Histoplasma Urinary Antigen Testing Obviates the Need for Coincident Serum Antigen Testing

Histoplasma Urinary Antigen Testing Obviates the Need for Coincident Serum Antigen Testing Abstract Objectives Serum and urine antigen (SAg, UAg) detection are common tests for Histoplasma capsulatum. UAg detection is more widely used and reportedly has a higher sensitivity. We investigated whether SAg detection contributes meaningfully to the initial evaluation of patients with suspected histoplasmosis. Methods We reviewed 20,285 UAg and 1,426 SAg tests ordered from 1997 to 2016 and analyzed paired UAg and SAg tests completed on the same patient within 1 week. We determined the positivity rate for each test. Results Of 601 paired specimens, 542 were concurrent negatives and 48 were concurrent positives (98% agreement). Medical records were available for eight of 11 pairs with discrepant results. UAg was falsely positive in six instances, truly positive once, and falsely negative once. Conclusions These findings support using a single antigen detection test, rather than both UAg and SAg, as an initial screen for suspected histoplasmosis. This aligns with the current practice of most physicians. Histoplasma, Histoplasmosis, Diagnosis, Antigen, Test utilization Histoplasma capsulatum is a dimorphic fungus that is found in bird or bat guano. It is endemic to the Ohio and Mississippi River valleys of North America and also present in Central America, parts of South America, and other locations around the world.1 Most patients acquire the infection through inhalation of airborne conidia. Infections are usually asymptomatic or unrecognized. The most common presentations for high-level and low-level exposures are acute and subacute pulmonary histoplasmosis, respectively.1-4 Chronic pulmonary histoplasmosis may occur in those with chronic lung disease, and mediastinal histoplasmosis syndromes are also possible. Immunocompromised hosts are at highest risk for disseminated histoplasmosis.5-7 Endocarditis, pericarditis, and central nervous system infections are less common manifestations of histoplasmosis. Accurate diagnosis of histoplasmosis is essential because patients often present with nonspecific symptoms. Furthermore, histoplasmosis generally responds well when antifungal therapy is given early in the course of disease but can progress to serious illness if undetected. Evaluation of patients with suspected histoplasmosis involves both clinical findings and laboratory studies. Several tests used for the diagnosis of histoplasmosis include culture,2,8 microscopic examination,9 antibody detection (ie, serology),10,11 antigen detection by enzyme-linked immunosorbent assay (ELISA),12,13 and molecular diagnostic methods, such as polymerase chain reaction–based assays.14,15 Histoplasma capsulatum grows slowly in culture, sometimes taking several weeks. Culture may be falsely negative in patients with a low fungal burden or if antifungal therapy had been given.16,17 Direct microscopic examination is rapid but may have low sensitivity in some forms of histoplasmosis; the specificity is limited by the similar morphology of other small yeasts (eg, Candida glabrata).16 Serologic tests are useful in some circumstances but may yield false-negative results in recently exposed or immunocompromised individuals; positive serologic results may be challenging to interpret, as these may represent remote exposure.17-19 Histoplasma antigen detection by ELISA is a widely performed test for the diagnosis of histoplasmosis, especially those with suspected acute pulmonary or disseminated histoplasmosis.2 First described in 1986 by Wheat et al,12 this assay involves the detection of Histoplasma polysaccharide antigen. This assay is most commonly performed on urine and/or serum samples. The sensitivity and specificity of urine antigen detection (UAg) and serum antigen detection (SAg) tests vary based on the type and extent of the Histoplasma infection. The sensitivity of UAg detection is greater than SAg detection in patients with disseminated histoplasmosis.17 The relative sensitivities of UAg and SAg detection in patients with acute pulmonary histoplasmosis (APH) vary depending on the study, with UAg detection demonstrating either a higher or comparable sensitivity compared with SAg detection.18,20,21 The sensitivities of both tests are considerably lower in chronic pulmonary infection.12,17 Generally, UAg detection is ordered as an initial screen for histoplasmosis, while in fewer instances, SAg detection has been ordered instead of or in addition to UAg detection. The goal of this retrospective analysis was to determine whether SAg detection plays a useful role in the initial evaluation of a patient with suspected histoplasmosis, with the view of reducing unnecessary testing in these patients. Materials and Methods The Cleveland Clinic electronic archives were searched for UAg and SAg tests, as well as corresponding fungal culture results from 1997 to 2016. Antigen detection tests performed on samples other than urine or serum were excluded, as were tests that were not performed or where performance status/sample type was unable to be determined. Most of the antigen detection assays were performed at MiraVista Diagnostics (Indianapolis, IN), except those ordered from January 23, 2014, to August 28, 2015, which were performed on site using the IMMY ALPHA Histoplasma Antigen Enzyme Immunoassay (Immuno-Mycologics, Norman, OK) for the diagnosis of histoplasmosis according to the package instructions and as described previously.22 This assay was performed on urine samples only. Results for both urine and serum Histoplasma antigen detection assays were reported in units between 0.4 and 19.0 ng/mL. Values below 0.4 ng/mL but above the limit of detection were reported as “positive, below the limit of quantification,” and results positive but above 19.0 ng/mL were reported as “positive, above the limit of quantification.” We analyzed pairs of UAg and SAg tests that were completed on the same patient within the 1 week (ie, paired specimens). UAg and SAg tests that were in agreement (ie, uniformly positive or negative) were noted. When both assays were positive, the result was considered a true positive, whereas when both assays were negative, the result was considered a true negative. The electronic medical record review was undertaken for any patients in whom the UAg and SAg results were discrepant to determine if the patient had histoplasmosis. The information reviewed in the medical record included medical history and the results of fungal culture, fungal serologic studies, and any histologic or cytologic findings. The final clinical impression by the attending physician was also noted, as well as consultations from infectious diseases specialists. This information was used to categorize (ie, true positive, false positive, etc) the discrepant UAg and SAg results. Proportions were compared as percentage for a χ2 determination of significance, where appropriate, using Epic Calc 2000 (available at http://www.brixtonhealth.com/epicalc.html). Results In total, 20,285 UAg and 1,426 SAg tests were ordered from 1997 to 2016, and 19,998 UAg and 1,250 SAg tests met the inclusion criteria for further study. There were 19,391 single UAg tests, 643 single SAg tests, and 601 paired UAg and SAg specimens. Single UAg tests yielded a higher positivity rate (17.0%) than paired tests (9.4%) (P < .001), while single and paired SAg tests had similar positivity rates of 7.8% and 8.3%, respectively (P = .623) (Supplemental Table 1; all supplemental materials can be found at American Journal of Clinical Pathology online). A total of 601 paired specimens were identified. This patient group was 44% female and had a mean (SD) age of 56 (18) years. Race was reviewed and found to be noncontributory (data not shown). A total of 106 (18%) paired specimens consisted of UAg and SAg tests ordered by two different physicians. The setting of the ordering physician was available in 417 of the 601 paired specimens, with 287 (69%) inpatient, 125 (30%) outpatient, and five (1%) in the emergency department. Of the total paired specimens reviewed, 542 (90%) were concurrent negatives and 48 (8%) were concurrent positives. Of the 37 concurrent positives with medical records available for review, there were 26 cases