moderate and severe keratoconus examples as mild forms, and our dataset was not skewed toward early forms of the disease. We specifically included as many severe cases of keratoconus as possible to detect any effects of the disease on corneal surface area. Normal corneas exhibited less than 1.5 diopters (D) of cylinder power measured by simulated keratometry. Astigmatic corneas had 1.5 D to 4.5 D of regular cylinder power. Keratoconus corneas were diagnosed by an ophthalmologist using traditional clinical signs such as corneal thinning and the Munson sign; our sample included mild, moderate, and severe forms of the disease as defined in a previous study.Keratoconus-suspectcorneas were defined as those that could not be clinically classified as keratoconus, although they showed topographic patterns resembling mild keratoconus.The photorefractive keratectomy cases were randomly selected from videokeratography acquired at 1, 3, 6, 12, 18, or 24 months after the surgical procedure was performed using the VISX 20/20 excimer laser system (VISX Inc, Santa Clara, Calif). None of the corneas in the study had a history of contact lens wear.As each TMS map was analyzed by the surface area algorithm, the chord radius and the computed surface area for each patch of the surface were written to a SigmaPlot 4.0 (SPSS Inc, Chicago, Ill) database file. If interpolated or missing values were encountered for any of the 7424 chord radius values from ring 1 to 29, the map was not used for surface area analysis. The dataset size of each category as described in the preceding paragraph indicates the final numbers of cases where TMS maps met the criterion of interpolation-free data.By averaging surface area and chord radius plots obtained for each cornea in a group, a single best-fitting function to describe the mean surface area as a function of mean chord radius was determined for each corneal group, including the subgroups of the 3 levels of keratoconus severity (TableCurve 2D; SPSS Inc). Additionally, the mean ± SE surface area was evaluated at every 0.5 mm of chord radius and at a theoretical limbus of 5.85 mm, except for keratoglobus, for which n = 1 (SigmaStat; SPSS Inc). All corneal groups except the single case of keratoglobus were curve-fitted with the following equation: X−1 = a + (b/Y2), where X signifies surface area and Y is the chord radius. The keratoglobus data were fitted better by the power function: X = aYb. The a and b parameters for all curves are presented in Table 2. The goodness of fit and predictive ability were demonstrated by a large overall F value and a coefficient of determination (R2) near the value of 1.Table 2. Results of Corneal Surface Area Measures at a Chord Radius of 5.85 mm by Corneal Group*43.84-D Sphere (Theoretical)Normal GroupKeratoconus GroupKeratoconus- Suspect GroupCorneal Astigmatism GroupPhotorefractive Keratectomy GroupKeratoglobus CaseNo. of cases129511017391Mean surface area129.1120.3116.2118.9119.8122.2129.995% Confidence interval. . .±2.2±3.4±6.9±6.5±1.9. . .Percent from normal+7.32. . .−3.41−1.16−0.42+1.58+7.98R21.01.01.01.01.01.01.0F statistic2 × 10 73 × 1061 × 1064 × 1067 × 1054 × 1051 × 107Fit SE0.2160.0860.4280.0760.1780.2320.103a−1.7 × 10−3−9.1 × 10−40−8.1 × 10−4−9.7 × 10−4−6.4 × 10−43.06b0.3250.3160.3040.3160.3190.3022.06*D indicates diopters; R2, the coefficient of determination; a and b, the parameters for all curves described in the table, corresponding with the equation X = aYb, where X is the surface area and Y is the chord radius.It was necessary to extrapolate the mean surface area plot for each group out to a chord radius of 5.85 mm to facilitate comparisons and calculate the total corneal surface area for a typical cornea. With the exception of rare cases of megalocornea, microcornea, and congenital glaucoma, corneas have an average diameter of 11.7 mm at the limbus.To facilitate comparisons and emphasize differences, we plotted the residual difference between the mean surface area for each keratoconus group and the normal group (which served as a control), as a function of chord radius.RESULTSMATHEMATICAL MODEL VERIFICATIONAs expected, there was no difference in surface area measurement between the iterative approach of the computer algorithm and manual calculation of surface area for any portion of a sphere or ellipsoidal surface.TMS-1 CALIBRATION TARGET VERIFICATIONThere was good correspondence between the TMS-1 calibration test sphere results and their corresponding computer spherical models as shown in Figure 2, top. When the functions were graphed as difference plots, the 2 flatter test surfaces had errors below 0.5% of the total surface area at any chord radius, while a maximum error of 2% was generated using the steeper 6.5-mm radius-of-curvature test sphere (Figure 2, bottom). The original test surfaces were not available for retesting; therefore, we could not check to see if the vertex curvature of the steepest surface was perhaps inaccurately recorded. Nevertheless, a maximum instrument error of 2% was considered sufficiently accurate to proceed with calculating surface area from TMS-1 maps of actual corneas.Figure 2.Top, Surface area as a function of chord radius for 3 spheres. Symbols indicate measured spherical calibration target surface area generated from data contained in TMS-1 map files. Legend values indicate the radius of curvature of each calibration sphere. Plotted lines indicate computed surface area for 3 hypothetical spheres with vertex power identical to the calibration targets, based on the mathematical formula for surface area of a sphere. Note that area measurement points for steeper spheres were shifted toward relatively lower chord radii, just as mire rings appear in the videokeratoscope display. Bottom, Surface area error (calibration target data minus the computed sphere value) as a function of chord radius for the data shown in the top section.MEAN SURFACE AREA BY CORNEAL CONDITIONResults for all corneal conditions are shown in Table 2. Normal corneas had a mean surface area of 120.3 ± 2.22 mm2, which was 8% to 25% less than previous estimates based on spherical, ellipsoidal, or complex shape modelsand 13.6% greater than a flat plane-projection area estimate (Table 1).When the normal cornea group was compared with other groups, the surface area difference ranged from −3.41% to +1.58% (−4.1 mm2to +1.9 mm2) of that of the normal cornea with the exception of keratoglobus, which was estimated to be 7.98% (+9.6 mm2) greater than normal (Table 2). Because of the rarity of keratoglobus, caution should be used when assuming our result is typical. Likewise, extrapolation for a single set of data to the theoretical limbus may induce some inaccuracy.The surface area difference plots for the corneas with keratoconus and keratoglobus as compared with normal corneas are shown in Figure 3. The keratoglobus data were consistently greater than those of the normal group, whereas the keratoconus data, subdivided by severity, tended to exhibit increasingly less surface area toward the periphery and were distributed such that mild keratoconus corneas were most similar to normal corneas and the most severe cases were most different from the normal cases. However, a 1-way analysis of variance test comparing normal and keratoconus surface area for each of the subcategories of keratoconus at the 5.85-mm chord radius and at every 0.5-mm chord radius step indicated no significant difference among any pairs of points at any chord radius (P= .5-1.0). It is interesting to note that although the results are not significant, the tendency toward the reduction in surface area in the periphery in keratoconus is entirely consistent with our hypothesis that subtle peripheral flattening has an important influence on the surface area measurement. It seems to follow that the greater the severity of the cone, the greater the peripheral flattening, and the lower the total surface area. With only a single case of keratoglobus, it is impossible to know if the result is typical. However, this example does estimate a much larger surface area than was measured for either the normal or the keratoconus groups, and the value is well above the 95% confidence interval for either of the other 2 groups (Table 2).Figure 3.Change in mean surface area compared with normal corneas as a function of chord radius. Error bars are the SEs of the difference. The vertical dashed line represents the average limbus. Raw data for the single case of keratoglobus is plotted along with a best-fit function. Keratoconus cases are subdivided into 3 severity subgroups. D indicates diopters.SURFACE AREA CHANGE DURING KERATOCONUS PROGRESSIONWe analyzed 8 videokeratography maps obtained during a period of 76 months from the left cornea of an individual with clinically diagnosed keratoconus situated in the inferior periphery. The disease progressed from a mild to a moderate state with as much as an 8-D increase in curvature at the steepest point on the topography map (Figure 4). The vertex curvature also increased to a lesser extent, but a point in the superior hemisphere directly opposite the steepest point of the cone was shown to decrease in curvature over time. In fact, the appearance of flattening covered a relatively large region of the superior hemisphere, but the visual impact was minimal when static images of the color-coded contours were viewed. The keratoconus prediction index (KPI)for this patient increased from 0.18 at month 0 to 0.36 at month 46, whereupon it remained stable to month 70 and then decreased to 0.33 at month 76. Although KPI is not specifically a severity index, it does correlate with keratoconus severity (R= 0.892; P<.001).The surface area at the limbus was plotted as a function of time (Figure 5). Despite the obvious trend toward worsening of keratoconus in this patient during a 6-year period, there was only a minor trend toward increasing corneal surface area, and this trend may be insignificant because over time; the data appeared to fluctuate about a mean value in a sinusoidal-like fashion. There was no statistical difference between the mean of the normal group and the mean of this patient's surface area measurements during the 76-month period (Figure 5). However, there was considerably more variance in this patient's data.Figure 4.Top, Surface curvature values of a keratoconus cornea with respect to time. As the disease progresses, the steepest surface location associated with this inferiorly situated cone increases in curvature. Center, The corneal vertex, lying off the cone, increases in curvature to a lesser amount. Bottom, A point in the superior hemisphere opposite the cone shows a slight but important decrease in curvature. D indicates diopters.Figure 5.Total surface area of a keratoconus cornea over time. The measurement appears to vary in a sinusoidal manner with no definite trend toward increasing surface area as the disease progresses (compare with Figure 4). The mean surface area for this cornea is not significantly different from the mean of the normal group (points beyond the axis break to the right).SURFACE AREA AND CORNEAL ASTIGMATISMMaps from the TMS-1 of corneal astigmatism with 1.8 D, 3.4 D, and 3.5 D of cylinder oriented at various axes were modeled using nonrotationally symmetric ellipsoids, and then compared with the measured surface areas. Simulated keratometry values from the TMS-1 examinations were used to model the curvature of the cornea in the x and y directions, while the TMS-1 vertex radius of curvature provided the z-axis shape parameter. Table 3shows that the 3 shape parameters (a, b, and c) of the theoretical ellipsoid models varied among the 3 examples. The surface area plots (Figure 6) were virtually identical in profile, however.Table 3. General Ellipsoid Models of Corneal Astigmatism*Shape ParametersCase 1Case 2Case 3Vertex radius, mm7.917.377.87Vertex power, D42.4445.5442.88Sim K cylinder, D3.43.51.8a8.097.647.95b7.487.087.62c7.917.377.80*D indicates diopters; a, b, and c, the axis intercepts in the model-testing equation x2/a2+ y2/b2+z2/c2= 1, where x, y, and z are coordinates of the surface points.Figure 6.Corneal surface area as a function of chord radius for 3 cases of corneal astigmatism with vertex sphere/cylinder powers of 42.88 D/1.8 D, 45.54 D/3.5 D, and 42.44 D/3.4 D. Symbols represent derived values obtained from the TMS-1 files. Lines represent the output of a general ellipsoid model based on the corneal vertex power and simulated keratometry values for each of the cases. Note that the right axis represents surface area as a percentage of the normal condition. D indicates diopters.The fitted ellipsoid surface area functions (Figure 6) tended to slightly overestimate surface area toward the limbus because the ellipsoid data were generated solely with simulated keratometry and vertex powers and these parameters tend to describe curvature of the central cornea. Fitting curves to all the data (as was performed in Table 2for all astigmatic corneas) would have resulted in a better solution. Our surface area data support the observation that the cornea is indeed flatter in the periphery than a true ellipsoid.The fact that corneas are not true ellipsoids also accounts for our normal corneal surface area calculation being approximately 12 mm2less than the theoretical corneal surface area estimated by an ellipsoid model (Table 1).During the course of fitting models to corneas of various shapes, we realized that only certain ellipsoids appeared to be suitable models of the basic shape of the central cornea while simultaneously fulfilling a conservation of surface area principle. In other words, a corneal shape that is relatively steep in one meridian must be compensated either by flattening in another meridian (typically the orthogonal meridian) or by altering the eccentricity of the overall shape to the extent that surface area remains near a value of 120 mm2. Corneas apparently do not exist in an infinite variety of shapes, but tend to exist only in forms that conserve surface area. This observation is analogous to the biomechanical concept of coupling, in which steepening in one meridian can induce flattening of the orthogonal meridian. Similarly, if surface area is conserved, then corneal volume must tend to remain constant.SURFACE AREA AND SPHERICITYThe mean vertex radius-of-curvature of the normal cornea group was 7.70 mm (43.84 D), which amounts to a surface area of 129.1 mm2at a limbus radius of 5.85 mm for a perfectly spherical cornea (Figure 3, Table 2). This value is only 0.69% less than the often-cited estimate of 130 mm2reported by Maurice,but is 7.32% larger than the measured surface area of the living cornea, which is aspherical.Spheres have relatively greater surface area in the periphery for a given chord radius, compared with actual corneas, particularly beyond 3.5 mm into the periphery. Nevertheless, the periphery contains a significantly greater proportion of the actual corneal surface; approximately 50% of the total surface area of the cornea is located within a 1.5-mm-wide annulus and adjacent to the limbus (Figure 6, Right scale). Therefore, centrally located deformations that invoke any increase in localized surface area could be totally compensated by a decrease in surface area produced by a flattening in the far periphery. Simultaneous cone steepening and peripheral flattening is known to occur in keratoconus.We have shown that the most severe form of keratoconus may have a greater than normal surface area within the central 4 mm, but this is offset by a tendency for relatively less area toward the far periphery (Figure 3).To better illustrate the relationship between radius-of-curvature, chord radius, and surface area, we used our algorithm to model surface area for spheres with vertex curvatures between 30 D and 90 D. We observed that surface area was surprisingly insensitive to even a 60-D increase in curvature within a central 3-mm chord radius (Figure 7). For a specific chord radius, surface area does increase toward the periphery as curvature is increased; however, low-curvature surfaces achieve far greater surface area than a highly curved surface if the chord radius is not limited. Thus, we can envision that even a small additional flattening in the already flat periphery can produce a large decrease in surface area. A slight peripheral flattening may arise out of a localized steepening elsewhere on the surface of the cornea.Figure 7.Surface area of spheres as a function of chord radius and vertex curvature. Note that for a region near the vertex, surface area is relatively insensitive to the curvature of the sphere. However, surface area becomes increasingly sensitive to a chord radius value that approaches the radius of curvature of the sphere, implying that peripheral curvature on the cornea can have an important effect on the computation of total surface area. NA indicates not applicable; D, diopters.COMMENTBased on our results, our hypothesis that keratoconus and normal corneas have virtually identical surface areas seems to be correct. We conclude that several factors contribute to this condition: (1) The closer a region of localized steepening is to the corneal vertex, the less effect it has on total surface area. (2) A localized region of steepening such as is seen with a well-defined cone tends to contribute to only a small portion of the total surface area. (3) Steepening in one localized region can produce flattening elsewhere on the cornea through biomechanical coupling. (4) A 2-dimensional map of the corneal surface greatly deemphasizes the contribution of the corneal periphery to surface area. (5) In an absolute color-contour map, a subtle, peripheral flattening effect producing only one color contour step or less spread over a large corneal area may be barely noticeable when compared with a well-defined cone exhibiting several color contour steps; however, the effects of flattening and steepening may counterbalance one another and produce no significant change in total surface area.Thus, the concept of a cone produced primarily by stromal stretching seems to be an entirely intuitive construct used to explain keratoconus deformation and the illusion of an increased surface area. Any stretching that does occur is probably highly localized and relatively insignificant when compared with the total area of the surface. Stromal thinning, which has often been associated with stretching, may be a direct result of the degenerative process, rather than linked to a stretching process.To emphasize that the topographic appearance of shape change is not necessarily linked to stretching, consider an example of contact lens–induced corneal warpage. It is well known that corneal curvature can change dramatically with contact lens–induced warpage, so much so that it can on occasion resemble early keratoconus. However, a warped cornea typically regains a normal topographic appearance after cessation of lens wear. It is difficult to conceive of a process whereby contact lens wear increases cornea surface area by stretching, and then returns it to a completely normal shape in a matter of days. It seems more reasonable that the topographic appearance is related not to stretching, but rather to deformations acting under the principle of surface area conservation.A literature review of the past 33 years indicated that more than 75% of corneas with keratoglobus spontaneously ruptured prior to treatment, but fewer than 1% of corneas with keratoconus were reported to have ruptured. This result may be indicative of differences in treatment for these conditions, but perhaps a better explanation is that corneas with keratoconus rarely burst simply because surface area is conserved and the stroma's elastic limit is never reached. Edmundcompared the elastic modulus of normal corneas and corneas with keratoconus reported in several previous studies, and found no difference in the immediate elastic response of the tissue, although viscoelastic tissue properties differed with keratoconus. Even if elasticity does increase in keratoconus, there is no conflict with the results of our study. Increased elasticity alone does not imply that the surface area must increase by stretching, provided that strain-inducing stresses are controlled. It is important to note that acute hydrops occurs in about 6% to 8% of all keratoconus cases; however, microscopic tears of the Descemet membrane alone would not be expected to significantly increase corneal surface area, even though they do suggest biomechanical events occurring throughout the corneal structure.In a simplistic sense, the decreasing radius-of-curvature of an emerging cone may act as a stress-reducing mechanism that also counteracts the effects of a diminishing cross-sectional thickness of a shell according to the following well-known relationship: S = P (r/2t), where S is the stress tension in the corneal shell, P is the intraocular pressure, t is the thickness of the tissue, and r is the radius of curvature.Similarly, the increasing amount of flattening (ie, a larger radius of curvature) in thicker corneal regions distant from the cone should allow higher stresses to be tolerated without requiring any concomitant increase in surface area from stress-induced stretching.On the other hand, untreated corneas with keratoglobus tend to rupture because the entire corneal surface is involved in the disease,which means that no other region can be recruited to deform in a way that will minimize stress and thus conserve surface area. Once the elastic threshold is reached in this completely thinned tissue, it can stretch no further and the tissue ruptures. From this standpoint, surface area measures may have some clinical benefit as a means of assessing structural integrity in these diseases, differentiating keratoglobus from keratoconus, and establishing a surface area threshold beyond which rupture is imminent.In this study, we have shown that keratoconus corneas do not have an appreciably increased surface area compared with normal corneas. Therefore, one must reconsider whether keratoconus is a true ectasia according to the standard definition, which implies stretching and an increased surface area, or if keratoconus may be more correctly considered as an extreme form of corneal warpage caused by a lack of structural integrity brought on by stromal degeneration and external forces. Warpage is generally considered to be a reversible condition and does not involve corneal thinning. Therefore, keratoconus is not a true warpage. However, our findings also suggest that keratoconus is not a true ectasia, particularly when compared to examples such as arterial aneurysms or keratoglobus that do show ample evidence of increased surface area and stretching. Unfortunately, ectasia is a term that is well ingrained in the ophthalmic vocabulary as being synonymous with keratoconus, and this is likely to remain so.In conclusion, the normal human corneal surface area was found to be approximately 120 mm2by measurements made on living eyes, and this value appears to be conserved in a variety of corneal shapes, including that of corneas with keratoconus. Whenever corneal surface area remains constant, the possibility to remold the surface into an optically useful shape remains feasible.JHKrachmerRSFederMWBelinKeratoconus and related noninflammatory corneal thinning disorders.Surv Ophthalmol.1984;28:293-322.LJMaguireEctatic corneal degenerations.In: Kaufman HE, Barron BA, McDonald MB, eds. The Cornea. 2nd ed. Boston, Mass: Butterworth-Heinemann; 1998:525-550.YSRabinowitzKeratoconus.Surv Ophthalmol.1998;42:297-319.KJBrownKeratoconus.Cornea.1988;7:163-169.AMBrooksIFRobertsonAMMahoneyOcular rigidity and intraocular pressure in keratoconus.Aust J Ophthalmol.1984;12:317-324.DMMauriceThe cornea and sclera.In: Davson H, ed. The Eye. New York, NY: Academic Press; 1969:490-491.NEhlersOn the size of the conjunctival sac.Acta Ophthalmol.1965;43:205-210.LSKwokCalculation and application of the anterior surface area of a model human cornea.J Theor Biol.1984;108:295-313.LSKwokThe surface area of an ellipsoid revisited.J Theor Biol.1989;139:573-574.KDHannaFEJouveGOWaringPreliminary computer simulation of the effects of radial keratotomy.Arch Ophthalmol.1989;107:911-918.WLotmarTheoretical eye model with aspherics.J Opt Soc Am.1971;61:1522-1529.MAWatskyMMJablonskiHFEdelhauserComparison of conjunctival and corneal surface area in rabbit and human.Curr Eye Res.1988;7:483-486.TTMcMahonJBRobinKMScarpullaJLPutzThe spectrum of topography found in keratoconus.CLAO J.1991;17:198-204.MKSmolekSDKlyceCinemakeratography using computer morphing.Optom Vis Sci.1997;74:970-975.MKSmolekThe Multimedia Atlas of Videokeratography.Woburn, Mass: Butterworth-Heinemann; 1999.WHAFinchamMHFreemanOptics.9th ed. London, England: Butterworths; 1980:62.CRobertsAnalysis of the inherent error of the TMS-1 topographic modeling system in mapping a radially aspheric surface.Cornea.1995;14:258-265.AWGoodmanAnalytical Geometry and the Calculus.2nd ed. New York, NY: Macmillan Publishing Co Inc; 1969:235.EJGarbocziJFSnyderJFDouglasMFThorpeGeometrical percolation threshold of overlapping ellipsoids.Phys Rev E.1995;52:819-828.MKSmolekSDKlyceCurrent keratoconus detection methods compared with a neural network approach.Invest Ophthalmol Vis Sci.1997;38:2290-2299.MJHoganJAAlvaradoJEWeddellHistology of the Human Eye.Philadelphia, Pa: WB Saunders; 1971:61.LMMatsudaCLWoldorffRTKameJKHayashidaClinical comparison of corneal diameter and curvature in Asian eyes with those of Caucasian eyes.Optom Vis Sci.1992;69:51-54.AAKiskisSNMarkowitzJDMorinCorneal diameter and axial length in congenital glaucoma.Can J Ophthalmol.1985;20:93-97.NMaedaSDKlyceMKSmolekHWThompsonAutomated keratoconus screening with corneal topography analysis.Invest Ophthalmol Vis Sci.1994;35:2749-2757.RBMandellJCCorzineSAKleinPeripheral corneal topography and the limbus.Invest Ophthalmol Vis Sci.1998;38(suppl):S1036.CEdmundCorneal elasticity and ocular rigidity in normal and keratoconic eyes.Acta Ophthalmol (Copenh).1988;66:134-140.ESjøntoftCEdmundIn vivo determination of Young's modulus for the human cornea.Bull Math Biol.1987;49:217-232.TAPooleCalculation of stress distribution in keratoconus.N Y State J Med.1973;73:1284-1288.AArciniegasLEAmayaMechanical behavior of the sclera.Ophthalmologica.1986;193:45-55.JACameronKeratoglobus.Cornea.1993;12:124-130.Accepted for publication March 21, 2000.This work was supported in part by US Public Health Service grants EY03311 (Dr Klyce) and EY02377 (LSU Eye Center core grant) from the National Eye Institute, National Institutes of Health, Bethesda, Md.Presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Fla, May 14, 1997.Corresponding author: Michael K. Smolek, PhD, LSU Eye Center, 2020 Gravier St, Suite B, New Orleans, LA 70112 (e-mail: [email protected]).

Keltner, John L.; Johnson, Chris A.; Quigg, Jacqueline M.; Cello, Kimberly E.; Kass, Michael A.; Gordon, Mae O.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1187pmid: 10980763