of disseminated histoplasmosis; three cases of acute, one case of subacute, and six cases of chronic pulmonary histoplasmosis (classification criteria described previously23); and one case of granulomatous mediastinitis. The remaining 11 of 601 pairs had discordant results. Of the nine specimens that were UAg positive/SAg negative, cultures were either negative (n = 6) or not performed (n = 3). Of the two specimens that were SAg positive/UAg negative, cultures were either negative (n = 1) or not performed (n = 1). This yields 98% agreement between UAg and SAg Histoplasma antigen detection. UAg and SAg detection tests are quantitative, and we correlated the values obtained in the concurrent positive and discordant tests Figure 1. Thirty-five of the 48 concurrent positives had values for UAg and SAg while the remaining 13 had at least one test with a value below or above the limit of quantification (Supplemental Table 2). Of the discordant pairs, only one pair, UAg positive below the limit of quantification and SAg negative, did not have a value for the positive test. Figure 1 View large Download slide Correlation between quantifiable urine antigen (UAg) and serum antigen (SAg) values. A, Thirty-five of 48 dual-positive UAg and SAg pairs had two quantifiable values. The remaining 13 are shown in Supplemental Table 2. Pairs are arranged by increasing difference between the two values. The minimum difference is 0.05 ng/mL, the maximum difference is 24.61 ng/mL, and mean difference is 4.4 ng/mL with a standard deviation of 5.3 ng/mL. B, Eight of the nine discordant pairs with positive UAg and negative SAg testing had reported values, whereas the ninth pair (*) was below the limit of quantification. C, Both discordant pairs with positive SAg and negative UAg testing had reported values. Figure 1 View large Download slide Correlation between quantifiable urine antigen (UAg) and serum antigen (SAg) values. A, Thirty-five of 48 dual-positive UAg and SAg pairs had two quantifiable values. The remaining 13 are shown in Supplemental Table 2. Pairs are arranged by increasing difference between the two values. The minimum difference is 0.05 ng/mL, the maximum difference is 24.61 ng/mL, and mean difference is 4.4 ng/mL with a standard deviation of 5.3 ng/mL. B, Eight of the nine discordant pairs with positive UAg and negative SAg testing had reported values, whereas the ninth pair (*) was below the limit of quantification. C, Both discordant pairs with positive SAg and negative UAg testing had reported values. To determine if the second test of the paired specimens was carried out in confirmation of the first, we performed chart review of the cases in which one result was viewable before the second was ordered. Of the 542 dual negatives, this occurred 28 times, 16 of which were available for review. In eight cases, there was no mention of a reason for the additional test; in the other eight cases, the repeat test was done for a documented reason such as conformation of the prior test. In the 48 dual positive cases, there were five cases in which one result was viewable before the other test was ordered, and two of these cases were available for review. In one case, the physician mentioned obtaining a Histoplasma SAg level to monitor therapy; in the other, no reason was mentioned. Among the discordant results, there were two pairs in which the result of the first test was viewable before the second was sent; in neither case was there mention of the reason for testing. The 11 discordant pairs resulted between October 22, 2012, and August 12, 2016. Of the 11 patients with discrepant UAg/SAg results, eight had available medical records. In the discordant cases, there was no documentation of one test occurring to monitor treatment or initiation of antifungal therapy between the two tests. Based predominantly on the infectious diseases consult notes and culture/serology data, we categorized six of the UAg results as false positives, one UAg result as a true positive, and one UAg result as a false negative Table 1. The sensitivity in this assessment of paired specimens was 98% for both the UAg and SAg tests. The specificity of the UAg test was 99%, whereas it was 100% for the SAg test. Table 1 Characteristics of Eight of the 11 Discordant Pairsa Diagnostic Test Result  Presenting Condition  Comorbidities  Criteria for Histoplasma Diagnosis  Treatment/Outcome  UAg: 1.3 ng/mL SAg: ND Fungal cx: 1/3 BAL cx positive  Pneumonia  CKD H/O prostate cancer Leukopenia  FP UAg in light of repeat testing, per ID note  Improved  UAg: 1.1 ng/mL SAg: ND Fungal blood cx negative  Fungemia  ESRD  FP UAg in light of negative SAg and fungal cx; per ID note  Improved, found to have Candida fungemia with endophthalmitis  UAg: 1.8 ng/mL SAg: ND Fungal blood cx negative  SBO, bacteremia, FTT, thrombocytopenia  CKD Malnutrition  FP UAg. Seriously ill with Escherichia coli bacteremia and CMV viremia  Death  UAg: 2.9 ng/mL SAg: ND Fungal blood cx negative  Respiratory distress  Liver transplant on tacrolimus, mycophenolate mofetil  FP UAg in light of negative SAg and fungal cx; per ID note  Treated for 3 days with amphotericin B, then halted after negative SAg; died of respiratory failure  UAg: below q SAg: ND Fungal cx (lung): Aspergillus  Evaluation of cavitary lesion found on imaging  ESRD  FP UAg in light of negative SAg and Aspergillus in cx  Treated with voriconazole 8 weeks, clinically improved  UAg: 0.9 ng/mL SAg: ND Fungal blood cx negative Serology negative  Swelling/erythema of right hand, left buttock  UC on adalimumab + multiple steroid courses  FP UAg considering incongruous clinical picture per ID note  Improved  UAg: 4.82 ng/mL SAg: ND H/O surgically proven pulmonary histoplasmosis  Right orbital mass- autoimmune vs meningioma    TP UAg, FN SAg; possibly residual UAg shedding per neurologist note  Lost to follow-up  UAg: ND SAg: 1.58 ng/mL Fungal blood cx negative Serology (H and M bands) positive  Fever of unknown origin  Marfans following mitral valve and ascending aorta repair H/O Histoplasma endocarditis  FN UAg, TP SAg in light of (+) SAg and (+) serology per ID note; diagnosed with disseminated histoplasmosis ± endocarditis  Clinically improved on itraconazole for 3 months, SAg negative 5 months later  Diagnostic Test Result  Presenting Condition  Comorbidities  Criteria for Histoplasma Diagnosis  Treatment/Outcome  UAg: 1.3 ng/mL SAg: ND Fungal cx: 1/3 BAL cx positive  Pneumonia  CKD H/O prostate cancer Leukopenia  FP UAg in light of repeat testing, per ID note  Improved  UAg: 1.1 ng/mL SAg: ND Fungal blood cx negative  Fungemia  ESRD  FP UAg in light of negative SAg and fungal cx; per ID note  Improved, found to have Candida fungemia with endophthalmitis  UAg: 1.8 ng/mL SAg: ND Fungal blood cx negative  SBO, bacteremia, FTT, thrombocytopenia  CKD Malnutrition  FP UAg. Seriously ill with Escherichia coli bacteremia and CMV viremia  Death  UAg: 2.9 ng/mL SAg: ND Fungal blood cx negative  Respiratory distress  Liver transplant on tacrolimus, mycophenolate mofetil  FP UAg in light of negative SAg and fungal cx; per ID note  Treated for 3 days with amphotericin B, then halted after negative SAg; died of respiratory failure  UAg: below q SAg: ND Fungal cx (lung): Aspergillus  Evaluation of cavitary lesion found on imaging  ESRD  FP UAg in light of negative SAg and Aspergillus in cx  Treated with voriconazole 8 weeks, clinically improved  UAg: 0.9 ng/mL SAg: ND Fungal blood cx negative Serology negative  Swelling/erythema of right hand, left buttock  UC on adalimumab + multiple steroid courses  FP UAg considering incongruous clinical picture per ID note  Improved  UAg: 4.82 ng/mL SAg: ND H/O surgically proven pulmonary histoplasmosis  Right orbital mass- autoimmune vs meningioma    TP UAg, FN SAg; possibly residual UAg shedding per neurologist note  Lost to follow-up  UAg: ND SAg: 1.58 ng/mL Fungal blood cx negative Serology (H and M bands) positive  Fever of unknown origin  Marfans following mitral valve and ascending aorta repair H/O Histoplasma endocarditis  FN UAg, TP SAg in light of (+) SAg and (+) serology per ID note; diagnosed with disseminated histoplasmosis ± endocarditis  Clinically improved on itraconazole for 3 months, SAg negative 5 months later  BAL, bronchoalveolar lavage; CKD, chronic kidney disease; CMV, cytomegalovirus; cx, culture; ESRD, end-stage renal disease; FN, false negative; FP, false positive; FTT, failure to thrive; H/O, history of; ID, infectious disease doctors; ND, not detected (negative result); q, level of quantitation; SAg, serum antigen; SBO, small bowel obstruction; TP, true positive; UAg, urine antigen; UC, ulcerative colitis. aCharacteristics of the patients with discordant paired UAg and SAg results include the following: diagnostic test results for histoplasmosis, including UAg and SAg detection, fungal culture, and serologies; the patient’s presenting condition per hospital discharge summary; and relevant patient comorbidities, including immunosuppressive conditions (cancer, autoimmune disease, human immunodeficiency virus/AIDS) and kidney disease, which may affect the quantity of Histoplasma antigen into the urine. In addition, immunosuppressive therapies (chemotherapy, disease-modifying antirheumatic drugs, biologic agents, and greater than 3 weeks of prednisone) are listed. The criteria presented for or against the diagnosis of Histoplasma infection as well as the patient’s outcome are depicted in the last two columns. View Large Discussion The current practice of most physicians in the evaluation of patients suspected to have histoplasmosis is to perform the antigen detection test on a urine specimen, accompanied by additional laboratory testing such as culture, serology, or histopathologic studies in certain cases. Sometimes, Histoplasma serum antigen (SAg) detection is performed concurrently with a urine antigen (UAg) test. Multiple studies of the sensitivity and specificity of UAg and SAg detection have found that the assay performance varies based on the nature of the infection Table 2. The sensitivity of UAg detection appears to be higher than SAg for disseminated infections, while there is not a significant difference between UAg and SAg detection for APH. A recent case report from the Histoplasma reference laboratory indicates that testing both urine and serum provides the highest sensitivity for Histoplasma detection.25 Current Centers for Disease Control and Prevention guidelines state that urine and/or serum antigen detection is the most sensitive and specific method to diagnose histoplasmosis.26 Due to the varied possible presentations of Histoplasma infection, head-to-head comparison of UAg and SAg detection on the same patients provides evidence to inform clinical practice—specifically, whether both UAg and SAg detection tests should be ordered, or whether one of these alone (most commonly UAg alone) would suffice. There are few such data, although Hage et al24 reported antigenemia in 100% of 31 patients with disseminated histoplasmosis, with all but one of these testing positive for antigenuria. Through this large retrospective analysis, we provide evidence that supports that a single antigen detection test is appropriate in the initial evaluation for histoplasmosis. Table 2 Sensitivities of UAg and SAg Detection Infection  UAg Sensitivity, %  SAg Sensitivity, %  Brief Description  Study  Disseminated histoplasmosis with AIDS  95  86  80 patients, 79 UAg and 63 SAg tests  Williams et al17  Disseminated histoplasmosis without AIDS  81  64  28 patients; 17 with immunosuppressive conditions besides AIDS; 27 UAg and 11 SAg tests  Williams et al17  Disseminated histoplasmosis with and without AIDS  91.8  —  158 patients  Hage et al24  APH  75  0  8 cavers  Davies18  APH  65  69  130 patients  Swartzentruber et al20  APH  83.3  —  6 patients  Hage et al24  Subacute pulmonary histoplasmosis  30.4  —  46 patients  Hage et al24  Chronic pulmonary histoplasmosis  87.5  —  7 patients  Hage et al24  Disseminated or pulmonary histoplasmosis  79  82  Meta-analysis of 1,029 patients  Fandiño-Devia et al21  Infection  UAg Sensitivity, %  SAg Sensitivity, %  Brief Description  Study  Disseminated histoplasmosis with AIDS  95  86  80 patients, 79 UAg and 63 SAg tests  Williams et al17  Disseminated histoplasmosis without AIDS  81  64  28 patients; 17 with immunosuppressive conditions besides AIDS; 27 UAg and 11 SAg tests  Williams et al17  Disseminated histoplasmosis with and without AIDS  91.8  —  158 patients  Hage et al24  APH  75  0  8 cavers  Davies18  APH  65  69  130 patients  Swartzentruber et al20  APH  83.3  —  6 patients  Hage et al24  Subacute pulmonary histoplasmosis  30.4  —  46 patients  Hage et al24  Chronic pulmonary histoplasmosis  87.5  —  7 patients  Hage et al24  Disseminated or pulmonary histoplasmosis  79  82  Meta-analysis of 1,029 patients  Fandiño-Devia et al21  APH, acute pulmonary histoplasmosis; SAg, serum antigen; UAg, urine antigen. View Large We found a 98% agreement between 601 paired urine and serum specimens, with 542 uniformly negative tests and 48 uniformly positive tests. There were only 11 instances wherein the paired results were discrepant, and medical records were available for review for eight of these patients. Six patients had positive UAg results but negative SAg results and negative corresponding cultures. Based on medical record review, which disclosed an extensive workup, we determined that each of these results was a false positive. We reviewed the medical records of eight of the 11 patients with discordant paired results (Table 1). As may be predicted, UAg detection, the test with higher sensitivity, had six results characterized as false positives after medical record review. The validity of the characterization of the remaining discordant cases is arguable; one was characterized as UAg true positive/SAg false negative, whereas the other was characterized as UAg false negative/SAg true positive. The UAg true-positive/SAg false-negative result occurred in a patient who had disseminated histoplasmosis 2 years before with a residual pulmonary lesion that was thought to be largely resolved; it was hypothesized that the pulmonary lesion was either chronically active or shedding antigen. The UAg false negative/SAg true positive involved a patient with fever of unknown origin who had negative fungal blood cultures but was treated with itraconazole based on a positive SAg detection result and slowly improved. This patient had a remote history of Histoplasma endocarditis but was not found to have infective endocarditis during this illness. Overall, UAg and SAg detection performed comparatively, with a greater number of false positives produced by the UAg test, which is considered a useful characteristic of a screening test. The discordant quantitative UAg and SAg results were not all below the limit of quantification or extremely low values compared with the dual positives, indicating that the actual value of a low positive test may serve as a guide but not a direct correlate of the likelihood of histoplasmosis (Figure 1). Our 11 discordant pairs resulted between 2012 and 2016. Our results extend to 1997, but nearly all of the paired specimens were ordered between 2009 and the present, indicating that paired testing, rather than discordant results, likely increased during this time period. The reason for the increase in ordering paired testing is likely multifactorial and may include increased awareness of both tests. In 2009, our institution transitioned to electronic ordering. There are conflicting results on the effect of electronic ordering systems on the number of duplicate orders; this timing is suspicious, and the ease of ordering the studies may have in part contributed to increased dual testing.27,28 Most of the Histoplasma antigen detection tests in this study were sent to MiraVista Diagnostics. However, from January 23, 2014, to August 28, 2015, a validated, laboratory-developed test was performed on site. During this time, 12 patients with moderately positive Histoplasma UAg test results subsequently received repeat UAg testing with SAg testing later in the same week. The paired SAg result and a repeat UAg test were negative upon repeat testing. Upon review of the medical records, we found that the infectious disease team recorded its impression that the on-site laboratory-developed assay had a higher false-positive rate and therefore conducted repeat testing on patients at a higher frequency. No similar patterns were found outside of the time period that the in-house assay was in use. The repeat UAg and subsequent SAg tests on these 12 patients were performed within the same week