ObjectiveTo determine the frequency with which visual field abnormalities observed on follow-up visual fields for patients in the Ocular Hypertension Treatment Study were confirmed on retest.MethodsBetween April 1, 1994, and March 1, 1999, 21,603 visual fields were obtained from 1637 patients in the Ocular Hypertension Treatment Study. When follow-up visual fields are outside the normal limits on the Glaucoma Hemifield Test, the Corrected Pattern Standard Deviation (P<.05), or both, subsequent follow-up visual fields are monitored to confirm the abnormality. Abnormalities are confirmed if they are again abnormal on the Glaucoma Hemifield Test, the Corrected Pattern Standard Deviation, or both; if the defect is not artifactual; and if the same index and location are involved. Reliability criteria used by the study consisted of a limit of 33% for false positives, false negatives, and fixation losses.ResultsOf the 21,603 regular follow-up visual fields, 1006 were follow-up retests performed because of an abnormality (n = 748) or unreliability (n = 258). We found that 703 (94%) of the 748 visual fields were abnormal and reliable, and 45 (6%) were abnormal and unreliable. On retesting, abnormalities were not confirmed for 604 (85.9%) of the 703 originally abnormal and reliable visual fields.ConclusionsMost visual field abnormalities in patients in the Ocular Hypertension Treatment Study were not verified on retest. Confirmation of visual field abnormalities is essential for distinguishing reproducible visual field loss from long-term variability.THE OCULAR Hypertension Treatment Study (OHTS) is a multicenter trial, funded by the National Eye Institute, National Institutes of Health, Bethesda, Md. The OHTS seeks to evaluate the safety and efficacy of topical ocular hypotensive medication in preventing or delaying the onset of visual field loss, optic nerve damage, or both in patients with ocular hypertension who are at moderate risk for developing primary open-angle glaucoma.Half of the patients receive topical ocular hypertensive medication, and half receive careful observation only. Automated static perimetry (Humphrey field analyzer program 30-2 full-threshold test; Humphrey Systems, Dublin, Calif) is used as one of the primary outcome measures for the OHTS.Progressive glaucomatous cupping, as determined from optic disc photographs, is also a primary outcome measurement for the OHTS.A detailed description of the OHTS protocol is available elsewhere.Automated static perimetry in patients with glaucomatous visual field loss exhibits large amounts of variability.For patients with ocular hypertension, the typical visual field has normal sensitivity, and variability has been reported to be lower than in glaucomatous visual fields.However, long-term longitudinal investigationsof patients with ocular hypertension are limited. The present study examines the reproducibility of visual field abnormalities observed in patients with ocular hypertension who were enrolled in the OHTS. We determined the frequency with which visual field abnormalities observed on follow-up visual fields in the OHTS were confirmed on retest.MATERIALS AND METHODSBefore testing patients in the OHTS, technicians are required to complete a certification process that covers all aspects of OHTS visual field testing. Technicians must demonstrate that they can perform visual fields and enter patient data according to OHTS protocol. Informed consent was obtained from each patient before the study.Visual field testing in the OHTS consists of automated static perimetry using program 30-2 on the Humphrey field analyzer. The visual field protocol used in the OHTS is a modification of the one used in the Optic Neuritis Treatment Trial.Testing includes a full-threshold test strategy, a 31.5-apostilb background, a size III target, a foveal threshold determination, and a short-term fluctuation determination. The limit for fixation losses, false positives, and false negatives is 33% for the OHTS. If a patient's pupils are less than 3 mm in diameter, the eyes are dilated before visual field testing. Dilation was needed in 562 (2.6%) of the cases. An appropriate lens correction is placed before the eye to be tested, and the nontested eye is occluded. If the lens correction exceeds 5 diopters, a soft contact lens correction is used to minimize trial lens rim artifacts.All visual fields obtained for the OHTS are sent to the Visual Field Reading Center, University of California, Davis, for processing and analysis. Each field is evaluated by a comprehensive quality control system to determine if all aspects of the OHTS protocol were followed. The approach is similar to that used in the Optic Neuritis Treatment Trial.The quality control system addresses 3 areas of performance by the OHTS clinic technician and clinic coordinator, as shown in Table 1: (1) whether the correct visual field testing parameters were used (mean test parameter errors), (2) whether the patient data were entered correctly (patient data errors), and (3) whether visual field handling instructions were followed (shipment errors). Continuous feedback on the quality control findings is provided to the visual field technicians and the clinic coordinators. This feedback is provided in 3 primary ways: (1) individual reports are sent to the clinics for each visual field monthly, (2) summary reports regarding overall clinical performance are sent quarterly, and (3) telephone calls are made when necessary. The aim of this feedback is to ensure that visual field quality in the OHTS is optimal. In addition, the patients' pupil sizes and refractions are monitored to minimize the possibility that they are causes of visual field abnormalities.Table 1. OHTS Visual Field Quality Control*Error MessagesPoint ValueTest variable errors (maximum of 60 points)Program 30-2 not used−60STATPAC 2 not used−60Stimulus size 3 not used−60Full-threshold strategy not used−60Short-term flucuation test turned off−60Foveal threshold test turned off−10Blind spot check turned off−10Inappropriate blind spot check size used−10Central fixation target not used−5Visual field data not saved on disk−20Patient data errors (maximum of 30 points)Pupil diameter <3 mm or not entered−10Visual acuity not entered−3Birth date incorrect−5Site IDNot entered in ID field−3Incorrect−2Technician IDNot entered in ID field−3Incorrect−2Visit codeNot entered in ID field−3Incorrect−2Dilation information incorrect−2Patient IDNot entered in name field−5Incorrect−4Distance RxNot entered in name field−10Entered incorrectly−3Rx used not entered−10Sphere error in patient's Rx−7A sphere of ± 0.25 D must be dropped for test−3Cylinder discrepancy in patient's Rx−3Axis discrepancy in patient's Rx−2A cylinder <1.00 D must be dropped and a spherical equivalent used−3A cylinder lens must be used for a cylinder ≥1.00 D−3Shipment errors (maximum of 10 points)Visual field printout not included in shipment−3Humphrey disk file shipped late−3Humphrey disk file directory not included in shipment−2Visual field not checked off on Humphrey disk directory−1ShipmentNot packaged properly−2Not addressed properly−2Humphrey diskNot labeled−2Not labeled correctly−1Prestudy formNot included with visual field−2Not completed correctly−1This field was not faxed to the VFRCAs required−3On time−2*Analyses were performed at the Visual Field Reading Center, University of California, Davis. OHTS indicates Ocular Hypertension Treatment Study; ID, identification; Rx, prescription; D, diopter; and VFRC, Visual Field Reading Center.To enroll in the study, in addition to meeting other entrance criteria,a patient needed 2 normal, reliable visual fields for each eye. A maximum of 3 visual field tests were allowed on each eye to obtain these 2 normal, reliable visual fields, and they had to be performed within a 3-month period. A technically acceptable visual field was considered to be normal if all visual field indexes were within normal limits and if there were no clusters of abnormal points that had low sensitivity and might be consistent with early glaucomatous damage. A clusteris defined as 2 or more horizontally or vertically contiguous abnormal points (P<.05), which could represent early stages of glaucomatous loss (eg, a subtle nasal step). A visual field was considered to be reliable if false positives, false negatives, and fixation losses were below 33%.Follow-up visual fields are obtained every 6 months. In the OHTS, the Glaucoma Hemifield Test and the Corrected Pattern Standard Deviation are the Humphrey indexes that are monitored to detect the development of possible glaucomatous visual field loss. Through 1997, if a technically acceptable follow-up visual field was abnormal on the Glaucoma Hemifield Test (outside normal limits or a general reduction of sensitivity), the Corrected Pattern Standard Deviation (P<.05), or both, a retest was performed on the eye in question within the same 6-month follow-up visit window, preferably within 8 weeks. A visual field abnormality was considered not confirmed if it was determined that it was artifactual (trial lens rim artifacts that disappear on retest or superior depression that disappears with taping of the eyelid) as judged by the Visual Field Reading Center readers. An abnormality was considered confirmed if the same index was involved on test and retest and if the abnormality was in the same general location (involving similar points as the previous visual field).During follow-up, a high percentage of first abnormal visual fields were found to be normal according to OHTS standards on retest. Accordingly, a more stringent criterion for confirmation of visual field abnormalities was adopted effective January 1, 1998, at the recommendation of the OHTS Data and Safety Monitoring Committee, the OHTS Steering Committee, and the OHTS Full Investigative Group. The protocol was changed so that confirmation of a visual field abnormality required 3 consecutive visual fields with a defect of the same character in the same general location. Thus, a patient with an abnormal visual field is tested at the next regularly scheduled follow-up visit in 6 months. If the Visual Field Reading Center considers the second visual field abnormal, it requests a third visual field to be completed in 1 day to 8 weeks. If the visual field abnormality is confirmed on the third visual field, the Visual Field Reading Center prepares a narrative description of the abnormality and sends all visual fields to the OHTS Coordinating Center for review by the OHTS End Point Committee. The OHTS End Point Committee, which is masked as to randomization assignment, determines whether the visual field abnormality is clinically relevant and can be attributed to primary open-angle glaucoma based on a review of all clinical information.Unreliable follow-up visual fields (false-positive errors, false-negative errors, or fixation losses exceeding 33%) are also retested. However, if the visual field is again unreliable on retest, no action is taken, and the patient will simply be tested again at the next regularly scheduled visit. For this study, only abnormal and reliable visual fields were examined.RESULTSWe report on 21,603 regular follow-up visual fields obtained between April 1, 1994, and March 11, 1999 (36 regular follow-up fields were not used because a Humphrey program 30-2 was not used), and 703 retests that were performed because the regular follow-up visual field was abnormal and reliable according to OHTS standards (abnormal on the Glaucoma Hemifield Test, the Corrected Pattern Standard Deviation, or both). Only 0.17% (36/21 603) of the regular follow-up visual fields at 11 centers were unusable. Nine centers had 3 or less unusable visual fields, and 2 had 6 or more unusable visual fields. Some of the abnormal visual fields were from the same eye at different follow-up visits. Visual fields that determined a glaucomatous or nonglaucomatous end point were included in the analysis; however, visual fields obtained after an end point was reached were not included.Figure 1shows the overall quality control performance of the OHTS clinical centers for a quarterly grading system during a 2-year period (January 1, 1997, to December 31, 1998). Each visual field is graded on a 100-point scale, as described in Table 1. On this scale, 0 represents a perfect score and 100 represents the maximum number of error points (Table 1). Fifty-nine percent (12 831/21 603) of the follow-up visual fields had perfect scores of 0 errors. As shown in Figure 1, the overall visual field performance at the OHTS clinical centers has been excellent, with a quarterly mean consistently around 2 error points per visual field and the mean number of error points declining over time.Figure 1.The overall quality control performance of the Ocular Hypertension Treatment Study (OHTS) clinical centers in following the OHTS visual field protocol during a 2-year period (January 1, 1997, to December 31, 1998) is shown. There are 3 general areas of clinic performance that are evaluated: (1) basic test parameters (mean test error points), (2) patient data (mean patient data error points), and (3) data shipment (mean shipment error points). Each visual field is graded on a 100-point scale on whether the protocol was followed in each of these 3 areas. On this scale, 0 represents a perfect score, with increasing point scores reflecting more severe quality control problems.Table 2shows the results for the retest visual fields that directly followed the 703 technically acceptable abnormal visual fields. Some of the abnormal visual fields were from the same eye at different follow-up visits. The initial abnormality was confirmed on 99 (14.1%) and not confirmed on 604 (85.9%) of the 703 originally abnormal visual fields. The 604 abnormalities that were not confirmed fell into 3 categories: (1) 467 (66.4%) tested within normal limits on all indexes; (2) 112 (15.9%) were normal according to OHTS standards but had a borderline result on at least one index; and (3) 25 (3.6%) were abnormal according to OHTS standards, but the defect was due to artifact, in a different location, or on a different index than on the preceding visual field. A review of refractive error and pupil size information revealed that they could not have been a contributory factor in any of the nonconfirmed visual field abnormalities. A few nonconfirmed visual field losses could be attributed to a heavy eyebrow or droopy eyelid (6 [0.9%] of 703) or to trial lens rim artifacts (3 [0.4%] of 703).Table 2. Results for the Retest Visual FieldsVariable*No. (%) of Visual FieldAbnormality replicated99 (14.1)Abnormality not replicated604 (85.9)Normal467 (66.4)Normal but with borderline results112 (15.9)GHT results are borderline39 (5.5)CPSD P<.0129 (4.1)MD P≤.0522 (3.1)GHT results are borderline, CPSD P<.1013 (1.8)GHT results are borderline, MD P≤.052 (0.3)CPSD P<.10 and MD P≤.054 (0.6)GHT results are borderline, CPSD P<.10 and MD P≤.053 (0.4)Initial abnormality not replicated25 (3.6)†Different index15 (2.1)Different location1 (0.1)ArtifactHeavy eyebrow or droopy eyelid6 (0.9)Lens rim3 (0.4)Fatigue or drowsiness0Total703(100.0)*GHT indicates Glaucoma Hemifield Test; CPSD, Corrected Pattern Standard Deviation; and MD, mean deviation.†Individual percentages may not sum to these totals because of rounding.In most cases, nonconfirmation of visual field abnormalities in patients in the OHTS does not appear to be related to uncooperative patients, protocol violations, or careless test administration. Unreliable visual fields were not included in the data analysis for this study. However, only 389 (1.8%) of the 21,603 regular follow-up visual fields exceeded the 33% limits for fixation losses, false-positive errors, or false-negative errors. As shown in Figure 1, the excellent quality control scores attest to the outstanding performance of visual field technicians and clinic coordinators in the OHTS (0.17% [36/21,603] of the follow-up visual fields were unusable due to a non–30-2 test strategy). In addition, the low rate of unreliable visual fields is related to our enrollment criteria requiring 2 reliable fields.Of the 9 cases of artifactual results, 2 are shown in Figure 2and Figure 3. These artifacts accounted for a small portion of abnormal test results and were not confirmed on retest. Figure 2provides an example of a probable trial lens rim artifact that disappears on retest. A high plus lens correction was used, increasing the likelihood that this was a trial lens rim artifact. Figure 3provides an example of a probable droopy upper eyelid producing a superior visual field loss that is not present on retest. As shown in Figure 4, Figure 5, Figure 6, and Figure 7, most cases included visual field loss that was typical of localized glaucomatous defects. Figure 8provides an atypical example of visual field loss (cause unclear) that resolves on retest.Figure 2.This example shows consecutive follow-up fields for the right eye of a patient in the Ocular Hypertension Treatment Study. Top, The first test, performed on May 20, 1997, shows trial lens rim artifact that might be misinterpreted as glaucomatous visual field loss. The prescription used was +5.00 + 0.50 × 180. The pupil diameter was 4.0 mm. The results of the Glaucoma Hemifield Test (GHT) were outside normal limits, and the Corrected Pattern Standard Deviation (CPSD) was P<.005. Bottom, The artifact is absent on the retest, performed on May 27, 1997. The prescription used was a +5.50-diopter sphere. The pupil diameter was 4.0 mm. The results of the GHT and the CPSD were normal.Figure 3.This example shows consecutive follow-up fields for the right eye of a patient in the Ocular Hypertension Treatment Study. Top, The first test, performed on May 3, 1996, shows evidence of a drooping eyelid or eyebrow. The prescription used was a +3.00-diopter sphere. The pupil diameter was 4.0 mm. The results of the Glaucoma Hemifield Test (GHT) were normal, and the Corrected Pattern Standard Deviation (CPSD) was P<.05. Bottom, Evidence of a drooping eyelid or eyebrow is absent on the retest (eyelid taped), performed on May 17, 1996. The prescription used was a +2.00 diopter sphere. The pupil diameter was 4.0 mm. The results of the GHT and the CPSD were normal.Figure 4.This example shows consecutive follow-up visual fields for the left eye of a patient in the Ocular Hypertension Treatment Study. Top, The first test, performed on April 2, 1997, shows an inferior arcuate visual field defect. The prescription used was a +4.00-diopter sphere. The pupil diameter was 4.5 mm. The results of the Glaucoma Hemifield Test (GHT) were outside normal limits, and the Corrected Pattern Standard Deviation (CPSD) was P<.02. Bottom, The inferior arcuate visual field defect completely cleared up on the retest, performed on April 9, 1997. The prescription used was a +4.00-diopter sphere. The pupil diameter was 4.5 mm. The results of the GHT and the CPSD were normal.Figure 5.This example shows consecutive follow-up fields for the right eye of a patient in the Ocular Hypertension Treatment Study. Top, The first test, performed on May 30, 1997, shows an inferior arcuate defect with central depression. The prescription used was a +3.25-diopter sphere. The pupil diameter was 4.0 mm. The results of the Glaucoma Hemifield Test (GHT) were outside normal limits, and the Corrected Pattern Standard Deviation (CPSD) was P<.10. Bottom, The inferior arcuate defect with central depression resolved on the retest, performed on June 6, 1997. The prescription used was a +3.25-diopter sphere. The pupil diameter was 4.5 mm. The results of the GHT and the CPSD were normal.Figure 6.This example shows consecutive follow-up fields for the right eye of a patient in the Ocular Hypertension Treatment Study. Top, The first test, performed on May 1, 1996, shows a superior partial arcuate defect with an inferior nasal step. The prescription used was −2.00 + 1.50 × 175. The pupil diameter was 4.0 mm. The results of the Glaucoma Hemifield Test (GHT) were outside normal limits, and the Corrected Pattern Standard Deviation (CPSD) was normal. Bottom, The superior arcuate defect with an inferior nasal step resolved on the retest, performed on May 28, 1996. The prescription used was −2.00 + 1.50 × 175. The pupil diameter was 4.0 mm. The results of the GHT and the CPSD were normal.Figure 7.This example shows consecutive follow-up visual fields for the right eye of a patient in the Ocular Hypertension Treatment Study. Top, The first test, performed on July 12, 1995, shows generalized diffuse loss superiorly and inferiorly, with a suggestion of a double arcuate visual field defect. The prescription used was a +2.50-diopter sphere. The pupil diameter was 5.0 mm. The result of the Glaucoma Hemifield Test (GHT) was a general reduction, and the results of the Corrected Pattern Standard Deviation (CPSD) were normal. Bottom, The generalized diffuse loss, along with the suggestion of a double arcuate visual field defect, completely cleared up on the retest, performed on July 28, 1995. The prescription used was a +3.00-diopter sphere. The pupil diameter was 4.0 mm. The results of the GHT and the CPSD were normal.Figure 8.This example shows consecutive follow-up fields for the left eye of a patient in the Ocular Hypertension Treatment Study. Top, The first test, performed on July 30, 1997, shows an inferior temporal vertical step (cause unclear). The prescription used was a −3.50-diopter sphere. The pupil diameter was 6.0 mm. The results of the Glaucoma Hemifield Test (GHT) were outside normal limits, and the Corrected Pattern Standard Deviation (CPSD) was P<.05. Bottom, The inferior temporal vertical step resolved on the retest, performed on September 15, 1997. The prescription used was a −3.50-diopter sphere. The pupil diameter was 5.0 mm. The results of the GHT and the CPSD were normal.COMMENTPrevious investigationshave reported that automated static perimetric threshold tests exhibit variability, within a test procedure and from one examination to another. In normal subjects, this variability is between 2 and 3 dB.Several factors have been shown to affect the amount of variability in normal subjects, including target size,visual field eccentricity,age,and threshold sensitivity.In patients with glaucomatous visual field loss, the amount of variability is much higher,with greater variability for locations with reduced sensitivity. The variability of threshold determinations for moderate visual field loss can be 3 to 4 times as large as for regions with normal sensitivity. Heijl and colleaguesfound that the 95% confidence limits for moderate visual field loss (8-18 dB of loss) encompassed nearly the entire measurement range (0-40 dB) of the Humphrey field analyzer. In patients with ocular hypertension, the visual field has normal sensitivity, and variability is much lower than in those with glaucoma.Werner et alhave reported that the variability of visual fields in patients with ocular hypertension is only slightly higher than in normal subjects. Thus, one might expect glaucomatous visual field changes in patients with ocular hypertension to be more reliable than moderate glaucomatous visual field defects.Variability in visual fields, which may be due to factors unrelated to optic nerve pathological features, is not limited to glaucoma. Similar results of high variability in areas of visual field loss have been reported for optic neuritis.Wall et alexamined the short- and long-term variability of automated perimetry in healthy subjects and in patients with optic neuritis who were thought to be stable and who had a residual visual field (Humphrey mean deviation of −3.00 to −20.00 dB). Patients with optic neuritis demonstrated variations in visual field sensitivity that were outside the entire range of the variability for normal controls. In the present study, the cause of the abnormality (whether it be early glaucomatous or not) can only be defined when the OHTS is completed.The 85.9% rate of visual field abnormality not being confirmed after a single abnormal visual field in this study indicates that there is considerable variability in the visual fields of patients with ocular hypertension as they begin to show early glaucomatous visual field loss. Previous studies have reported that increased visual field variability may be an early sign of glaucomatous damage. Hart and Beckerbelieve that glaucomatous visual fields go through 3 phases: The first is the initial stage, with no defect demonstrable despite the fact that occult damage is occurring. The second is a period in which shallow defects are often transient and are barely detectable. The chronological course of initial visual field defects was marked in 22 of the 98 eyes by a phenomenon of transiently appearing defects. In the third phase, visual field defects progress at an uneven pace to become dense. The transient nature of initial visual field defects (at the threshold stage) and the invariant findings of their greater density on recurrence was considered by Hart and Becker to be the best evidence of progressive damage occurring to the visual system before its detection by light-sense perimetry. The OHTS should help us to understand whether abnormal visual fields that return to normal are the first stage of progressive glaucomatous field loss or are simply long-term variability. In addition, the OHTS should also help to determine the significance of a single abnormal test result. The OHTS will be able to determine if patients with abnormal or borderline visual fields that return to normal have a different long-term prognosis than patients who do not have any abnormal visual fields.In patients with progressive glaucoma, there are difficulties in distinguishing between truly progressive glaucoma and long-term variability unless several visual fields are obtained over time. Wernerconcluded that a minimum of 6 visual fields were needed to make informed clinical judgments as to whether a patient's visual field was stable or progressing. Quantitative approaches using linear regression have come to similar conclusions, indicating that approximately 7 visual fields obtained over several years are needed to reliably distinguish progression from intratest variability.For studies using a discrete measure (change from baseline) rather than regression techniques, it has been found that confirmation of changes are necessary to avoid "overcalling" progression of visual field loss. In the Normal-Tension Glaucoma Study, Schulzerfound that 4 to 6 confirming visual field tests (2 of 3 tests performed within 1 to 4 weeks showing change, followed by 2 of 3 tests performed 3 months later) were needed to reliably determine visual field progression. Chauhan and colleaguesdefined progression as at least 4 nonedge test locations that were beyond the 5% probability level on the Glaucoma Change Probability program (significant change from baseline) and a confirming field with complete overlap of at least 4 of these locations. The Early Manifest Glaucoma Trialuses a similar strategy, except the Glaucoma Change Probability is based on the pattern deviation values rather than on the total deviation values and 3 locations beyond the 5% level need to be confirmed on 3 successive tests.Several strategies have been attempted to reduce the variability associated with conventional automated perimetry.Because of the strict OHTS quality control system that provides regular feedback to the clinical centers and the visual field technicians about their performance and handling of the visual fields, visual field quality has not been a factor in the variability. Only 1.8% of the 21,603 regular follow-up visual fields (fixation losses, false-positive errors, or false-negative errors) were beyond the 33% limit. In addition, neither pupil size nor refractive errors contributed to the variability. Unreliable visual fields were not included in the present study. We will have a better understanding of whether these abnormalities are due to early transient pathological features, unreliable fields, or a combination of both with further follow-up evaluations. Because of the variability demonstrated in the present study, the OHTS has changed its visual field protocol for confirming abnormality. Three consecutive abnormal visual fields are required, for which the defect is not artifactual, the same index is involved, and the abnormality is in the same location.Our results indicate that most of the initial visual field abnormalities in the patients in the OHTS are not reproduced on retest. Confirmation of visual field abnormalities through retesting is essential for distinguishing the development or progression of glaucomatous visual field loss from long-term variability.MOGordonMAKassThe Ocular Hypertension Treatment Study: design and baseline description of the participants.Arch Ophthalmol.1999;117:573-583.MOGordonMAKassand the Ocular Hypertension Treatment Study GroupManual of Procedures.Washington, DC: National Technical Information Services; 1997. Publication PB97-148308NZ.CHolminCEKrakauVariability of glaucomatous visual field defects in computerized perimetry.Albrecht Von Graefes Arch Klin Exp Ophthalmol.1979;210:235-250.AHeijlALindgrenGLindgrenTest-retest variability in glaucomatous visual fields.Am J Ophthalmol.1989;108:130-135.BCChauhanJDTompkinsRPLeBlancTAMcCormickCharacteristics of frequency-of-seeing curves in normal subjects, patients with suspected glaucoma, and patients with glaucoma.Invest Ophthalmol Vis Sci.1993;34:3534-3540.DBHensonMNDarlingDetecting progressive visual field loss.Ophthalmic Physiol Opt.1995;15:387-390.MWallRJMawKEStanekBCChauhanThe psychometric function and reaction times of automated perimetry in normal and abnormal areas of the visual field in patients with glaucoma.Invest Ophthalmol Vis Sci.1996;37:878-885.JFlammerSMDranceFFankhauserLAugustinyDifferential light threshold in automated static perimetry.Arch Ophthalmol.1984;102:876-879.JFlammerSMDranceMZulaufDifferential light threshold.Arch Ophthalmol.1984;102:704-706.JFlammerFluctuations in the visual field.In: Drance SM, Anderson DR, eds. Automatic Perimetry in Glaucoma: A Practical Guide.New York, NY: Grune & Stratton Inc; 1985:161-173.EBWernerNSahebDThomasVariability of static visual threshold responses in patients with elevated IOPs.Arch Ophthalmol.1982;100:1627-1631.CAJohnsonAJAdamsEJCassonJDBrandtBlue-on-yellow perimetry can predict the development of glaucomatous visual field loss.Arch Ophthalmol.1993;111:645-650.PASampleGAMartinezRNWeinrebColor visual fields: a five-year prospective study in suspect eyes and eyes with primary open-angle glaucoma.In: Mills RP, ed. Perimetry Update 1992/1993.Amsterdam, the Netherlands: Kugler; 1993:473-486.JLKeltnerCAJohnsonRWBeckPAClearyJOSpurrQuality control functions of the Visual Field Reading Center (VFRC) for the Optic Neuritis Treatment Trial (ONTT).Control Clin Trials.1993;14:143-159.RALewisCAJohnsonJLKeltnerPKLabermeierVariability of quantitative automated perimetry in normal observers.Ophthalmology.1986;93:878-881.AHeijlGLindgrenJOlssonNormal variability of static perimetric threshold values across the central visual field.Arch Ophthalmol.1987;105:1544-1549.JKatzASommerA longitudinal study of the age-adjusted variability of automated visual fields.Arch Ophthalmol.1987;105:1083-1086.LBGilpinWCStewartHHHuntCDBroomThreshold variability using different Goldmann stimulus sizes.Acta Ophthalmol (Copenh).1990;68:674-676.MWallKEKutzkoBCChauhanVariability in patients with glaucomatous visual field damage is reduced using size V stimuli.Invest Ophthalmol Vis Sci.1997;38:426-435.BCChauhanPHHouseIntratest variability in conventional and high-pass resolution perimetry.Ophthalmology.1991;98:79-83.AHeijlALindgrenGLindgrenMPatellaInter-test threshold variability in glaucoma: importance of censored observations and general field estimate.In: Mills RP, Heijl A, eds. Perimetry Update 1988/89.Amsterdam, the Netherlands: Kugler; 1989:313-324.MWallCAJohnsonKEKutzkoRNguyenCBritoJLKeltnerLong- and short-term variability of automated perimetry results in patients with optic neuritis and healthy subjects.Arch Ophthalmol.1998;116:53-61.WMHartBBeckerThe onset of evolution of glaucomatous visual field defects.Ophthalmology.1982;89:268-279.EBWernerIn discussion of: Schulzer M, and the Normal-Tension Glaucoma Study Group. Errors in the diagnosis of visual field progression in normal-tension glaucoma.Ophthalmology.1994;101:1595.MKBirchPKWishartNPO'DonnellDetermining progressive visual field loss in serial Humphrey visual fields.Ophthalmology.1995;102:1227-1234.SDSmithJKatzHAQuigleyAnalysis of progressive change in automated visual fields in glaucoma.Invest Ophthalmol Vis Sci.1996;37:1419-1428.JKatzDGilbertHAQuigleyASommerEstimating progression of visual field loss in glaucoma.Ophthalmology.1997;104:1017-1025.JMWildNHutchingsMKHusseyJGFlanaganGETropePointwise univariate linear regression of perimetric sensitivity against follow-up time in glaucoma.Ophthalmology.1997;104:808-815.MSchulzerand the Normal-Tension Glaucoma Study GroupErrors in the diagnosis of visual field progression in normal-tension glaucoma.Ophthalmology.1994;101:1589-1594.BCChauhanPHHouseTAMcCormickRPLeBlancComparison of conventional and high-pass resolution perimetry in a prospective study of patients with glaucoma and healthy controls.Arch Ophthalmol.1999;117:24-33.MCLeskeAHeijlLHymanBBengtssonMHusseinand the Early Manifest Glaucome Trail GroupThe Early Manifest Glaucoma Trial: baseline results [abstract].Invest Ophthalmol Vis Sci.1999;40(suppl):S173. Association for Research in Vision and Ophthalmology abstract 926.JGFlanaganJMWildGETropeEvaluation of Fast-Pac, a new strategy for threshold estimation with the Humphrey field analyzer, in a glaucomatous population.Ophthalmology.1993;100:949-954.RPMillsHSBarnebeyCVMigliazzoYLiDoes saving time using FASTPAC or suprathreshold testing reduce quality of visual fields?Ophthalmology.1994;101:1596-1603.MSchaumbergerBSchaferBJLachenmayrGlaucomatous visual fields.Invest Ophthalmol Vis Sci.1995;36:1390-1397.CAJohnsonBCChauhanLRShapiroProperties of staircase procedures for estimating thresholds in automated perimetry.Invest Ophthalmol Vis Sci.1992;33:2966-2974.MWallJLefanteMConwayVariability of high-pass resolution perimetry in normals and patients with idiopathic intracranial hypertension.Invest Ophthalmol Vis Sci.1991;32:3091-3095.PHouseMSchulzerSDranceGDouglasCharacteristics of the normal central visual field measured with resolution perimetry.Graefes Arch Clin Exp Ophthalmol.1991;229:8-12.EGramerDKonticGKKrieglsteinComputer perimetry of glaucomatous visual field defects at different stimulus sizes.Ophthalmologica.1981;183:162-167.DBHensonJEvansBCChauhanCLaneInfluence of fixation accuracy on threshold variability in patients with open angle glaucoma.Invest Ophthalmol Vis Sci.1996;37:444-450.RJStaritaJPiltzJRLynnRLFellmanTotal variance of serial Octopus visual fields in glaucomatous eyes.Doc Ophthalmol Proc Ser.1987;49:85-90.EBWernerBPetrigTKrupinKIBishopVariability of automated visual fields in clinically stable glaucoma patients.Invest Ophthalmol Vis Sci.1989;30:1083-1089.BCChauhanSMDranceGRDouglasThe use of visual field indices in detecting changes in the visual field in glaucoma.Invest Ophthalmol Vis Sci.1990;31:512-520.RJBoeglinJCaprioliMZulaufLong-term fluctuation of the visual field in glaucoma.Am J Ophthalmol.1992;113:396-400.SDSmithJKatzHAQuigleyAnalysis of progressive change in automated fields in glaucoma.Invest Ophthalmol Vis Sci.1996;37:1419-1428.Accepted for publication March 17, 2000.This study was supported in part by grants EY09307 and EY09341 from the National Eye Institute, Bethesda, Md; an unrestricted research support grant from Research to Prevent Blindness Inc, New York, NY; and a grant from Merck & Co, Inc, Whitehouse Station, NJ.We thank John Spurr, MA, MBA, and Peter Gunther, BA, for their assistance in the visual field data analysis and in the preparation of the manuscript.Ocular Hypertension Treatment Study (OHTS) GroupClinical Centers, Investigators, and Clinic Coordinators and StaffBascom Palmer Eye Institute, University of Miami, Miami, Fla:Richard K. Parrish II, MD; Donald L. Budenz, MD; Francisco E. Fantes, MD; Steven J. Gedde, MD (investigators); Madeline L. Del Calvo, BS. M Angela Vela, MD, PC, Atlanta, Ga:M. Angela Vela, MD; Thomas S. Harbin, Jr, MD; Paul McManus, MD; Charles J. Patorgis, OD; Ron Tilford, MD (investigators); Montana L. Hooper, COT; Stacey S. Goldstein, COMT; June M. LaSalle, COA; Debbie L. Lee, COT; Michelle D. Mondshein; Emily J. Reese Smith; Julie M. Wright, COT. Cullen Eye Institute, Baylor College of Medicine, Houston, Tex:Ronald L. Gross, MD; Silvia Orengo-Nania, MD (investigators); Pamela M. Frady, COMT, CCRC; Benita D. Slight, COT, EMT-P. Devers Eye Institute, Portland, Ore:George A. (Jack) Cioffi, MD; Elizabeth Donohue, MD; Steven Mansberger, MD; E. Michael Van Buskirk, MD (investigators); Kathryn Sherman; JoAnne M. Fraser, COT. Emory University Eye Center, Atlanta:Reay H. Brown, MD; Allen D. Beck, MD (investigators); Donna Leef, MMSc, COMT; Jatinder Bansal, COT; David Jones, COT. Henry Ford Medical Center, Troy, Mich:G. Robert Lesser, MD; Deborah Darnley-Fisch, MD; Monica Gibson, MD; Nauman R. Imami, MD; James Klein, MD; Talya Kupin, MD; Rhett Schiffman, MD (investigators); Melanie Gutkowski, COMT, CO; Jim Bryant, COT; Ingrid Crystal Fugmann, COMT; Jeannine Gartner; Wendy Gilroy, COMT; Melina Mazurk, COT; Colleen Wojtala. Johns Hopkins University School of Medicine, Baltimore, Md:Donald J. Zack, MD, PhD; Donald A. Abrams, MD; Robert A. Copeland, MD; Ramzi Hemady, MD; Eve J. Higginbotham, MD; Henry D. Jampel, MD, MHS; Omofolasade B. Kosoko, MD; Scott LaBorwit, MD; Stuart J. McKinnon, MD, PhD; Irvin P. Pollack, MD; Sreedhar V. Potarazu, MD; Harry A. Quigley, MD; Alan L. Robin, MD (investigators); Rachel Scott, BS, COA; Rani Kalsi; Felicia Keel, COA; Robyn Priest-Reed, MMSc. Charles R. Drew University, Jules Stein Eye Institute, UCLA, Los Angeles, Calif:Anne L. Coleman, MD; Richard S. Baker, MD; Hyong S. Choe, MD; Y. P. Dang, MD; Ricky Hou, MD; Francis La Rosa, MD (investigators); Jackie R. Sanguinet, BS, COT; Bobbie Ballenberg, COMT; Salvador Murillo; Manju Sharma. W. K. Kellogg Eye Center, Ann Arbor, Mich:Terry J. Bergstrom, MD; Sayoko E. Moroi, MD, PhD (investigators); Carol J. Pollack-Rundle, BS, COT; Michelle A. Tehranisa, COA. Kresge Eye Institute, Wayne State University, Detroit, Mich:Dong H. Shin, MD, PhD; Bret A. Hughes, MD; Mark S. Juzych, MD; John M. O'Grady, MD; John M. Ramocki, MD; Stephen Y. Reed, MD; Dian Shi, MD (investigators); Beverly D. McCarty, LPN, ST, COA; Mary B. Hall; Laura L. Schulz, CNA; Linda A. Van Conett, COT. University of Louisville, Louisville, Ky:Robert D. Fechtner, MD; Judit Ambrus, MD; Robb Shrader, MD; Joern Soltau, MD; Gil Sussman, MD; Thom Zimmerman, MD, PhD (investigators); Sandy Lear, RN; Kathleen Coons, COT. Mayo Clinic/Foundation, Rochester, Minn:David C. Herman, MD; Douglas H. Johnson, MD; Paul H. Kalina, MD (investigators); Becky A. Nielsen, LPN; Nancy J. Tvedt. New York Eye & Ear Infirmary, New York, NY:Jeffrey M. Liebmann, MD; Robert O. Ritch, MD; Robert F. Rothman, MD; Celso Tello, MD (investigators); Kim A. Barget; Eugenie Hartman, PhD; Melissa X. Perez; Jean L. Walker, COA. Ohio State University, Columbus:Robert J. Derick, MD; N. Douglas Baker, MD; David Lehmann, MD; Paul Weber, MD (investigators); Lori Black; Mary Cassady, COA; Crystal Hendricks, COT; Tammy Lauderbaugh; Kathyrne McKinney, COMT; Diane Moore, COA. Pennsylvania College of Optometry/Allegheny University of the Health Sciences, Philadelphia:G. Richard Bennett, MS, OD; Elliot Werner, MD; Myron Yanoff, MD (investigators); Lindsay C. Bennett, BA; Mary Jameson, Opt, TR. Scheie Eye Institute, University of Pennsylvania, Philadelphia:Jody R. Piltz-Seymour, MD; Oneca Heath-Phillip, MD (investigators); Jane L. Anderson, MS; Janice T. Petner, COA. University of California, Davis, Sacramento:James D. Brandt, MD; Craig Bindi, MD; Jeffrey J. Casper, MD; Janet Han, MD; Denise Kayser, MD; Sooyung Kim, MD; Alan M. Roth, MD; Ivan R. Schwab, MD (investigators); Ingrid J. Clark, COA; Vachiraporn X. Jaicheun, COA; Denise M. Owensby, BS, COA. University of California, San Diego, La Jolla:Robert N. Weinreb, MD; J. Rigby Slight, MD (investigators); Rivak Hoffman, COT; Dawn D. Frasier; Barbara Brunet; Julia Williams. University of California–San Francisco:Michael V. Drake, MD; Allan J. Flach, MD; Robert Stamper, MD (investigators); Lou Anne Aber, COA; Peggy Yamada, COT. University Suburban Health Center, South Euclid, Ohio:Kathleen A. Lamping, MD; Laurence D. Kaye, MD (investigators); Sheri Burkett-Porter, COA; Carla De La Rosa Valenti; Angela K. McKean; Laura Brevard, COT; Susan Van Huss. Washington OHTS Center, Washington, DC:Douglas E. Gaasterland, MD; Frank S. Ashburn, MD; Arthur Schwartz, MD; Howard S. Weiss, MD (investigators); Anne M. Boeckl, MS; Robin Montgomery; Donna Claggett; Deanne Griffin; Karen D. Schacht, COT. Washington University School of Medicine, St. Louis, Mo:Martin B. Wax, MD; Edward Barnett, MD; Michael A. Kass, MD; Allan E. Kolker, MD; Carla J. Siegfried, MD (investigators); Arnold D. Jones, COA; Lori A. Clark, COT; Forunata Darmody, COT.CommitteesExecutive/Steering Committee:Douglas R. Anderson, MD; Anne Coleman, MD; Michael Drake, MD; Donald F. Everett, MA; Mae E. Gordon, PhD; Dale K. Heuer, MD; Eve J. Higginbotham, MD; Chris A. Johnson, PhD; Michael A. Kass, MD; John L. Keltner, MD; Richard K. Parrish II, MD; Arthur Shedden, MD; M. Roy Wilson, MD (investigators); Carol J. Pollack-Rundle, COT; Patricia A. Morris; Ann K. Wilder, RN, BSN.Data and Safety Monitoring Committee:Roy Beck, MD, PhD; John Connett, PhD; Claude Cowan, MD; Barry Davis, MD, PhD; Donald F. Everett, MA (non-voting); Mae O. Gordon, PhD (non-voting); Michael A. Kass, MD (non-voting); Ronald Munson, PhD; Arthur Shedden, MD (non-voting); Mark Sherwood, MD; Gregory L. Skuta, MD (investigators).End Point Committee:Dale Heuer, MD; Eve Higginbotham, MD; Richard K. Parrish II, MD; Mae O. Gordon, PhD (investigators).Resource CentersWashington University School of Medicine, Coordinating Center:Mae O. Gordon, PhD; J. Philip Miller (investigators); Joel Achtenberg, MSW; Mary Bednarski, MAS; Julia Beiser, MS; Karen Clark; Christopher Ewing; Ellen Long, CCRA; Patricia Morris; Denise Randant; Ann K. Wilder, RN, BSN; Chairman's Office:Michael A. Kass, MD (investigator); Deborah Dunn; Carolyn Miles.Project Office, National Eye Institute, Rockville, Md:Donald F. Everett, MA (investigator).Optic Disc Reading Center, Bascom Palmer Eye Institute, University of Miami:Richard K. Parrish II, MD; Douglas R. Anderson, MD; Donald L. Budenz, MD (investigators); Maria-Cristina Wells, MPH; William Feuer, MS; Ditte Hess, CRA; Heather Johnson; Joyce Schiffman, MS; Ruth Vandenbroucke.Visual Field Reading Center: University of California, Davis, Sacramento:John L. Keltner, MD (investigator), and Discoveries in Sight, Devers Eye Institute:Chris A. Johnson, PhD (investigator); Kimberly E. Cello, BS; Bhupinder S. Dhillon, BSc; Denise M. Owensby, BS; Jacqueline M. Quigg, BS.Ancillary Study Reading CentersConfocal Scanning Laser Ophthalmoscopy Reading Center, University of California, San Diego:Robert N. Weinreb, MD; Linda Zangwill, PhD (investigators); Keri Dirkes, MPH; Chris Asvar.Short Wave Length Automated Perimetry Reading Center, Devers Eye Institute, Legacy Portland Hospitals:Chris A. Johnson, PhD (investigator); Erna Hibbitts.Corneal Endothelial Cell Density Reading Center, Mayo Clinic/Foundation:William M. Bourne, MD (investigator); Becky Nielsen, LPN; Thomas P. Link, CRA, BA; Jay A. Rostvold.Reprints: John L. Keltner, MD, Department of Ophthalmology, University of California, Davis, 4860 Y St, Suite 2400, Sacramento, CA 95817 (e-mail: [email protected]).

Holbrook, Janet T.; Davis, Matthew D.; Hubbard, Larry D.; Martin, Barbara K.; Holland, Gary N.; Jabs, Douglas A.; Gilpin, Adele Kaplan; Meinert, Curtis; Reshef, Daniel S.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1196pmid: 10980764

ObjectiveTo identify ocular and systemic factors that predict advancement of cytomegalovirus (CMV) retinitis during treatment.MethodsPatients with acquired immunodeficiency syndrome were enrolled in a multicenter clinical trial designed to evaluate foscarnet sodium and ganciclovir sodium as therapy for newly diagnosed CMV retinitis. Ocular characteristics at baseline and measurements of retinitis were assessed from fundus photographs by graders at a fundus photograph reading center. The following measures of advancement were assessed: (1) lesion border movement of at least 750 µm or development of a new lesion in involved eyes; (2) rate of increase in retinal area with CMV in involved eyes; and (3) development of retinitis in uninvolved eyes of patients with unilateral disease at baseline.ResultsIn eyes with retinitis, risk factors at baseline for advancement while receiving treatment included smaller area involved, active margins of retinitis, and posterior location. Risk factors for development of retinitis in uninvolved fellow eyes included blood and urine cultures positive for CMV and lower CD8+T-lymphocyte count.ConclusionsLesion characteristics can be used to predict advancement of preexisting disease, whereas only systemic factors are associated with development of bilateral disease. These analyses describe retinitis activity before the introduction of potent antiretroviral therapies but provide an important reference point for patients in whom CMV retinitis develops after failure or intolerance of antiretroviral agents.BENEFICIAL effects of therapy for cytomegalovirus (CMV) retinitis in patients with acquired immunodeficiency syndrome (AIDS) have been identified in several clinical trials.However, little information has been published on factors that might predict early progression or more rapid spread of retinitis during therapy. Identification of risk factors for early enlargement of existing lesions or development of a new lesion in uninvolved eyes might assist in planning examination schedules and treatment regimens for individual patients.The Foscarnet-Ganciclovir CMV Retinitis Trial (FGCRT) was designed to compare the safety and efficacy of intravenous foscarnet sodium and ganciclovir sodium for treatment of newly diagnosed retinitis. The protocol for the trial was suspended in 1991 due to excess mortality among patients assigned to ganciclovir, but no difference in retinitis progression or visual outcomes was noted between patients assigned to foscarnet and those assigned to ganciclovir. Previously reported patient characteristics associated with greater risk of retinitis progression in this study included lower CD4+T-lymphocyte counts, blood cultures positive for CMV, and bilateral disease.For our report, data from the FGCRT were analyzed to identify ocular characteristics present at diagnosis of CMV retinitis that were associated with earlier progression or more rapid spread of retinitis in involved eyes. In addition, we attempted to identify ocular and systemic risk factors specific for development of retinitis in fellow eyes of patients with unilateral disease at initial diagnosis. The information presented herein applies to patients in whom potent antiretroviral therapies fail or who are intolerant of them and to patients in whom CMV retinitis develops as a result of renewed or persistent immune dysfunction.PATIENTS AND METHODSPatients at 11 clinical centers with newly diagnosed, untreated CMV retinitis were eligible for the FGCRT. Descriptions of the design, procedures, and findings from the FGCRT are published elsewhere.To evaluate changes in the manifestation of retinitis over time, baseline data from patients with newly diagnosed, untreated retinitis enrolled in the later Studies of Ocular Complications of AIDS Research Group trial, the Monoclonal CMV Retinitis Trial (MACRT) at 15 clinical centers also underwent analysis.The protocols for the trials and the consent forms were reviewed and approved by institutional review boards at the coordinating center and at the clinical centers.For evaluations of retinitis, the retina was divided into 3 zones. Zone 1 included the area within 1500 µm of the margin of the optic disc or within 3000 µm of the center of the macula; zone 2 extended from the limits of zone 1 to a circle passing through the ampullae of the vortex veins; and zone 3 extended peripherally from zone 2 to the ora serrata.For our analyses, the macula was defined as the area within 1000 µm of the center of the fovea.Patients were assigned randomly to treatment with intravenous ganciclovir or foscarnet (allocation ratio, 1:1). Study visits were scheduled at enrollment, every 2 weeks for the first 8 weeks, then every month for up to 6 months, and every 2 months after that. For all patients who required multiple courses of induction therapy, additional follow-up visits were scheduled at the beginning of each course of induction therapy. Indirect ophthalmoscopic examinations and fundus photography were required at every study visit.The fundus photography protocol specified that photographs of 9 fields be taken with a wide-angle fundus camera, ie, stereoscopic photographs of the disc and macula surrounded by nonstereoscopic photographs of 8 peripheral fields.Fundus photographs were evaluated at the fundus photograph reading center (FPRC) by graders unaware of treatment assignment. The photograph grading protocol included assessments of various features of CMV retinitis, principally, location of lesions and appearance of their borders, degree of activity, extent of satellites (small foci of retinitis surrounded by normal-appearing retina), prominence of hemorrhage, and, for follow-up photographs, progression of retinitis.RETINITIS MEASURESThe extent of retinitis in each eye was determined by FPRC graders with planimetric measurements of a digitized mosaic of the retina created from fundus photographs. This measure was defined as the percentage of total retinal area in zones 1 and 2 with retinitis, ie, area with retinitis divided by the total area of zones 1 and 2 of the retina. Zone 3 was not evaluated because it could not be depicted completely in the fundus photographs. Measurements of percentage of retina with retinitis were made at trial enrollment (baseline) and at 3 and 6 months of follow-up. Data on area with retinitis in zones 1 and 2 were included if baseline and follow-up photograph sets were available in which 85% or more of the area within these zones could be evaluated. For 18 patients who participated in a substudy examining deferral of anti-CMV therapy, measurements of retinal area with retinitis were determined at the time of enrollment rather than at the start of treatment (median length of deferral was 21 days). The FPRC graders assessed progression or occurrence of a new lesion by comparison of photographs taken at follow-up visits with those taken during the 5 days before randomization. The FPRC graders were not informed of clinician assessments of retinitis, but were aware when patients underwent additional courses of induction therapy (by the visit identification code). Laterality of retinitis at baseline and during follow-up was determined by evaluation of the FPRC and clinician assessment of retinitis. In discrepant cases, baseline and follow-up data from ophthalmologic examinations and photograph evaluations were reviewed to determine whether discrepancies were due to false-positive or false-negative observations.The following retinitis outcomes were defined: (1) progression in an involved eye, defined as the movement of a border at least 750 µm along a front 750 µm or greater in length, or the occurrence of a new lesion at least 750 µm in diameter and separated from a previous lesion by 750 µm or more; (2) rate of increase in the area of retina with retinitis (ie, spread) in the same eye, defined as the difference between the total percentage of retinal area with retinitis at the previous measurement time and the percentage at the current visit divided by the number of months between measurements; and (3) development of CMV retinitis in the uninvolved fellow eye.For observations with an apparent decrease of 5% or less in the area of retina with retinitis (40 observations from 39 eyes of 35 patients), the rate of increase was assigned a value of 0% per month. In cases where there was an apparent decrease of more than 5% in retinal area involved at 3 or 6 months of follow-up (15 observations from 14 eyes of 14 patients), photographic mosaics were reviewed again. In all of these cases, there were errors in the evaluation of the baseline or follow-up photographs that resulted in the apparent decrease in retinal area. These errors were corrected.ANALYSISFor most analyses, data from eyes with retinitis at baseline as determined by FPRC graders were used. Categorical and ordinal baseline variables were collapsed and combined according to their meaning and their frequency distributions. Missing data were imputed to the median or most common category if less than 10 observations had missing data; otherwise, a dummy variable was created to represent a separate category for missing data. Percentage area of retina with retinitis at baseline was logarithm-transformed for analysis. Associations between ocular characteristics at baseline were examined with linear regression models, Kruskal-Wallis tests, and χ2tests.Baseline characteristics were examined for associations with first progression or development of retinitis in previously uninvolved fellow eyes with the Kaplan-Meier procedure and proportional hazards regression.Variables for which there was evidence that a linear model was not appropriate were treated as categorical variables for further evaluations; otherwise, the continuous variable was used for model-building analysis. Event rates were calculated as the number of events divided by the number of person-years.Variables were selected for inclusion in multivariate models with stepwise regression (forward selection) and best-subset selection procedures.Variables relating to very infrequent characteristics (ie, <4% of eyes exhibited the characteristic) were not included in model selection procedures. The stability of the final models was checked by excluding observations with large residuals (≥2), which may have had relatively strong influence on overall model results, and by repeating the stepwise regression procedure. For presentation of results, continuous variables were transformed into categorical variables.For the continuous outcome of rate of increase in retinal area with retinitis, variance of the median was calculated with the following formula: π/2 × &b.sigma;2/n, where π equals 3.141 and &b.sigma;2is the variance associated with the rate of increase.Associations between baseline characteristics and rate of increase in retinal area with retinitis were investigated with Kruskal-Wallis tests and linear regression models.The stepwise procedure for linear regression was used to select baseline variables for a multivariate model; ranks of the rate of increase in retinal area involved were used as the outcome for the selection procedure. The selection procedure was repeated on the subsets of data, with exclusion of outlier observations, observations with high leverage, or observations with rates of increase equal to zero.For all models, a variable was included in a multivariate regression model if the Pvalue for inclusion and the Pvalue for the Wald test for the covariate was .10 or less. Robust techniques for variance estimation were used.Interactions between variables selected for the model, zone of disease, and bilateral status were checked. In the final models, CD4+T-lymphocyte counts and CMV blood culture status at baseline were included because these characteristics were shown in previous analyses to be related to time to first progression.RESULTSOf the 234 patients randomized to treatment in the FGCRT, 224 had baseline and follow-up photographic data available for analysis. Baseline and follow-up data were available for 318 eyes with retinitis at enrollment and 120 fellow eyes with no retinitis at enrollment. Selected baseline characteristics of the 224 patients included in these analyses are presented in Table 1.Table 1. Baseline Characteristics of Patients With CMV Retinitis*No. (%)Patients224 (100.0)Trial treatment assignmentFoscarnet sodium104 (46.4)Ganciclovir sodium120 (53.6)Retinitis characteristicsBilateral retinitis†104 (46.4)Zone 1 involved148 (66.1)CMV culture resultsBlood and urine positive59 (26.3)Blood positive only22 (9.8)Urine positive only61 (27.2)Blood and urine negative52 (23.2)Both missing30 (13.4)CD4+T-lymphocyte counts<0.014 × 109/L76 (33.9)≥0.014 × 109/L73 (32.6)Missing75 (33.5)CD8+T-lymphocyte counts≤0.20 × 109/L55 (24.6)0.201 × 10/L to 0.38 × 109/L45 (20.1)>0.38 × 109/L44 (19.6)Missing80 (35.7)*CMV indicates cytomegalovirus. Percentages have been rounded and may not sum 100.†Based on fundus photograph reading center and clinician assessments.CHARACTERISTICS OF EYES WITH RETINITISSelected characteristics of the 318 eyes with retinitis at baseline are presented in Table 2. Associations between retinitis characteristics and other ocular abnormalities were examined. The median percentage of zones 1 and 2 with retinitis was 10% in the 307 eyes with these data. Area involved in zones 1 and 2 was larger in eyes with hemorrhagic lesion borders, perivascular cuffing, vitreous haze, or retinal surface abnormalities. Lesion borders were active along at least 90% of their visible extent in 80.5% of 318 eyes. In eyes with more narrow (≤2250 µm) borders or without border satellites, the visible border exhibited less activity (<90%) more frequently. Overall, 77.7% of eyes had retinal hemorrhage at the lesion borders. Perivascular cuffing, present in 20.8% of all involved eyes, was more frequent when lesion borders were wide and hemorrhagic.Table 2. Baseline Characteristics From FPRC Evaluations of Photographs of Eyes Involved With CMV Retinitis*No. (%)All involved eyes318 (100.0)LocationZone 1 involved168 (52.8)Zone 2 involved304 (95.6)Zone 3 involved266 (83.6)Macular involvementWithin 1000 µm of center42 (13.2)Within 250 µm of center15 (4.7)Optic disc involvement30 (9.4)ActivityBorder activity<90% of lesion border active51 (16.0)≥90% of lesion border active256 (80.5)Missing data11 (3.5)Width of solid border≤2250 µm124 (39.0)>2250 µm194 (61.0)SatellitesNo satellites23 (7.2)Satellites present283 (89.0)Missing data12 (3.8)Hemorrhagic lesion border247 (77.7)Perivascular cuffing66 (20.8)Vitreous haze50 (15.7)OtherRetinal surface abnormalities24 (7.5)Subretinal hemorrhage8 (2.5)Vitreous hemorrhage7 (2.2)Hemorrhage or microaneurysms128 (40.3)Cotton wool spots145 (45.6)*FPRC indicates fundus photograph reading center; CMV, cytomegalovirus.To evaluate secular trends in the characteristics of retinitis at diagnosis, ocular characteristics of involved eyes were compared with those of 73 patients (100 involved eyes) with newly diagnosed retinitis who enrolled in the MACRT from September 14, 1995, through July 30, 1996. Both sets of eyes had similar characteristic profiles except that eyes of patients enrolled in the MACRT were less likely to have zone 1 involvement (41.4% vs 52.8%; P= .05), and to exhibit perivascular cuffing (8.0% vs 20.8%; P= .004). In eyes with retinitis at baseline, the median percentage of zones 1 and 2 with retinitis was similar to that for patients enrolled in the FGCRT (9% vs 10%; P= .93).PROGRESSION IN INVOLVED EYESProgression, while receiving treatment, in involved eyes before first reinduction was observed by the FPRC in 232 (73.0%) of 318 eyes involved at baseline (Table 3). Median time to progression was 59 days (95% confidence interval [CI], 52-70 days). Most progression determinations (223/232 [96.1%]) were based on border movement of at least 750 µm. In 18 (7.8%) of 232 eyes, a new lesion in the same eye was observed at the same time border movement was observed. Only 9 (3.9%) of 232 progressions were based on the appearance of new lesions alone. In 6 (22.2%) of 27 eyes in which a new lesion was observed at the time of first progression, the lesion was in zone 1.Table 3. Baseline Risk Factors for Progression and Rate of Increase in Area in Involved Eyes*FPRC-Assessed ProgressionRate of IncreaseNo. of Events/ No. of EyesRate†RR‡P§No. of Eyes&par;% of Area/ mo¶β#P**All involved eyes232/3183.2NANA2252.5NANAArea involved (zones 1, 2)<5%††85/1045.21.0NI733.40NI5%-15%69/923.60.7.01622.7−.6.16>15%71/1112.10.5<.001902.2−1.7.002Missing7/112.60.6.330NININIZone 1 involvedNot involved††114/1503.5NI‡‡NI1062.1NI‡‡NIInvolved118/1683.2NINI1193.1NINIDistance from optic diskAt disk††55/842.7NI‡‡NI553.30NI<1500 µm52/713.2NINI533.1−.9.351500-4300 µm64/873.2NINI692.3−1.4.11>4300 µm51/644.2NINI421.9−2.1.003Missing10/113.4NINI62.5−1.5.32Border activity<90% Border active††36/512.11.0NI412.0NI‡‡NI≥90% Border active186/2563.71.61.01792.5NINIMissing7/111.90.7.2752.2NINIBorder satellitesAbsent††16/232.3NI‡‡NI170.80NIPresent209/2833.5NINI2002.72.3<.001Missing7/121.7NINI80.8−.1.94CMV blood culture statusNegative††95/1362.61.0NI922.10NIPositive97/1224.01.4.03963.1.4.61Missing40/604.01.2.51372.8.1.33CD4+T-Lymphocyte count<0.014 × 109/L‡81/1024.61.0NI753.50NI≥0.014 × 109/L68/1042.00.6.009761.5−1.5<.001Missing83/1124.40.9.71743.5−.1.29Treatment assignmentFoscarnet sodium††112/1543.21.0NI1082.30NIGanciclovir sodium120/1643.31.0.951172.50.96*FPRC indicates fundus photograph reading center; RR, relative risk; NA, not applicable; and NI, not included.†Indicates rate of events per person-year of follow-up.‡Estimated with proportional hazards regression models adjusted for ocular and patient characteristics listed.§Robust Pvalues were based on Wald tests.&par;There were 356 observations for 225 eyes: 197 at 3 months and 159 at 6 months.¶Median rate of change.#Coefficients were estimated with a linear regression model and adjusted for the ocular and patient characteristics listed; they estimate the adjusted difference in the rate of increase in retinitis area (percentage of area per month) between the indicated subgroup and the reference group.**Robust Pvalues were calculated with general estimating equations procedures.The ranks of the rate of increase were used as the outcome for estimating Pvalues.††Reference group for estimation of RR or β for each subgroup; by definition, the RR of the reference group is 1 and the β is 0.‡‡The covariate did not meet the inclusion criteria for the mutivariate model for this outcome.RISK FACTORS FOR FIRST PROGRESSION IN INVOLVED EYESAll the characteristics of involved eyes listed in Table 2were examined for association with progression. In univariate analyses, smaller area involved (as measured by total percentage of zones 1 and 2 with retinitis), absence of zone 3 involvement, and more active borders (defined as activity along ≥90% of the border surrounding the lesion) were associated with earlier first progression. The association between area involved and time to first progression is shown in Figure 1. Based on the results from multivariate models, retinitis characteristics independently associated with earlier progression were smaller area involved and activity along at least 90% of lesion borders (Table 3). On average, eyes with more than 15% of zones 1 and 2 involved were at half the risk of progression, as were eyes with less than 5% involved. Eyes with at least 90% of lesion borders graded as active were 1.5 times as likely to have progression than were eyes with less active borders. Among the systemic variables included in the multivariate model, higher CD4+T-lymphocyte counts and blood culture negative for CMV at baseline were associated with a lower risk of progression. Exclusion from the data set of 10 observations with large residuals (≥2) did not change the results of the model selection procedures.Figure 1.Kaplan-Meier estimates for time to first progression of cytomegalovirus retinitis in involved eyes, stratified by area in zones 1 and 2 with retinitis at baseline (small, <5%; medium, 5%-15%; and large, >15%). Bottom, Number of eyes, number of progressions (events), and progressions per person year (rate) for 6-month intervals. P<.001, log rank test.We checked for evidence that unmeasured area with retinitis in zone 3, the presence of retinitis satellites, or bilateral disease status might explain the relative risk of progression associated with area involved. There was no evidence that the presence or absence of retinitis in zone 3, the presence or absence of satellites, or bilateral disease status influenced the relative risk of progression associated with area of retina with retinitis (P= .72, P= .13, and P= .20, respectively).RATE OF INCREASE IN RETINAL AREA WITH RETINITISBaseline and follow-up data to estimate rates of increase in area of retina involved for eyes with retinitis at baseline were available for 225 eyes of 164 patients, which was 70.8% of eyes and 73.2% of patients with baseline and follow-up fundus photographs. Eyes from patients with longer follow-up times were more likely to have these data available, because follow-up assessments of area were made after 3 and 6 months of follow-up.The median percentage of retinal area (zones 1 and 2) with retinitis was 10% at baseline (95% CI, 8%-13% [n = 225]), 21% at 3 months (95% CI, 16%-25% [n = 197]), and 30% at 6 months (95% CI, 21%-39% [n = 159]). The median rate of increase per month in the percentage of retina with retinitis during all follow-up was 2.5% per month (95% CI, 2.3%-2.6% per month). At 3 months, the median rate of increase was 2.6% per month (95% CI, 2.3%-2.8% per month), and at 6 months, it was 2.3% per month (95% CI, 2.0%-2.6% per month), and these rates were not significantly different (P= .23).RISK FACTORS FOR RATE OF INCREASE IN RETINAL AREA INVOLVEDAll the photographic and systemic characteristics examined for an association with first progression also were examined for an association with rate of increase in area in involved eyes. Results of univariate analyses were similar to those for time to first progression, except that the presence of satellites and border hemorrhage were significant risk factors and extent of border activity was not. Based on multivariate models, retinitis characteristics independently associated with a more rapid rate of increase in retinal area involved while receiving treatment were smaller area involved, shorter distance between the optic disc and the closest lesion border, and the presence of satellites along the retinitis borders (Table 3). The β coefficient associated with each characteristic is the average adjusted difference in rate of increase in retinal area involved (percentage per month) between the indicated category and the reference category. For example, the average rate of increase for eyes with greater than 15% of retinal area with retinitis at baseline was 1.8% per month less than the average rate in eyes with less than 5% involved (Table 3). The rate of increase in eyes in which the closest lesion to the optic disc was more than 4300 µm away was, on average, 2.0% per month less than in eyes with lesions adjacent to the disc. The presence of border satellites was associated with an average increase in that rate of 2.1% per month.The association of retinal area with the rate of increase was not different in eyes with zone 3 involvement vs those without (P= .18), nor was it different in eyes with satellites vs those without (P= .38). However, there was weak evidence that the negative association between area involved and rate of increase was stronger in eyes of patients with unilateral disease at baseline than in those of patients with bilateral disease (P= .05).After adjustment for ocular characteristics found to be associated with rate of increase, patients with higher CD4+T-lymphocyte counts at baseline had, on average, lower rates of increase. Treatment assignment (foscarnet or ganciclovir) and CMV blood culture status at baseline were not associated with the rate of retinitis increase during follow-up.We assessed the stability of the final model by deleting outliers for the outcome, observations with high leverage, and observations for which the rate of increase was zero. Exclusion of 2 observations with outlier values for rate of increase from the data set did not change the results of the model selection procedures. Exclusion of the 13 observations with outlier values for leverage resulted in the selection of optic disc involvement in addition to the previously selected covariates. Exclusion of the 40 observations (25 at 3 months and 15 at 6 months) from 39 eyes (35 patients) in which no retinitis advancement was observed resulted in the omission of area involved and distance from the optic disc from the selected covariate set and the selection of optic disc involvement. The median percentage of retinal area involved at baseline was larger for the omitted 39 eyes than for the other 214 eyes (17% vs 10%; P= .02).DEVELOPMENT OF RETINITIS IN AN UNINVOLVED FELLOW EYEOf the 224 patients included in these analyses, 120 had unilateral disease at randomization to treatment. Excluded from this analysis were 7 patients with unilateral disease based on FPRC assessment but bilateral disease based on clinician assessment, 1 patient in whom bilateral disease developed during deferral of treatment, and 1 patient for whom the FPRC was uncertain as to the disease status at baseline. Cytomegalovirus retinitis developed in 31 (25.8%) of the 120 uninvolved eyes during follow-up (Figure 2). The median time to development of a lesion in the uninvolved eye was 23 months (95% CI, 13 months-∞). There was a suggestion that patients in whom bilateral disease developed during follow-up had a substantially longer median survival than patients in whom it did not (16 vs 9 months; P= .13). New lesions in eyes not involved at baseline involved zone 1 in 4 eyes (12.9%); zone 2, but not zone 1, in 16 eyes (51.6%); and zone 3 only in 11 eyes (35.5%).Figure 2.Kaplan-Meier estimates of time to development of cytomegalovirus retinitis in fellow eyes not involved at baseline (fundus photograph reading center and clinician assessments). Bottom, Number of eyes, number with new retinitis (events), and the number of events per person-year (rate) for 6-month intervals.RISK FACTORS FOR DEVELOPMENT OF A LESION IN AN UNINVOLVED FELLOW EYECharacteristics examined for an association with more rapid retinitis spread in an involved eye also were examined for an association with development of retinitis in an uninvolved eye while receiving treatment. In addition, the following systemic characteristics were examined: months since the diagnosis of AIDS, number of opportunistic infections diagnosed before retinitis diagnosis, CD4+and CD8+T-lymphocyte counts and other hematologic measures, predicted creatinine clearance, and blood and urine CMV culture status at baseline. Karnofsky score was not included in this analysis because the variable appeared to violate the proportional hazards assumption (P= .009). Those patients with higher Karnofsky scores at baseline (≥90) were at higher risk for development of retinitis in an uninvolved eye than were patients with lower Karnofsky scores in the first 7 months of follow-up but at lower risk after that time.In univariate analyses, blood and urine cultures both positive for CMV, urine cultures positive for CMV with negative or missing results for blood culture, lower CD8+T-lymphocyte counts, and presence of satellites at retinitis borders were associated with a greater risk of development of bilateral disease. The CD4+T-lymphocyte counts were not associated with development of bilateral disease (P= .72). In the multivariate model, no ocular characteristics of involved eyes of patients with unilateral retinitis were associated with the development of retinitis in previously uninvolved fellow eyes. Systemic characteristics at baseline identified as independent risk factors for development of bilateral disease were blood and urine cultures positive for CMV, lower CD8+T-lymphocyte count, and no previous AIDS-related opportunistic infections at diagnosis of retinitis (Table 4).Table 4. Risk Factors for Development of Bilateral CMV Retinitis in Patients With Unilateral Disease at Baseline*No. of Events/ No. of PatientsRate†RR‡P§All31/1200.4NANACD8+T-lymphocyte count≤0.20 × 109/L&par;15/360.81.0. . .0.201 × 109/L to 0.38 × 109/L9/310.51.1.88>0.38 × 109/L6/350.20.2.01Missing1/180.10.1.03CD4+T-Lymphocyte count≤0.014 × 109/L&par;18/550.61.0. . .>0.014 × 109/L13/500.30.7.43Missing0/150.0&par;&par;Blood and urine CMV culture resultsBoth negative¶6/330.21.0. . .Blood and urine positive11/260.75.0.006Blood positive1/90.22.0.55Urine positive9/350.43.2.06Missing4/170.42.7.16No. of opportunistic infectionsNone&par;21/650.51.0. . .≥110/550.30.3.002Treatment assignmentFoscarnet sodium¶14/510.41.0. . .Ganciclovir sodium17/690.41.4.38*CMV indicates cytomegalovirus; FPRC, fundus photograph reading center; and RR, relative risk. Assessments were performed by FPRC graders or clinicians.†Rates were calculated as events per person-year.‡Estimated with proportional hazards regression procedures and were adjusted for all listed covariates.§Pvalues were based on Wald tests.&par;RR and Pvalue could not be calculated.¶Reference group; RR for reference group is 1.0 by definition and noPvalue was calculated.In the final model that included these variables and treatment assignment, patients with the highest CD8+T-lymphocyte counts (>0.38 × 109/L) were at lowest risk for development of retinitis in the other eye; their risk was one fifth that of patients with the lowest CD8+counts (≤0.20 × 109/L). Patients with blood and urine cultures both positive for CMV at baseline were almost 5 times as likely to have development of retinitis in the other eye as were patients with negative results of cultures. Patients with urine cultures positive for CMV and a negative or missing result of blood culture at diagnosis were more than 3 times as likely to have development of retinitis in the other eye as were patients with negative results of cultures. The risk of bilateral retinitis in patients with a diagnosis of at least 1 opportunistic infection before the diagnosis of CMV retinitis was about one third that of patients with no other opportunistic infections at diagnosis. Treatment assignment was not associated with development of bilateral disease.COMMENTThe objectives of these analyses were (1) to identify characteristics of patients with newly diagnosed CMV retinitis that were predictive of retinitis progression or spread in involved eyes, and (2) to identify characteristics predictive of development of retinitis in uninvolved eyes in patients being treated for CMV retinitis.The baseline risk factors identified for both measures of retinitis advancement in involved eyes while receiving treatment, progression and the rate of increase in area affected by retinitis, were smaller area of retina with retinitis and signs of greater border activity, as well as lower CD4+T-lymphocyte counts. More posterior location of retinitis was only associated with rate of increase in area involved. Inclusion of these baseline risk factors in the models with treatment assignment did not alter the overall conclusion of the trial that intravenous foscarnet and ganciclovir were equally efficacious for treatment of CMV retinitis (Table 3and Table 4).The associations of border activity, posterior location, and CD4+T-lymphocyte counts were not surprising, but the association of smaller area of retinitis with more rapid advancement of retinitis initially seems counterintuitive. The apparent effect of the size of the area involved may be related to the disruption of retinal circulation caused by the full-thickness retinal necrosis characteristic of CMV retinitis.Larger lesions may enhance drug delivery because of more extensive circulatory disruption. Hence, replication of virus may be inhibited more quickly in these eyes. It is also possible that smaller area of involvement also may be a marker for new disease, which may spread more rapidly because an immune response, albeit small, has just begun. The evidence of this hypothesis was limited; some evidence (P= .05) suggested that the association between smaller area of involvement and faster rate of increase in area was greater for eyes of patients with unilateral disease, which we used as a marker for new disease, than for eyes of patients with bilateral disease. However, there was less evidence to suggest that the association of area of involvement and time to first progression were influenced by laterality of disease (P= .20).The association of smaller area involved at baseline with faster retinitis progression or more rapid rates of spread could be an artifact of measurement procedures. An identical amount of border movement may be easier to detect for a smaller lesion. Observations in which no increase in area was noted during follow-up tended to be from eyes with a larger area of involvement at baseline. Eyes with larger areas of retina involved were more likely to have zone 3 involvement, where border advancement could not be measured reliably, but we found little evidence that retinitis progression rate or rates of increase in area differed in eyes with retinitis that extended into zone 3 than in eyes with similar areas of involvement without zone 3 involvement.A faster rate of increase in retinal area with retinitis was weakly associated with each of the (highly correlated) measures of posterior location (involvement of zone 1, involvement of the macula or disc, and shorter distance from the disc); shorter distance from the disc was selected for inclusion in the multivariate model (Table 3). Posterior location also was associated with clinician-assessed progression (data not shown). The association of closer proximity to the optic disc with faster rates of increase in retinal area with retinitis may have been affected by the poorer quality of peripheral photographs. Alternatively, faster spread of retinitis located near the disc in treated patients might be related to greater thickness and vascularity of the retina here, which might provide more opportunity for spread along vessels.A retrospective study of patients with untreated retinitis or active retinitis despite treatment with intravenous ganciclovir suggested that border advancement toward the anterior retina was more rapid than advancement toward the posterior retina.Our study did not analyze the direction of border advancement. Both studies found that the location of the retinitis lesions at baseline was not associated with retinitis progression and that border activity was associated.Development of new lesions in a previously uninvolved eye was an infrequent and late-stage event (median time to event, 23 months) occurring in about a quarter of the patients with unilateral CMV retinitis. The overall median survival of patients with unilateral disease was substantially shorter (10.3 months), so the 23-month estimate is likely to be influenced by a survival bias. Regardless, longevity is probably an important risk factor for development of bilateral disease. Other risk factors independently associated with development of bilateral disease were blood and urine CMV cultures both positive for CMV, urine culture only positive for CMV, lower CD8+T-lymphocyte counts, and fewer AIDS-related opportunistic infections (Table 4). These risk factors were not related to development of bilateral disease via an association with longer survival. Cultures positive for CMV and lower CD8+T-lymphocyte counts at baseline were associated with shorter survival, and the number of AIDS-related opportunistic infections was not associated with survival.Blood cultures that are positive for CMV have been shown to be associated with poorer prognosis for CMV retinitis.Positive cultures may indicate higher viral load in the blood and more severe immune suppression, which may increase the likelihood of hematogenous spread of virus to an uninvolved eye. Decrease in CD8+T-lymphocyte counts is a marker for immunosuppression, as indicated by the increased risk of CMV retinitis and shorter survival.The finding that fewer AIDS-related opportunistic infections is a risk factor for bilateral disease is paradoxical. In view of the small number of patients included in these analyses and the many statistical tests performed, this could be a chance finding.The rate of increase in retinal area in zones 1 and 2 with retinitis was constant during the first 6 months of follow-up; the median was 2.6% of retinal area per month. This finding seems to conflict with a previous report from this trial that indicated a trend toward shorter times to progression from the first to the second progression (48 to 41 days) and from the second to the third progression (41 to 35 days).Those findings indicate that the rate of increase in retinitis area might be expected to accelerate during follow-up. The conflict may, in part, be due to different data sets, analysis approaches, and measures. The previously reported retinitis progression results were based on analysis of patients' time to progression in either eye. The current analyses used eyes involved at baseline as the unit of analysis. Progression is a binary outcome based on a relatively small amount of linear movement along a segment of a lesion border and can be due to irregular border advancement that does not correlate with overall border movement, whereas rate of increase in area is a continuous outcome that integrates the border movement along the entire visible circumference of lesions during longer periods of follow-up.There are limitations to our study that should be kept in mind when interpreting these results. Data were collected as part of a clinical trial comparing treatments for patients with previously untreated CMV retinitis, ie, data collection was not designed with these analyses in mind. The current analyses were limited to factors for which data were available and assumed that there were no unknown confounders such as viral strain heterogeneity or host immunologic factors. The analyses were exploratory in nature with few a priori hypotheses. There was low statistical power to detect associations for some variables, such as optic disc involvement. These patients experienced a high mortality rate, reducing the number who could be observed for longer-term retinitis outcomes; the results therefore may be influenced by a survival bias. These data were collected before the availability of potent antiretroviral agents and combination antiretroviral therapies in 1996. Therefore, they may not reflect the course of retinitis in patients receiving these medications. However, the ocular and retinitis characteristics were quite similar to those of patients with newly diagnosed retinitis in 1995 and 1996, some of whom received potent antiretroviral therapy, and severe immunodeficiency is still required for the development of CMV retinitis.Despite these limitations, these analyses identified the following baseline risk factors for faster progression or spread of retinitis in involved eyes in patients being treated for CMV retinitis: smaller area of involvement, more posterior involvement, and greater border activity. Our findings support more aggressive treatment of such eyes, and we have no reason to think that this suggestion should not be applied to patients in whom CMV retinitis develops before potent antiretroviral treatment is initiated or after it fails. Aggressive treatment of small early lesions remains important. With increased survival, if patients develop CMV retinitis before immune reconstitution or early during antiretroviral treatment, aggressive treatment of CMV retinitis will limit the area with retinitis and thereby decrease the risk of vision-limiting complications, eg, foveal destruction or retinal detachment, in long-term survivors.RJWhitleyMAJacobsonDNFriedbergGuidelines for the treatment of cytomegalovirus diseases in patients with AIDS in the era of potent antiretroviral therapy.Arch Intern Med.1998;158:957-969.Studies of the Ocular Complications of AIDS Research Group in collaboration with the AIDS Clinical Trials GroupMortality in patients with the acquired immune deficiency syndrome treated with either foscarnet or ganciclovir for cytomegalovirus retinitis.N Engl J Med.1992;326:213-220.Studies of Ocular Complications of AIDS Research Group in collaboration with the AIDS Clinical Trials GroupFoscarnet-Ganciclovir Cytomegalovirus Retinitis Trial 4: visual outcomes.Ophthalmology.1994;101:1250-1261.Studies of the Ocular Complications of AIDS (SOCA) in collaboration with the AIDS Clinical Trial GroupCytomegalovirus (CMV) culture results, drug resistance, and clinical outcome in patients with AIDS and CMV retinitis treated with ganciclovir or foscarnet.J Infect Dis.1997;176:50-58.Studies of Ocular Complications of AIDS (SOCA) Research Group in collaboration with the AIDS Clinical Trials Group (ACTG)Studies of Ocular Complications of AIDS Foscarnet-Ganciclovir Cytomegalovirus Retinitis Trial, I: rationale, design and methods.Control Clin Trials.1992;13:22-39.Studies of Ocular Complications of AIDS Research Group, AIDS Clinical Trials Group.Assessment of cytomegalovirus retinitis: clinical evaluation vs centralized grading of fundus photographs.Arch Ophthalmol.1996;114:791-805.The Studies of Ocular Complications of AIDS Research Group in collaboration with the AIDS Clinical Trials GroupMSL-109 adjuvant therapy for cytomegalovirus retinitis in patients with AIDS: the Monoclonal Antibody Cytomegalovirus Retinitis Trial.Arch Ophthalmol.1997;115:1528-1536.GNHollandWCBuhlesBMastreHJKaplanUCLA CMV Retinopathy Study GroupA controlled retrospective study of ganciclovir treatment for cytomegalovirus retinopathy: use of a standardized system for the assessment of disease outcome.Arch Ophthalmol.1989;107:1759-1766.SPAzenARIrvineMDDavisThe validity and reliability of photographic documentation of proliferative vitreoretinopathy.Ophthalmology.1989;96:352-357.FMostellarJWTukeyData Analysis and Regression.New York, NY: Addison-Wesley Longley Inc; 1977:17, 387-390.PArmitageGBerryStatistical Methods in Medical Research.Malden, Mass: Blackwell Publishers; 1987:143-150, 205-211, 306, 412.ELKaplanPMeierNonparametric estimation from incomplete observations.J Am Stat Assoc.1958;53:457-481.DRCoxRegression models and life-tables.J R Stat Soc (B).1972;34:187-220.Not AvailableSAS/STAT Software: Change and Enhancement Through Rels 6.12.Chicago, Ill: SAS Institute; 1997.DCollettModelling Survival Data in Medical Research.New York, NY: Chapman & Hall; 1994:78-87.KYLiangSLZegerLongitudinal data analysis using generalized linear models.Biometrika.1986;73:13-22.DOLinLSWeiThe robust inference for the Cox proportional hazards model.J Am Stat Assoc.1989;84:1074-1077.GNHollandATufailCMJordonCytomegalovirus diseases.In: Pepose JS, Holland GN, Wilhelmus KR, eds. Ocular Infections and Immunity.St Louis, Mo: Mosby–Year Book Inc; 1995:1088-1128.GNHollandJDShulerProgression rates of cytomegalovirus retinopathy in ganciclovir-treated and untreated patients.Arch Ophthalmol.1992;110:1435-1442.DAJabsCEngerJPDunnMFormanLHubbardfor the CMV Retinitis and Viral Resistance Study GroupCytomegalovirus retinitis and viral resistance: 3. culture results.Am J Ophthalmol.1998;126:543-549.CASabinAMocroftMBofillChanges in lymphocyte subsets in human immunodeficiency virus–positive persons with <5 CD4 T lymphocytes/mm3.J Infect Dis.1998;178:1166-1169.SOkaYNagataYFujinoCD8+T lymphocyte counts as an adjunctive predictor or cytomegalovirus retinitis in patients with acquired immunodeficiency syndrome.Intern Med.1997;36:461-465.MLTay-KearneyCEngerRDSembaWRoyalJPDunnDAJabsT cell subsets and cytomegalovirus retinitis in human immunodeficiency virus–infected patients.J Infect Dis.1997;176:790-794.Accepted for publication February 12, 2000.Supported by cooperative agreements and grants U10 EY 08057, 1 R03 EY10731-01, NRSA EY07127 (The Johns Hopkins University School of Hygiene and Public Health, Baltimore, Md), U10 EY 08052 (The Johns Hopkins School of Medicine), and U10 EY 08067 (University of Wisconsin School of Medicine, Milwaukee) from the National Eye Institute, Bethesda, Md. Additional support was provided by grants 5M01 RR 00350 (Baylor College of Medicine, Houston, Tex), 5M01 RR 00035 and 5M01 RR 00722 (The Johns Hopkins University), 5M01 RR 05096 (Louisiana State University/Tulane, New Orleans), 5M01 RR 00071 (Mt Sinai Medical Center, New York, NY), 5M01 RR 00047 (New York Hospital–Cornell Medical Center, New York), 5M01 RR 00096 (New York University, New York), 5M01 RR 00048 (Northwestern University, Evanston, Ill), 5M01 RR 00865 (University of California–Los Angeles), 5M01 RR 00083 (University of California–San Francisco), and 5M01 RR 05280 (University of Miami, Miami, Fla) from the National Center for Research Resources through General Clinical Research Center, Bethesda. Support also was provided by cooperative agreements U01 AI 27668 (The Johns Hopkins University), U01 AI 27674 (Louisiana State University/Tulane), U01 AI 27669 (Memorial Sloan-Kettering, New York), U01 AI 25917 (New York Hospital–Cornell Medical Center), U01 AI 27667 (Mount Sinai Medical Center), U01 AI 27665 (New York University), U01 AI 25915 (Northwestern University), U01 AI 27660 (University of California–Los Angeles), U01 AI 27670 (University of California–San Diego), and U01 AI 27663 (University of California–San Francisco). Funding also was provided by Astra Pharmaceutical Products, Inc, Westborough, Mass.Drugs were provided by Astra Pharmaceutical Products, Inc; Burroughs Wellcome, Co, Research Triangle Park, NC; and Syntex Research, Palo Alto, Calif.Corresponding author: Janet T. Holbrook, PhD, MPH, 615 N Wolfe St, Room 5010, Baltimore, MD 21205 (e-mail: [email protected]).Reprints: Studies of Ocular Complications of AIDS Chairman's Office, 550 N Broadway, Suite 700, Baltimore, MD 21205.