and therefore were also included as concurrent negative paired specimens in our analysis. These results do not show an absolute superiority of UAg over SAg, but the quantitative results of the dually positive specimens (Figure 1) suggest a superior sensitivity of the UAg test over the SAg test. In addition, this review demonstrates that there was no apparent benefit of ordering both tests simultaneously, since the results usually agreed. A practical and cost-effective clinical approach would be to limit initial studies to UAg testing, as most physicians currently do, and to follow up positive tests with a confirmatory SAg study, as well as reviewing concomitantly ordered (if the clinical suspicion warrants) additional studies, such as culture, serology, and histopathologic/cytologic assessments. Further analysis of the 601 paired specimens reveals that a minority of the second tests, nine of 601, were ordered for a documented clinical reason such as confirmation of the first result. We found that most (82%) of the paired tests were ordered by the same physician, and most (69%) were ordered in the inpatient setting. This provides initial guidance on how to target efficient test utilization awareness, but future studies should characterize the population on which paired specimens are ordered to better direct these efforts. Test overutilization leads to increased strain on the health care system, a decrease in patient satisfaction, and even physical and psychological harm to the patient.29,30 Based on the Medicare reimbursement rate for both UAg and SAg (ie, $17.52 each), a $10,687 savings could have been achieved if one rather than both of the 601 paired tests had been ordered; this reduction would not have been associated with misdiagnoses or patient harm. Systematically adopting the already widely used policy of performing a UAg detection test alone rather than combined UAg/SAg testing is both clinically acceptable and fiscally responsible. References 1. Wheat LJ, Azar MM, Bahr NC et al.   Histoplasmosis. Infect Dis Clin North Am . 2016; 30: 207- 227. Google Scholar CrossRef Search ADS PubMed  2. Kauffman CA. Histoplasmosis: a clinical and laboratory update. Clin Microbiol Rev . 2007; 20: 115- 132. Google Scholar CrossRef Search ADS PubMed  3. Gustafson TL, Kaufman L, Weeks R et al.   Outbreak of acute pulmonary histoplasmosis in members of a wagon train. Am J Med . 1981; 71: 759- 765. Google Scholar CrossRef Search ADS PubMed  4. Ward JI, Weeks M, Allen D et al.   Acute histoplasmosis: clinical, epidemiologic and serologic findings of an outbreak associated with exposure to a fallen tree. Am J Med . 1979; 66: 587- 595. Google Scholar CrossRef Search ADS PubMed  5. Assi M, Martin S, Wheat LJ et al.   Histoplasmosis after solid organ transplant. Clin Infect Dis . 2013; 57: 1542- 1549. Google Scholar CrossRef Search ADS PubMed  6. Nacher M, Adenis A, Mc Donald S et al.   Disseminated histoplasmosis in HIV-infected patients in South America: a neglected killer continues on its rampage. Plos Negl Trop Dis . 2013; 7: e2319. Google Scholar CrossRef Search ADS PubMed  7. Wheat LJ, Slama TG, Norton JA et al.   Risk factors for disseminated or fatal histoplasmosis: analysis of a large urban outbreak. Ann Intern Med . 1982; 96: 159- 163. Google Scholar CrossRef Search ADS PubMed  8. Wheat LJ, Wass J, Norton J et al.   Cavitary histoplasmosis occurring during two large urban outbreaks: analysis of clinical, epidemiologic, roentgenographic, and laboratory features. Medicine (Baltimore) . 1984; 63: 201- 209. Google Scholar CrossRef Search ADS PubMed  9. Kauffman CA. Histoplasmosis. In: Crow MK, Doroshow JH, Drazen JM, et al, eds. Goldman-Cecil Medicine . 2nd ed. Philadelphia, PA: Elsevier/Saunders; 2016:2070-2072. 10. Kaufman L. Serodiagnosis of fungal disease. In: Rose NR, Friedman H, eds. Manual of Clinical Immunology . Washington, DC: American Society for Microbiology; 1976: 363-3 81. 11. Picardi JL, Kauffman CA, Schwarz J et al.   Detection of precipitating antibodies to Histoplasma capsulatum by counterimmunoelectrophoresis. Am Rev Respir Dis . 1976; 114: 171- 176. Google Scholar PubMed  12. Wheat LJ, Kohler RB, Tewari RP. Diagnosis of disseminated histoplasmosis by detection of Histoplasma capsulatum antigen in serum and urine specimens. N Engl J Med . 1986; 314: 83- 88. Google Scholar CrossRef Search ADS PubMed  13. Wheat LJ. Improvements in diagnosis of histoplasmosis. Expert Opin Biol Ther . 2006; 6: 1207- 1221. Google Scholar CrossRef Search ADS PubMed  14. Koepsell SA, Hinrichs SH, Iwen PC. Applying a real-time PER assay for Histoplasma capsulatum to clinically relevant formalin-fixed paraffin-embedded human tissue. J Clin Microbiol . 2012; 50: 3395- 3397. Google Scholar CrossRef Search ADS PubMed  15. Guedes HL, Guimarães AJ, Muniz Mde M et al.   PCR assay for identification of Histoplasma capsulatum based on the nucleotide sequence of the M antigen. J Clin Microbiol . 2003; 41: 535- 539. Google Scholar CrossRef Search ADS PubMed  16. Guimarães AJ, Nosanchuk JD, Zancopé-Oliveira RM. Diagnosis of histoplasmosis. Braz J Microbiol . 2006; 37: 1- 13. Google Scholar CrossRef Search ADS PubMed  17. Williams B, Fojtasek M, Connolly-Stringfield P et al.   Diagnosis of histoplasmosis by antigen detection during an outbreak in Indianapolis, Ind. Arch Pathol Lab Med . 1994; 118: 1205- 1208. Google Scholar PubMed  18. Davies SF. Serodiagnosis of histoplasmosis. Semin Respir Infect . 1986; 1: 9- 15. Google Scholar PubMed  19. Wheat J, French ML, Kohler RB et al.   The diagnostic laboratory tests for histoplasmosis: analysis of experience in a large urban outbreak. Ann Intern Med . 1982; 97: 680- 685. Google Scholar CrossRef Search ADS PubMed  20. Swartzentruber S, Rhodes L, Kurkjian K et al.   Diagnosis of acute pulmonary histoplasmosis by antigen detection. Clin Infect Dis . 2009; 49: 1878- 1882. Google Scholar CrossRef Search ADS PubMed  21. Fandiño-Devia E, Rodríguez-Echeverri C, Cardona-Arias J et al.   Antigen detection in the diagnosis of histoplasmosis: a meta-analysis of diagnostic performance. Mycopathologia . 2016; 181: 197- 205. Google Scholar CrossRef Search ADS PubMed  22. LeMonte A, Egan L, Connolly P et al.   Evaluation of the IMMY ALPHA histoplasma antigen enzyme immunoassay for diagnosis of histoplasmosis marked by antigenuria. Clin Vaccine Immunol . 2007; 14: 802- 803. Google Scholar CrossRef Search ADS PubMed  23. Wheat LJ, Conces D, Allen SD et al.   Pulmonary histoplasmosis syndromes: recognition, diagnosis, and management. Semin Respir Crit Care Med . 2004; 25: 129- 144. Google Scholar CrossRef Search ADS PubMed  24. Hage CA, Ribes JA, Wengenack NL et al.   A multicenter evaluation of tests for diagnosis of histoplasmosis. Clin Infect Dis . 2011; 53: 448- 454. Google Scholar CrossRef Search ADS PubMed  25. Wheat LJ. Diagnosing histoplasmosis by antigen detection . Available at http://miravistalabs.com/diagnosing-histoplasmosis-by-antigen-detection/. 26. Centers for Disease Control and Prevention. Information for health professionals about histoplasmosis. 2017. Available at https://www.cdc.gov/fungal/diseases/histoplasmosis/health-professionals.html. Accessed July 4, 2017. 27. Stewart BA, Fernandes S, Rodriguez-Huertas E et al.   A preliminary look at duplicate testing associated with lack of electronic health record interoperability for transferred patients. J Am Med Inform Assoc . 2010; 17: 341- 344. Google Scholar CrossRef Search ADS PubMed  28. Kerwin TC, Leighton H, Buch K et al.   The effect of adoption of an electronic health record on duplicate testing. Cardiol Res Pract . 2016; 2016: 1950191. Google Scholar CrossRef Search ADS PubMed  29. Procop GW, Keating C, Stagno P et al.   Reducing duplicate testing: a comparison of two clinical decision support tools. Am J Clin Pathol . 2015; 143: 623- 626. Google Scholar CrossRef Search ADS PubMed  30. Procop GW, Yerian LM, Wyllie R et al.   Duplicate laboratory test reduction using a clinical decision support tool. Am J Clin Pathol . 2014; 141: 718- 723. Google Scholar CrossRef Search ADS PubMed  © American Society for Clinical Pathology, 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png American Journal of Clinical Pathology Oxford University Press

Histoplasma Urinary Antigen Testing Obviates the Need for Coincident Serum Antigen Testing

Loading next page...