Plummer, Daniel J.; Azen, Stanley P.; Freeman, William R.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1205pmid: 10980765

ObjectiveTo determine the effectiveness of scanning laser entoptic perimetry as a noninvasive platform for screening for retinal damage in visually asymptomatic patients within the central 120° (diameter) of vision.DesignA masked study comparing entoptic perimetry with fundus photographs.SettingThe Shiley Eye Center and the AIDS Ocular Research Unit at the University of California, San Diego.PatientsFifty-eight patients recruited during ophthalmologic visits for treatment or follow-up of ocular disease.MeasurementsFor each testing session, we compared the presence of a disturbance in the entoptic stimulus with the presence of retinal disease within the central 120° of vision, centered on the fovea.ResultsScanning laser entoptic perimetry has a sensitivity and specificity of more than 90%, a positive predictive value of 100%, and a negative predictive value of 89% for screening retinal lesions within the central 120° diameter of vision.ConclusionScanning laser entoptic perimetry may be an effective and inexpensive screening test for diagnosing retinal disease in hospitals and community clinics.ONE OF the most challenging problems in ophthalmology is the development of effective retinal screening tests for peripheral retinal disease. Visual dysfunction is unique in that patients experiencing peripheral retinal damage often remain visually asymptomatic. Patients generally do not notice any disturbance of the visual field until retinal damage occurs close to the fovea. Repression of peripheral scotomas is related to the Troxler phenomenon. In the Troxler phenomenon,a fixed spot of light above threshold presented to the peripheral visual field will begin to slowly disappear from view. This phenomenon applies primarily outside 12° from fixation. This is likely due to neural mechanisms in the brain, and has the adaptive value in human vision of allowing structures in a constant position in the visual field (eg, blood vessels) to be repressed and not interfere with visual function. Scotomas due to retinal damage are also repressed by the Troxler phenomenon, and therefore are not perceived by patients, especially if they are outside the fovea.Early detection of potentially treatable infectious retinopathies, ocular melanomas, and other retinal diseases is essential for the prevention of severe vision loss and potentially fatal systemic diseases. Damage due to many of these diseases (eg, cytomegalovirus [CMV] retinitis) may be insidious, particularly because it often affects the peripheral retina first and patients are often asymptomatic until irreversible destruction of central retina (macula) and loss of visual acuity occurs.A procedure that can measure the extent and locations of retinal scotomata would therefore be expected to give a relatively precise determination of the retinal damage.Until now, there have been no rapid, noninvasive screening techniques for peripheral visual dysfunction. The Amsler grid is a perimetric tool administered to the patient for central and paracentral scotomas.It is commonly used by ophthalmologists as a diagnostic technique for measuring changes in visual function. Observation of a grid of lines on a paper that subtend a 10° radius from fixation will detect central retinal scotomas. However, the Amsler grid may be ineffective for detecting disturbances in vision even within the central 10° radius of vision,and certainly cannot detect defects peripheral to this. As a result, there currently exist no noninvasive techniques for detecting peripheral disturbances in vision that can be administered to the patient without specialized equipment. Teich and Saltzmanhave increased the effectiveness of the Amsler grid with a new stimulus out to a 22.5° radius (45° field), but they reported only a 65% sensitivity, which is insufficient for large-scale screening of patient populations. Visual field perimetry is an alternative that accurately measures the location and extent of peripheral visual field scotomata, but is impractical to primary care physicians as a screening test because of cost, testing time, and need for trained technicians.Entoptic, or snow-field perimetry, is based on the work of Aulhorn and Köstand later work by Ramachandran.A simple computer monitor filled with random particle motion, when viewed by someone with a normal visual system, will appear as "visual noise." Aulhorn and Köst noted that patients with peripheral retinal lesions were able to outline their scotomas. Areas corresponding to the damaged retina appeared to have no random motion, and were "gray" or "motionless" in appearance. Areas where patients reported no random particle motion corresponded to retinal lesions.Based on these results, our group first developedand then enhanceda clinically useful screening test for retinal damage in patients with the acquired immunodeficiency syndrome caused by CMV retinitis with the use of a computer monitor. We found that with this test, patients with CMV retinitis can see scotomas in the eye(s) with the infection. We demonstrated that entoptic perimetry had a high sensitivity and specificity to detect lesions due to CMV retinitis within a 30° radius from the fovea (60° field).While early detection of retinal diseases within the central 30° radius of vision is critical to preserving macular function, a test that only screens within this area will not be useful for the majority of the peripheral retina. Practical considerations limit the amount of retina that can be screened by flat-panel technology. As one attempts to present the entoptic stimulus to more peripheral areas of the retina using a computer monitor, several problems arise. One potential solution is simply moving the patient closer to the screen. However, patients quickly reach their accommodative limit (particularly in geriatric populations who need to be frequently screened), showing that this technique has limited usefulness. Another option is having patients view a very large display (eg, a video display projected on a wall or a very large-screen television). This suffers from several problems including (1) loss of contrast and lack of lighting control, which is critical in entoptic perimetry; (2) the requirement for a large amount of space to present and store the equipment; and most important, (3) a distortion of the stimulus as one gets close to the large screen (but outside the accommodative limit) while attempting to view the image in the peripheral retina.We have overcome the limitations of flat-screen technology presentation of entoptic perimetry by using a virtual reality device in the form of the Microvision Virtual Retinal Display™ system (VRD) (Microvision Inc, Seattle, Wash). There are several advantages to this technology over monitors. Images are projected directly into the eye, presented at virtual infinity, and can be imaged over the peripheral retina. This compensates for all but the most severe refractive errors, and also eliminates peripheral image distortion and the quality of the image allows for extremely high contrast. The scanning laser equipment is portable, easily fitting within a briefcase, allowing mobility within a clinical setting. A narrow exit pupil in our device ensured that patients were fixated centrally, greatly reducing error rates due to inappropriate fixation.In this study, we evaluate wide field scanning laser entoptic perimetry in assessing retinal damage from various pathological conditions, resulting in dense retinal scotomata within the central 120° diameter of the visual field.PATIENTS AND METHODSPATIENTSWe recruited 58 patients from the Shiley Eye Center and from the AIDS Ocular Research Unit at the University of California, San Diego. Patients were selected from the general retina clinic of one of us (W.R.F.). There was no minimum requirement for visual acuity. A total of 110 eyes were tested using scanning laser entoptic perimetry. (Six patients had only 1 eye tested due to disease causing complete blindness in the fellow eye.) All patients were recruited during ophthalmologic visits for treatment or follow-up of ocular disease. Participation was voluntary and we received informed consent.STIMULUSScanning laser entoptic perimetry consisted of a monocular presentation on a VRD of monochromatic random particle motion. Each "pixel value" could be either on at 635 nm or off. The VRD delivered the entoptic stimulus through a narrow exit pupil (1 mm), which was then viewed by the patient.The stimulus presented to the patient through the VRD was also "mirrored" by virtue of a video signal splitter that displayed the identical stimulus on a computer monitor. This allowed the experimenter to view the identical stimulus as the patient and control the entoptic perimetry program without interfering with the view of the patient in the VRD.PROCEDURESPatients' eyes were initially dilated, an ophthalmologic examination was performed, and fundus photographs taken. Fundus photographs were taken to include all areas of retinal disease as previously described.A diagram of lesion locations was made by a qualified ophthalmologist (W.R.F.). Presence and locations of these lesions was confirmed by fundus photography, thereby effectively providing documentation of true location of any lesions on the retina within 1 hour of testing. In all cases, lesions observed by indirect ophthalmoscopy were in complete concordance with fundus photographs.Next, patients were shown the computer monitor that mirrored the stimulus inside the VRD, and were shown an example of the entoptic stimulus. Patients were given instructions on how to use the virtual pen. Patients were explained that they would view the identical stimulus (but for the red color) within the VRD.The entoptic program has 2 modes of display. The stimulus mode displays the entoptic stimulus. As the virtual pen was brought into close proximity to a touch-sensitive pad, the stimulus mode ended and the program entered the "recording" mode, where patients were presented with a blank workspace for drawing. The recording mode had several options. Placing the pen on the pad and moving it (keeping a firm, light pressure on the stylus) produced a black line against the background. Removing the pen from the pad but keeping it in close proximity to the pad (ie, close than 1 cm) allowed the patients to move the cursor on the screen without drawing. Pulling the pen away from the pad further than 1 cm returned the viewer to the stimulus mode. Placing the pen close to the pad would again return the patient to the drawing screen, and previously drawn scotomas would remain. In this way, participants were able to turn the stimulus on and off under their own control. All actions were monitored by the technician who viewed the computer monitor during testing. This instructional phase rarely took longer than 2 minutes.After instructions, patients were seated in front of the VRD and asked to view the VRD with 1 eye (an eyepatch was provided). They were asked to fixate in specific locations within the visual field, and while remaining fixated, to report any perceptual changes.Unlike computer monitors that can be viewed from a wide variety of angles, by virtue of the narrow exit pupil, patients had to concentrate on fixating within the VRD to see the entoptic stimulus. If their gaze wandered, the stimulus disappeared from view and they saw a black field. Thus, unless the patient was fixated centrally within the VRD, patients were unable to see the stimulus.SCREENING TO 120°The VRD we used had a capability for screening out to 30° radius when the patient was fixated centrally on a fixation crosshair. However, as there was no peripheral image distortion by having the patient fixate on the corners of the virtual screen, we placed crosshairs at the 4 corners of the screen as well as halfway along the vertical and horizontal edges of the screens. By having patients fixate on a corner of the virtual image (eg, lower left), we were effectively able to screen out 60° from fixation for a given quadrant. This procedure was repeated for the 3 other corners in a random order, therefore screening the entire central 120° of the retina.SCORING OPHTHALMOLOGIC FINDINGSPresence or absence of retinal damage was determined by an expert ophthalmologist (W.R.F.) using indirect ophthalmoscopy and confirmed by fundus photography. For each of the diseases listed in Table 1, we determined areas of damage to the retina using the following rules:Table 1. Frequency Distribution and Diagnostic Description of Study Eyes*DiagnosisNo. of Patients With Diagnosis (n = 58)No. of Eyes With Diagnosis (n = 80)No. of Companion Eyes Normal (n = 30)No. of Companion Eyes Not Tested (n = 6)Sensitivity Central 120°Specificity Central 120°Mean No. of Lesions in Eyes With DiagnosisMean No. of Entoptic DisturbancesCMV retinitis1929810.961.001.11.2Age-related macular degeneration914310.861.001.01.0Retinal detachment or tear1113721.001.001.21.3Diabetic retinopathy†47 (1 with RD)010.71. . .1.01.0Macular hole44401.001.001.01.0Branch retinal vein occlusion33301.00. . .1.31.0Ocular melanoma33301.001.001.01.0AMPPE12001.00. . .1.01.0Vitritis1200. . .1.0000Drusen alone22111.001.001.52Non-HIV–related toxoplasmosis11101.001.001.01.0*CMV indicates cytomegalovirus; AMPPE, acute posterior multifocal placoid pigment epitheliopathy; HIV, human immunodeficiency virus; RD, retinal detachment; and ellipses, measures cannot be calculated.†As determined by areas of nonperfusion.Diabetic retinopathy, branch retinal vein occlusion.Areas of nonperfusion and edema as seen with fluorescein angiography, confirmed by fundus photography. For each patient, there were also areas that had undergone panretinal laser photocoagulation.CMV retinitis.Areas of retinal destruction by fundus photography seen as "healed" retinitis.Ocular melanoma.Areas of the retina corresponding to the location of the tumor.Macular hole.Areas relating to the hole and surrounding cuff of fluid.Acute posterior multifocal placoid pigment epitheliopathy.Areas of the retinal pigment epithelium disturbed despite excellent visual acuity.Age-related macular degeneration and drusen.Areas in eyes of patients without laser surgery, both wet and dry, as confirmed by fundus photography.Retinal detachment/tear.Area of retinal detachment as confirmed by fluorescein angiography and fundus photography.Toxoplasmosis.Area of retinal scar as confirmed by fundus photography.The ophthalmologist also classified lesions as within the central 10° (radius) of the visual field, between 10° and 30°, from 30° to 60°, or outside the central 60° radius, as measured from the fovea. The ophthalmologic examination was performed before entoptic perimetry testing.SCORING PERIMETRIC FINDINGSPresence or absence of entoptic perimetry visual field disturbance was determined by an expert psychophysicist (D.J.P.). If a patient drew an area using the computer interface that corresponded to a localized change in the entoptic stimulus, the eye was classified as having a visual field disturbance. As with the ophthalmologic findings, we classified visual field disturbances as within the central 10° (radius) of the visual field, between 10° and 30°, from 30° to 60°, or outside the central 60° radius, as measured from the fovea. The psychophysicist was masked to the outcome of the ophthalmologic findings.STATISTICAL ANALYSISFor each study eye, we computed the sensitivity, specificity, and positive and negative predictive values of scanning laser entoptic perimetry. Sensitivity was calculated as the ratio of the number of eyes scored positive by scanning laser entoptic perimetry to the number of eyes scored positive by fundus photography. Specificity was calculated as the ratio of the number of eyes scored negative by perimetry to the number of eyes scored negative by fundus photography. Positive predictive value was calculated as the ratio of the number of eyes scored positive by scanning laser entoptic perimetry diagnosed as having retinal damage to the number of eyes with entoptic disturbances. Negative predictive value was calculated as the ratio of the number of eyes scored negative by scanning laser entoptic perimetry diagnosed as having retinal damage to the number of eyes without entoptic disturbances. We calculated these summary statistics for the following 6 regions: (1) lesions within the central 10° radius (perimacular area), (2) within 30° (SOCA [Studies of the Ocular Complications of AIDS] zone 1), (3) within 60°, (4) from 10° to 30°, (5) from 30° to 60°, and (6) from 10° to 60° (peripheral retinal imaging area).RESULTSFifty-eight patients (41 men and 17 women) underwent funduscopic examination and scanning laser entoptic perimetry testing for a total of 110 eyes. Table 1provides a breakdown of the numbers of patients and eyes in order of the frequency of diagnosis.Table 2summarizes the mean ±SD sensitivity and specificity stratified by retinal location along with both the positive and negative predictive values stratified by retinal location. Overall, we found that scanning laser entoptic perimetry had sensitivities ranging from 87% to 93% and specificities ranging from 91% to 100%, while positive predictive values ranged from 80% to 100% and negative predictive values ranged from 89% to 97%. In particular, we found that scanning laser entoptic perimetry has a sensitivity of 93% ± 6%, a specificity of 100% ± 0%, a positive predictive value of 100% ± 0%, and a negative predictive value of 89% ± 7% for detecting retinal lesions within the entire 120° visual field tested. Within the perimacular area (central 10° radius of vision), where patients are usually symptomatic, we find that scanning laser entoptic perimetry had a sensitivity of 93% ± 9%, a specificity of 91% ± 6%, a positive predictive value of 80% ± 13%, and a negative predictive value of 97% ± 5%. For areas where patients generally remain asymptomatic to retinal lesions (ie, from 10° to 60° radius from the fovea), scanning laser entoptic perimetry had a sensitivity of 92% ± 8%, a specificity of 95% ± 5%, a positive predictive value of 94% ± 7%, and a negative predictive value of 94% ± 7%.Table 2. Sensitivity, Specificity, and Positive and Negative Predictive Values of Scanning Laser Entoptic Perimetry, by Retinal RegionRegionMean ± SD, %SensitivitySpecificityPositive Predictive ValueNegative Predictive ValueWithin 20° diameter (perimacular)93 ± 991 ± 680 ± 1397 ± 5From 20° to 60° diameter90 ± 1193 ± 581 ± 1496 ± 6From 60° to 120° diameter87 ± 1199 ± 289 ± 1094 ± 8Within central 60° diameter90 ± 893 ± 692 ± 792 ± 8Within central 120° diameter93 ± 6100 ± 0100 ± 089 ± 7Between 20° and 120° diameter (peripheral)92 ± 895 ± 594 ± 794 ± 7STIMULUS SIZE AND VISUAL ACUITYAs previously reported, the optimal sensitivity for patients with visual acuities of 20/40 or better was obtained by using a high-frequency stimulus.However, we found that patients who have poor central visual acuity (eg, ≤20/100) often cannot perceive the fine stimulus. In this study, there were 8 eyes that required a larger stimulus size to perceive the entoptic stimulus. For each of these cases, the patients had a poor central visual acuity. We performed sensitivity and specificity analyses using the minimum pixel size that the patients could perceive.CASE REPORTSThe study cohort included 1 control patient with Behçet disease but no retinal damage. Despite the opacification of the optic media, this patient was able to view the entoptic stimulus and reported no visual disturbances to the entoptic field.The study cohort also included a patient with a new retinal detachment (3 days). The detachment involved nearly the entire hemifield from the far periphery nearly up to the fovea. Upon viewing the stimulus, the patient clearly saw entoptic visual field disturbance extending into the far periphery. The following day, the detachment was successfully repaired surgically. The patient underwent a vitrectomy without scleral buckle, had a long-acting gas injection and laser application anterior to the equator to the retinal breaks. No procedures that would have caused retinal destruction occurred within the visual field. We tested the patient 1 day postoperatively (2 days after the initial testing session) and despite the high refractive error introduced by the surgical procedure (due to the gas), the patient was able to view the entoptic stimulus (using a pixel stimulus size of 10) and found that the entoptic disturbance had disappeared. We followed up this patient at biweekly intervals for a period of 2 months and found no further visual field disturbances, and the stimulus size required to perceive entoptic perimetry decreased with the reduction of the refractive error due to the decrease in size of the gas bubble. These follow-up visits were not included in the sensitivity and specificity analyses presented above.COMMENTSCANNING LASER ENTOPTIC PERIMETRY AS A GENERAL SCREENING DEVICEWe previously demonstratedthat entoptic perimetry was effective in screening for full-depth scotomas from peripheral human immunodeficiency virus–related CMV retinitis. This study shows that scanning laser entoptic perimetry is sensitive and specific for screening for complete scotomas that are the result of retinal diseases. These results demonstrate that scanning laser entoptic perimetry is a viable possibility for a screening test to be administered by physicians, particularly primary care providers, and in underserved communities, where rapid, noninvasive screening procedures can be administered by support staff inexpensively. As entoptic perimetry screening takes less than 1 minute per eye, patients could potentially be routinely screened during annual physical checkups. This would not only allow asymptomatic patients with potentially sight-threatening diseases to be referred to ophthalmologists before central vision is impacted, early detection of diseases such as ocular melanoma will allow early treatment before other organs are affected.ANALYSIS BY LOCATION WITHIN THE RETINAIn our previous studies of CMV retinitis,we specifically did not include patients with central or optic nerve damage. One of the reasons for performing subgroup analyses within different regions based on distance from the fovea is that only the central portion of vision (within 10° radius from the fovea) is affected in patients with diseases such as macular holes, acute posterior multifocal placoid pigment epitheliopathy, and age-related macular degeneration, and these patients, usually symptomatic, are artificially increasing our sensitivity. We maintained a sensitivity and specificity over 90% in our subgroup analyses, which included only those areas where retinal damage would cause patients to generally remain asymptomatic (from 10° to 60° radius from the fovea).This study also presents the first data using scanning laser entoptic perimetry to screen for lesions due to retinal disease outside the central 30° of vision. These results show that scanning laser entoptic perimetry is as sensitive and specific for the peripheral retina (from 30° to 60°) as we previously demonstrated for the retina out to 30°.Furthermore, this method requires no more time for screening, unlike current standard perimetric methods such as threshold perimetry.ADVANTAGES OF THE VRDOur previous studies demonstrated that, using a large computer monitor, we could screen the central 30° radius of vision rapidly and inexpensively. With the VRD as a hardware platform, we have now demonstrated that scanning laser entoptic perimetry can screen for retinal disease with a high sensitivity and specificity within the central 120° diameter field of vision. This is a significant improvement over previous rapid screening methods such as the Amsler grid, presenting the image over 75% more retinal area. Goldmann and Humphrey visual field perimetry can be used as screening tools for mapping of retinal scotomas out to 180° from the fovea, but requires not only a significant investment in technicians and overhead for the provider, but also requires considerable time from the patient. As a result, the standard perimetric tests currently available are not good candidates for large-scale, community-based screening programs.Entoptic perimetry is not intended to replace the current uses of visual field perimetry, but instead can provide a valuable tool for the primary care provider in detecting retinal disease early. Furthermore, the VRD is portable, and with optimization, could become part of school-based or other screening programs.The case reports we present also suggest that entoptic perimetry can be used by ophthalmologists to rapidly assess visual function in patients with opacities of the optic media that might prevent clear views of the retina, especially in patients with cataracts or vitritis. Furthermore, we were able to evaluate the success of retinal detachment repair in a patient who was tested both preoperatively and postoperatively. Despite the fact that the visual acuity was assessed as hand motion, the patient was able to see the entoptic stimulus and report that the previous visual disturbance had disappeared. The fact that we had a diabetic patient who was able to view laser burns suggests that in screening tests, patients will be able to detect very small lesions throughout the visual field. These results warrant further detailed investigation.DTroxlerÜber das Verschwinden gegebener Gegenstände innerhalb unseres Gesichtskreises.Ophthalm Bibliothek.1804;2:51-53.GNHollandOphthalmic disorders associated with the acquired immune deficiency syndrome.In: Insler MS, ed. AIDS and Other Sexually Transmitted Diseases and the Eye.Orlando, Fla: Grune & Stratton; 1987.DAJabsTreatment of cytomegalovirus retinitis in patients with AIDS [editorial; comment].Ann Intern Med.1996;125:144-145.SJRyanRetina.St Louis, Mo: Mosby; 1989.AMFineMJElmanJEEbertPAPrestiaJSStarrSLFineEarliest symptoms caused by neovascular membranes in the macula.Arch Ophthalmol.1986;104:513-514.RASchuchardValidity and interpretation of Amsler grid reports.Arch Ophthalmol.1993;111:776-780.SATeichBRSaltzmanEvaluation of a new self-screening chart for cytomegalovirus retinitis in patients with AIDS.J Acquir Immune Defic Syndr Hum Retrovirol.1996;13:336-342.EAulhornGKöst[White noise field campimetry: a new form of perimetric examination].Klin Monatsbl Augenheilkd.1988;192:284-288.VSRamachandranRLGregoryPerceptual filling in of artificially induced scotomas in human vision [see comments].Nature.1991;350:699-702.VSRamachandranFilling in the blind spot [letter; comment].Nature.1992;356:115.VSRamachandranBlind spots.Sci Am.1992;266:86-91.DJPlummerJFArevaloWFramJIQuiceroPASampleWRFreemanEffectiveness of entoptic perimetry for locating peripheral scotomas caused by cytomegalovirus retinitis.Arch Ophthalmol.1996;114:828-831.DJPlummerABankerITaskintunaThe utility of entoptic perimetry as a screening test for cytomegalovirus retinitis.Arch Ophthalmol.1999;117:202-207.JGGrossSABozzetteWCMathewsLongitudinal study of cytomegalovirus retinitis in acquired immune deficiency syndrome.Ophthalmology.1990;97:681-686.SSBylsmaCLAchimCAWileyThe predictive value of cytomegalovirus retinitis for cytomegalovirus encephalitis in acquired immunodeficiency syndrome.Arch Ophthalmol.1995;113:89-95.Accepted for publication January 9, 2000.This work was supported by grants NEI EY11961 (Dr Plummer), NEI EY07366 (Dr Freeman), and Core Grant for Vision Research NEI EY-03040 (Dr Azen) from the National Institutes of Health, Bethesda, Md, and a departmental grant from Research to Prevent Blindness, New York, NY (Dr Freeman).Corresponding author: Daniel J. Plummer, PhD, Shiley Eye Center, Department of Ophthalmology, School of Medicine, University of California, San Diego, La Jolla, CA 92093-0946 (e-mail: [email protected]).

Chappelow, Aimee V.; Marmor, Michael F.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1211pmid: 10980766

ObjectiveTo examine results of the multifocal electroretinogram (MERG) after spontaneous resolution of central serous chorioretinopathy (CSC) detachments.MethodsMultifocal electroretinograms were recorded from both eyes of 5 recovered patients with CSC and 10 age-matched healthy subjects. All patients with CSC had bilaterally subnormal MERG amplitudes during a first attack of CSC occurring 7 to 23 months earlier.ResultsAfter recovery from CSC, MERG A-wave and B-wave amplitudes increased markedly where the detachment resolved, and moderately elsewhere in the posterior pole of both eyes. However, the signals from both eyes remained either subnormal or low-normal relative to controls. Multifocal electroretinogram B-wave latencies improved from prolonged to mid-normal values in both eyes.ConclusionsBoth eyes of patients with active unilateral CSC exhibit diminished MERG amplitudes. Although MERG response amplitudes increased modestly after recovery from CSC, they remained statistically subnormal throughout the posterior pole of both eyes. These findings support the theory that subretinal fluid retention in CSC is secondary to diffuse pathologic changes in the choroid and/or retinal pigment epithelium. They also suggest that the underlying or predisposing abnormalities of CSC resolved only partially in our patients. Components of the MERG may have value as a prognostic tool for judging the risk of developing symptomatic CSC.CENTRAL SEROUS chorioretinopathy (CSC) is a disease of unknown pathogenesis in which fluid enters the potential space between the retinal pigment epithelium (RPE) and the photoreceptors. Fluorescein angiography shows a focal source of dye leakage into the area of fluid accumulation. Individuals with high-stress lifestyles and "Type A" personalities are most often affected,and men are more frequently affected than women. There is an increasing body of clinical and experimental evidence that the retention of fluid that enters from a focal leak is secondary to an underlying diffuse choroidal or RPE dysfunction.For instance, the fellow (unaffected) eye often shows abnormalities of the RPE and may sometimes develop a serous detachment. Areas much larger than the locus of detachment have shown choriocapillary insufficiency and capillary hyperfusion using indocyanine green angiography.Multifocal electroretinography (MERG) during active disease has demonstrated subnormal macular cone responses broadly in the posterior pole of both the affected and fellow eyes.Experimental work has shown the RPE to be so efficient at removing subretinal fluid that fluid is not likely to accumulate from a small focal leak if the surrounding RPE is healthy.Finally, animals given repeated systemic injections of adrenaline and corticosteroids have developed multifocal serous detachments.To the extent that changes in the MERG reflect underlying pathologic changes in the choroidal vasculature and/or RPE, the MERG may provide a clinical index of susceptibility to detachment in CSC. However, further information is needed to determine which MERG parameters are most indicative of the active state of CSC, and whether these parameters change in relation to clinical findings. Two previous studieslooked at the focal electroretinogram (ERG) from the central macula of affected eyes with CSC after the resolution of detachments and found variable degrees of ERG recovery between 2 weeks and 5 months. We are not aware of any comparable studies using the MERG, which gives more localized information about cone function in different regions of the macula. In this study, we recorded the MERG in both eyes of 5 patients 7 to 23 months after spontaneous clinical recovery from CSC, and compared the results with recordings obtained during the active phase of the disease.SUBJECTS AND METHODSSUBJECTSFive patients studied previously with MERG recordings during an initial (and only) episode of unilateral CSCwere reexamined (Table 1). None of the patients had significant abnormalities in the fellow eye at initial examination. Two were women and 3 were men; their ages ranged from 37 to 52 years (mean age, 42 years). All of the patients were in good general health, and none were taking corticosteroids. Subject 2 had Reiters syndrome, for which he used nonsteroidal anti-inflammatory medications and misoprostol. In all subjects, the initial serous detachments resolved without laser photocoagulation between 1 and 3 months after initial visit. Follow-up MERGs were recorded 7 to 23 months after the initial visit, and no less than 4 months after the resolution of fluid. Visual acuity at the time of this study was 20/25 or better in both eyes, and fundus examination revealed no serous detachments. There were no RPE alterations other than mild granularity in the affected eyes, but the fellow eye of subject 5 had developed 2 tiny pigment epithelial detachments temporal to the macula.Table 1. Characteristics of Patients With CSC*Subject No./ Sex/Age, yVA During Active CSCVA After Resolution1/F/4020/2020/202/M/5220/2020/253/M/3720/1520/204/M/3720/5020/205/F/4220/3020/20*CSC indicates central serous chorioretinopathy; VA, the corrected visual acuity of the affected eye.Ten control subjects of similar age (age range, 31-55 years; mean age, 40 years) had an MERG recorded during the course of this study. All were in good general health, with 20/20 OU visual acuity and no evidence of retinal disease. The tenets of the Declaration of Helsinki were followed, and all subjects gave written informed consent following a full explanation of procedures.MERG RECORDINGMultifocal electroretinogram recordings were obtained with a VERIS instrument (Tomey Corp, Nagoya, Japan) and analyzed using the VERIS Science software (Electro-Diagnostic Imaging Inc, San Mateo, Calif) developed by Sutter and Tran.The stimulus consisted of an achromatically flickering 103-hexagon array with an average photopic luminance of 100 candelas (cd)/m2. Each subject was refracted using spectacle lenses and adjusted to the appropriate viewing distance. Room lights were dimmed so that luminance was approximately 1 cd/m2at a calibration spot 6 in (15.2 cm) from an off-white wall.Patients' pupils were fully dilated with 1% tropicamide and 2.5% phenylephrine hydrochloride and corneas were anesthetized with 0.5% proparacaine hydrochloride. The MERG of each eye was recorded with the same Burian-Allen electrode (Hansen Instruments Inc, Iowa City, Iowa). The MERG recording time was 7 minutes and 17 seconds, divided into eight 54.6-second intervals.DATA ANALYSISSignals were band-pass filtered (3-300 Hz) and amplified using a gain of 10 (Amplifier Model 12C; Grass Instrument Co, Quincy, Mass).Each MERG data file was run once through an artifact rejection process included in the software, and responses were density scaled to account for variation in hexagon area. To compare MERG waveforms at different eccentricities within the macula, data were averaged from 6 concentric rings of MERG responses from the fovea outward to 20°. The negative and positive deflections of the MERG waveform will be referred to as A- and B-waves, respectively, as they are generally analogous to the A- and B-wave components of a full-field ERG.The MERG data from our previous report,during the active phase of CSC, included 2 recordings made with 241 stimulus hexagons. To obtain averages that incorporate these data with the data from recordings made with 103 hexagons, the 241-hexagon waveforms were grouped into 6 concentric rings that approximated the eccentricities of the 6-ring groupings used with 103-hexagon arrays. Since the eccentricities are not exactly equal, we used weighted averages of the 103- and 241-hexagon array ring eccentricity values on the x-axis when plotting these results from the active phase of CSC.Differences between the means of controls and CSC patients were evaluated statistically using the 2-tailed Student ttest, with P<.05 considered statistically significant. These results are presented in Table 2, whereas the figures show a range (envelope) of values from our control subjects.Table 2. MERG Parameters in Eyes Affected by CSC and Fellow Eyes*Retinal Eccentricity†Control (n = 10)With Active CSCAfter RecoveryFellow Eye (n = 5)Affected Eye (n = 5)Fellow Eye (n = 5)Affected Eye (n = 5)A-wave amplitude, nV/deg2044.5 ± 9.322.9 ± 8.4‡19.4 ± 7.7‡29.4 ± 10.0‡29.1 ± 12.83.5-3.626.5 ± 6.514.8 ± 5.3‡11.5 ± 3.2‡18.5 ± 3.8‡20.4 ± 5.9§6.517.8 ± 4.210.7 ± 2.8‡11.5 ± 2.6‡13.1 ± 2.0‡14.0 ± 3.010.5-10.913.7 ± 2.79.4 ± 2.6‡8.8 ± 1.4‡10.6 ± 2.1‡10.3 ± 1.7‡15-15.411.9 ± 2.18.4 ± 2.5‡7.9 ± 1.3‡9.3 ± 1.7‡9.2 ± 1.3‡20-20.811.6 ± 2.37.6 ± 1.9‡7.4 ± 1.7‡9.3 ± 1.7‡9.4 ± 1.3‡B-wave amplitude, nV/deg20102.8 ± 18.250.7 ± 18.2‡39.9 ± 7.0‡66.8 ± 14.8‡67.7 ± 20.8‡§3.5-3.658.8 ± 12.330.8 ± 11.5‡24.3 ± 7.7‡43.3 ± 8.8‡42.0 ± 10.9‡§6.539.2 ± 8.922.2 ± 4.8‡22.8 ± 6.2‡28.4 ± 3.3‡29.4 ± 5.0‡§10.5-10.930.6 ± 6.919.3 ± 5.1‡18.7 ± 4.2‡23.3 ± 3.4‡22.4 ± 3.7§15-15.426.4 ± 5.817.1 ± 4.6‡16.4 ± 3.9‡20.7 ± 3.5‡20.1 ± 3.0‡§20-20.825.4 ± 6.216.1 ± 3.6‡16.1 ± 4.8‡20.4 ± 3.820.8 ± 2.7‡§B-wave latency, ms029.7 ± 1.733.3 ± 2.0‡37.3 ± 2.7‡30.0 ± 1.0§31.0 ± 2.0§3.5-3.629.6 ± 2.031.4 ± 1.0‡33.2 ± 2.2‡29.5 ± 0.9§30.3 ± 0.9§6.529.2 ± 1.233.3 ± 0.9‡33.8 ± 2.8‡29.7 ± 0.9§29.9 ± 1.1§10.5-10.929.0 ± 1.532.8 ± 1.0‡35.2 ± 0.7‡29.8 ± 1.6§30.7 ± 2.2§15-15.429.7 ± 1.632.5 ± 0.5‡33.6 ± 1.4‡30.5 ± 1.630.7 ± 1.620-20.830.2 ± 1.632.2 ± 1.933.2 ± 0.9‡30.7 ± 1.130.3 ± 0.4§*Data are presented as mean ± SD unless otherwise indicated. MERG indicates multifocal electroretinogram; CSC, central serous chorioretinopathy.†Radius (in degrees) at the approximate center of the stimulus hexagons for each concentric grouping of waveforms. Some of the eccentricities are expressed as a range of values because the data were derived from recordings made with both 103- and 241-hexagon stimulus arrays, which yield slightly different waveform arrays.‡Results differ significantly from control eyes (P<.05).§Results differ significantly from the same eyes during active CSC (P<.05).RESULTSTable 2shows averaged MERG waveform parameters (A-wave amplitude, B-wave amplitude, and B-wave latency) at different eccentricities from the fovea, before and after recovery from CSC. The A-wave and B-wave amplitudes improved markedly after recovery in the area of detachment (approximated by the center 2 rings of the affected eye), and to a moderate degree elsewhere in the posterior pole of both eyes. Even after recovery, the MERG amplitudes in both eyes remained statistically subnormal (relative to controls) at most of the eccentricities.These A-wave and B-wave amplitude data are shown graphically in Figure 1and Figure 2to facilitate the comparison of prerecovery and postrecovery signals relative to the envelope of normal values. Values are plotted on a log scale so that the range of normal values is similar in the central and peripheral areas of the macula. There is an improvement after recovery but also a persistence of subnormal or borderline amplitude at all eccentricities of the posterior fundus. The effect of the detachment during the active phase of disease is evident in affected eyes as a dip in A-wave and B-wave amplitudes at 0° and 3.5° eccentricity. After recovery, the amplitudes in this central area returned to a level comparable with the fellow eye.Figure 1.Multifocal electroretinogram A-wave amplitude before and after recovery from central serous chorioretinopathy (CSC), relative to eccentricity from the fovea. The dashed lines enclose the range of normal values. Amplitudes during active CSC (diamonds) are subnormal at all eccentricities. Amplitudes after recovery (squares) are improved but still fall in a low-normal to subnormal range. Standard deviations are given in Table 2.Figure 2.Multifocal electroretinogram B-wave amplitude before and after recovery from central serous chorioretinopathy. The data are plotted as in Figure 1. The pattern of B-wave recovery was similar to that of the A-wave, shown in Figure 1.The changes in MERG B-wave latency with recovery are shown in Table 2and Figure 3. During active disease, the average latencies were significantly prolonged (although just outside the normal range) in both affected and fellow eyes. After recovery, both eyes showed response latencies in the middle of the normal range.Figure 3.Multifocal electroretinogram B-wave latency before and after recovery from central serous chorioretinopathy. Data are plotted as in Figure 1. Latencies were delayed in both eyes during active central serous chorioretinopathy, but improved to normal on resolution of the serous detachments.Table 3presents MERG results in relation to patient age and to the time between the resolution of detachment and the final MERG. No obvious correlations are seen, but the sample is too small to draw firm conclusions.Table 3. MERG Response Parameters Relative to Clinical History*Subject No./Age, yTime of Final MERG, moComposite Macular B-Wave†After Resolution of CSCAfter First MERG During CSCAmplitude, nV/deg2Latency, msDuring CSCAfter ResolutionDuring CSCAfter Resolution1/404722.925.733.3302/52161915.819.93529.23/37202114.925.130.8304/377811.518.635305/42232321.421.43530.8Healthy‡/40NANA26.626.629.729.7*MERG indicates multifocal electroretinogram; CSC, central serous chorioretinopathy; NA, not applicable.†Values are the composite average of all MERG responses.‡Values are the average of the healthy subjects.COMMENTOur results show that local ERG responses improved after resolution of clinical CSC, but remained borderline or subnormal in amplitude throughout the posterior pole of both affected and fellow eyes. As might be expected, the greatest improvement occurred in the area of the detachment (the central retina of affected eyes). However, the persistence of MERG recording abnormalities beyond the area of detachment and in the fellow eye is consistent with the concept that CSC is pathophysiologically a diffuse rather than a focal disorder, and suggests that some of these pathological changes can persist even after resolution of the detachment.We are unaware of any other ERG studies that have differentiated between regions of the macula in CSC, or followed abnormalities in the fellow eye of CSC patients. Nagata and Honda,using a 4° stimulus spot to record a focal ERG, found persistent low amplitudes in affected eyes 2 to 7 weeks after xenon arc photocoagulation and resolution of the detachments. Miyake et al,using a 10° stimulus spot, observed recovery of focal ERG B-wave amplitudes and latencies 2 to 5 months after recovery from CSC. However, the macular oscillatory potentials remained subnormal. Our results show that the A-wave and B-wave amplitudes, which were initially reduced broadly across the posterior pole of both eyes, improved in both eyes after resolution of the detachment, but not to normal. While close comparison between these studies is difficult because of methodological differences, they are consistent to the extent that there seems to be both some degree of recovery and some degree of persistent abnormality in the affected macula after the resolution of fluid.Since the ERG abnormalities we have observed involve both the A wave and B wave, the retinal changes must originate at the level of the photoreceptors. It is possible that systemic factors such as adrenergic and corticosteroid activity are affecting the photoreceptors directly. However, we suspect that this photoreceptor dysfunction reflects alterations in the underlying choroidal vasculature and/or RPE, although there is no multifocal test of RPE function available to confirm this directly. We have postulated elsewherehow factors such as adrenergic and corticosteroid stress can alter choroidal vascular function and secondarily (if not primarily) diminish the water transport capability of the RPE, creating a susceptibility to subretinal fluid accumulation that results in CSC. Our previous study of the active stage of CSCshowed that MERG abnormalities were diffuse and bilateral, which supports the concept that the disease represents broad dysfunction of the choroid and RPE rather than just a focal event (such as a leak). The present data lend further support to this view. On one hand, the MERG responses (which represent a broad area of macular dysfunction) changed in accordance with the clinical course of disease (they improved with recovery). On the other hand, a degree of MERG abnormality remained, which is not surprising since many of the stress factors or other systemic causes that affect CSC patients are likely to persist even after the attack.The MERG may in time prove useful in assessing the degree of susceptibility to a CSC attack, especially in high-risk patients with stressful lifestyles, patients using corticosteroids, or patients with a history of prior attacks. It may also aid in determining whether or not drug treatment or stress reduction can reduce the predisposition to detachment. However, further studies are needed to determine which parameters (eg, MERG amplitude or latency) are most relevant. The results from this study suggest that the finding of increased B-wave latency, along with reduced A-wave and B-wave amplitudes, might turn out to be the most useful measure of clinical risk. Reduced MERG amplitudes persisted during clinical remission, but reduced amplitudes in combination with waveform delays were observed only during the active phase of the disease. This proposal remains hypothetical since we have not yet observed any recurrent detachments, our patient sample is small, and we have not had the opportunity to follow patients over longer periods to find out whether full recovery of the MERG amplitudes might eventually occur.LAYanuzziType A behavior and central serous chorioretinopathy.Retina.1987;7:111-130.MFMarmorNew hypothesis on the pathogenesis and treatment of serous retinal detachment.Graefes Arch Clin Exp Ophthalmol.1988;226:548-552.MFMarmorOn the cause of serous detachments and acute central serous chorioretinopathy.Br J Ophthalmol.1997;81:812-813.CPrunteJFlammerChoroidal capillary and venous congestion in central serous chorioretinopathy.Am J Ophthalmol.1996;121:26-34.TIidaKMuraokaNHagimuraKTakahashiChoroidal lesions of central serous chorioretinopathy by indocyanine green angiography.Jpn J Clin Ophthalmol.1994;48:1583-1593.MFMarmorFTanCentral serous chorioretinopathy: bilateral multifocal ERG abnormalities.Arch Ophthalmol.1999;117:184-188.HYoshiokaYKatsumeHAkuneExperimental central serous chorioretinopathy in monkey eyes: fluorescein angiographic findings.Ophthalmologica.1982;185:168-178.MNagataYHondaStudies on local electric response of the human retina. VI. Macular ERGs of a typical central serous retinopathy [In Japanese].Nippon Ganka Gakkai Zasshi.1970;74:957-964.YMiyakeNShiroyamaIOtaMHoriguchiLocal macular electroretinographic responses in idiopathic central serous chorioretinopathy.Am J Ophthalmol.1988;106:546-550.EESutterDTranThe field topography of ERG components in man, I. The photopic luminance response.Vision Res.1992;32:433-446.DCHoodWSeipleKHolopigianVGreensteinA comparison of the components of the multifocal and full-field ERGs.Vis Neurosci.1997;14:533-544.Accepted for publication February 24, 2000.Corresponding author: Michael F. Marmor, MD, Department of Ophthalmology, Stanford University Medical Center, Stanford, CA 94305-5308 (e-mail: [email protected]).

Shields, Carol L.; Shields, Jerry A.; Cater, Jacqueline; Gündüz, Kaan; Miyamoto, Curtis; Micaily, Bizhan; Brady, Luther W.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1219pmid: 10980767