 
/lp/ou_press/histoplasma-urinary-antigen-testing-obviates-the-need-for-coincident-8Xi3t5Aqlq
Publisher
Oxford University Press
Copyright
© American Society for Clinical Pathology, 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0002-9173
eISSN
1943-7722
D.O.I.
10.1093/ajcp/aqx169
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives Serum and urine antigen (SAg, UAg) detection are common tests for Histoplasma capsulatum. UAg detection is more widely used and reportedly has a higher sensitivity. We investigated whether SAg detection contributes meaningfully to the initial evaluation of patients with suspected histoplasmosis. Methods We reviewed 20,285 UAg and 1,426 SAg tests ordered from 1997 to 2016 and analyzed paired UAg and SAg tests completed on the same patient within 1 week. We determined the positivity rate for each test. Results Of 601 paired specimens, 542 were concurrent negatives and 48 were concurrent positives (98% agreement). Medical records were available for eight of 11 pairs with discrepant results. UAg was falsely positive in six instances, truly positive once, and falsely negative once. Conclusions These findings support using a single antigen detection test, rather than both UAg and SAg, as an initial screen for suspected histoplasmosis. This aligns with the current practice of most physicians. Histoplasma, Histoplasmosis, Diagnosis, Antigen, Test utilization Histoplasma capsulatum is a dimorphic fungus that is found in bird or bat guano. It is endemic to the Ohio and Mississippi River valleys of North America and also present in Central America, parts of South America, and other locations around the world.1 Most patients acquire the infection through inhalation of airborne conidia. Infections are usually asymptomatic or unrecognized. The most common presentations for high-level and low-level exposures are acute and subacute pulmonary histoplasmosis, respectively.1-4 Chronic pulmonary histoplasmosis may occur in those with chronic lung disease, and mediastinal histoplasmosis syndromes are also possible. Immunocompromised hosts are at highest risk for disseminated histoplasmosis.5-7 Endocarditis, pericarditis, and central nervous system infections are less common manifestations of histoplasmosis. Accurate diagnosis of histoplasmosis is essential because patients often present with nonspecific symptoms. Furthermore, histoplasmosis generally responds well when antifungal therapy is given early in the course of disease but can progress to serious illness if undetected. Evaluation of patients with suspected histoplasmosis involves both clinical findings and laboratory studies. Several tests used for the diagnosis of histoplasmosis include culture,2,8 microscopic examination,9 antibody detection (ie, serology),10,11 antigen detection by enzyme-linked immunosorbent assay (ELISA),12,13 and molecular diagnostic methods, such as polymerase chain reaction–based assays.14,15 Histoplasma capsulatum grows slowly in culture, sometimes taking several weeks. Culture may be falsely negative in patients with a low fungal burden or if antifungal therapy had been given.16,17 Direct microscopic examination is rapid but may have low sensitivity in some forms of histoplasmosis; the specificity is limited by the similar morphology of other small yeasts (eg, Candida glabrata).16 Serologic tests are useful in some circumstances but may yield false-negative results in recently exposed or immunocompromised individuals; positive serologic results may be challenging to interpret, as these may represent remote exposure.17-19 Histoplasma antigen detection by ELISA is a widely performed test for the diagnosis of histoplasmosis, especially those with suspected acute pulmonary or disseminated histoplasmosis.2 First described in 1986 by Wheat et al,12 this assay involves the detection of Histoplasma polysaccharide antigen. This assay is most commonly performed on urine and/or serum samples. The sensitivity and specificity of urine antigen detection (UAg) and serum antigen detection (SAg) tests vary based on the type and extent of the Histoplasma infection. The sensitivity of UAg detection is greater than SAg detection in patients with disseminated histoplasmosis.17 The relative sensitivities of UAg and SAg detection in patients with acute pulmonary histoplasmosis (APH) vary depending on the study, with UAg detection demonstrating either a higher or comparable sensitivity compared with SAg detection.18,20,21 The sensitivities of both tests are considerably lower in chronic pulmonary infection.12,17 Generally, UAg detection is ordered as an initial screen for histoplasmosis, while in fewer instances, SAg detection has been ordered instead of or in addition to UAg detection. The goal of this retrospective analysis was to determine whether SAg detection plays a useful role in the initial evaluation of a patient with suspected histoplasmosis, with the view of reducing unnecessary testing in these patients. Materials and Methods The Cleveland Clinic electronic archives were searched for UAg and SAg tests, as well as corresponding fungal culture results from 1997 to 2016. Antigen detection tests performed on samples other than urine or serum were excluded, as were tests that were not performed or where performance status/sample type was unable to be determined. Most of the antigen detection assays were performed at MiraVista Diagnostics (Indianapolis, IN), except those ordered from January 23, 2014, to August 28, 2015, which were performed on site using the IMMY ALPHA Histoplasma Antigen Enzyme Immunoassay (Immuno-Mycologics, Norman, OK) for the diagnosis of histoplasmosis according to the package instructions and as described previously.22 This assay was performed on urine samples only. Results for both urine and serum Histoplasma antigen detection assays were reported in units between 0.4 and 19.0 ng/mL. Values below 0.4 ng/mL but above the limit of detection were reported as “positive, below the limit of quantification,” and results positive but above 19.0 ng/mL were reported as “positive, above the limit of quantification.” We analyzed pairs of UAg and SAg tests that were completed on the same patient within the 1 week (ie, paired specimens). UAg and SAg tests that were in agreement (ie, uniformly positive or negative) were noted. When both assays were positive, the result was considered a true positive, whereas when both assays were negative, the result was considered a true negative. The electronic medical record review was undertaken for any patients in whom the UAg and SAg results were discrepant to determine if the patient had histoplasmosis. The information reviewed in the medical record included medical history and the results of fungal culture, fungal serologic studies, and any histologic or cytologic findings. The final clinical impression by the attending physician was also noted, as well as consultations from infectious diseases specialists. This information was used to categorize (ie, true positive, false positive, etc) the discrepant UAg and SAg results. Proportions were compared as percentage for a χ2 determination of significance, where appropriate, using Epic Calc 2000 (available at http://www.brixtonhealth.com/epicalc.html). Results In total, 20,285 UAg and 1,426 SAg tests were ordered from 1997 to 2016, and 19,998 UAg and 1,250 SAg tests met the inclusion criteria for further study. There were 19,391 single UAg tests, 643 single SAg tests, and 601 paired UAg and SAg specimens. Single UAg tests yielded a higher positivity rate (17.0%) than paired tests (9.4%) (P < .001), while single and paired SAg tests had similar positivity rates of 7.8% and 8.3%, respectively (P = .623) (Supplemental Table 1; all supplemental materials can be found at American Journal of Clinical Pathology online). A total of 601 paired specimens were identified. This patient group was 44% female and had a mean (SD) age of 56 (18) years. Race was reviewed and found to be noncontributory (data not shown). A total of 106 (18%) paired specimens consisted of UAg and SAg tests ordered by two different physicians. The setting of the ordering physician was available in 417 of the 601 paired specimens, with 287 (69%) inpatient, 125 (30%) outpatient, and five (1%) in the emergency department. Of the total paired specimens reviewed, 542 (90%) were concurrent negatives and 48 (8%) were