ObjectiveTo identify clinical predictive factors for visual outcome in a large series of patients who underwent plaque radiotherapy for uveal melanoma.DesignClinical factors, including patient data, tumor features, and radiation variables, were analyzed for their impact on visual acuity using Cox proportional hazards regression models.ParticipantsPatients with uveal melanoma and initial visual acuity of 20/100 or better in the affected eye who were treated with plaque radiotherapy between July 1976 and June 1992.Main Outcome MeasuresTwo end points were used to evaluate posttreatment visual acuity: (1) final visual acuity (good [20/20-20/100] vs poor [20/200 to no light perception]) and (2) loss of visual acuity (minimal [<5 lines Snellen visual acuity] vs moderate [≥5 lines Snellen visual acuity]).ResultsOf 1300 consecutive patients with uveal melanoma treated by plaque radiotherapy, 1106 had a visual acuity of 20/100 or better at the time of treatment. In this group, poor visual acuity was found in 34% at 5 years and 68% at 10 years of follow-up. From multivariable analysis, clinical factors that best predicted poor visual acuity were increasing tumor thickness, proximity to foveola of less than 5 mm, notched plaque shape, tumor recurrence, patient age 60 years or older, subretinal fluid, cobalt isotope, anterior tumor margin posterior to equator, and worse initial visual acuity. Moderate loss of visual acuity of 5 Snellen lines or more was found in 33% at 5 years and 69% at 10 years of follow-up. From multivariable analysis, clinical factors that best predicted moderate visual acuity loss included increasing tumor thickness, worse initial visual acuity, notched plaque shape, tumor recurrence, proximity to foveola of less than 5 mm, patient age of 60 years or older, subretinal fluid, and diabetes mellitus or hypertension. When analyzing visual outcome with regard to tumor thickness, ultimate poor visual acuity of 20/200 or worse at 5 years was found in 24% with a small melanoma (≤3.0 mm), 30% with a medium melanoma (3.1-8.0 mm), and 64% with a large melanoma (>8.0 mm). When analyzing visual outcome with regard to tumor proximity to visually important structures, tumors less than 5 mm from the optic disc or foveola demonstrated poor visual acuity in 35% at 5 years, whereas those 5 mm or more from the optic disc and foveola showed poor visual acuity in 25% at 5 years.ConclusionsUltimate visual acuity after plaque radiotherapy for uveal melanoma depends on many factors, including patient age and general health, initial visual acuity, tumor location and size, subretinal fluid, radioactive isotope, and final tumor control. At 10 years' follow-up, 68% of patients demonstrate poor visual acuity. Visual acuity is most effectively preserved in eyes with small tumors outside a radius of 5 mm from the optic disc and foveola.PLAQUE RADIOTHERAPY continues to be an important treatment method for patients with uveal melanoma.Previous studieshave provided information on local and systemic tumor control using this method and have indicated that overall life prognosis is comparable to other management techniques such as enucleation. In addition, the ongoing prospective Collaborative Ocular Melanoma Study (COMS)will possibly provide information regarding systemic tumor control with plaque radiotherapy compared with enucleation.With any organ preservation technique comes the practical question of organ function. Regarding the eye, organ function is usually measured as visual acuity. Results of previous studies,using many different analytical techniques and visual acuity end points, have indicated that visual acuity is generally preserved in patients with smaller uveal melanomas situated farther from the optic disc and foveola. In 1984, Cruess et alreported visual acuity results in a group of 77 patients treated with cobalt 60 plaque radiotherapy and found that a radiation dose of 5000 cGy or more has a toxic effect on the optic disc and fovea. They found final visual acuity of 20/200 or better with this dose in only 65% of eyes at the optic disc and 52% at the foveola.Similar results with proton beam radiotherapyand helium ion radiotherapyidentified factors related to poor visual outcome, including greater tumor thickness, closer proximity to optic disc and foveola, submacular fluid, worse pretreatment vision, and increasing radiation dose to optic disc, foveola, and lens.In this article, we analyze our experience with 1106 consecutive patients with visual acuity of 20/100 or better managed on the Oncology Service at Wills Eye Hospital, Philadelphia, Pa, with plaque radiotherapy over a period of 16 years. The impact of clinical and treatment factors on ultimate visual acuity was analyzed, and practical interpretation of this data is provided.PATIENTS AND METHODSThe clinical records of all patients with the diagnosis of uveal melanoma treated on the Oncology Service between July 1976 and June 1992 were reviewed. Patients with initial visual acuity of 20/100 or better in the affected eye were selected for inclusion in this study. Clinical data were gathered regarding patient and tumor features and radiation variables and then were analyzed for outcome of final visual acuity.In the following paragraphs, reference categories used in subsequent statistical analyses are marked with an asterisk (*). Patient features at initial examination included age, race (African American, Hispanic, Asian, or white*), sex (female or male*), medical problems (none,* diabetes mellitus, hypertension, or hypercholesterolemia), and previous or present chemotherapy. Ocular data included best-corrected Snellen visual acuity measurement at 20 feet (20/20, 20/25, 20/30, 20/40, 20/50, 20/60, 20/70, 20/80, 20/100, 20/200, 20/400, counting fingers, hand motions, light perception, or no light perception), lens (normal,* cataract, pseudophakia, or aphakia), glaucoma, and intraocular pressure.Tumor data included anatomical location (iris, ciliary body, or choroid*), meridional location of tumor epicenter (superior, superotemporal, temporal, inferotemporal, inferior, inferonasal,* nasal, superonasal, or macula), proximity to optic nerve and foveola (in millimeters), anterior and posterior tumor margins (iris, ora serrata, between ora serrata and equator, or posterior to equator*), largest basal dimension (based on ophthalmoscopy, in millimeters), largest thickness (based on ultrasonography, in millimeters), shape (dome,* mushroom, diffuse, or plateau), and subretinal fluid (absent* or present)Radiation plaque data included radioisotope (iodine 125,* ruthenium 106, cobalt 60, or iridium 192); plaque shape (round,* notched, curvilinear, or rectangular); plaque size; hours of radiation exposure; radiation dose (in centigray) to the tumor apex, tumor base, optic disc, foveola, and lens; and radiation rate (in centigray per hour) to the tumor apex, tumor base, optic disc, foveola, and lens. Adjuvant treatment of laser photocoagulation or thermotherapy was not applied in any case.Follow-up examinations were generally made at 3- to 6-month intervals for up to 5 years and at 6- to 12-month intervals thereafter. Follow-up data included the date and treatment of tumor recurrence. Tumor recurrence was defined as any amount of documented tumor growth in thickness or base detected by ophthalmoscopy or ultrasonography. At date last seen, the final best-corrected Snellen visual acuity was noted.STATISTICAL ANALYSISThe main outcome in this study was final visual acuity. Visual acuity was analyzed for 2 end points: final visual acuity (good [20/20-20/100] vs poor [20/200 to no light perception]) and loss of visual acuity (minimal [<5 Snellen lines] vs moderate [≥5 Snellen lines]).In the final visual acuity analyses, patients who ultimately underwent enucleation were combined with the poor visual acuity group (20/200 or worse and visual acuity decrease of ≥5 Snellen lines). In the loss of visual acuity analysis, pretreatment and posttreatment visual acuity at date last seen were analyzed for a decrement of at least 5 Snellen lines of acuity by a Cox proportional hazards model using time to event as the end point. The loss of a "line of acuity" was defined as a decrease from one visual acuity level to the next, such as from 20/100 to 20/200 or counting fingers to hand motions.The effect of individual clinical variables on the development of each outcome was analyzed by a series of univariate Cox porportional hazards regressions.Correlation among the variables was determined using Pearson correlations. All variables were analyzed as discrete variables except for patient age, intraocular pressure, tumor base, tumor thickness, proximity to the optic disc, proximity to the foveola, percentage overhang of the optic disc, radiation dose, and radiation rate, which were analyzed as continuous variables and later grouped into discrete categories to derive cutoff values. Variables that were significant on a univariable level (P<.05) were entered into a stepwise regression analysis. For variables that showed a high degree of correlation, only one variable from the set of associated variables was entered at a time in subsequent multivariate models. A final multivariable model fitted variables identified as significant predictors (P<.05) in the stepwise model and variables deemed clinically important for the visual acuity outcome.We also analyzed individually the 5 most important risk factors and the combined impact of 2, 3, 4, and all 5 predictive factors on ultimate visual acuity. Empirical data were tabulated regarding the number and percentage of patients demonstrating poor visual acuity (analysis 1) or moderate loss of visual acuity (analysis 2) for individual factors and a combination of factors. Using Kaplan-Meier estimates,the raw percentages were then adjusted for differing lengths of follow-up among the patients, and 5-year estimates of percentage of poor visual acuity (analysis 1) or moderate visual acuity loss (analysis 2) with various combinations of risk factors were calculated.Relative risks (RRs) were calculated for poor visual acuity (analysis 1) or moderate loss of visual acuity (analysis 2) given a single factor or a constellation of factors. All of the covariates were fit simultaneously into a final multivariable model, and the risk estimates were computed for each covariate. The formula for computing the RR for combinations of factors was as follows:RR = Exponentiation (fl1+ fl2+ . . . fln)For a 2-factor model in analysis of combined risks for poor visual acuity, fl1was the parameter estimate for the first factor (eg, tumor thickness; fl1= 0.642, RR = 1.9) and fl2was the parameter estimate for the second factor (eg, proximity to foveola <5 mm; fl2= 0.405, RR = 1.4). The combined risk was then exponentiation (0.642 + 0.405) = exponentiation (1.047) = 2.9.Kaplan-Meier survival estimates were used to analyze the development of poor visual acuity (20/200 to no light perception) and loss of more than 5 Snellen visual acuity lines as a function of time. Additional Kaplan-Meier estimates were performed to assess the previous 2 visual acuity end points over time as a function of tumor size graded as small (≤3.0 mm), medium (3.1-8.0 mm), and large (>8.0 mm) as well as proximity to the visually important structures of the optic disc and foveola (>5.0 mm from both).RESULTSGENERAL DATAThere were 1300 consecutive patients with uveal melanoma managed with plaque radiotherapy during the 16 years of this study. The mean age of the 1106 patients who had an initial visual acuity of 20/100 or better at the time of plaque treatment was 58 years (median, 59 years; range, 10-91 years). There were 1092 whites (99%), 11 African Americans (1%), 2 Hispanics (<1%), and 1 Asian (<1%); 548 patients (50%) were male and 558 (50%) were female.Initial visual acuity was 20/20 to 20/30 in 696 patients (63%) and 20/40 to 20/100 in 410 (37%). The anatomical location of the tumor was iridic in 10 patients (<1%), iridociliary in 7 (<1%), iridociliochoroidal in 2 (<1%), ciliochoroidal in 101 (9%), and choroidal in 986 (89%). The tumor meridional location was superior in 63 patients (6%), superotemporal in 247 (22%), temporal in 123 (11%), inferotemporal in 263 (24%), inferior in 69 (6%), inferonasal in 139 (13%), nasal in 59 (5%), superonasal in 133 (12%), and macula in 10 (<1%). Mean proximity to the optic nerve was 5.1 mm (median, 5 mm; range, 0-20 mm) and to the foveola was 4.7 mm (median, 4 mm; range, 0-22 mm). In 26 eyes (2%), the tumor overhung the optic nerve. The mean largest tumor basal dimension was 10.3 mm (median, 10 mm; range, 1.5-20.0 mm) and the mean largest tumor thickness was 4.7 mm (median, 4 mm; range, 0.3-12.2 mm). The tumor shape was dome in 991 eyes (90%), mushroom in 111 (10%), diffuse in 2 (<1%), and plateau in 2 (<1%). Subretinal fluid was present in 750 patients (68%).The radioisotope used in the plaque was iodine 125 in 649 patients (59%), ruthenium 106 in 60 (5%), cobalt 60 in 300 (27%), and iridium 192 in 97 (9%). The plaque shape was round in 852 patients (77%), notched in 234 (21%), curvilinear in 15 (1%), and rectangular in 5 (<1%). The most common plaque size was 15 mm. Mean time of radiation exposure was 141.2 hours (median, 121.9 hours; range, 8.5-511.3 hours). Mean radiation dose to the tumor apex was 9517 cGy (median, 9090 cGy; range, 1200-9780 cGy), to the tumor base was 34,128 cGy (median, 33,000 cGy; range, 9000-70,300 cGy), to the optic disc was 7064 cGy (median, 4118 cGy; range, 47-62,055 cGy), to the foveola was 8694 cGy (median, 5276 cGy; range, 95-42,577 cGy), and to the lens was 2408 cGy (median, 1470 cGy; range, 33-31,450 cGy). Mean radiation rate to the tumor apex was 80 cGy/h (median, 78 cGy/h; range, 9-126 cGy/h), to the tumor base was 280 cGy/h (median, 261 cGy/h; range, 41-412 cGy/h), to the optic disc was 66 cGy/h (median, 40 cGy/h; range, 3-418 cGy/h), to the foveola was 82 cGy/h (median, 52 cGy/h; range, 1-462 cGy/h), and to the lens was 23 cGy/h (median, 14 cGy/h; range, 0.3-282 cGy/h).At date last seen, the final best-corrected Snellen visual acuity was good (20/20-20/100) in 539 patients (49%) and poor (20/200 to no light perception) in 567 (51%). Loss of less than 5 Snellen lines of visual acuity was noted in 515 patients (46%) and loss of 5 or more lines was found in 591 (54%).ANALYSIS 1: FINAL VISUAL ACUITYOn univariable analysis, factors related to poor visual acuity outcome (20/200 to no light perception) included patient age of 60 years or older; tumor location in superior, temporal, inferotemporal, or superonasal portions of the fundus; tumor base of 10 mm or more; tumor thickness greater than 8 mm; mushroom tumor shape; proximity to the optic disc (decreasing); proximity to the foveola less than 5 mm; tumor overhang of the optic disc by 50% or more; anterior tumor margin posterior to the equator; presence of subretinal fluid; use of the radioactive isotope iridium; notched plaque shape; a radiation dose at the tumor apex of 9000 cGy or higher; a radiation dose at the tumor base of 33,300 cGy or higher; increasing radiation dose and rate at optic disc; increasing radiation rate at lens and tumor recurrence (Table 1).Table 1. Plaque Radiotherapy for 1106 Patients With Uveal Melanoma: Univariable Analyses of the Significant Clinical Factors Related to Poor Visual Acuity of 20/200 to No Light Perception*Variable†Description‡Variable Significance Level§PRR (95% CI)Age>60 vs ≤60 yb<.0011.3 (1.1-1.6)Radiation dose at tumor apex<9000 vs ≥9000 cGyb.0011.3 (1.1-1.6)Radiation dose at tumor base≥33,300 vs <33,300 cGyb<.0011.3 (1.1-1.6)Radiation dose at discPer 10-Gy increasec.011.03 (1.01-1.05)Radiation rate at discPer 100x (1 cGy/h) increasec.0091.4 (1.1-1.8)Radiation rate at lensPer 100x (1 cGy/h) increasec.031.7 (1.1-2.8)Tumor base≥10 mm vs <10 mmb<.0011.4 (1.2-1.6)Thickness>3-≤8 vs ≤3 mmb.0081.3 (1.1-1.6)>8 vs ≤3 mmd<.0013.0 (2.3-4.0)Tumor shapeMushroom vs domed<.0011.8 (1.4-2.3)Subretinal fluidPresent vs absentd.0041.3 (1.1-1.6)Tumor quadrantSuperior vs inferonasald.031.6 (1.1-2.4)Superotemporal vs inferonasald.241.2 (0.9-1.7)Temporal vs inferonasald.041.5 (1.1-2.1)Inferotemporal vs inferonasald.041.4 (1.1-1.9)Inferior vs inferonasald.091.5 (0.9-2.3)Nasal vs inferonasald.171.4 (0.9-2.1)Superonasal vs inferonasald.021.5 (1.1-2.2)Macula vs inferonasald.062.2 (1.0-4.9)IsotopeIridium vs iodined.031.4 (1.1-1.8)Ruthenium vs iodined.300.8 (0.5-1.2)Cobalt vs iodined.060.8 (0.7-1.1)Proximity to optic discPer 1-mm increase in proximityc.0011.04 (1.02-1.07)Proximity to foveola<5 vs ≥5 mmb<.0011.4 (1.2-1.7)Tumor overhang of disc≥50% vs <50%d.032.0 (1.1-3.8)Anterior tumor marginPosterior vs ora to equatord.031.3 (1.0-1.6)Posterior vs ora serratad.290.9 (0.7-1.1)Posterior vs irisd.630.8 (0.4-1.8)Plaque shapeNotched vs roundd<.0011.5 (1.2-1.8)Tumor recurrence&par;Present vs absentd<.0013.7 (2.8-4.9)*Includes enucleation as an end point for poor vision. RR indicates relative risk; CI, confidence interval.†All variables were analyzed in a series of bivariate Cox proportional hazards models that controlled for initial visual acuity.‡Reference variable is listed second.§Variable significant on a continuous (c), discrete (d), or both continuous and discrete (b) level. For discrete variables, the significant cutoff value is stated.&par;First tumor recurrence analyzed as a time-dependent covariate.On multivariable analysis, the best combination of factors related to poor visual acuity outcome (20/200 to no light perception) were patient age of 60 years or older, poor initial visual acuity, increasing tumor thickness, proximity to foveola less than 5 mm, anterior tumor margin posterior to the equator, presence of subretinal fluid, radioactive isotope (ruthenium, cobalt, and iridium), notched plaque shape, and tumor recurrence (Table 2).Table 2. Plaque Radiotherapy for 1106 Patients With Uveal Melanoma: Multivariable Analyses of the Significant Clinical Factors Related to Poor Visual Acuity of 20/200 to No Light Perception*Variable†Description‡Variable Significance Level§PRR (95% CI)Initial visual acuityPer 1 Snellen line worsec.031.04 (1.01-1.08)Age>60 vs ≤60 yd<.0011.4 (1.1-1.6)ThicknessPer 1-mm increaseb<.0011.19 (1.14-1.24)IsotopeRuthenium vs iodined.021.7 (1.1-2.6)Cobalt vs iodined.0061.3 (1.1-1.6)Iridium vs iodined.031.4 (1.1-1.9)Subretinal fluidPresent vs absentd.0021.3 (1.1-1.6)Proximity to foveola<5 vs ≥5 mmb<.0011.5 (1.3-1.9)Anterior tumor marginOra to equator vs posteriord.011.4 (1.1-1.7)Ora serrata vs posteriord.280.8 (0.7-1.1)Iris vs posteriord.560.6 (0.1-3.3)Plaque shapeNotched vs roundd<.0011.5 (1.3-1.9)Tumor recurrence&par;Present vs absentd<.0014.4 (3.2-6.0)*Includes enucleation as an end point for poor vision. RR indicates relative risk; CI, confidence interval.†All variables were analyzed in a series of bivariate models that controlled for initial visual acuity.‡The reference variable is listed second.§Variable significant on a continuous (c), discrete (d), or both continuous and discrete (b) level. For discrete variables, the significant cutoff value is stated.&par;First tumor recurrence analyzed as a time-dependent covariate.Using Kaplan-Meier estimates, 3% of patients had poor visual acuity at 1 year, 34% at 5 years, 68% at 10 years, and 87% at 15 years. When evaluating final visual acuity as a function of tumor thickness using Kaplan-Meier estimates, eyes with a small melanoma demonstrated poor vision in less than 1% of patients at 1 year, 24% at 5 years, and 60% at 10 years. Eyes with a medium melanoma demonstrated poor vision in 3% of patients at 1 year, 31% at 5 years, and 69% at 10 years. Eyes with a large melanoma displayed poor vision in 8% of patients at 1 year and 64% at 5 years (too few patients were available for a reliable 10-year estimate) (Figure 1).Figure 1.Kaplan-Meier estimates showing the proportion of patients free of poor visual acuity (20/200 to no light perception) over time according to tumor thickness for 1106 patients with uveal melanoma treated with plaque radiotherapy.When evaluating final visual acuity as a function of tumor proximity to visually vital structures, eyes with uveal melanoma within 5 mm of the optic disc or foveola showed poor visual acuity in 2% of patients at 1 year, 35% at 5 years, and 73% at 10 years. Eyes with uveal melanoma 5 mm or more from the optic disc and foveola showed poor visual acuity in 4% of patients at 1 year, 25% at 5 years, and 57% at 10 years (Figure 2).Figure 2.Kaplan-Meier estimates showing the proportion of patients free of poor visual acuity (20/200 to no light perception) over time according to proximity of the tumor to the optic disc and foveola for 1106 patients with uveal melanoma treated with plaque radiotherapy.Analysis of the combined effects of the predictive factors at the time of treatment revealed ultimate poor visual acuity in 19% of patients with no risk factors and in a mean of 39% of patients with 1 factor, 49% with 2 factors, 58% with 3 factors, 64% with 4 factors, and 50% with 5 factors (Table 3). The combination of clinical factors most predictive of poor visual outcome were an age of 60 years or older, tumor thickness greater than 8 mm, and the presence of subretinal fluid (80% of patients developed poor visual acuity by 5 years). Analysis of the combined effects of the RRs of the predictive factors at the time of treatment revealed a RR for poor visual acuity of 1.3 to 2.6 with 1 factor, 1.8 to 3.6 with 2 factors, 2.5 to 5.1 with 3 factors, 3.6 to 7.1 with 4 factors, and 9.3 with 5 factors.Table 3. Plaque Radiotherapy for 1106 Patients With Uveal Melanoma: Estimates of Relative Risk* and Kaplan-Meier Probability† of Poor Visual Acuity of 20/200 to No Light Perception for Combinations of Clinical FactorsClinical FactorDescriptionKaplan-Meier Probability (95% Confidence Interval)RR (95% Confidence Interval)Patients at Risk (No. of Events/No. With Factor)None0.19 (0.08-0.30)1.018/64A: Age≥60 vs <60‡ y0.37 (0.33-0.42)1.4 (1.2-1.6)301/540B: Proximity to foveola<5 vs ≥5‡ mm0.35 (0.31-0.40)1.4 (1.2-1.7)368/616C: Thickness>8 vs ≤8‡ mm0.64 (0.54-0.75)2.6 (2.0-3.3)75/100D: Subretinal fluidPresent vs absent0.34 (0.30-0.38)1.3 (1.1-1.6)412/758E: Anterior tumor marginPosterior to equator vs other0.26 (0.19-0.32)1.4 (1.1-1.7)94/217AB0.42 (0.36-0.49)2.0 (1.4-2.7)190/293AC0.70 (0.57-0.83)3.6 (2.3-5.4)46/58AD0.42 (0.36-0.48)1.8 (1.3-2.6)209/349AE0.32 (0.22-0.41)2.0 (1.3-2.8)53/114BC0.64 (0.49-0.79)3.6 (2.3-5.5)36/46BD0.38 (0.33-0.43)1.8 (1.3-2.7)284/465BE0.36 (0.25-0.47)2.0 (1.3-2.8)50/101CD0.68 (0.56-0.80)3.4 (2.2-5.3)51/69CE0.66 (0.40-0.92)3.6 (2.2-5.6)10/16DE0.29 (0.21-0.38)1.8 (1.2-2.7)70/163ABC0.66 (0.47-0.85)5.1 (2.7-9.0)22/27ABD0.46 (0.38-0.53)2.5 (1.5-4.4)140/209ABE0.45 (0.30-0.60)2.7 (1.5-4.6)32/56ACD0.80 (0.67-0.94)4.7 (2.6-8.7)31/38ACE0.66§ (0.27-1.00)5.1 (2.5-9.1)5/7ADE0.38 (0.26-0.50)2.5 (1.4-4.5)39/78BCD0.73 (0.56-0.89)4.7 (2.6-8.9)28/35BCE0.69 (0.39-0.98)5.1 (2.5-9.3)8/12BDE0.38 (0.26-0.50)2.5 (1.4-4.6)41/84CDE0.62 (0.30-0.94)4.7 (2.4-9.0)6/12ABCD0.75 (0.53-0.96)6.6 (3.0-14.5)15/19ABCE0.60§ (0.17-1.00)7.1 (2.9-15.3)4/6ABDE0.50 (0.30-0.70)3.6 (1.6-7.5)24/41ACDE0.63&par; (0.06-1.00)6.6 (2.8-14.8)2/4BCDE0.70 (0.35-1.00)6.6 (2.8-15.1)5/9ABCDE0.5&par; (0.01-1.00)9.3 (3.3-24.7)1/3*Combinations of relative risk (RR) were computed from the final multivariable model containing only these variables.†Kaplan-Meier estimates were computed for 5-year outcome.‡The reference category.§Only the 4-year estimate was available.&par;Only the 2-year estimate was available.ANALYSIS 2: LOSS OF VISUAL ACUITYOn univariable analysis, factors related to loss of visual acuity (≥5 lines of Snellen visual acuity) included patient age of 60 years or older, underlying diabetes mellitus or hypertension, tumor location in superonasal fundus, tumor base of 10 mm or more, tumor thickness of 3 mm or more, mushroom tumor shape, retinal invasion by uveal melanoma, proximity to the optic disc less than 5 mm, proximity to the foveola less than 5 mm, tumor overhang of the optic disc by 50% or more, anterior tumor margin posterior to the equator, presence of subretinal fluid, notched plaque shape, a radiation dose at the tumor apex of 9000 cGy or higher, a radiation dose at the tumor base of 33,300 cGy or higher, increasing radiation dose and rate at optic disc, and tumor recurrence (Table 4).Table 4. Plaque Radiotherapy for 1106 Patients With Uveal Melanoma: Univariable Analyses of the Significant Clinical Factors Related to Loss of at Least 5 Lines of Snellen Visual Acuity*Variable†Description‡Variable Significance Level§PRR (95% CI)Age>60 vs ≤60 yb<.0011.1 (1.1-1.6)Radiation dose at tumor apex<9000 vs ≥9000 cGyb<.0011.2 (1.1-1.5)Radiation dose at tumor base≥33,300 vs <33,300 cGyb.0091.3 (1.1-1.5)Radiation dose at discPer 10-Gy increasec.0021.02 (1.01-1.04)Radiation rate at discPer 100x (1 cGy/h) increasec.041.29 (1.01-1.65)Tumor base≥10 mm vs <10 mmb.041.4 (1.2-1.7)Thickness>3-≤8 vs ≤3 mmb.0021.3 (1.1-1.6)>8 vs ≤3 mmb<.0013.0 (2.2-3.9)Medical problemsDiabetes/hypertension vs noned.0051.4 (1.1-1.7)Retinal invasion by tumorPresent vs absentd.022.5 (1.2-5.3)Tumor shapeMushroom vs domed<.0011.7 (1.4-2.2)Subretinal fluidPresent vs absentd<.0011.4 (1.2-1.6)Tumor quadrantSuperior vs inferonasald.081.5 (1.0-2.2)Superotemporal vs inferonasald.281.2 (0.9-1.6)Temporal vs inferonasald.101.3 (1.0-1.9)Inferotemporal vs inferonasald.161.2 (0.9-1.7)Inferior vs inferonasald.091.5 (1.0-2.2)Nasal vs inferonasald.431.2 (0.8-1.8)Superonasal vs inferonasald.041.5 (1.1-2.0)Macula vs inferonasald.082.0 (0.9-4.4)Proximity to optic disc<5 vs ≥5 mmb.021.2 (1.1-1.4)Proximity to foveola<5 vs ≥5 mmb<.0011.4 (1.1-1.6)Tumor overhang of disc≥50% vs <50%d.032.1 (1.1-4.1)Anterior tumor marginOra to equator vs posteriord.251.1 (0.9-1.4)Ora serrata vs posteriord.260.9 (0.7-1.1)Iris vs posteriord.042.0 (1.1-4.0)Plaque shapeNotched vs roundd<.0011.5 (1.2-1.8)Tumor recurrence&par;Present vs absentd<.0013.6 (2.7-4.8)*Includes enucleation as an end point for poor vision. RR indicates relative risk; CI, confidence interval.†All variables were analyzed in a series of bivariate models that controlled for initial visual acuity.‡The reference variable is listed second.§Variables significant on a continuous (c), discrete (d), or both continuous and discrete (b) level. For discrete variables, the significant cutoff value is stated.&par;First tumor recurrence analyzed as a time-dependent covariate.On multivariable analysis, the best combination of factors related to loss of visual acuity (≥5 lines of Snellen visual acuity) were patient age of 60 years or older, medical problems of diabetes mellitus or hypertension, worse initial visual acuity, tumor thickness (>8 mm), proximity to foveola less than 5 mm, presence of subretinal fluid, notched plaque shape, and tumor recurrence (Table 5).Table 5. Plaque Radiotherapy for 1106 Patients With Uveal Melanoma: Multivariate Analyses of the Significant Clinical Factors Related to Loss of at Least 5 Lines of Snellen Visual Acuity*Variable†Description‡Variable Significance Level§PRR (95% CI)Initial visual acuityPer 1 Snellen line worseb<.0011.4 (1.2-1.6)Age>60 vs ≤60 yb<.0011.4 (1.1-1.6)Thickness>8 vs ≤8 mmb<.0011.17 (1.13-1.21)Medical problemsDiabetes/hypertension vs noned.0041.4 (1.1-1.7)Subretinal fluidPresent vs absentd.0011.4 (1.1-1.7)Proximity to foveola<5 vs ≥5 mmb<.0011.4 (1.2-1.7)Plaque shapeNotched vs roundd<.0011.5 (1.2-1.8)Tumor recurrence&par;Present vs absentd<.0014.1 (3.0-5.5)*Includes enucleation as an end point for poor vision. RR indicates relative risk; CI, confidence interval.†All variables analyzed in series of bivariate models that controlled for initial visual acuity.‡The reference variable is listed second.§Variables significant on a continuous (c), discrete (d), or both continuous and discrete (b) level. For discrete variables, the significant cutoff value is stated.&par;First tumor recurrence analyzed as time-dependent covariate.Using Kaplan-Meier estimates, 3% of patients had loss of visual acuity at 1 year, 33% at 5 years, 69% at 10 years, and 88% at 15 years. When evaluating loss of visual acuity as a function of tumor thickness using Kaplan-Meier estimates, eyes with a small melanoma demonstrated loss of visual acuity in less than 1% of patients at 1 year, 26% at 5 years, and 62% at 10 years. Eyes with a medium melanoma demonstrated loss of visual acuity in 3% of patients at 1 year, 32% at 5 years, and 70% at 10 years. Eyes with a large melanoma displayed loss of visual acuity in 7% of patients at 1 year, 61% at 5 years, and 87% at 10 years (Figure 3).Figure 3.Kaplan-Meier life table curves showing the proportion of patients free of moderate loss of visual acuity (≥5 Snellen lines) over time according to tumor thickness for 1106 patients with uveal melanoma treated with plaque radiotherapy.When evaluating loss of visual acuity as a function of tumor proximity to visually vital structures, eyes with uveal melanoma within 5 mm of the optic disc or foveola showed loss of visual acuity in 2% of patients at 1 year, 36% at 5 years, and 74% at 10 years. Eyes with uveal melanoma 5 mm or more from the optic disc and foveola showed loss of visual acuity in 3% of patients at 1 year, 26% at 5 years, and 59% at 10 years (Figure 4).Figure 4.Kaplan-Meier life table curves showing the proportion of patients free of moderate loss of visual acuity (≥5 Snellen lines) over time according to proximity of the tumor to the optic disc and foveola for 1106 patients with uveal melanoma treated with plaque radiotherapy.Analysis of the combined effects of the predictive factors at the time of treatment revealed loss of visual acuity (≥5 lines of Snellen visual acuity) in 15% of patients with no clinical risk factors and in a mean of 42% of patients with 1 factor, 53% with 2 factors, 63% with 3 factors, and 64% with 4 factors (Table 6). The combination of clinical factors with greatest predictive value for loss of visual acuity were tumor thickness greater than 8 mm, presence of subretinal fluid, tumor proximity within 5 mm of foveola, and underlying medical problems of diabetes mellitus or hypertension (83% of patients developed loss of visual acuity). Analysis of the combined effects of the RRs of the predictive factors at the time of treatment revealed a RR for loss of visual acuity of 1.3 to 2.5 with 1 factor, 1.7 to 3.5 with 2 factors, 2.4 to 4.9 with 3 factors, 3.3 to 6.4 with 4 factors, and 8.3 with 5 factors present.Table 6. Plaque Radiotherapy for 1106 Patients With Uveal Melanoma: Estimates of Relative Risk* and Kaplan-Meier Probability† of Loss of at Least 5 Lines of Snellen Visual Acuity for Combinations of Clinical FactorsClinical FactorDescriptionKaplan-Meier Probability (95% Confidence Interval)RR (95% Confidence Interval)Patients at Risk (No. of Events/No. With Factor)None0.15 (0.06-0.25)1.021/67A: Age≥60 vs <60‡ y0.39 (0.35-0.44)1.4 (1.2-1.6)311/536B: Thickness>8 vs ≤8‡ mm0.61 (0.51-0.72)2.5 (1.9-3.2)70/100C: Subretinal fluidPresent vs absent0.36 (0.32-0.40)1.4 (1.2-1.7)418/750D: Proximity to foveola<5 vs ≥5‡ mm0.37 (0.33-0.41)1.3 (1.1-1.6)365/609E: Medical problemsDiabetes/hypertension vs none0.39 (0.31-0.47)1.3 (1.1-1.7)107/179AB0.69 (0.56-0.82)3.5 (2.3-5.3)43/58AC0.44 (0.38-0.50)2.0 (1.4-2.8)212/347AD0.44 (0.37-0.50)1.8 (1.3-2.6)190/290AE0.43 (0.32-0.53)1.8 (1.3-2.7)63/110BC0.65 (0.52-0.78)3.5 (2.3-5.5)49/69BD0.62 (0.46-0.77)3.3 (2.2-5.1)35/46BE0.71 (0.47-0.95)3.3 (2.1-5.4)12/15CD0.40 (0.35-0.45)1.8 (1.3-2.7)278/459CE0.48 (0.37-0.58)1.8 (1.3-2.9)69/113DE0.43 (0.33-0.54)1.7 (1.2-2.7)68/103ABC0.80 (0.66-0.95)4.9 (2.7-9.0)30/38ABD0.67 (0.48-0.86)4.6 (2.5-8.4)21/27ABE0.49§ (0.09-0.89)4.6 (2.5-8.8)4/7ACD0.47 (0.39-0.54)2.5 (1.5-4.5)138/207ACE0.52 (0.38-0.66)2.5 (1.5-4.7)39/68ADE0.52 (0.38-0.65)2.4 (1.4-4.4)44/66BCD0.71 (0.55-0.88)4.6 (2.5-8.7)28/35BCE0.87 (0.63-1.00)4.6 (2.5-9.1)8/9BDE0.71 (0.47-0.95)4.2 (2.4-8.5)12/15CDE0.53 (0.40-0.66)2.4 (1.4-4.5)45/68ABCD0.78 (0.57-0.98)6.4 (3.0-14.3)15/19ABCE0.33&par; (0.01-0.87)6.4 (2.9-15.0)2/3ABDE0.67§ (0.13-1.00)5.9 (2.7-14.0)3/4ACDE0.61 (0.43-0.78)3.3 (1.7-7.5)27/42BCDE0.83 (0.54)5.9 (2.8-14.6)6/7ABCDE. . .8.3 (3.2-23.9)1/2*Combinations of relative risk (RR) were computed from the final multivariable model containing only these variables.†Kaplan-Meier estimates computed for 5-year outcome (unadjusted for initial vision).‡The reference category.§Only the 4-year estimate was available.&par;Only the 1-year estimate was available.COMMENTThe ideal treatment for uveal melanoma is not certain. The primary goal of treatment is to eradicate the tumor and provide the best life prognosis for the patient. Secondary goals of treatment include retention of the globe and visual function. Options for management include enucleation, plaque radiotherapy, charged particle radiotherapy, local resection, thermotherapy, and combinations of these methods.These techniques, in various previous studies,have been found to be equivalent in securing life prognosis. In addition, the added benefit of possible visual function is provided by the conservative techniques that spare the eye.Plaque radiotherapy continues to be one of the most popular methods of conservative management for uveal melanoma. Plaque radiotherapy results in ocular salvage in 94% of patients.However, ocular radiation complications remain problematic and vary depending on tumor location, size, and other factors.In an extensive reviewof 136 patients with ciliary body melanoma treated with plaque radiotherapy, radiation complications at 5 years included cataract in 48% of patients, neovascular glaucoma in 21%, retinopathy in 20%, scleral necrosis in 12%, vitreous hemorrhage in 11%, and papillopathy in 3%. Most of these complications contributed to associated visual acuity loss. In contrast, a reviewof 630 patients with choroidal melanoma in the visually sensitive macular region revealed complications at 5 years of cataract in 32% of patients, neovascular glaucoma in 8%, retinopathy (maculopathy) in 40%, scleral necrosis in less than 1%, vitreous hemorrhage in 9%, and papillopathy in 13%. Reasons for visual acuity loss differed substantially between these 2 groups. However, visual acuity decrement by 3 or more Snellen lines was similar at 40% at 5 years in both groups of patients.Several studies during the past 2 decades have focused specifically on visual results after radiotherapy of choroidal melanoma. Cruess and associatesin 1984 evaluated 77 eyes with posterior uveal melanoma treated with cobalt plaque radiotherapy. They evaluated patients with initial visual acuity of 20/25 or better only. Therefore, their study was designed to assess visual results in a subset of patients with only excellent initial visual acuity. At 36 months, 81% had visual acuity better than 20/200 and 62% had visual acuity better than 20/50. When assessing visual acuity with regard to tumor size, 95% of eyes with a small or medium posterior uveal melanoma had visual acuity better than 20/200 at 36 months, whereas only 48% of eyes with a large tumor demonstrated visual acuity better than 20/200. They also found that worse visual results occur with tumors near the optic disc or fovea, especially when the radiation dose to these structures exceeds 5000 cGy.In 1986, Seddon and coworkersreported visual outcome in 440 eyes with posterior uveal melanoma treated with proton beam radiation. The scope of their study was slightly greater than that of Cruess et alin that theyincluded all patients with initial visual acuity of 20/200 or better. They identified important risk factors for poor visual outcome, including greater tumor thickness, proximity to optic disc and fovea, and pretreatment visual acuity. They noted that tumors with a thickness of 5.0 mm or less, distance from the disc and fovea greater than 3.0 mm, and pretreatment visual acuity of 20/40 or better were most likely to offer the best visual prognosis.In 1996, Char and associatesreported on 426 eyes with ciliochoroidal melanoma that were eligible for randomization (enucleation vs plaque radiotherapy) in the COMSbut were treated outside the study with radiotherapy (helium ion or plaque). They found that 36 months after treatment, 36% of eyes had visual acuity of 6/12 or better. The length of visual retention depended most on tumor thickness; tumor location with respect to the optic nerve, fovea, and ciliary body; and patient age. They concluded that some patients eligible for COMS randomization to enucleation vs plaque radiotherapy can retain excellent long-term vision, and this should be explained to each candidate prior to randomization.The previous studies added to our understanding of visual outcome after plaque radiotherapy for uveal melanoma. In our analysis, we attempted to be comprehensive and included all 1106 patients who were not legally blind (20/100 or better) and assessed their risk to become legally blind (20/200 or worse) in the affected eye. We did not analyze only tumors at certain sites in the eye with certain thicknesses, as in the COMS eligibility criteria, but we assessed all tumors at all sites in the eye to provide a realistic view of visual acuity after ocular radiotherapy.Many of the previously identified risk factors for poor visionwere confirmed in our analysis of this larger group of patients. In the multivariable analysis we found that factors that significantly predict poor visual outcome of 20/200 or worse were numerous and included older patient age (≥60 years), poor initial visual acuity, increasing tumor thickness, posteriorly situated tumor less than 5.0 mm to the foveola, presence of associated subretinal fluid, notched plaque shape (because of the proximity of the tumor ≤2.0 mm from the optic disc), use of a radioisotope other than iodine 125 (eg, cobalt 60, iridium 192, or ruthenium 106), and ultimate tumor recurrence (Table 2). When assessing the most important of these factors, the presence of 1 of these factors leads to a 39% chance for ultimate poor vision and the presence of 2 or more factors leads to an approximately 50% or greater chance for poor ultimate vision (Table 3). Five years after treatment, 24% of patients with a small melanoma had poor vision, whereas 31% with a medium tumor and 64% with a large tumor had poor visual acuity (Figure 1).Similarly, risks for loss of 5 Snellen lines of visual acuity paralleled the above findings. In the multivariable analysis, risks for visual acuity loss of 5 Snellen lines included older patient age (≥60 years), medical problems of diabetes mellitus or hypertension, poor initial visual acuity, greater tumor thickness, presence of subretinal fluid, proximity of the tumor less than 5.0 mm to the foveola and 2.0 mm or less to the optic disc (notched plaque), and ultimate tumor recurrence (Table 5). When assessing the most important of these factors, the presence of 1 of these factors leads to a 42% chance for an ultimate 5 lines of visual loss, and 2 or more factors leads to approximately 50% or greater risk for visual acuity loss (Table 6). Five years after treatment, 26% of patients with a small melanoma had visual acuity loss of at least 5 Snellen lines, whereas 32% with a medium tumor and 61% with a large tumor showed similar visual acuity loss (Figure 3).Factors that offered the worst ultimate visual results were patient age of 60 years or older, tumor thickness greater than 8 mm, and presence of subretinal fluid, and these clinical factors were associated with poor visual acuity in 80% of patients (Table 3). Factors that provided the greatest risk for loss of at least 5 lines of Snellen visual acuity were tumor thickness greater than 8 mm, presence of subretinal fluid, tumor within 5 mm of the foveola, and underlying medical problems such as diabetes mellitus and hypertension, causing visual loss in 83% of affected eyes (Table 6). The combinations of risk factors affecting vision were calculated in our study and will be important in counseling of patients for plaque radiotherapy, allowing them reasonable expectations for visual outcome.Our study has potential limitations. This retrospective study reflects our experience over a 16-year period as a tertiary care referral center. Thus, we may have managed more difficult cases, which would engender poorer visual outcome. Included in this study are patients with large tumors who refused enucleation and patients with large uveal melanoma in their only seeing eye, situations that lead to referral to our service and subsequent irradiation of the tumor that would otherwise have been managed with enucleation. Also included are patients who refused or were not eligible for randomization into the COMS.Many of those not eligible had juxtapapillary choroidal melanoma that we subsequently treated with custom-designed plaque radiotherapy, realizing that the visual prognosis would be guarded.In addition, there may be a bias in this study toward management of uveal melanoma with irradiation, whereas other surgeons would have preferred enucleation or local surgical resection.The subset of patients with hope for long-term good vision after plaque radiotherapy for choroidal melanoma include younger patients with small tumors at sites remote from the optic disc and foveola. It is important for all patients treated with plaque radiotherapy for choroidal melanoma to realize that the globe is usually salvaged, in 94% of cases,using this organ-sparing technique, but the visual function of the eye is limited and ultimate visual outcome is generally poor. However, even with relatively poor visual outcome, Tunc and associatesfound that quality of life was best in patients who received irradiation for choroidal melanoma compared with enucleation and tumor resection. Othershave reported that the choice of treatment for choroidal melanoma does not seem to be associated with large differences in quality of life in the long term.The results of our study may be useful for more accurate counseling of patients with uveal melanoma with regard to visual expectations. It is hoped that tumor control and visual results can be improved with newer combination techniques of plaque radiotherapy and adjunctive thermotherapy. Visual results after these newer therapeutic modifications will not be available for several years. Providing a lower radiation dosein tumors treated with combined modalities is possible, with the goal to avoid long-term radiation complications. Newer methods of transpupillary thermotherapy alone may also reasonably spare vision for smaller tumors in the paramacular and peripapillary region.With these advances, it is possible that patients will have a better visual outcome after conservative treatment for uveal melanoma.JAShieldsCLShieldsIntraocular Tumors: A Text and Atlas.Philadelphia, Pa: WB Saunders Co; 1992:171-205.JAShieldsCLShieldsAtlas of Intraocular Tumors.Philadelphia, Pa: Lippincott Williams & Wilkins; 1999:113-140.JAShieldsCLShieldsLADonosoManagement of posterior uveal melanoma.Surv Ophthalmol.1991;36:161-195.PTFingerRadiation therapy for choroidal melanoma.Surv Ophthalmol.1997;42:215-232.CLShieldsJAShieldsKGündüzRadiation therapy for uveal malignant melanoma.Ophthalmic Surg Lasers.1998;29:397-409.PKLommatzschResults after β-irradiation (106Ru/106Rh) of choroidal melanomas: 20 years' experience.Br J Ophthalmol.1986;70:844-851.SPackerSStollerMLLesserLong-term results of 125I irradiation of uveal melanoma.Ophthalmology.1992;99:767-774.SSeregardETrampeILaxResults following episcleral ruthenium plaque radiotherapy for posterior uveal melanoma: the Swedish experience.Acta Ophthalmol Scand.1997;75:11-16.JJAugsburgerJWGamelKLauritzenLWBradyCobalt 60 plaque radiotherapy vs enucleation for posterior uveal melanoma.Am J Ophthalmol.1990;109:585-592.JJAugsburgerZMCorreaJFriereLWBradyLong-term survival in choroidal and ciliary body melanoma after enucleation versus plaque radiation therapy.Ophthalmology.1998;105:1670-1678.Collaborative Ocular Melanoma StudyDesign and methods of a clinical trial for a rare condition: the Collaborative Ocular Melanoma Study.Control Clin Trials.1993;14:362-391.AFCruessJJAugsburgerJAShieldsVisual results following cobalt plaque radiotherapy for posterior uveal melanomas.Ophthalmology.1984;91:131-136.JSeddonESGragoudasLPolivogianisVisual outcome after proton beam irradiation of uveal melanoma.Ophthalmology.1986;93:666-674.JMSeddonESGragoudasKMEganUveal melanomas near the optic disc and fovea: visual results after proton beam irradiation.Ophthalmology.1987;94:354-361.DHCharSKrollJMQuiveyJCastroLong-term visual outcome of radiated uveal melanomas in eyes eligible for randomization to enucleation versus brachytherapy.Br J Ophthalmol.1996;80:117-124.DRGuyerSMukaiKMEganJMSeddonSMWalshESGragoudasRadiation maculopathy after proton beam irradiation for choroidal melanoma.Ophthalmology.1992;99:1275-1285.KGündüzCLShieldsJAShieldsPlaque radiotherapy for uveal melanoma with predominant ciliary body involvement.Arch Ophthalmol.1999;117:170-177.KGündüzCLShieldsJAShieldsJCaterJEFreireLWBradyRadiation complications and tumor control after plaque radiotherapy of choroidal melanoma with macular involvement.Am J Ophthalmol.1999;127:579-589.PDePotterCLShieldsJAShieldsPlaque radiotherapy for juxtapapillary choroidal melanoma: visual acuity and survival outcome.Arch Ophthalmol.1996;114:1357-1365.FBacinFKwiatkowskiHDalensResultats a long terme de la curietherapie par le cobalt 60 des melanomes de l'uvee.J Fr Ophtalmol.1998;21:333-344.PSummanenIImmonenTKivelaVisual outcome of eyes with malignant melanoma of the uvea after ruthenium plaque radiotherapy.Ophthalmic Surg Lasers.1995;26:449-460.GCBrownJAShieldsGSanbornRadiation retinopathy.Ophthalmology.1982;89:1494-1501.GCBrownJAShieldsGSanbornRadiation optic neuropathy.Ophthalmology.1982;89:1489-1493.KGündüzCLShieldsJAShieldsRadiation retinopathy following plaque radiotherapy of posterior uveal melanoma.Arch Ophthalmol.1999;117:609-614.ETLeeStatistical Methods for Survival Data Analysis.2nd ed. New York, NY: John Wiley & Sons Inc; 1992:258.EKaplanPMeierNonparametric estimation from incomplete observation.J Am Stat Assoc.1958;53:457-481.CLShieldsJAShieldsUKarlssonReasons for enucleation following plaque radiotherapy: clinical findings.Ophthalmology.1989;96:919-924.MTuncDCharSKrollUveal melanoma therapies: quality of life effects.Eyenet.October 1997:16.KJCruickshanksDGFrybackDMNondahlTreatment choice and quality of life in patients with choroidal melanoma.Arch Ophthalmol.1999;117:461-467.PTFingerMicrowave thermoradiotherapy for uveal melanoma: results of a 10-year study.Ophthalmology.1997;104:1794-1803.CLShieldsJAShieldsJCaterTranspupillary thermotherapy for choroidal melanoma: tumor control and visual outcome in 100 consecutive cases.Ophthalmology.1998;105:581-590.Accepted for publication February 11, 2000.This study was supported by the Macula Foundation, New York, NY (Dr C. L. Shields); the Paul Kayser International Award of Merit in Retina Research, Houston, Tex (Dr J. A. Shields); The Lions Eye Foundation, Philadelphia, Pa (Dr J. A. Shields); and the Eye Tumor Research Foundation, Philadelphia (Dr C. L. Shields).Reprints: Carol L. Shields, MD, Oncology Service, Wills Eye Hospital, 900 Walnut St, Philadelphia, PA 19107.

Girkin, Christopher A.; Emdadi, Alireza; Sample, Pamela A.; Blumenthal, Eytan Z.; Lee, Alex C.; Zangwill, Linda M.; Weinreb, Robert N.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1231pmid: 10980768

ObjectiveTo compare progression in short-wavelength automated perimetry (SWAP) and white-on-white (standard) perimetry in eyes with progressive glaucomatous changes of the optic disc detected by serial stereophotographs.MethodsForty-seven glaucoma patients with at least 2 disc stereophotographs more than 2 years apart, along with standard perimetry and SWAP examinations within 6 months of each disc photo of the same eye, were included in the study. The mean follow-up time was 4.1 years (range, 2.0-8.9 years). Baseline and follow-up stereophotographs were then graded and compared for the presence of progression. Progression in standard perimetry and SWAP, using the Advanced Glaucoma Intervention Study scoring system and a clinical scoring system, was compared between eyes with progressive change on stereophotographs and those without.ResultsTwenty-two of 47 eyes showed progressive change by stereophotographs. There was a statistically significant difference in the mean change in Advanced Glaucoma Intervention Study scores for both standard perimetry (P<.004) and SWAP (P<.001) between the progressed and nonprogressed groups. The sensitivity, specificity, and area under the receiver operator characteristic curve were higher using SWAP than standard perimetry when evaluated by either algorithm. This was statistically significant only in the area under the receiver operator characteristic curve for the Advanced Glaucoma Intervention Study scoring system (P= .04).ConclusionsShort-wavelength automated perimetry identified more patients than standard perimetry as having progressive glaucomatous changes of the optic disc. Compared with standard perimetry, SWAP may improve the detection of progressive glaucoma.CURRENT METHODS for detection of progressive glaucomatous optic nerve damage rely on clinical examination, including serial stereoscopic examinations of the optic disc, as well as serial evaluations using standard achromatic visual fields. Media opacities and small pupils limit both direct stereoscopic disc evaluation and stereophotographs. Additionally, evaluation of the optic disc is subjective, with significant interexaminer variability.Assessment of visual fields to detect progression is limited by several factors. These include poor sensitivity,intertest variability,patient experience, testing fatigue,media opacities,pupil size,test set-up, and the subject's level of attention.Because of the limitations of white-on-white (standard) perimetry in the diagnosis of early glaucoma, there is considerable interest in developing more sensitive measures of visual function. One such technique is short-wavelength automated perimetry (SWAP), which selectively isolates the S-cone responses in the central visual field.The short-wavelength–sensitive pathways mediate the S-cone signal via the bistratified subpopulation of ganglion cells.Short-wavelength automated perimetry uses a 2-color increment threshold procedure that presents a blue stimulus, which preferentially stimulates the short-wavelength–sensitive pathway, against a yellow background. This background saturates the rods and suppresses the sensitivity of the long- and medium-wavelength pathways.Short-wavelength automated perimetry has been shown to be more sensitive than standard perimetry in detecting early glaucomatous optic nerve damage.The selective nature of the SWAP stimuli may increase its ability to detect progression, even in advanced disease.Additionally, visual field defects detected using SWAP were found to predict defects eventually detected with standard perimetry.The purpose of the current study was to compare SWAP with standard perimetry for detection of progressive glaucomatous changes in the optic disc.PATIENTS AND METHODSPATIENTSWe reviewed 382 records of patients diagnosed as having an optic disc appearance consistent with glaucoma, based on masked grading of optic disc stereophotographs (cup-disc ratio >0.7, cup-disc ratio asymmetry ≥0.2, rim thinning, excavation, nerve fiber layer defects, or disc hemorrhages) from a longitudinal study of glaucoma patients at the Glaucoma Center of the University of California–San Diego (La Jolla). Forty-seven eyes of 47 patients met the following inclusion/exclusion criterion: Each patient had at least 2 stereophotographs 2 or more years apart, and at least 1 reliable standard and 1 reliable SWAP examination within 6 months of each disc photograph in the same eye. The first stereophotograph that showed signs of progression that fulfilled the inclusion criterion was used. If there was no progression, the latest photograph from our longitudinal database was used. Additionally, patients with high refractive error (defined as > ± 5.00 spherical equivalent or ± 3.00 cylinder), lens changes (defined as loss of >1 line of visual acuity with a nuclear sclerotic cataract, or the development of any degree of posterior subcapsular cataract), or who underwent cataract extraction with or without trabeculectomy during the follow-up period were excluded.VISUAL FIELDSStandard perimetry was performed using the 24-2 full threshold algorithm using the commercially available Humphrey Field Analyzer (Humphrey-Zeiss, San Leandro, Calif). The Humphrey Field Analyzer was also used to perform SWAP examinations, in which a 440-nm narrowband size V blue stimulus is presented against a broadband 500- to 700-nm yellow background for 200 milliseconds to maximize spatial and temporal summation, further enhancing isolation of the short-wavelength–sensitive pathway.All patients had at least 1 additional visual field for both SWAP and standard perimetry, prior to the baseline field and after the follow-up field, that confirmed the degree of visual field defect using the Ocular Hypertension Study visual field criteria for an abnormal field.Baseline and follow-up standard perimetry and SWAP visual fields closest to the time of the stereophotographs, were assessed using a clinical scoring system (CSS) and the Advanced Glaucoma Intervention Study (AGIS) scoring system. Using the CSS, visual field progression was based on the presence of at least 1 of the following criteria: (1) development of a new scotoma, defined as 2 adjacent points in a previously normal area at the .01 probability level on the pattern deviation plot, or 1 point within the central 10° that decline by 10 dB or more; (2) expansion of an existing scotoma, defined as 2 contiguous points adjacent to the scotoma that decline by 10 dB or more; or (3) deepening of an existing scotoma, defined as 2 points in the scotoma that decline by 10 dB or more. Using AGIS, progression was defined as an increase in the AGIS score by greater than or equal to 4 points.Changes in standard perimetry and SWAP, using both AGIS and the CSS, were compared with progression detected by serial stereophotographs.STEREOPHOTOGRAPHSSimultaneous baseline and follow-up stereophotographs of each patient were graded for the presence of progression in a masked fashion by 2 experienced stereophotograph graders (C.A.G., E.Z.B.). A third experienced stereophotograph grader (R.N.W.) resolved all cases of disagreement. All graders were glaucoma specialists. Patients were divided into 2 groups: progressive and nonprogressive. Glaucomatous progression by stereophotographs was defined as either a decrease in the neuroretinal rim due to focal notching, undermining, or diffuse rim thinning, or the expansion of a preexisting nerve fiber layer defect or the development of a new nerve fiber layer defect. Changes in the neuroretinal rim and the parapapillary nerve fiber layer were determined by direct visual comparison of each quadrant of the disc. Measurements of disc area were obtained for each patient using a confocal scanning laser ophthalmoscope (Heidelberg Retinal Tomograph, Heidelberg Engineering, Heidelberg, Germany).STATISTICAL ANALYSESStatistical analyses were completed using JMP software (SAS Institute, Cary, NC). Differences between the progressed and nonprogressed groups were compared using a ttest or χ2test. Comparisons of the mean change in AGIS scores using SWAP and standard perimetry in the progressed group were evaluated using a paired ttest. The sensitivity, specificity, and the area under the receiver operating characteristic (ROC) curve were determined for each technique. The area under the ROC curve reflects the ability of a test to provide a result that corresponds with the dependant variable. A value of 1 is a perfect correlation, whereas a value of 0.5 indicates no correlation. A McNemar χ2test was used to compare the sensitivity and specificity found using each technique. To compare standard perimetry and SWAP with respect to sensitivity, patients who were deemed to have progressed (by stereophotographs) were tabulated according to whether standard perimetry and/or SWAP showed progression or stable disease, resulting in a 2 × 2 table. The McNemar χ2is calculated from this table in the usual manner. A similar table was prepared for patients deemed to have stable disease (by stereophotographs) and this was used to test the differences in specificity. The areas under the ROC curve for each technique were compared using the method of Delong et al,which uses the fraction of concordant pairs, with each tied pair counted as half. In a dichotomous classifying variable, the area under the ROC curve turns out to be equal to half the value of the sum of sensitivity and specificity. A Pvalue of less than .05 was considered statistically significant.RESULTSThere was agreement between the initial 2 stereophotograph graders on identification of progression of glaucomatous optic disc damage by stereophotograph for 36 (76%) of 47 subjects, with 15 (32%) graded as progressed, and 21 (45%)of 47 graded as nonprogressed. The 2 graders disagreed on the remaining 11 subjects (23%). Seven of these 11 subjects were considered progressed by the third grader, who resolved the cases of disagreement. Thus, a total of 22 patients (47%) were considered progressed by serial optic nerve stereophotographs, with 25 patients (53%) considered nonprogressed.The ocular characteristics of both the progressed and nonprogressed groups are outlined in Table 1. The 2 groups did not differ significantly with respect to disc area, refraction, or the change in pupil size between the baseline and follow-up standard perimetry. The mean highest documented intraocular pressure in the ophthalmic record was 5.4 mm Hg higher in the progressed group than the nonprogressed group (P<.04). The disc area was smaller, and the baseline mean deviation and corrected pattern SD for standard perimetry and vertical cup-disc ratio were worse in the progressed group than in the nonprogressed group, but these differences were not statistically significant.Table 1. Ocular Characteristics*Nonprogressed Patients (n = 25)Progressed Patients (n = 22)PMean change in pupil size, mm−0.014 (1.25)−0.045 (0.860).95Mean sphere, D−1.61 (2.15)−1.85 (3.79).79Mean cylinder, D+1.17 (1.39)+1.26 (0.89).90Mean peak IOP, mm Hg24.0 (5.8)29.4 (11.3).04Mean disc area, mm22.56 (1.97)2.02 (0.51).22Baseline standard, mean deviation, dB−2.96 (3.65)−4.24 (4.92).31Baseline standard, corrected pattern SD, dB3.77 (4.38)4.53 (4.21).55Baseline vertical cup-disc ratio0.69 (0.191)0.75 (0.131).15*Data are expressed as mean (SD) unless otherwise indicated. D indicates diopters; I0P, intraocular pressure.The baseline AGIS score for SWAP was slightly higher (mean, 2.5; SD, 2.8) than the baseline AGIS score for standard perimetry (mean, 2.3; SD, 2.7). However, this difference was not statistically significant (P= .81).For the baseline fields, 5 patients had SWAP and standard perimetry prior to stereophotographs (average of 2.5 months for SWAP and 2.6 months for standard perimetry), and 42 patients had SWAP and standard perimetry after the stereophotographs (average of 1.6 months for SWAP and 1.3 months for standard perimetry). For the follow-up fields, 5 patients had SWAP prior to stereophotographs (average of 2.6 months), and 42 had SWAP after (average of 1.2 months). Twenty-six patients had standard perimetry prior to stereophotographs (average of 0.9 months) and 21 had it after (average of 1.4 months).The patient characteristics of the progressed and nonprogressed groups are summarized in Table 2. There was no significant difference in mean age, sex, or race. Both groups were predominantly white. The average length of follow-up was 4.2 years (range, 2.0-8.9 years) for the nonprogressed group and 4.0 years (range, 2.0-8.9 years) for the progressed group. This difference was not significant.Table 2. Patient Characteristics*Nonprogressed Patients (n = 25)Progressed Patients (n = 22)PAge, y64.3 (14.5)66.9 (11.4).49SexMale9 (36)12 (55).20Female16 (64)10 (45)RaceWhite21 (84)17 (77)Black1 (4)3 (13).66Asian1 (4)1 (5)Hispanic2 (8)1 (5)Mean follow-up period (range), y4.2 (2.0-8.5)4.0 (2.0-8.9).70*Data are expressed as mean (SD) unless otherwise indicated.Among eyes showing progression based on assessment of stereophotographs, the number of eyes progressed by each visual field test using both AGIS and CSS criteria is shown in Figure 1. The distribution of patients who demonstrated stable visual fields by each visual field test and the evaluation by each algorithm among eyes stable by photographic assessment is shown in Figure 2.Figure 1.Number of patients with progressive optic disc cupping found to have progressed using visual fields. SWAP indicates short-wavelength automated perimetry; CSS, clinical scoring system; and AGIS, Advanced Glaucoma Intervention Study Criteria.Figure 2.Number of patients with stable optic discs found stable using visual fields. SWAP indicates short-wavelength automated perimetry; CSS, clinical scoring system; and AGIS, Advanced Glaucoma Intervention Study Criteria.In the progressed group, standard perimetry showed progression in 7 (32%) of 22 patients while SWAP showed progression in 12 (55%) of 22 patients using AGIS criteria for visual field progression. The 2 types of visual fields agreed on the identification of progression for 6 patients. In the progressed group by photographic criteria, the mean difference in AGIS score was 4.53 (SD, 4.21) for SWAP and 3.62 (SD, 5.14) for standard perimetry, which was not statistically different (P<.10). In the nonprogressed group by photographic criteria, neither standard perimetry nor SWAP detected progressive field changes by AGIS criteria. In this group, the mean difference in AGIS score was 0.48 (SD, 0.96) for standard perimetry and 0.64 (SD, 0.75) for SWAP, which was not statistically different (P<.44). There was a statistically significant difference in the mean difference of AGIS scores for both standard perimetry (P<.004) and SWAP (P<.001) between the progressed and nonprogressed groups.Using the CSS for visual field progression, standard perimetry progressed in 13 (59%) of 22 patients, while SWAP progressed in 16 (73%) of 22 patients in the progressed group by photographic criteria. Twelve patients were identified as progressed. In the nonprogressed group by photographic criteria, standard perimetry was stable in 22 (88%) of 25 patients, while SWAP was stable in 23 (92%) of 25 patients, with 20 patients identified as stable. Among patients identified as progressed by review of stereophotographs, SWAP identified more patients as progressed than did standard perimetry with each algorithm. The sensitivity and specificity for progression and the area under the ROC curve for each visual field technique evaluated by each algorithm are shown in Table 3. The sensitivity and specificity were judged for both visual field techniques compared with disc progression as demonstrated by stereophotographs. While the sensitivity, specificity, and the area under the ROC curve for detecting progression was higher with SWAP than with standard perimetry, this difference was significant only for the AGIS scoring system.Table 3. Sensitivity and Specificity of Serial SWAP and Standard Perimetry Fields in the Detection of Stereophotograph-Defined Progression*CSSPAGIS Scoring SystemPSWAPStandard PerimetrySWAPStandard PerimetrySensitivity72.759.1.3854.533.8.13Specificity92.088.01.001001001.00Area under ROC curve0.8410.735.190.7730.659.04*Data are presented as percentages unless otherwise indicated. SWAP indicates short-wavelength automated perimetry; CSS, clinical scoring system; AGIS, Advanced Glaucoma Intervention Study; and ROC, receiver operator characteristic.COMMENTOur study compared baseline and follow-up standard perimetry and SWAP in glaucoma patients who had stable baseline and follow-up optic nerve stereophotographs with those patients with progressive glaucomatous change documented with stereophotography, using 2 different scoring algorithms for visual field progression.The rate of progression defined by stereophotographs in our study was 46.8%. The mean duration of follow-up was 4.1 years (range, 2.0-8.9), yielding a rate of progression of 11.4% per year. Previously published reports have estimated a 7% per year rate of progression based on serial stereophotographs.Ours was not a population study but a study of higher-risk individuals, so meaningful information regarding rates of progression cannot be generalized to a more diverse population.Using both the AGIS and CSS, SWAP showed a higher sensitivity for progressive optic disc changes than standard perimetry, without loss of specificity, and thus a larger area under the ROC curve. However, this difference was only statistically significant using the AGIS scoring system.Evaluation using the CSS with SWAP showed the highest sensitivity at 72.7%, with a specificity of 92% and the largest area under the ROC curve (0.841) (Table 3). The relatively low sensitivity of both types of visual fields to stereophotograph-defined changes in the optic disc is consistent with previous reports of structural changes occurring without demonstrable functional defects.This supports the hypothesis that large numbers of ganglion cell populations may lose function before detection of progression is possible by standard perimetry.However, it is important to emphasize that there is no "gold standard" in evaluating progression in glaucoma and that evaluation of optic disc stereophotographs to assess progression is subject to interexaminer interpretation.While both standard perimetry and SWAP showed statistically significant differences in the mean differences in AGIS scores between the progressed and nonprogressed groups, the mean change in AGIS scores was higher in SWAP than in standard perimetry. However, this result did not achieve statistical significance (P<.11).Using the AGIS scoring system, neither standard perimetry nor SWAP detected progression in the nonprogressed group based on photographic assessment. With the CSS, SWAP detected progression in 2 patients while standard perimetry detected progression in 3 in the nonprogressed group. Thus, while the CSS showed a higher sensitivity for the detection of structural changes, it was less specific than the AGIS scoring system. This probably reflects a more lax cut-off criterion for progression in the CSS compared with the AGIS scoring system. However, it is also possible that the lower specificity of the CSS is due to changes in visual function that may have occurred prior to observable changes in optic disc structure.The redundancy inherent in the visual system due to multiple subpopulations of ganglion cells with overlapping receptive fields may provide one explanation for the lack of sensitivity of standard perimetry to detect early axonal loss in glaucoma. Different ganglion cell types respond to specific, yet overlapping, components of the visual image and project this information along distinct interconnected visual channels, at least at the level of the lateral geniculate nucleus.Standard perimetry stimulates several of these visual channels simultaneously. By developing selective stimuli to target specific ganglion cell populations, the responses of an isolated visual channel may be maximized and consequently may provide a more sensitive test to detect axonal loss in glaucoma.Short-wavelength automated perimetry is used in clinical practice to detect early glaucomatous damage. Both standard perimetry and SWAP have been shown to correlate with structural changes of the optic disc.One study demonstrated a higher correlation between structural changes demonstrated by scanning laser ophthalmoscopy in early glaucoma and the mean deviation of SWAP than for the mean deviation of standard perimetry.Additionally, defects detected using SWAP tended to be deeper and more extensive than defects found with standard perimetry. Two 5-year prospective studies performed independently have each found that functional deficits demonstrated with SWAP in some patients with ocular hypertension and glaucoma are detectable several years earlier than with standard perimetry and, thus, may predict the development of functional loss with standard perimetry.This increased sensitivity might provide a better marker for progression and thus show a higher correlation with structural changes in the optic nerve over time. On the other hand, increased variability,greater sensitivity to cataractous changes of the crystalline lens,and loss of the smaller subpopulation of bistratified ganglion cells in extremely advanced disease might limit the usefulness of SWAP in the detection of progressive glaucomatous damage.These previous studies have compared progressive scotomas using SWAP with those using standard perimetry. To evaluate the ability of visual function tests such as standard perimetry and SWAP to detect progression in glaucoma, these tests should be compared with structural changes of the optic disc.In summary, this study has demonstrated that serial SWAP corresponded better than serial standard perimetry with glaucomatous changes of the optic disc, yielding a higher sensitivity and similar specificity. Using the AGIS scoring system, there was a significant difference in the area under the ROC curve between standard perimetry and SWAP. This evidence seems to indicate that serial SWAP may improve the detection of glaucomatous progression. However, a larger study population and evaluation of new progression algorithms are required to determine if there is a clinically significant difference between standard perimetry and SWAP using all grading algorithms.PRLichterVariability of expert observers in evaluating the optic disc.Trans Am Ophthalmol Soc.1977;74:532-572.LZangwillSShakibaJCaprioliRNWeinrebAgreement between clinicians and a confocal scanning laser ophthalmoscope in estimating cup/disk ratios.Am J Ophthalmol.1995;119:415-421.RNWeinrebAssessment of optic disc topography for diagnosing and monitoring glaucoma [editorial].Arch Ophthalmol.1998;116:1229-1231.TGZeyenJCaprioliProgression of disc and field damage in early glaucoma.Arch Ophthalmol.1993;111:62-65.EBWernerBPetrigTKrupinKIBishopVariability of automated visual fields in clinically stable glaucoma patients.Invest Ophthalmol Vis Sci.1989;30:1083-1089.CAJohnsonJMNelson-QuiggA prospective three-year study of response properties of normal subjects and patients during automated perimetry.Ophthalmology.1993;100:269-274.EBWernerTKrupinAAdelsonMEFeitlEffect of patient experience on the results of automated perimetry in glaucoma suspect patients.Ophthalmology.1990;97:44-48.UGuthauserJFlammerQuantifying visual field damage caused by cataract.Am J Ophthalmol.1988;106:480-484.KALindenmuthGLSkutaRRabbaniDCMuschEffects of pupillary constriction on automated perimetry in normal eyes.Ophthalmology.1989;96:1298-1301.Not AvailableAdvanced Glaucoma Intervention Study 2: visual field test scoring and reliabilityOphthalmology.1994;101:1445-1455.PASampleRNWeinrebColor perimetry for assessment of primary open-angle glaucoma.Invest Ophthalmol Vis Sci.1990;31:1869-1875.PRMartinAJWhiteAKGoodchildHDWilderAESeftonEvidence that blue-on cells are part of the third geniculocortical pathway in primates.Eur J Neurosci.1997;9:1536-1541.PASampleRNWeinrebProgressive color visual field loss in glaucoma.Invest Ophthalmol Vis Sci.1992;33:2068-2071.CAJohnsonAJAdamsEJCassonJDBrandtBlue-on-yellow perimetry can predict the development of glaucomatous visual field loss.Arch Ophthalmol.1993;111:645-650.PASampleRNWeinrebStandard achromatic perimetry vs. short-wavelength automated perimetry for following glaucoma.In: GK K, ed. Glaucoma Update V. Germany: Springer Verlag; 1995:197-204.CAJohnsonAJAdamsEJCassonJDBrandtProgression of early glaucomatous visual field loss as detected by blue-on-yellow and standard white-on-white automated perimetry.Arch Ophthalmol.1993;111:651-656.MKassThe Ocular Hypertension Treatment Study.J Glaucoma.1994;3:97-100.Not AvailableThe Advanced Glaucoma Intervention Study (AGIS) 1: study design and methods and baseline characteristics of study patients.Control Clin Trials.1994;15:299-325.ERDeLongDMDeLongDLClarke-PearsonComparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach.Biometrics.1988;44:837-845.PJAiraksinenATuulonenHIAlankoRate and pattern of neuroretinal rim area decrease in ocular hypertension and glaucoma.Arch Ophthalmol.1992;110:206-210.BEPrum JrEvaluation of progressive structural damage in glaucoma.Chibret Int J Ophthalmol.1994;60-67.ASommerJKatzHAQuigleyClinically detectable nerve fiber atrophy precedes the onset of glaucomatous field loss.Arch Ophthalmol.1991;109:77-83.ASommerIPollackAEMaumeneeOptic disc parameters and onset of glaucomatous field loss, I: methods and progressive changes in disc morphology.Arch Ophthalmol.1979;97:1444-1448.HAQuigleyEMAddicksWRGreenOptic nerve damage in human glaucoma, III: quantitative correlation of nerve fiber loss and visual field defect in glaucoma, ischemic neuropathy, papilledema, and toxic neuropathy.Arch Ophthalmol.1982;100:135-146.ALColemanASommerCEngerInterobserver and intraobserver variability in the detection of glaucomatous progression of the optic disc.J Glaucoma.1996;5:384-389.PHSchillerNKLogothetisThe color-opponent and broad-band channels of the primate visual system.Trends Neurosci.1990;13:392-398.PASampleCFBosworthRNWeinrebShort-wavelength automated perimetry and motion automated perimetry in patients with glaucoma.Arch Ophthalmol.1997;115:1129-1133.PTeesaluKVihanninjokiPJAiraksinenATuulonenHemifield association between blue-on-yellow visual field and optic nerve head topographic measurements.Graefes Arch Clin Exp Ophthalmol.1998;236:339-345.NYamagishiAAntonPASampleMapping structural damage of the optic disk to visual field defect in glaucoma.Am J Ophthalmol.1997;123:667-676.RNWeinrebSShakibaPASampleAssociation between quantitative nerve fiber layer measurement and visual field loss in glaucoma.Am J Ophthalmol.1995;120:732-738.RBathijaLZangwillCCBerryPASampleRNWeinrebDetection of early glaucomatous structural damage with confocal scanning laser tomography.J Glaucoma.1998;7:121-127.AAntonNYamagishiLZangwillPASampleRNWeinrebMapping structural to functional damage in glaucoma with standard automated perimetry and confocal scanning laser ophthalmoscopy.Am J Ophthalmol.1998;125:436-446.PTeesaluKVihanninjokiPJAiraksinenATuulonenELaaraCorrelation of blue-on-yellow visual fields with scanning confocal laser optic disc measurements.Invest Ophthalmol Vis Sci.1997;38:2452-2459.PASampleJDTaylorGAMartinezMLuskyRNWeinrebShort-wavelength color visual fields in glaucoma suspects at risk.Am J Ophthalmol.1993;115:225-233.JMWildRPCubbidgeIEPaceyRRobinsonStatistical aspects of the normal visual field in short-wavelength automated perimetry.Invest Ophthalmol Vis Sci.1998;39:54-63.PASampleCAJohnsonGHaegerstrom-PortnoyAJAdamsOptimum parameters for short-wavelength automated perimetry.J Glaucoma.1996;5:375-383.Accepted for publication February 11, 2000.This study was supported in part by grant NEI EY08208 from the National Eye Institute, National Institutes of Health, Bethesda, Md (Dr Sample), the Glaucoma Research Foundation, San Francisco, Calif (Dr Sample), grant EY11008 from the National Eye Institute, National Institutes of Health, Bethesda, Md (Dr Zangwill), the Heed Ophthalmic Foundation, Chicago, Ill (Dr Girkin), and the Joseph Drown Foundation, Los Angeles, Calif (Dr Weinreb).Presented in part at the annual meeting of the Association of Research and Vision and Ophthalmology, Ft Lauderdale, Fla, May 1999.Corresponding author: Robert N. Weinreb, MD, Glaucoma Center, University of California–San Diego, 9500 Gilman Dr, La Jolla, CA 92093-0946 (e-mail: [email protected]).

Jacobson, Daniel M.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1237pmid: 10980769

Nork, T. Michael; Poulsen, Gretchen L.; Nickells, Robert W.; Hoeve, James N. Ver; Cho, Nam-Chun; Levin, Leonard A.; Lucarelli, Mark J.

2000 JAMA Ophthalmology

doi: 10.1001/archopht.118.9.1242pmid: 10980770

ObjectiveTo test a hypothesis of photoreceptor involvement in retinal ganglion cell (RGC) death in chronic glaucoma.MethodsLaser spots were applied to 6 eyes of 3 rhesus monkeys, causing focal destruction of the outer retina, including the photoreceptors. After 3 to 4 weeks, experimental glaucoma was induced in the right eyes of each monkey using argon laser trabecular destruction (ALTD). The intraocular pressures in these eyes were elevated for 3 to 7 months. As a control, 1 additional monkey underwent retinal laser photocoagulation followed by optic nerve transection instead of ALTD. Following enucleation, the retinas were embedded and sectioned for histologic evaluation.ResultsThere was extensive loss of RGCs in the eyes with ALTD except over the large retinal laser spots, where there was an increased survival of RGCs. The RGC protection was not observed in the monkey that had undergone optic nerve transection.ConclusionPhotocoagulation of the outer retina that completely destroys the photoreceptors results in survival of the overlying RGCs in experimental glaucoma in monkey eyes.Clinical RelevanceAlthough this is an experimental model and not a therapeutic option, these results suggest that treatments other than lowering intraocular pressure may be potential therapies for preventing RGC death in glaucomatous eyes.CHRONIC GLAUCOMA is one of the leading causes of blindnessdespite advances in pharmacologic and surgical treatment to lower intraocular pressure (IOP). Elevated IOP is traditionally thought to injure the retinal ganglion cell (RGC) axons as they pass through the optic nerve head, either by mechanical deformation of the lamina cribrosaor by reducing the blood supply to the nerve.The RGCs are then presumed to die by retrograde degeneration owing to interruption of axonal transport that brings neurotrophins from the brain to the eye.However, this hypothesis of RGC death in glaucoma, referred to in this article as the retrograde hypothesis, does not explain several observations.For example, it has long been known that glaucoma results in a blue-yellow color confusion—a type of color vision deficit that is more characteristic of outer retinal disease than optic nerve injury.Findings from some electrophysiological studies have suggested photoreceptor involvement,including a recent finding of increased latency of the early multifocal electroretinographic response in an arcuate distribution in human glaucoma.Furthermore, Yancey and Linsenmeierfound decreased choroidal PO2and changes seen in the C wave and standing potential of the full-field electroretinogram (which they ascribed to outer retinal ischemia) with even moderate short-term elevations of IOP in cat eyes.Photoreceptor involvement in glaucoma has also been observed at the histopathologic and molecular levels. In humans with glaucoma and in experimental glaucoma in monkeys, we found that the photoreceptors, especially the red and green cones, were swollen.Tezel et alreported that heat shock protein HSP 60 is elevated in the photoreceptors of human eyes with glaucomatous damage. Preliminary work in our laboratory also suggests that the messenger RNA for red- and green-cone opsin is selectively reduced in humans with glaucoma as well as in experimental glaucoma in monkeys.Photoreceptor injury in glaucoma might be an epiphenomenon or a response to dying ganglion cells. Alternatively, it could be that the photoreceptors play an important role in either causing or exacerbating ganglion cell injury in this disease. We will refer to this latter possibility as the anterograde hypothesisof RGC death in glaucomatous eyes.Retinal ganglion cells do not depend on anterograde neurotrophic stimuli, as evidenced by their survival in such conditions as retinal laser photocoagulation,retinitis pigmentosa,and the Royal College of Surgeons rat.Therefore, the anterograde hypothesis requires that glaucoma cause some specific malfunction of viable photoreceptors that in turn destroys the RGCs. If the hypothesis is correct, focal ablation of photoreceptors before induction of experimental glaucoma should protect the overlying RGCs, an effect that would not be predicted by the retrograde hypothesis (Figure 1).Figure 1.Design of experiment to discriminate between the retrograde and anterograde hypotheses of ganglion cell death in chronic glaucoma. First, a retinal laser is applied (light gray), which is absorbed by the pigment epithelium and converted to heat (dark gray) that kills the photoreceptors (upper right). Experimental glaucoma is then produced by argon laser trabecular destruction. The 2 hypotheses predict different outcomes with respect to survival of the subset of ganglion cells overlying the healed laser spot (lower half). NF indicates nerve fiber layer; GC, ganglion cell layer; IN, inner nuclear layer; ON, outer nuclear layer; R/C, rod/cone layer; BM, basement membrane; and Ch, choroid.We present the results of such an experiment in which rhesus monkey photoreceptors were focally destroyed by retinal laser photocoagulation prior to the induction of experimental glaucoma. This treatment provided a protective effect to the RGCs. Although this result is consistent with the anterograde hypothesis, it is also possible that the laser treatment causes nonspecific effects that promote RGC survival under these experimental conditions. Possible mechanisms for this protective effect are discussed.MATERIALS AND METHODSRETINAL LASER PHOTOCOAGULATION FOLLOWED BY INDUCTION OF EXPERIMENTAL GLAUCOMAAll animal procedures adhered to the Association for Research in Vision and Ophthalmology, Rockville, Md, statement on the use of animals in vision research. Eight eyes of 4 rhesus monkeys underwent retinal laser photocoagulation. The laser spots, using the argon green wavelength (514.5 nm) (Coherent, Santa Clara, Calif), were applied to the region of the posterior pole that corresponds to early visual field loss in humans with glaucoma, ie, 5 to 20 arc degrees eccentric to fixation. Various spot sizes were created, from 100 to 4000 µm (spots more than 1000 µm were made by applying confluent 200-µm individual burns) with energies from 50 to 130 mW and exposure times from 0.1 to 0.5 seconds.After about 1 month (enough time to permit the debris of photocoagulation to be cleared by the retina), argon laser trabecular destruction (ALTD) was performed in 1 eye of 3 animals.Briefly, a Kaufman-Wallow single-mirror monkey gonio lens (Ocular Instruments Inc, Bellevue, Wash)was used to deliver the laser to the trabecular meshwork. During each treatment session, either 270° or 360° of the anterior (nonpigmented) meshwork was photocoagulated with 514.5 nm of argon green laser using either 50- or 100-µm spots of 1.0-W intensity and a 0.5-seconds' duration. Seventy-five to 200 such spots were applied per session. The IOP, which was measured twice weekly with a handheld digital tonometer (Tono-Pen XL; Mentor O & O, Nowell, Mass), usually took 3 weeks to rise, and the elevation lasted variable lengths of time. (This device was found to underestimate the IOP at higher pressures in the cynomolgus monkey.A similar study has not been done for the rhesus, although its eye being intermediate in size between the human and the cynomolgus monkey, it might be expected to have a smaller error than that found in the cynomolgus monkey.) Additional laser treatments were carried out as needed to maintain elevated pressures. Two sessions of ALTD were administered to monkeys No. 1 and No. 3, and 4 sessions were used for monkey No. 2. In 1 animal (monkey No. 1), 1 drop of 0.5% timolol maleate (Timoptic; Merck & Co, Whitehouse Station, NJ) was applied to the cornea of the treated eye on 2 occasions to lower the IOP.OPTIC NERVE TRANSECTIONTo control for the possibility that retinal laser photocoagulation may promote RGC survival by mechanisms other than photoreceptor removal, the optic nerve was transected (instead of inducing experimental glaucoma) in 1 eye of 1 animal. As with the other monkeys, the fellow eye underwent retinal laser photocoagulation only. The optic nerve transection (ONT) was accomplished using the technique described by Gonnering et al.Briefly, a lateral orbitotomy was first performed. The optic nerve was then transected 6 to 8 mm behind the globe posterior to the entry of the retinal artery and vein. On recovery (approximately 1 week), a fluorescein angiogram with the animal receiving ketamine anesthesia was obtained to demonstrate patency of the retinal artery.HISTOLOGIC PREPARATIONThe animals were euthanized and the eyes were removed within 10 minutes and promptly chilled in 0.1-mol/L sodium phosphate buffer (pH 7.4) at 4°C. Within 5 minutes, the eyes were removed from the buffer and cut coronally just anterior to the equator. The posterior pole was fixed in 4% paraformaldehyde at 4°C for 24 hours and then stored in 0.1-mol/L sodium phosphate buffer (pH 7.4) at 4°C. Segments of retina containing individual laser spots were removed and embedded in glycol methacrylate, sectioned at a thickness of 1.9 µm, and stained with thionin.TWO-DIMENSIONAL RECONSTRUCTIONSSerial sections were cut through some of the retinal laser spots. For each histologic section, the nucleoli of cells in the ganglion cell layer that had the anatomical characteristics of RGCs (ie, large round nuclei, prominent nucleoli, abundant cytoplasm, and metachromatic thionin staining consistent with Nissl substance) were identified. The position of each nucleolus of cells meeting these criteria was measured with respect to a given point (either the mid point of the smaller laser burns or the edge of the large, confluent burns) using a micrometer in the ocular of the microscope. Two-dimensional (D) reconstructions of the laser spots were then plotted.RESULTSRETINAL LASER PHOTOCOAGULATION FOLLOWED BY EXPERIMENTAL GLAUCOMAInitially, the laser spots had a white appearance. After 3 weeks, they became darker with mottled pigmentation. The IOPs were elevated to variable degrees in the treated eyes (Table 1).Intraocular Pressures*MonkeyTreatmentDuration, wk†Mean IOP ± SD, mm HgIOP Range, mm Hg‡1ALTD20.041 ± 1212-642ALTD22.045 ± 1516-693ALTD33.039 ± 1119-634ONT5.512 ± 47-18*IOP indicates intraocular pressure; ALTD, argon laser trabecular destruction; and ONT, optic nerve transection.†Duration of IOP elevation for monkeys Nos. 1-3 and the time between ONT and sacrifice for monkey No. 4.‡Ranges of pressures. The IOP means do not include the IOPs during the intervals between the ALTD applications and the initial pressure rises for monkeys No. 1-3.In the RGC layers of the 3 monkeys treated with ALTD, increased densities of cells were seen over the large laser spots (Figure 2) relative to the surrounding retina. (The effect was not seen over the small laser spots.) These cells were located in a layer contiguous with the adjacent RGC layer. The inner plexiform layer and in some cases the inner nuclear layer were intact. Few cells were present in the inner plexiform layer. The cells in the ganglion cell layer over the laser spots had large round nuclei, prominent nucleoli, abundant cytoplasm, and metachromatic thionin staining that was consistent with Nissl substance. Only a few cells in this layer had spindle-shaped nuclei or contained cytoplasmic pigment granules (characteristic of activated glial cells, retinal pigment epithelium [RPE] cells, or macrophages), which were readily distinguishable from the putative RGCs.Figure 2.A, Fundus photograph immediately following retinal laser photocoagulation. After about 1 month, argon laser trabecular destruction was performed. The eye was enucleated after 142 days of elevated pressure. The fovea is indicated by the plus sign (bar = 500 µm). B, Low magnification, and C, medium magnification of the large laser spot (asterisk) shown by the black arrow in part A. Ganglion cells are seen to survive primarily over the laser spot. The black arrows in part B indicate the border between viable and absent photoreceptors (toluidine blue; B, bar = 50 µm; C, bar = 20 µm). D, High magnification of ganglion cell layer over laser spot, and E, in control eye (toluidine blue; D and E, bar = 5 µm). Note similar anatomy of cells. Histologic section of the retinal laser spot indicated by the white arrow in part A is shown in the bottom half of Figure 5.To better illustrate the phenomenon, serial sections were cut 1.9 µm thick through some of the laser spots. The locations of the putative RGC nucleoli were then measured and plotted on 2-D graphs. Figure 3shows such a reconstruction comparing sections of retina from the same location of the temporal maculae in the left (control) and right (glaucomatous) eyes of monkey No. 1. Marked reduction in numbers of ganglion cells in the glaucomatous eye were seen everywhere except over the laser spot. Within the region where the photoreceptors had been ablated by previous retinal laser photocoagulation, the overlying putative RGCs are present in densities approaching that of the control eye (Figure 4).Figure 3.Two-dimensional reconstruction of the ganglion cell layer from serial histologic sections. Each section is oriented vertically and was cut 1.9 µm thick. The ordinate represents distance along a single section for any given position on the abscissa. The points represent ganglion cell nucleoli. Control shows an area of retina in the control left eye centered 2 mm temporal to the fovea along the horizontal midline. There is no retinal laser spot in this region. Glaucoma shows an area of the retina from the eye with experimental glaucoma corresponding to the large laser spot indicated by the black arrow in Figure 2, A. This portion of the retina has the same temporal eccentricity as the control. The thick black lines denote the edge of the laser spot in terms of the surviving photoreceptors. The ordinates are oriented in a superior (larger numbers)/inferior direction. The fovea is located to the left of the left plot and to the right of the right plot.Figure 4.Three-dimensional density plots of the ganglion cells shown in Figure 3. Note that the peak density over the laser spot in the eye with experimental glaucoma (red band) is similar to the average density in the control eye.At somewhat greater eccentricities than the spot shown in Figure 2, there are fewer RGCs normally present. Even so, the effect was still present, as shown in Figure 5. Unlike the more intense laser spot shown in Figure 2, the cells in the inner nuclear layer over these spots were present in apparently normal numbers; however, a downward stretching of this layer was seen. In addition, a few macrophages and/or activated RPE cells were present.Figure 5.Comparison of 2 retinal laser spots of similar intensity and eccentricity in monkey No. 1. The arrows indicate the border between viable and absent photoreceptors. A, Control eye. Spot is 3 mm superior to foveal center. One or 2 layers of nuclei are present in the ganglion cell layer. The density of cells in this layer over the ablated photoreceptors is similar to the surrounding, normal retina (toluidine blue, bar = 50 µm). B, Glaucomatous eye. Spot is 3 mm superotemporal to foveal center (indicated by white arrow in Figure 2, A). As in part A, 1 or 2 layers of nuclei are present in the ganglion cell layer over the ablated photoreceptors. However, the density of cells in this layer are greatly reduced in the surrounding, untreated retina.A similar effect was also found over the large, confluent laser spot in monkey No. 3 (Figure 6) where the average density of cells in the RGC layer over the lasered area in the glaucomatous eye was greater than the adjacent unlasered region (Figure 7). Unlike the more intense laser spot shown in Figure 2, the bipolar cells over these spots were present in normal numbers (not shown).Figure 6.Fundus photographs of both eyes for monkey No. 3 immediately following retinal laser photocoagulation. Note the confluent patches in the inferior maculae. A, Experimental glaucoma was subsequently produced in the right eye. The white rectangles represent the areas used for 2-dimensional reconstruction (Figure 7). B, Note that the sampled area in the control left eye is somewhat closer to the fovea than the treated right eye.Figure 7.Two-dimensional reconstruction of ganglion cells over the edge of 2 symmetrically placed, large, confluent laser spots in monkey No. 3. The positive numbers on the left axis denote distances over the burns. The thick black lines mark the edges of the burns—no photoreceptors are present over the burns (positive numbers). A distinct reduction in ganglion cell density is evident in the eye with glaucoma but not in the control eye. Larger distances along ordinates indicate decreasing foveal eccentricities.The mean densities as a function of foveal eccentricity for the sampled areas shown in Figure 7were plotted in Figure 8. The densities were greater everywhere in the control eye (note the different vertical scales). Since the specimens were sampled from regions near the foveas (about 1 mm inferotemporal to the foveal centers, where ganglion cell density changes rapidly with eccentricity), these greater apparent densities in the control eye could be the result of the samples not being taken from corresponding areas. However, a generalized reduction in ganglion cell density in the eye with glaucoma could not be ruled out. The transition of densities across the edge of the laser spot was a smooth one for the control, with a gradual increase in ganglion cell density with decreasing foveal distances. By contrast, the glaucomatous eye showed a marked reduction in densities from the lasered to the unlasered portions of the retina.Figure 8.Plots of average ganglion cell density as a function of foveal eccentricity. Positive numbers (to the right of the dotted lines) indicate ganglion cells overlying the laser spots (destroyed photoreceptors). The larger positive distances represent decreasing foveal eccentricity. Bars denote SEMs.One large spot (≥200-µm diameter of ablated photoreceptors) was quantitatively analyzed by measuring nucleolar positions and creating 2-D reconstructions for each of 3 eyes with glaucoma. All 3 of these spots showed greater densities of ganglion cells than the surrounding retinas. In addition, a fourth large spot in the glaucomatous eye of monkey No. 1 was examined that subjectively appeared to have a greater RGC density than the surrounding retina (Figure 5). Several small spots (≤100 µm) were visually inspected in all of the glaucomatous eyes. There was no qualitative increase in ganglion cells over these spots. Two-dimensional reconstructions were not made for the small spots.Three large spots were also examined by nucleolar measurement and 2-D reconstruction in nonglaucomatous eyes. No difference in ganglion cell density was apparent compared with the surrounding retina (Figure 7and Figure 9).Figure 9.Two-dimensional reconstruction of ganglion cells over laser spots (no photoreceptors are present between the thick black lines) in a control eye and one that had undergone optic nerve transection (ONT) approximately 1 month following the laser application. The 2 eyes are from monkey No. 4. Both spots are located about 2.5 mm superonasal to the foveal centers. The plots are oriented such that the larger negative distances on the ordinates are closest to the foveas. There is no apparent change in density of ganglion cells seen over the laser spots in either eye, although the putative retinal ganglion cell density in the eye with the ONT is only about 10% of that in the control eye. The bipolar cells are present in normal numbers over the laser spots (not shown). The linear strip of lower density just above the upper thick black line in the control eye corresponds to a retinal vein.RETINAL LASER PHOTOCOAGULATION FOLLOWED BY OPTIC NERVE TRANSECTIONMonkey No. 4 underwent retinal laser photocoagulation in both eyes followed 4 weeks later by ONT in the right eye. A fluorescein angiogram was obtained 4 weeks after the ONT. This showed normal retinal blood flow, indicating that the site of the ONT had been posterior to the insertion of the central retinal artery and vein.Five ½ weeks after the transection, the eyes were enucleated and processed for histologic evaluation as described earlier. Symmetrically located large laser spots were removed from the eyes, embedded, and serially sectioned. The locations of the RGC nucleoli were measured, and 2-D graphical reconstructions were made (Figure 9). In the control eye, the RGCs were neither increased nor decreased compared with the surrounding retina. The same was true for the eye that had undergone ONT, except that the average density of RGCs was greatly reduced in all areas.COMMENTTYPE OF CELLS OVER LASER SPOTSAlthough a few cells had features consistent with macrophages, activated RPE cells, or glial cells, most of the cells present over the retinal laser spots were probably RGCs. They were located at the level of and were continuous with the adjacent RGC layer. Few cells were seen in the inner plexiform layer, which would not have been the case had a great number of other cells (eg, glial or RPE) migrated in from elsewhere. Furthermore, a similar concentration of cells was not found either over the laser spots in the control eyes or over the retinal laser spots in the eye with ONT—again suggesting that most of the cells over the laser spots in the glaucomatous eyes were not other cell types that might have formed or congregated as part of the wound-healing process.Morphologically, the cells resembled RGCs in that they had large round nuclei, prominent nucleoli, abundant cytoplasm, and metachromatic thionin staining consistent with Nissl substance. Although anatomically similar in appearance, displaced amacrine cells are far too rare at these eccentricities to account for the density of cells found.Only a few cells in this layer had the spindle-shaped nuclei or contained cytoplasmic pigment granules characteristic of activated glial cells, RPE cells, or macrophages. Also, the distribution of these cells seen on 2-D reconstruction was similar to (albeit of greater density than) the surrounding RGC layer (see the large laser patch in Figure 7).OTHER POTENTIAL CAUSES OF ALTERED RGC DENSITY OVER LASER SPOTSWe cannot completely rule out some effect on the RGCs over the laser spots in the nonglaucomatous eyes. A few histologic sections of some of the spots appeared to have a subtle increase in RGC-like cells. However, when the positions of all nucleoli in all the cells with morphologic characteristics of RGCs were measured and plotted, no average increase in density was apparent (see controls in Figure 5, Figure 7, and Figure 9). The cause of this apparent increase in RGC concentration in some of the sections may have been owing to normal random fluctuation, which is also seen over the nonlasered areas, or to slight shrinkage of the tissue over the laser spots. Increased RGC density has not been noted in the numerous publications describing the histologic features of laser spots in the eyes of either humans or monkeys.In any event, the effect, if it exists, is small compared with the markedly higher RGC concentrations over the laser spots relative to the surrounding retinas seen in the glaucomatous eyes.POSSIBLE MECHANISMS FOR THE PROTECTIVE EFFECT OF RETINAL LASER PHOTOCOAGULATION IN GLAUCOMANonspecific Protective EffectAlthough, these results are predicted by the anterograde hypothesis, it is also possible that they could be owing to some as yet unknown nonspecific effect not necessarily related to photoreceptor injury per se, such as thinning of the retina, which would allow more oxygen to reach the RGC layer from the choroid; releasing of neurotrophic or neuroprotective factors, such as basic fibroblastic growth factor; or activating synthesis of heat shock proteins in the RGCs.Activated retinal pigment epithelial cells, macrophages, or Müller cells could serve as mediators for these products. However, no protective effect was seen in the case of retinal laser photocoagulation preceding ONT. Therefore, a nonspecific effect, if there is one, is not sufficient to permit ganglion cell survival under these circumstances.Owing to heat dissipation at the edges of the laser spots, the area of absent photoreceptors is smaller than the area of disturbed pigment in the RPE cells. This effect can be seen in Figure 2, where the diameter of whitening of the original spot is about 500 µm compared with a diameter of only about 250 µm of photoreceptor destruction (this disparity is notable even considering the approximately 15% tissue shrinkage on embedment in glycol methacrylate). Yet the extent of this protection seems to be limited precisely to the region of photoreceptor destruction. This, too, is consistent with the hypothesis that RGC death is mediated specifically by photoreceptor injury and not the result of laser damage to the RPE. Similarly, we believe that the small laser spots killed too few photoreceptors to produce a measurable protective effect.The presence or absence of bipolar cells in the laser spot does not seem to be critical in affecting protection of the RGCs. Furthermore, as noted previously, simply damaging the RPE cells without killing the photoreceptors is insufficient to protect the RGCs. Therefore, it may be that elimination of the photoreceptors is a critical aspect of the protective effect.Finally, nonspecific forms of RGC neuroprotection have been observed in a rat model of ONT. Yip et alreport a preliminary study showing that a short period of low-level laser applied to the retina enhances RGC survival after optic nerve axotomy, and Di Polo et alshowed that RGC death could be delayed by prolonged delivery of brain-derived neurotrophic factor. In both of these studies, the effects were transient, while our laser treatment provides a long-term (at least for the several months' duration of these experiments) rescue in primates with experimental glaucoma. The differences between these studies could be accounted for by several factors. First, axotomy may be too damaging for any nonspecific neuroprotective environment to overcome. This is evidenced in part by failure of the retinal laser photocoagulation to prevent RGC death in the axotomized primate eye. Alternatively, the loss of photoreceptors would provide a permanent break in the anterograde chain of events that leads to RGC death in glaucomatous eyes (see following section), and we would predict that the protective effect would be permanent. Both these models are consistent with our results and warrant further study.Interruption of the Transsynaptic Glutamate CascadeThe protective effect on RGCs described here could be specific to a mechanism of RGC death in glaucomatous eyes (ie, consistent with the anterograde hypothesis). One possible mechanism by which this could occur might be by interruption of the transsynaptic glutamate cascade between the photoreceptors and the RGCs. Elevated levels of glutamate have been found in the vitreous body of eyes of humans and monkeys with glaucoma.The source of this excess glutamate has not been demonstrated in glaucomatous eyes, although Yoles and Schwartzshowed that intraocular glutamate is increased in a rat model of optic nerve crush. However, considering that glutamate is the primary excitatory neurotransmitter of the photoreceptors and bipolar cells and that the outer retina shows signs of injury in glaucomatous eyes, this may be another possible source of the intravitreal glutamate. Although axonal injury can result in RGC death, these cells also contain N-methyl-D-aspartate glutamate receptorsand could possibly be damaged by chronically elevated glutamate levels.Considering that the photoreceptors show signs of stress, which is perhaps the result of decreased choroidal blood flow that is known to occur with elevated IOP,these injured cells might either release excess or fail to reuptake their glutamate. Increased levels of glutamate would, in turn, overstimulate the bipolar cells, which would release excess glutamate and thus chronically overexcite the RGCs.We have shown that retinal laser photocoagulation protects the overlying RGCs for extended periods in experimental glaucoma. Whether these cells are functional remains to be elucidated. The precise mechanisms behind this phenomenon are not known. It may be owing to nonspecific neuroprotective effects caused by the laser treatment or an interruption of a transsynaptic signal, which is toxic to RGCs originating from photoreceptors. Challenges remain to determine what the mechanism is and if it can be made applicable to the treatment of human glaucoma.HAQuigleyNumber of people with glaucoma worldwide.Br J Ophthalmol.1996;80:389-393.HAQuigleyWRGreenThe histology of human glaucoma cupping and optic nerve damage: clinicopathologic correlation in 21 eyes.Ophthalmology.1979;86:1803-1830.SSHayrehBlood supply of the optic nerve head and its role in optic atrophy, glaucoma, and oedema of the optic disc.Br J Ophthalmol.1969;53:721-748.MEPeaseSJMcKinnonHAQuigleyLAKerrigan-BaumrindDJZackObstructed axonal transport of BDNF and its receptor TrkB in experimental glaucoma.Invest Ophthalmol Vis Sci.2000;41:764-774.RWNickellsDJZackApoptosis in ocular disease: a molecular overview.Ophthalmic Genet.1996;17:145-165.JFrançoisGVerriestLes dyschromatopsies acquises dans le glaucome primaire.Ann Ocul (Paris).1959;192:191-199.PGrutznerSSchleicherAcquired color vision defects in glaucoma patients.Mod Probl Ophthalmol.1972;11:136-140.GAFishmanAEKrillMFishmanAcquired color vision defects in patients with open-angle glaucoma and ocular hypertension.Mod Probl Ophthalmol.1974;13:335-338.RLakowskiSMDranceAcquired dyschromatopsias: the earliest functional losses in glaucoma.Doc Ophthalmol Proc Ser.1979;19:159-165.DPoinoosawmySNagasubramanianJGlosterColor vision in patients with chronic simple glaucoma and ocular hypertension.Br J Ophthalmol.1980;64:852-857.DLAlvisElectroretinographic changes in controlled chronic open-angle glaucoma.Am J Ophthalmol.1966;61:121-131.DTFazioJRHeckenlivelyDAMartinREChristensenThe electroretinogram in advanced open-angle glaucoma.Doc Ophthalmol.1986;63:45-54.KHolopigianWSeipleCMayronRKotyMLorenzoElectrophysiological and psychophysical flicker sensitivity in patients with primary open-angle glaucoma and ocular hypertension.Invest Ophthalmol Vis Sci.1990;31:1863-1868.JVOdomJGFeghaliJCJinGWWeinsteinVisual function deficits in glaucoma: electroretinogram pattern and luminance nonlinearities.Arch Ophthalmol.1990;108:222-227.SLGrahamNot AvailableVaeganIGoldbergLBucklandFCHollowsFlash and pattern electroretinogram changes with optic atrophy and glaucoma.Exp Eye Res.1995;60:697-706.JNVer HoeveTJMurdockGAHeatleyTMNorkIncreased latency of early multifocal ERG response in glaucoma [abstract].Invest Ophthalmol Vis Sci.2000;41:S520.CMYanceyRALinsenmeierThe electroretinogram and choroidal PO2in the cat during elevated intraocular pressure.Invest Ophthalmol Vis Sci.1988;29:700-707.TMNorkJNVer HoeveGLPoulsenSwelling and loss of photoreceptors in chronic human and experimental glaucomas.Arch Ophthalmol.2000;118:235-245.GTezelRHernandezMBWaxImmunostaining of heat shock proteins in the retina and optic nerve head of normal and glaucomatous eyes.Arch Ophthalmol.2000;118:511-518.GLPoulsenCLSchlampTMNorkRWNickellsChanges in photoreceptor gene expression in glaucoma [abstract].Invest Ophthalmol Vis Sci.1999;40:S788.IHWallowRepair of the pigment epithelial barrier following photocoagulation.Arch Ophthalmol.1984;102:126-135.IHWallowWESponselTSStevensClinicopathologic correlation of diode laser burns in monkeys.Arch Ophthalmol.1991;109:648-653.JLStoneWEBarlowMSHumayunEde Juan JrAHMilamMorphometric analysis of macular photoreceptors and ganglion cells in retinas with retinitis pigmentosa.Arch Ophthalmol.1992;110:1634-1639.AJEisenfeldMMLaVailJHLaVailAssessment of possible transneuronal changes in the retina of rats with inherited retinal dystrophy: cell size, number, synapses, and axonal transport by retinal ganglion cells.J Comp Neurol.1984;223:22-34.DGaasterlandCKupferExperimental glaucoma in the rhesus monkey.Invest Ophthalmol.1974;13:455-457.HAQuigleyRMHohmanLaser energy levels for trabecular meshwork damage in the primate eye.Invest Ophthalmol Vis Sci.1983;24:1305-1307.JEPedersonDEGaasterlandLaser-induced primate glaucoma, I: progression of cupping.Arch Ophthalmol.1984;102:1689-1692.PLKaufmanIHWallowMinified diagnostic contact lenses for biomicroscopic examination and photocoagulation of the anterior and posterior segment in small primates [letter].Exp Eye Res.1985;40:883-885.JAPetersonJAKilandMACroftPLKaufmanIntraocular pressure measurement in cynomolgus monkeys: Tono-Pen versus manometry.Invest Ophthalmol Vis Sci.1996;37:1197-1199.RSGonneringRKDortzbachKAEricksonPLKaufmanThe cynomolgus monkey as a model for orbital research, II: anatomic effects of lateral orbitotomy.Curr Eye Res.1984;3:541-555.CACurcioKAAllenTopography of ganglion cells in human retina.J Comp Neurol.1990;300:5-25.IHLWallowClinicopathologic correlation of retinal photocoagulation in the human eye.In: Weingeist TA, Sneed SR, eds. Laser Surgery in Ophthalmology. Norwalk, Conn: Appleton & Lange; 1992:15-27.KRDiddieJTErnestThe effect of photocoagulation on the choroidal vasculature and retinal oxygen tension.Am J Ophthalmol.1977;84:62-66.MXiaoSMSastryZYLiEffects of retinal laser photocoagulation on photoreceptor basic fibroblast growth factor and survival.Invest Ophthalmol Vis Sci.1998;39:618-630.HKYipLWLamKFSoShort period of low level laser enhances retinal ganglion cell (RGC) survival after optic nerve axotomy in adult SD rats [abstract].Invest Ophth Vis Sci.1999;40:S132.ADi PoloLJAignerRJDunnGMBrayAJAguayoProlonged delivery of brain-derived neurotrophic factor by adenovirus-infected Muller cells temporarily rescues injured retinal ganglion cells.Proc Natl Acad Sci U S A.1998;95:3978-3983.EBDreyerDZurakowskiRASchumerSMPodosSALiptonElevated glutamate levels in the vitreous body of humans and monkeys with glaucoma.Arch Ophthalmol.1996;114:299-305.EYolesMSchwartzElevation of intraocular glutamate levels in rats with partial lesion of the optic nerve.Arch Ophthalmol.1998;116:906-910.MWatanabeMMishinaYInoueDifferential distributions of the NMDA receptor channel subunit mRNAs in the mouse retina.Brain Res.1994;634:328-332.EBDreyerDZhangSALiptonTranscriptional or translational inhibition blocks low dose NMDA-mediated cell death.Neuroreport.1995;6:942-944.YOtoriJYWeiCJBarnstableNeurotoxic effects of low doses of glutamate on purified rat retinal ganglion cells.Invest Ophthalmol Vis Sci.1998;39:972-981.AAlmABillBlood flow and oxygen extraction in the cat uvea at normal and high intraocular pressures.Acta Physiol Scand.1970;80:19-28.AAlmABillOcular and optic nerve blood flow at normal and increased intraocular pressures in monkeys (Macaca irus): a study with radioactively labeled microspheres including flow determinations in brain and some other tissues.Exp Eye Res.1973;15:15-29.PRojanapongpunSMDranceBJMorrisonOphthalmic artery flow velocity in glaucomatous and normal subjects.Br J Ophthalmol.1993;77:25-29.GMichelsonMJGrohMEGrohAGrundlerAdvanced primary open-angle glaucoma is associated with decreased ophthalmic artery blood-flow velocity.Ger J Ophthalmol.1995;4:21-24.MCelliniGLPossatiNCaramazzaRCaramazzaColour Doppler analysis of the choroidal circulation in chronic open-angle glaucoma.Ophthalmologica.1996;210:200-202.YYamazakiSMDranceThe relationship between progression of visual field defects and retrobulbar circulation in patients with glaucoma.Am J Ophthalmol.1997;124:287-295.ZButtCO'BrienGMcKillopPAspinallPAllanColor Doppler imaging in untreated high- and normal-pressure open-angle glaucoma.Invest Ophthalmol Vis Sci.1997;38:690-696.HJKaiserASchoetzauDStumpfigJFlammerBlood-flow velocities of the extraocular vessels in patients with high-tension and normal-tension primary open-angle glaucoma.Am J Ophthalmol.1997;123:320-327.Accepted for publication July 6, 2000.This study was supported in part by unrestricted gifts from Research to Prevent Blindness Inc, New York, NY (Dr Nork), the Glaucoma Research Foundation, San Francisco, Calif (Dr Nork), the Wisconsin Lions Eye Research Fund, Stevens Point (Dr Nork), and a Career Development Award from Research to Prevent Blindness Inc (Dr Nickells).This article is an abridgment of a thesis submitted in partial fulfillment of requirements for membership in the American Ophthalmological Society, May 2001 (Dr Nork).We thank Paul L. Kaufman, MD, and Charlene B. Y. Kim, PhD, for critically reviewing this manuscript and Hugh D. Wabers, BS, for photographic assistance.Reprints: T. Michael Nork, MD, Department of Ophthalmology and Visual Sciences, University of Wisconsin Medical School, 600 Highland Ave, F4/336, Madison, WI 53792-3220 (e-mail: [email protected]).