concurrent positives. Of the 37 concurrent positives with medical records available for review, there were 26 cases of disseminated histoplasmosis; three cases of acute, one case of subacute, and six cases of chronic pulmonary histoplasmosis (classification criteria described previously23); and one case of granulomatous mediastinitis. The remaining 11 of 601 pairs had discordant results. Of the nine specimens that were UAg positive/SAg negative, cultures were either negative (n = 6) or not performed (n = 3). Of the two specimens that were SAg positive/UAg negative, cultures were either negative (n = 1) or not performed (n = 1). This yields 98% agreement between UAg and SAg Histoplasma antigen detection. UAg and SAg detection tests are quantitative, and we correlated the values obtained in the concurrent positive and discordant tests Figure 1. Thirty-five of the 48 concurrent positives had values for UAg and SAg while the remaining 13 had at least one test with a value below or above the limit of quantification (Supplemental Table 2). Of the discordant pairs, only one pair, UAg positive below the limit of quantification and SAg negative, did not have a value for the positive test. Figure 1 View large Download slide Correlation between quantifiable urine antigen (UAg) and serum antigen (SAg) values. A, Thirty-five of 48 dual-positive UAg and SAg pairs had two quantifiable values. The remaining 13 are shown in Supplemental Table 2. Pairs are arranged by increasing difference between the two values. The minimum difference is 0.05 ng/mL, the maximum difference is 24.61 ng/mL, and mean difference is 4.4 ng/mL with a standard deviation of 5.3 ng/mL. B, Eight of the nine discordant pairs with positive UAg and negative SAg testing had reported values, whereas the ninth pair (*) was below the limit of quantification. C, Both discordant pairs with positive SAg and negative UAg testing had reported values. Figure 1 View large Download slide Correlation between quantifiable urine antigen (UAg) and serum antigen (SAg) values. A, Thirty-five of 48 dual-positive UAg and SAg pairs had two quantifiable values. The remaining 13 are shown in Supplemental Table 2. Pairs are arranged by increasing difference between the two values. The minimum difference is 0.05 ng/mL, the maximum difference is 24.61 ng/mL, and mean difference is 4.4 ng/mL with a standard deviation of 5.3 ng/mL. B, Eight of the nine discordant pairs with positive UAg and negative SAg testing had reported values, whereas the ninth pair (*) was below the limit of quantification. C, Both discordant pairs with positive SAg and negative UAg testing had reported values. To determine if the second test of the paired specimens was carried out in confirmation of the first, we performed chart review of the cases in which one result was viewable before the second was ordered. Of the 542 dual negatives, this occurred 28 times, 16 of which were available for review. In eight cases, there was no mention of a reason for the additional test; in the other eight cases, the repeat test was done for a documented reason such as conformation of the prior test. In the 48 dual positive cases, there were five cases in which one result was viewable before the other test was ordered, and two of these cases were available for review. In one case, the physician mentioned obtaining a Histoplasma SAg level to monitor therapy; in the other, no reason was mentioned. Among the discordant results, there were two pairs in which the result of the first test was viewable before the second was sent; in neither case was there mention of the reason for testing. The 11 discordant pairs resulted between October 22, 2012, and August 12, 2016. Of the 11 patients with discrepant UAg/SAg results, eight had available medical records. In the discordant cases, there was no documentation of one test occurring to monitor treatment or initiation of antifungal therapy between the two tests. Based predominantly on the infectious diseases consult notes and culture/serology data, we categorized six of the UAg results as false positives, one UAg result as a true positive, and one UAg result as a false negative Table 1. The sensitivity in this assessment of paired specimens was 98% for both the UAg and SAg tests. The specificity of the UAg test was 99%, whereas it was 100% for the SAg test. Table 1 Characteristics of Eight of the 11 Discordant Pairsa Diagnostic Test Result  Presenting Condition  Comorbidities  Criteria for Histoplasma Diagnosis  Treatment/Outcome  UAg: 1.3 ng/mL SAg: ND Fungal cx: 1/3 BAL cx positive  Pneumonia  CKD H/O prostate cancer Leukopenia  FP UAg in light of repeat testing, per ID note  Improved  UAg: 1.1 ng/mL SAg: ND Fungal blood cx negative  Fungemia  ESRD  FP UAg in light of negative SAg and fungal cx; per ID note  Improved, found to have Candida fungemia with endophthalmitis  UAg: 1.8 ng/mL SAg: ND Fungal blood cx negative  SBO, bacteremia, FTT, thrombocytopenia  CKD Malnutrition  FP UAg. Seriously ill with Escherichia coli bacteremia and CMV viremia  Death  UAg: 2.9 ng/mL SAg: ND Fungal blood cx negative  Respiratory distress  Liver transplant on tacrolimus, mycophenolate mofetil  FP UAg in light of negative SAg and fungal cx; per ID note  Treated for 3 days with amphotericin B, then halted after negative SAg; died of respiratory failure  UAg: below q SAg: ND Fungal cx (lung): Aspergillus  Evaluation of cavitary lesion found on imaging  ESRD  FP UAg in light of negative SAg and Aspergillus in cx  Treated with voriconazole 8 weeks, clinically improved  UAg: 0.9 ng/mL SAg: ND Fungal blood cx negative Serology negative  Swelling/erythema of right hand, left buttock  UC on adalimumab + multiple steroid courses  FP UAg considering incongruous clinical picture per ID note  Improved  UAg: 4.82 ng/mL SAg: ND H/O surgically proven pulmonary histoplasmosis  Right orbital mass- autoimmune vs meningioma    TP UAg, FN SAg; possibly residual UAg shedding per neurologist note  Lost to follow-up  UAg: ND SAg: 1.58 ng/mL Fungal blood cx negative Serology (H and M bands) positive  Fever of unknown origin  Marfans following mitral valve and ascending aorta repair H/O Histoplasma endocarditis  FN UAg, TP SAg in light of (+) SAg and (+) serology per ID note; diagnosed with disseminated histoplasmosis ± endocarditis  Clinically improved on itraconazole for 3 months, SAg negative 5 months later  Diagnostic Test Result  Presenting Condition  Comorbidities  Criteria for Histoplasma Diagnosis  Treatment/Outcome  UAg: 1.3 ng/mL SAg: ND Fungal cx: 1/3 BAL cx positive  Pneumonia  CKD H/O prostate cancer Leukopenia  FP UAg in light of repeat testing, per ID note  Improved  UAg: 1.1 ng/mL SAg: ND Fungal blood cx negative  Fungemia  ESRD  FP UAg in light of negative SAg and fungal cx; per ID note  Improved, found to have Candida fungemia with endophthalmitis  UAg: 1.8 ng/mL SAg: ND Fungal blood cx negative  SBO, bacteremia, FTT, thrombocytopenia  CKD Malnutrition  FP UAg. Seriously ill with Escherichia coli bacteremia and CMV viremia  Death  UAg: 2.9 ng/mL SAg: ND Fungal blood cx negative  Respiratory distress  Liver transplant on tacrolimus, mycophenolate mofetil  FP UAg in light of negative SAg and fungal cx; per ID note  Treated for 3 days with amphotericin B, then halted after negative SAg; died of respiratory failure  UAg: below q SAg: ND Fungal cx (lung): Aspergillus  Evaluation of cavitary lesion found on imaging  ESRD  FP UAg in light of negative SAg and Aspergillus in cx  Treated with voriconazole 8 weeks, clinically improved  UAg: 0.9 ng/mL SAg: ND Fungal blood cx negative Serology negative  Swelling/erythema of right hand, left buttock  UC on adalimumab + multiple steroid courses  FP UAg considering incongruous clinical picture per ID note  Improved  UAg: 4.82 ng/mL SAg: ND H/O surgically proven pulmonary histoplasmosis  Right orbital mass- autoimmune vs meningioma    TP UAg, FN SAg; possibly residual UAg shedding per neurologist note  Lost to follow-up  UAg: ND SAg: 1.58 ng/mL Fungal blood cx negative Serology (H and M bands) positive  Fever of unknown origin  Marfans following mitral valve and ascending aorta repair H/O Histoplasma endocarditis  FN UAg, TP SAg in light of (+) SAg and (+) serology per ID note; diagnosed with disseminated histoplasmosis ± endocarditis  Clinically improved on itraconazole for 3 months, SAg negative 5 months later  BAL, bronchoalveolar lavage; CKD, chronic kidney disease; CMV, cytomegalovirus; cx, culture; ESRD, end-stage renal disease; FN, false negative; FP, false positive; FTT, failure to thrive; H/O, history of; ID, infectious disease doctors; ND, not detected (negative result); q, level of quantitation; SAg, serum antigen; SBO, small bowel obstruction; TP, true positive; UAg, urine antigen; UC, ulcerative colitis. aCharacteristics of the patients with discordant paired UAg and SAg results include the following: diagnostic test results for histoplasmosis, including UAg and SAg detection, fungal culture, and serologies; the patient’s presenting condition per hospital discharge summary; and relevant patient comorbidities, including immunosuppressive conditions (cancer, autoimmune disease, human immunodeficiency virus/AIDS) and kidney disease, which may affect the quantity of Histoplasma antigen into the urine. In addition, immunosuppressive therapies (chemotherapy, disease-modifying antirheumatic drugs, biologic agents, and greater than 3 weeks of prednisone) are listed. The criteria presented for or against the diagnosis of Histoplasma infection as well as the patient’s outcome are depicted in the last two columns. View Large Discussion The current practice of most physicians in the evaluation of patients suspected to have histoplasmosis is to perform the antigen detection test on a urine specimen, accompanied by additional laboratory testing such as culture, serology, or histopathologic studies in certain cases. Sometimes, Histoplasma serum antigen (SAg) detection is performed concurrently with a urine antigen (UAg) test. Multiple studies of the sensitivity and specificity of UAg and SAg detection have found that the assay performance varies based on the nature of the infection Table 2. The sensitivity of UAg detection appears to be higher than SAg for disseminated infections, while there is not a significant difference between UAg and SAg detection for APH. A recent case report from the Histoplasma reference laboratory indicates that testing both urine and serum provides the highest sensitivity for Histoplasma detection.25 Current Centers for Disease Control and Prevention guidelines state that urine and/or serum antigen detection is the most sensitive and specific method to diagnose histoplasmosis.26 Due to the varied possible presentations of Histoplasma infection, head-to-head comparison of UAg and SAg detection on the same patients provides evidence to inform clinical practice—specifically, whether both UAg and SAg detection tests should be ordered, or whether one of these alone (most commonly UAg alone) would suffice. There are few such data, although Hage et al24 reported antigenemia in 100% of 31 patients with disseminated histoplasmosis, with all but one of these testing positive for antigenuria. Through this large retrospective analysis, we provide evidence that supports that a single antigen detection test is appropriate in the initial evaluation for histoplasmosis. Table 2 Sensitivities of UAg and SAg Detection Infection  UAg Sensitivity, %  SAg Sensitivity, %  Brief Description  Study  Disseminated histoplasmosis with AIDS  95  86  80 patients, 79 UAg and 63 SAg tests  Williams et al17  Disseminated histoplasmosis without AIDS  81  64  28 patients; 17 with immunosuppressive conditions besides AIDS; 27 UAg and 11 SAg tests  Williams et al17  Disseminated histoplasmosis with and without AIDS  91.8  —  158 patients  Hage et al24  APH  75  0  8 cavers  Davies18  APH  65  69  130 patients  Swartzentruber et al20  APH  83.3  —  6 patients  Hage et al24  Subacute pulmonary histoplasmosis  30.4  —  46 patients  Hage et al24  Chronic pulmonary histoplasmosis  87.5  —  7 patients  Hage et al24  Disseminated or pulmonary histoplasmosis  79  82  Meta-analysis of 1,029 patients  Fandiño-Devia et al21  Infection  UAg Sensitivity, %  SAg Sensitivity, %  Brief Description  Study  Disseminated histoplasmosis with AIDS  95  86  80 patients, 79 UAg and 63 SAg tests  Williams et al17  Disseminated histoplasmosis without AIDS  81  64  28 patients; 17 with immunosuppressive conditions besides AIDS; 27 UAg and 11 SAg tests  Williams et al17  Disseminated histoplasmosis with and without AIDS  91.8  —  158 patients  Hage et al24  APH  75  0  8 cavers  Davies18  APH  65  69  130 patients  Swartzentruber et al20  APH  83.3  —  6 patients  Hage et al24  Subacute pulmonary histoplasmosis  30.4  —  46 patients  Hage et al24  Chronic pulmonary histoplasmosis  87.5  —  7 patients  Hage et al24  Disseminated or pulmonary histoplasmosis  79  82  Meta-analysis of 1,029 patients  Fandiño-Devia et al21  APH, acute pulmonary histoplasmosis; SAg, serum antigen; UAg, urine antigen. View Large We found a 98% agreement between 601 paired urine and serum specimens, with 542 uniformly negative tests and 48 uniformly positive tests. There were only 11 instances wherein the paired results were discrepant, and medical records were available for review for eight of these patients. Six patients had positive UAg results but negative SAg results and negative corresponding cultures. Based on medical record review, which disclosed an extensive workup, we determined that each of these results was a false positive. We reviewed the medical records of eight of the 11 patients with discordant paired results (Table 1). As may be predicted, UAg detection, the test with higher sensitivity, had six results characterized as false positives after medical record review. The validity of the characterization of the remaining discordant cases is arguable; one was characterized as UAg true positive/SAg false negative, whereas the other was characterized as UAg false negative/SAg true positive. The UAg true-positive/SAg false-negative result occurred in a patient who had disseminated histoplasmosis 2 years before with a residual pulmonary lesion that was thought to be largely resolved; it was hypothesized that the pulmonary lesion was either chronically active or shedding antigen. The UAg false negative/SAg true positive involved a patient with fever of unknown origin who had negative fungal blood cultures but was treated with itraconazole based on a positive SAg detection result and slowly improved. This patient had a remote history of Histoplasma endocarditis but was not found to have infective endocarditis during this illness. Overall, UAg and SAg detection performed comparatively, with a greater number of false positives produced by the UAg test, which is considered a useful characteristic of a screening test. The discordant quantitative UAg and SAg results were not all below the limit of quantification or extremely low values compared with the dual positives, indicating that the actual value of a low positive test may serve as a guide but not a direct correlate of the likelihood of histoplasmosis (Figure 1). Our 11 discordant pairs resulted between 2012 and 2016. Our results extend to 1997, but nearly all of the paired specimens were ordered between 2009 and the present, indicating that paired testing, rather than discordant results, likely increased during this time period. The reason for the increase in ordering paired testing is likely multifactorial and may include increased awareness of both tests. In 2009, our institution transitioned to electronic ordering. There are conflicting results on the effect of electronic ordering systems on the number of duplicate orders; this timing is suspicious, and the ease of ordering the studies may have in part contributed to increased dual testing.27,28 Most of the Histoplasma antigen detection tests in this study were sent to MiraVista Diagnostics. However, from January 23, 2014, to August 28, 2015, a validated, laboratory-developed test was performed on site. During this time, 12 patients with moderately positive Histoplasma UAg test results subsequently received repeat UAg testing with SAg testing later in the same week. The paired SAg result and a repeat UAg test were negative upon repeat testing. Upon review of the medical records, we found that the infectious disease team recorded its impression that the on-site laboratory-developed assay had a higher false-positive rate and therefore conducted repeat testing on patients at a higher frequency. No similar patterns were found outside of the time period that the in-house assay was in use. The repeat UAg and subsequent SAg tests on these 12 patients were performed within the same week and therefore were also included as concurrent negative paired specimens in our analysis. These results do not show an absolute superiority of UAg over SAg, but the quantitative results of the dually positive specimens (Figure 1) suggest a superior sensitivity of the UAg test over the SAg test. In addition, this review demonstrates that there was no apparent benefit of ordering both tests simultaneously, since the results usually agreed. A practical and cost-effective clinical approach would be to limit initial studies to UAg testing, as most physicians currently do, and to follow up positive tests with a confirmatory SAg study, as well as reviewing concomitantly ordered (if the clinical suspicion warrants) additional studies, such as culture, serology, and histopathologic/cytologic assessments. Further analysis of the 601 paired specimens reveals that a minority of the second tests, nine of 601, were ordered for a documented clinical reason such as confirmation of the first result. We found that most (82%) of the paired tests were ordered by the same physician, and most (69%) were ordered in the inpatient setting. This provides initial guidance on how to target efficient test utilization awareness, but future studies should characterize the population on which paired specimens are ordered to better direct these efforts. Test overutilization leads to increased strain on the health care system, a decrease in patient satisfaction, and even physical and psychological harm to the patient.29,30 Based on the Medicare reimbursement rate for both UAg and SAg (ie, $17.52 each), a $10,687 savings could have been achieved if one rather than both of the 601 paired tests had been ordered; this reduction would not have been associated with misdiagnoses or patient harm. Systematically adopting the already widely used policy of performing a UAg detection test alone rather than combined UAg/SAg testing is both clinically acceptable and fiscally responsible. References 1. Wheat LJ, Azar MM, Bahr NC et al.   Histoplasmosis. Infect Dis Clin North Am . 2016; 30: 207- 227. Google Scholar CrossRef Search ADS PubMed  2. Kauffman CA. Histoplasmosis: a clinical and laboratory update. Clin Microbiol Rev . 2007; 20: 115- 132. Google Scholar CrossRef Search ADS PubMed  3. Gustafson TL, Kaufman L, Weeks R et al.   Outbreak of acute pulmonary histoplasmosis in members of a wagon train. Am J Med . 1981; 71: 759- 765. Google Scholar CrossRef Search ADS PubMed  4. Ward JI, Weeks M, Allen D et al.   Acute histoplasmosis: clinical, epidemiologic and serologic findings of an outbreak associated with exposure to a fallen tree. Am J Med . 1979; 66: 587- 595. Google Scholar CrossRef Search ADS PubMed  5. Assi M, Martin S, Wheat LJ et al.   Histoplasmosis after solid organ transplant. Clin Infect Dis . 2013; 57: 1542- 1549. Google Scholar CrossRef Search ADS PubMed  6. Nacher M, Adenis A, Mc Donald S et al.   Disseminated histoplasmosis in HIV-infected patients in South America: a neglected killer continues on its rampage. Plos Negl Trop Dis . 2013; 7: e2319. Google Scholar CrossRef Search ADS PubMed  7. Wheat LJ, Slama TG, Norton JA et al.   Risk factors for disseminated or fatal histoplasmosis: analysis of a large urban outbreak. Ann Intern Med . 1982; 96: 159- 163. Google Scholar CrossRef Search ADS PubMed  8. Wheat LJ, Wass J, Norton J et al.   Cavitary histoplasmosis occurring during two large urban outbreaks: analysis of clinical, epidemiologic, roentgenographic, and laboratory features. Medicine (Baltimore) . 1984; 63: 201- 209. Google Scholar CrossRef Search ADS PubMed  9. Kauffman CA. Histoplasmosis. In: Crow MK, Doroshow JH, Drazen JM, et al, eds. Goldman-Cecil Medicine . 2nd ed. Philadelphia, PA: Elsevier/Saunders; 2016:2070-2072. 10. Kaufman L. Serodiagnosis of fungal disease. In: Rose NR, Friedman H, eds. Manual of Clinical Immunology . Washington, DC: American Society for Microbiology; 1976: 363-3 81. 11. Picardi JL, Kauffman CA, Schwarz J et al.   Detection of precipitating antibodies to Histoplasma capsulatum by counterimmunoelectrophoresis. Am Rev Respir Dis . 1976; 114: 171- 176. Google Scholar PubMed  12. Wheat LJ, Kohler RB, Tewari RP. Diagnosis of disseminated histoplasmosis by detection of Histoplasma capsulatum antigen in serum and urine specimens. N Engl J Med . 1986; 314: 83- 88. Google Scholar CrossRef Search ADS PubMed  13. Wheat LJ. Improvements in diagnosis of histoplasmosis. Expert Opin Biol Ther . 2006; 6: 1207- 1221. Google Scholar CrossRef Search ADS PubMed  14. Koepsell SA, Hinrichs SH, Iwen PC. Applying a real-time PER assay for Histoplasma capsulatum to clinically relevant formalin-fixed paraffin-embedded human tissue. J Clin Microbiol . 2012; 50: 3395- 3397. Google Scholar CrossRef Search ADS PubMed  15. Guedes HL, Guimarães AJ, Muniz Mde M et al.   PCR assay for identification of Histoplasma capsulatum based on the nucleotide sequence of the M antigen. J Clin Microbiol . 2003; 41: 535- 539. Google Scholar CrossRef Search ADS PubMed  16. Guimarães AJ, Nosanchuk JD, Zancopé-Oliveira RM. Diagnosis of histoplasmosis. Braz J Microbiol . 2006; 37: 1- 13. Google Scholar CrossRef Search ADS PubMed  17. Williams B, Fojtasek M, Connolly-Stringfield P et al.   Diagnosis of histoplasmosis by antigen detection during an outbreak in Indianapolis, Ind. Arch Pathol Lab Med . 1994; 118: 1205- 1208. Google Scholar PubMed  18. Davies SF. Serodiagnosis of histoplasmosis. Semin Respir Infect . 1986; 1: 9- 15. Google Scholar PubMed  19. Wheat J, French ML, Kohler RB et al.   The diagnostic laboratory tests for histoplasmosis: analysis of experience in a large urban outbreak. Ann Intern Med . 1982; 97: 680- 685. Google Scholar CrossRef Search ADS PubMed  20. Swartzentruber S, Rhodes L, Kurkjian K et al.   Diagnosis of acute pulmonary histoplasmosis by antigen detection. Clin Infect Dis . 2009; 49: 1878- 1882. Google Scholar CrossRef Search ADS PubMed  21. Fandiño-Devia E, Rodríguez-Echeverri C, Cardona-Arias J et al.   Antigen detection in the diagnosis of histoplasmosis: a meta-analysis of diagnostic performance. Mycopathologia . 2016; 181: 197- 205. Google Scholar CrossRef Search ADS PubMed  22. LeMonte A, Egan L, Connolly P et al.   Evaluation of the IMMY ALPHA histoplasma antigen enzyme immunoassay for diagnosis of histoplasmosis marked by antigenuria. Clin Vaccine Immunol . 2007; 14: 802- 803. Google Scholar CrossRef Search ADS PubMed  23. Wheat LJ, Conces D, Allen SD et al.   Pulmonary histoplasmosis syndromes: recognition, diagnosis, and management. Semin Respir Crit Care Med . 2004; 25: 129- 144. Google Scholar CrossRef Search ADS PubMed  24. Hage CA, Ribes JA, Wengenack NL et al.   A multicenter evaluation of tests for diagnosis of histoplasmosis. Clin Infect Dis . 2011; 53: 448- 454. Google Scholar CrossRef Search ADS PubMed  25. Wheat LJ. Diagnosing histoplasmosis by antigen detection . Available at http://miravistalabs.com/diagnosing-histoplasmosis-by-antigen-detection/. 26. Centers for Disease Control and Prevention. Information for health professionals about histoplasmosis. 2017. Available at https://www.cdc.gov/fungal/diseases/histoplasmosis/health-professionals.html. Accessed July 4, 2017. 27. Stewart BA, Fernandes S, Rodriguez-Huertas E et al.   A preliminary look at duplicate testing associated with lack of electronic health record interoperability for transferred patients. J Am Med Inform Assoc . 2010; 17: 341- 344. Google Scholar CrossRef Search ADS PubMed  28. Kerwin TC, Leighton H, Buch K et al.   The effect of adoption of an electronic health record on duplicate testing. Cardiol Res Pract . 2016; 2016: 1950191. Google Scholar CrossRef Search ADS PubMed  29. Procop GW, Keating C, Stagno P et al.   Reducing duplicate testing: a comparison of two clinical decision support tools. Am J Clin Pathol . 2015; 143: 623- 626. Google Scholar CrossRef Search ADS PubMed  30. Procop GW, Yerian LM, Wyllie R et al.   Duplicate laboratory test reduction using a clinical decision support tool. Am J Clin Pathol . 2014; 141: 718- 723. Google Scholar CrossRef Search ADS PubMed  © American Society for Clinical Pathology, 2018. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

American Journal of Clinical PathologyOxford University Press

Published: Apr 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial