journal article
LitStream Collection
Home
Offner, Patrick J.; Moore, Ernest E.; Biffl, Walter L.
doi: 10.1001/archsurg.134.9.935pmid: 10487586
BackgroundAccumulating clinical and epidemiological evidence suggests significant gender differences in the incidence of and outcome following major infection. In a rodent model of hemorrhagic shock, investigators have shown that males manifest depressed cell-mediated immunity that is reversed by castration or pharmacologic testosterone receptor blockade. Female rats, in contrast, show enhanced immune function that is reduced to male levels by testosterone administration. This sexual dimorphism is believed responsible for the improved outcome in female mice following septic challenge.HypothesisMale gender is a risk factor for major infections following severe injury.DesignFive-year prospective cohort study ending October 1998.SettingUrban level I regional trauma center.Patients and MethodsA total of 545 trauma patients older than 15 years with an Injury Severity Score greater than 15 and survival more than 48 hours were prospectively identified and studied. Collected data included age, injury mechanism, and Injury Severity Score. Major infections, defined as pneumonia, abdominal and pelvic abscess, wound infection requiring operative debridement, and meningitis, were tabulated. The occurrence of major infections in males and females was compared using multiple logistic regression analysis.Main Outcome MeasurePostinjury major infectious complications.ResultsOf the 545 patients, 135 (24.8%) were female and 410 (75.2%) were male. Major infections occurred in 219 (40.2%) patients. Logistic regression confirmed that male gender is an independent risk factor for major infections (P=.04) after controlling for age and Injury Severity Score. Males had a 58% greater risk of developing a major infection (odds ratio, 1.58; 95% confidence interval, 1.01-2.48).ConclusionsMale gender is associated with a dramatically increased risk of major infections following trauma. This effect is most significant following injuries of moderate severity (Injury Severity Score 16-25) and persists in all age groups.DESPITE improvements in technology and critical care, sepsis and subsequent multiple organ failure (MOF) continue to be the most common cause of late postinjury death in the surgical intensive care unit (SICU).Several studies have documented depressed cell-mediated and humoral immune function following trauma and hemorrhage.This immune suppression may, in part, reflect a compensatory anti-inflammatory response attempting to control potentially destructive postinjury hyperinflammation.Recently, experimental studies have documented hormonal influences on the immune response in animals as well as in humans.Females have been observed to have more prominent hormonal and cell-mediated immune responses compared with males.Angele et alhave implicated high testosterone and low estradiol levels in the pathogenesis of immune suppression following trauma and hemorrhage using their well-established rodent model.Several clinical and epidemiological studies have corroborated better outcome from sepsis in females compared with males.Most recently, Schröder et alobserved a marked reduction in hospital mortality among female patients with surgical sepsis (26% vs 70% in males). The purpose of this study was to determine if gender differences exist in the occurrence of major infections following severe mechanical trauma using our prospective database. We hypothesized that males are at greater risk for postinjury major infections compared with females.PATIENTS AND METHODSPATIENT POPULATIONDuring the 5-year period ending October 31, 1998, data from 545 trauma patients admitted to the trauma SICU at Denver (Colo) Health Medical Center who met inclusion criteria were entered into our prospective MOF database. Inclusion criteria included an Injury Severity Score (ISS) greater than 15, age older than 15 years, and survival greater than 48 hours. Patients transferred from another facility were excluded if the transfer was made more than 24 hours after injury. The care of these patients was directed by existing protocols and supervised by 5 general surgeons with expertise in trauma and critical care. All patients were prospectively followed up until hospital discharge or death.PRIMARY OUTCOMEPatients were monitored for the development of infectious complications, which were categorized as either major or minor. Major infections included pneumonia, empyema, lung abscess, abdominal and pelvic abscess, extensive wound infection, meningitis, and other major infections. Pneumonia was diagnosed based on the following criteria: (1) infiltrate on plain radiograph persistent for more than 48 hours, (2) temperature greater than 38°C, (3) sputum Gram stain showing many polymorphonuclear leukocytes, (4) leukocytosis (white blood cell count >12.0×109/L) or leukopenia (white blood cell count <4.0×109/L), (5) blood culture positive for the same pathogen noted on sputum culture, (6) bronchoalveolar lavage quantitative culture with pathogen growth of more than 103colony-forming units per milliliter, and (7) histopathologic diagnosis (autopsy or open lung biopsy). Pneumonia was defined as one of the following combinations of these criteria: 1 + 5; 1 + at least two of 2, 3, or 4; 1 + 6 + at least one of 2, 3, 4, or 7. Pneumonia was excluded if there was clinical resolution without antimicrobial therapy or when an alternative diagnosis was established (clinically or at autopsy). Lung abscess was diagnosed on the basis of clinical and radiographic evidence. Empyema and abdominal abscesses were defined as purulent fluid collections requiring drainage. Major wound infections were those that required operative debridement. Meningitis was diagnosed using Centers for Disease Control and Prevention criteria.Other infections were classified as major if they were associated with septic shock (eg, urosepsis).STATISTICAL ANALYSISThe MOF database is maintained on an IBM PC using Microsoft Access 97 software (Microsoft, Inc, Redmond, Wash). Statistical analysis was performed using SPSS for Windows 9.0 software (SPSS Inc, Chicago, Ill). Univariate analysis was performed using the χ2test with Yates correction or Fisher exact test for categorical data and Student ttest for continuous variables. P<.05 was considered significant.Because infection risk is possibly related to injury severity as well as the hormonal milieu (which changes with age), data were analyzed following stratification by both age and ISS. Multiple logistic regression analysis was performed to assess gender as an independent risk factor for postinjury infectious complications after controlling for age and ISS.RESULTSDuring the 5-year study period, 545 patients were identified. There were 410 male and 135 female patients. Demographic and outcome data are presented in Table 1.Table 1. Patient Demographics and Outcome*All Patients (N = 545)Male Patients (n = 410)Female Patients (n = 135)PAge, y36 ± 0.735 ± 0.740 ± 1.6.001ISS27 ± 0.426 ± 0.528.9 ± 1.0.03Mechanism of injury (blunt)73.567.888.9<.001Base deficit, mEq/L8.0 ± 0.28.1 ± 0.37.6 ± 0.5.30RBC transfusion, U in first 12 h5.3 ± 0.45.8 ± 0.54.2 ± 0.7.10Major infections404132.55MOF16.015.916.3.89Hospital length of stay, d17.7 ± 0.717.9 ± 0.817.3 ± 1.4.74SICU length of stay, d14.3 ± 1.514 ± 1.716 ± 3.63Mortality9.49.39.6.87*Data are given as mean ± SEM or percentage of patients. ISS indicates Injury Severity Score; RBC, red blood cell; MOF, multiple organ failure; and SICU, surgical intensive care unit.Female patients were older, more severely injured, and had a higher incidence of blunt injury. Major infections occurred in 41% of males compared with 37.8% of females. There were no significant gender differences in base deficit, early transfusion requirement, frequency of MOF, or mortality. Similarly, hospital and SICU lengths of stay were not significantly different in men and women.The frequency (percentage of patients) of major infections stratified by age and gender is summarized in the following tabulation.Age CategoryISS 16-25ISS >2516-45 yearsMale2655Female145646-65 yearsMale4365Female2543>65 yearsMale5064Female2571In moderately injured patients (ISS 16-25), males experienced nearly double the rate of major infection compared with females. This effect was consistent across all age categories. In severely injured patients (ISS >25), infection rates were similar regardless of gender and age category.Multiple logistic regression analysis was performed with age, ISS, and gender as independent risk factors for major infections (Table 2). Each of the variables was a significant independent risk factor. Males were noted to have a 58% greater risk of developing a major infection compared with females (odds ratio, 1.58; P=.04). There was no significant interaction between age and gender. In addition, mechanism of injury was not a significant risk factor for postinjury infections.Table 2. Logistic Regression Analysis of Risk Factors for Major Infections Following TraumaVariableCoefficientOdds Ratio (95% Confidence Interval)PAge0.01591.02 (1.00-1.03).008Injury Severity Score0.08251.09 (1.06-1.11)<.001Male sex0.45901.58 (1.01-2.48).04COMMENTImmunosuppression following trauma has been well described and increases susceptibility to sepsis and subsequent MOF.Although the exact order of mechanistic events is not known, it is clear that some aspects of the immune system are stimulated while others are depressed. Moreover, dysfunctional immune response involves virtually every known participating cell line. We have demonstrated neutrophil priming for superoxide release within 6 hours of injury that returns to baseline by 24 hours.Subsequently, neutrophil superoxide production is decreased below baseline for up to 5 days. Elastase release, however, is preserved during this period. The result is that neutrophil oxidative defense mechanisms are impaired, placing the patient at risk for invasive infection. The capacity for neutrophil-mediated tissue injury via elastase release, however, is intact. Monocytes and macrophages demonstrate immediate excessive release of proinflammatory cytokines followed by substantial paralysis of certain cellular responses.In a well-established model of hemorrhagic shock, Wichmann and colleagueshave shown impaired production of splenocyte interleukin (IL) 2 and IL-3 in response to concanavalin A, and reduced IL-1 and IL-6 production by lipopolysaccharide-stimulated peritoneal and splenic macrophages. The depressed splenocyte IL-2 and IL-3 response is consistent with a decrease in the ratio of type 1 (TH1) and type 2 (TH2) helper T cells. Such a change in the relative contributions of TH1 (proinflammatory cytokine release) and TH2 (immunosuppressive cytokine release, such as IL-4 and IL-10) subpopulations would be expected to shift the balance toward immunosuppression. In fact, the authorshave previously implicated IL-4 and IL-10 in the immunosuppression secondary to hemorrhagic shock. Similarly, Mack et alreported depressed IL-2 and augmented IL-4 and IL-10 production from splenocytes isolated from mice subjected to hemorrhagic shock and femoral fractures. Recently, Lyons et aldocumented exaggerated IL-10 release from peripheral blood mononuclear cells harvested from patients 7 to 10 days following mechanical or thermal injury, and confirmed the pivotal role of IL-10 in predisposition to infection in a murine burn model. Much attention has also focused on the excessive release of prostaglandin E215 and down-regulation of monocyte class II major histocompatibility complex HLA-DR.The impaired release of IL-6 from tissue macrophages is intriguing, particularly considering that the same environmental conditions induce a shift to the TH2 helper T cell that responds with enhanced IL-6 production. Moreover, Faist et alreported that peripheral blood mononuclear cells from severely injured patients had enhanced IL-6 release but impaired IL-1 following lipopolysaccharide exposure. The precise role of IL-6 in regulating local and/or systemic inflammation remains unclear.Interleukin 6 is recognized as an integral mediator of the acute-phase response, but excessive or prolonged increases of circulating IL-6 concentration have been associated with infectious complications and MOF in patients after trauma, burns, or elective surgery.Cell-mediated immune responses appear to exhibit sexual dimorphism. A century ago, it was observed that castration in adult male rabbits increased thymic mass, suggesting that sex hormones influenced the immune system.Subsequently, experimental studies have shown that humoral and cell-mediated immune responses are more prominent in females.It has been proposed that maintenance of immune integrity in females following injury is due, at least in part, to the absence of immunosuppressive effects by androgenic hormones.In a murine model of trauma-hemorrhage, Wichmann et aldemonstrated that castration prior to injury prevented posttrauma immune depression. Castration 2 weeks prior to trauma-hemorrhage reduced plasma testosterone levels and preserved splenocyte IL-2 and IL-3 release compared with sham-operated animals. This laboratory also observed that pretreatment of female mice with 5α-dihydrotestosterone resulted in immunosuppression following trauma-hemorrhage similar to that occurring in males.The authors proposed that the balance between testosterone and estradiol may be responsible for the immune depression seen in male mice after injury.Several clinical studies suggest gender differences in the incidence and outcome from sepsis.In an epidemiological study of bacteremia, McGowan et alnoted a higher incidence of sepsis in males compared with females. In a pediatric population with severe burn injury, Barrow and Herndonnoted increased mortality in boys compared with girls. Boneretrospectively reviewed 4 severe sepsis studies and noted a predominance of male patients (60%-65%). Most recently, Schröder et alprospectively observed 52 patients (19 female, 33 male) with surgical sepsis admitted to a university hospital SICU. The male and female patients were well matched for age, cause of sepsis, Acute Physiology and Chronic Health Evaluation II (APACHE II) score, and multiple organ dysfunction score. Male patients experienced a significantly higher mortality rate (70% vs 26%; P<.008). The authors did not comment on the incidence of sepsis in these patients.Our prospective study adds documentation that males are at increased risk for developing major infections following injury. The male and female patients in our study were well matched for transfusion requirements and early base deficit. Mortality, MOF, and hospital and SICU lengths of stay were not affected by gender. Although female patients were older and more severely injured, multivariate analysis confirmed the increased risk in males after adjusting for these differences. The lack of a difference in severely injured patients (ISS >25) is not unexpected. Severely injured patients with massive soft tissue injury are at such high risk of infectious complications that gender differences become less important. We were surprised, however, that the increased risk of infection in males persisted in the postmenopausal age group, when hormonal differences were believed to be less operative. An explanation for this unexpected finding may be provided by Schröder et al,who noted increased estradiol levels in both men and postmenopausal women with sepsis. The source of estradiol in these patients was postulated to be from conversion of testosterone or decreased hepatic estrogen catabolism related to sepsis.In summary, we have shown that males are at increased of infection following trauma. This sexual dimorphism has therapeutic implications and must be taken into account when evaluating and planning interventional sepsis trials.LManshipRDMcMillinJJBrownThe influence of sepsis and multi-system organ failure on mortality in the surgical intensive care unit.Am Surg.1984;50:94-101.ASauaiaFAMooreEEMooreEpidemiology of trauma deaths: a reassessment.J Trauma.1995;38:185-193.IHChaudryAAyalaWErtelRNStephenHemorrhage and resuscitation: immunological aspects.Am J Physiol.1990;259:R663-R678.RNStephanTSKupperASGehaASBaueIHChandryHemorrhage without tissue trauma produces immunosuppression and enhances susceptibility to sepsis.Arch Surg.1987;122:62-68.RZellwegerAAyalaCMDeMasoIHChaudryTrauma-hemorrhage causes prolonged depression in cellular immunity.Shock.1995;4:149-153.ASAnsarWJPenhaleNTalalSex hormones, immune responses and autoimmune diseases: mechanisms of sex hormone action.Am J Pathol.1985;121:531-551.CJGrossmanInteractions between the gonadal steroids and the immune response.Science.1985;227:257-261.CJGrossmanGARoselleThe control of immune response by endocrine factors and the clinical significance of such regulation.Proc Clin Biochem Med.1986;4:9-56.MKAngeleAAyalaBAMonfilsWGCioffiKIBlaudIHChaudryTestosterone and/or low estradiol: normally required but harmful immunologically for males after trauma-hemorrhage.J Trauma.1998;44:78-84.JEMcGowanMWBarnesNFinlandBacteremia at Boston City Hospital: occurrence and mortality during 12 selected years (1935-1972) with special reference to hospital-acquired cases.J Infect Dis.1975;132:316-335.RCBoneToward an epidemiology and natural history of SIRS (systemic inflammatory response syndrome).JAMA.1992;268:3452-3455.JSchröderVKahlkeKHStaubachPZabelFStüberGender differences in human sepsis.Arch Surg.1998;133:1200-1205.JSGarnerWRJarvisTGEmoriTCHoranJMHughesCDC definitions for nosocomial infections, 1988.Am J Infect Control.1988;16:128-140.FAMooreASauaiaEEMooreJBHaenelJMBurchDCLezottePostinjury multiple organ failure: a bimodal phenomenon.J Trauma.1996;40:501-512.EFaistCSchinkelSZimmerUpdate on the mechanisms of immune suppression of injury and immune modulation.World J Surg.1996;20:454-459.AJBothaFAMooreEEMooreFJKimABanerjeeVMPetersonPostinjury neutrophil priming and activation: an early vulnerable window.Surgery.1995;118:358-365.EFaistMStorchLHueltnerFunctional analysis of monocyte (Mø) activity via synthesis patterns of interleukin 1, 6, 8 (IL-1, IL-6, IL-8) and neopterin in surgical intensive care patients.Surgery.1992;112:562-572.MWWichmannAAyalaIHChaudrySevere depression of host immune functions following closed-bone fracture, soft-tissue trauma, and hemorrhagic shock.Crit Care Med.1988;26:1372-1378.AAyalaDLLehmanCDHerdonMechanism of enhanced susceptibility to sepsis following hemorrhage: interleukin-10 suppression of T-cell response is mediated by eicosanoid-induced interleukin-4 release.Arch Surg.1994;129:1172-1178.VEMackMDMcCarterHANaamaDominance of T-helper 2-type cytokines after severe injury.Arch Surg.1996;131:1303-1309.ALyonsJLKellyMLRodrickMajor injury induces increased production of interleukin-10 by cells of the immune sytem with a negative impact on resistance to infection.Ann Surg.1997;226:450-460.WGCheadleMJHershmanSRWellhausenHLA-DR antigen expression on peripheral blood monocytes correlates with surgical infection.Am J Surg.1991;161:639-645.WLBifflEEMooreFAMooreInterleukin-6 in the injured patient—marker of injury or mediator of inflammation?Ann Surg.1996;224:647-664.ACalzolariRecherches experimentales sur un rapport probable entre la function du thymus et celle des testicules.Arch Ital Biol.1898;30:71-77.MWWichmannRZellwegerCMDeMasoAAyalaIHChaudryMechanism of immunosuppression in males following trauma-hemorrhage.Arch Surg.1996;131:1186-1192.REBarrowDNHerndonIncidence of mortality in boys and girls after severe thermal burns.Surgery.1990;170:295-298.Philip S. Barie, MD, New York, NY:Laboratory animal studies indicate that male animals are more likely to die after standardized injury plus infection models, and in vitro work suggests that androgens may have immunosuppressive effects. Clinical studies, however, are conflicting, and the issue must be resolved if we are to make progress with future clinical trial design, or even to prognosticate rationally about our own patients. Offner and colleagues report that male trauma patients are at increased risk to develop infection. However, an analysis of nearly 19,000 patients by Greco et al from the shock trauma unit at the University of Maryland, presented last January before the Eastern Association for the Surgery of Trauma, showed no difference in mortality despite a higher incidence of pneumonia in male patients. The present study shows no difference in mortality as well, raising the questions as to whether truly virulent infections are being observed or whether colonization generally is often misattributed to invasive infection in studies of this type. I have several questions for the authors.First, most large series of infection in trauma patients indicate an overall incidence of about 25%, yet yours was 40%. Is the difference attributable entirely to severity of illness in your selected population or are there other factors?Have you validated your definitions of pneumonia in other studies? In particular, I am concerned about 1 of your 2 sets of criteria in which you make the diagnosis of pneumonia based on infiltrate plus 2 of fever, Gram stain, or leukocytosis. This has been shown in studies by Meduri et al and Croce et al to yield an incidence of pneumonia of only 40% when bronchoscopic confirmation methods are employed.It appears that bacteremia may not have been accounted for, especially if shock was absent. Considering how major the problem of catheter-related infection is for all of us, why not include all infections without exception in your analysis? Can you tell us whether the infection case mix was equal between your groups?How did you construct your multivariate model to evaluate 3 independent variables only? Why was blunt trauma mechanism excluded when it was highly significant in the univariate analysis? What was the univariate Pvalue for the incidence of major infections that were, in fact, more than 20% more common in your male patients? If that Pvalue was .15 or less, many would have included it in the multivariate model as well. Inclusion of another significant independent variable might change your result, considering that the number of independent variables examined was only 3.Dr Offner:Your first question regarding our increased incidence of infections: This is due to the fact that we have a highly selective patient population, with a fairly severely injured patient population, compared with other studies, which include a greater range of injury severity in their patient population.Our pneumonia definition has been used in a variety of subsequent studies. As far as independent validation, that has not actually been done; however, I do think it is a fairly standard definition based on a number of facets as elucidated in the presentation. I didn't present all infections; however, I did analyze all infections. Male gender is still a risk factor for all infections, including minor infections as well.As for the infection case mix, I assume you mean whether or not it was pneumonia vs intra-abdominal abscess, etc. Those data are available, however, because of the nature of how our database is constructed. I have had great difficulty actually sorting that out. I can tell you that the predominant major infection that we do have in our intensive care unit is pneumonia, and again, that definition of pneumonia is equally applied to men and women; it is equally applied to all patients in that SICU.And finally, as far as a logistic regression model, more variables than those that were shown in the presentation were included; however, they were not significant, and that includes mechanism of injury, among others, that in the interest of time I chose not to present.Edwin Deitch, MD, Newark, NJ:This will relate to your outcome data. We know that it is very difficult to tie mortality to infection in the ICU patient, because some patients are very severely injured and they will die whether or not they get infected, while in others with more moderate injuries, the infection may tip them one way or the other. Your data show that it's in the intermediate level of injury that there is an association with a reduction in infection. You didn't show any data in that subgroup about the effect of infection on outcome or survival, so your survival studies were just the entire populations.In the subgroup of patients with ISSs between 16 and 25, where infection was reduced, did you see a concomitant improvement in outcome?Secondly, as a related question, people die of different things. Have you stratified your data to look at infectious death vs noninfectious death to determine if there is a difference between the sexes?And lastly, although we think it is androgens, some of your patients were postmenopausal and therefore they had different hormonal milieu. How do you account for that in your results?Dr Offner:We did not do subgoup analysis within the strata in terms of a statistical analysis. I did look at the numbers, and they would not be statistically significant. You are right, a subgroup analysis would be appropriate in that subgroup that appears to have the potential effect.Jan K. Horn, MD, San Francisco, Calif:I was wondering about the use of antibiotics in your patients. Can you assure us, for instance, that they all have the same preinfection regimen of antibiotics? That would have a significant impact on the incidence, since you did have a higher rate of infection.Dr Offner:That is a very good point, and I cannot document that with any surety. What I can tell you is that the care of the patients in the ICU is done on a fairly rigid protocol basis under the direction of a few surgeons who think alike, so most of these patients are taken care of in a uniform fashion, with rounds made on a daily basis by us in order to assure that that is being done.Mark A. Malangoni, MD, Cleveland, Ohio:You had a group of patients that would consist of older men and postmenopausal women who had the same infection risk differences as your younger groups of patients. We are looking for an explanation as to why that same relative risk exists in that group of people.In addition, your relative risk 95% confidence intervals overlap. You have got a few decimal points out there that aren't convincing, and I wonder if you could comment on how much you believe these statistics?Can you tell us about the incidence of penetrating injuries in these patients, or, for that matter, contaminated injuries? I would expect that one of the things that your data may show is that men may be more likely to have injuries that result in some degree of contamination and perhaps it is the contamination that leads them to develop later infection.Dr Offner:Let me try to address the first question first before I forget it. We were surprised when we found that the infection risk in the moderately injured patients was persistent through all age groups, because we expected that if this was related to hormonal differences, we would not see a difference when we reached that postmenopausal age group. So we were surprised by that. I don't have a good explanation for it except if you go back to the report by Schroder that was presented here last year, they documented elevated estradiol levels in both men and women in that postmenopausal age group and surmised that it was either related to septic-related hepatic dysfunction or some other factor.There was also an epidemiologic study of postmenopausal women that was performed about 10 years ago and published in the American Journal of Epidemiologythat looked at hormone levels, and what they found was hormone levels could not be predicted by age, rather, they seemed to be affected by other factors, such as obesity, whether someone smoked or not, and other concurrent problems. That may potentially explain some of these differences.The reason why we don't see an effect in the more severely injured patients is that it is possible that at some threshold of injury severity, tissue injury effects outweigh any gender effects that occur.There was a significant difference in penetrating injuries between the men and women; however, in the logistic regression analysis, a variable that included mechanism of injury was included and was not found to be significant in that regard. However, I did not individually look at penetrating injuries and the body region of injury to see if there was a difference in that regard; it is probably a good thing to do.David D. Dunn, MD, Minneapolis, Minn:This paper obviously deflates many of our societal and personal views regarding testosterone, unfortunately. Having said that, I would like to challenge your conclusion regarding whether this is the important variable in sepsis and infection, and point out that there are probably 3 other important papers that have appeared in the last 4 to 5 years—one telling us that tumor necrosis factor polymorphism is critical, another telling us that the IgG response to anticore antibody, coupled with IL-6, is important, and another telling us that resistant organisms are associated with higher mortality. What is the right answer here, and do you plan to measure other variables, as Dr Barie mentioned?Dr Offner:Clearly, this is a complicated area, and there probably are other effects that aren't accounted for, including genetic effects, as you mentioned. We are currently trying to collate some data on neutrophil function between men and women, as well as cytokine levels, but that is in progress, and really we have few patients right now. Hopefully, we will have more as time progresses.Irshad H. Chaudry, MD, Providence, RI:With regard to the lack of difference in survival rates between the males and females, can you tell us about the state of the menstrual cycle in the females? In this regard, our experimental studies have shown that females in the proestrus state tolerate trauma and sepsis better than males. However, the tolerance to trauma and sepsis was not different between male and females in the metestrus cycle. It is important to subgroup female patients in the different states of menstrual cycle and determine whether the different cycle states provide any difference in tolerance to trauma and sepsis.My second question deals with the older postmenopausal female patients. Could you please tell us if any of them were receiving estrogen replacement therapy? Did you stage the female patients with regard to what cycle?Dr Offner:These are essentially the same questions you asked Dr Schroder a year ago. Unfortunately, I don't have a good answer. When these major injured patients come in, we usually don't try to get a menstrual history from them, and it is not something that we typically put in our database, so we really don't have a good idea as far as where they were in their cycle, nor do we have a good idea as to whether or not they were taking estrogen replacement. We try to take a good medical history on these patients, but, unfortunately, it is not one of the fields that we include in our database, and so I don't have that information for you.Presented at the 19th Annual Meeting of the Surgical Infection Society, Seattle, Wash, April 29, 1999.Corresponding author: Patrick J. Offner, MD, MPH, Department of Surgery, MC 0206, Denver Health Medical Center, 777 Bannock St, Denver, CO 80204 (e-mail: [email protected]).
Di Fronzo, L. Andrew; Cymerman, Judith; O'Connell, Theodore X.
doi: 10.1001/archsurg.134.9.941pmid: 10487587
HypothesisIf factors accounting for the inability to tolerate early postoperative feeding after elective open colon resection can be identified, then perhaps these factors can be modified to decrease future failures.DesignConsecutive case series.SettingTertiary referral center.PatientsFrom 1993 to 1998, 200 consecutive patients undergoing elective open colon resection.InterventionEarly postoperative feeding protocol consisting of clear liquids on the evening of postoperative day 2, regular diet on postoperative day 3, and discharged home as tolerated. A subgroup of patients was treated with metoclopramide.Main Outcome MeasuresThe ability to tolerate early feeding. Postoperative complications. Length of hospitalization.ResultsTwenty-seven (13.5%) of the 200 patients failed to tolerate early feeding. 16 patients (8.0%) were immediately unable to tolerate oral intake, whereas 11 patients (5.5%) initially tolerated early postoperative feeding but required hospital readmission due to emesis. There were no abdominal abscesses or anastomotic leaks. In patients who failed early feeding, no significant differences were noted for age, comorbid medical illness, operative time, or additional surgical procedures, when compared with patients who tolerated early oral intake. However, 18 (20.9%) of the 86 men failed early feeding, compared with 5 (6.8%) of the 73 women (P=.01). Additionally, patients undergoing total abdominal colectomy or total proctocolectomy (n=11) failed 45.5% of the time, compared with 12.2% of the patients undergoing other types of colectomy (n=189) (P=.01). The addition of metoclopramide therapy did not significantly improve the ability to tolerate early feeding.ConclusionsIn patients undergoing elective open colon resection, early postoperative feeding is safe and effective, and produces a brief hospital stay compared with patients fed by traditional means. However, men and patients undergoing total abdominal colectomy are more likely to be intolerant of early postoperative feeding.THE TRADITIONAL approach to start postoperative feeding following elective colon resection has been to await the resolution of postoperative adynamic ileus, as indicated by the presence of bowel sounds and the passage of flatus and/or bowel movement. However, recent clinical trials of patients undergoing laparoscopic or laparoscopic-assisted colectomy, with feeding initiated by protocol rather than by objective signs of return of bowel function, resulted in earlier feeding and shortened hospital stay.These advantages initially were believed to be unique to laparoscopic colectomy, owing to smaller incisions and less manipulation of the gastrointestinal tract.More recently, numerous clinical trials examining the feasibility of early postoperative feeding following open colon resection,and randomized trials comparing early postoperative feeding in open and laparoscopic colectomy,have demonstrated that early postoperative feeding is equally safe and effective following open colon resection. However, questions remained: Can early postoperative feeding after open colon resection be applied to a larger group of patients when performed by multiple surgeons? What factors predisposed patients to the inability to tolerate early oral intake? In addition, we hypothesized that the addition of a gastrointestinal motility agent, metoclopramide, may improve the the ability to tolerate early feeding. This study was undertaken to answer these questions.PATIENTS AND METHODSFrom 1993 to 1998, an early postoperative feeding protocol was used for patients undergoing elective, open colon resections at Kaiser Permanente Medical Center, Los Angeles, Calif. This protocol consisted of no routine use of a postoperative nasogastric (NG) tube, patient-controlled analgesia using morphine, and initiation of a clear liquid diet on the evening of postoperative day (POD) 2. If the patient tolerated clear liquids, a regular diet was provided on POD 3, the patient-controlled analgesia was discontinued, and the patient was given oral narcotic analgesics. Patients were discharged home after 3 regular meals were tolerated, regardless of the passage of flatus and/or bowel movements. The first 50 consecutive patients were operated on by a single senior surgeon (T.X.O'C.); however, in the subsequent 150 patients, the operations were performed by multiple surgeons. For all patients, immediate postoperative complications, the ability to tolerate early feeding, and length of hospitalization were recorded. Daily interviews and examinations of patients were performed with special attention to nausea, emesis, and other indications of intolerance of early postoperative diet. Patients who experienced 2 episodes of emesis underwent NG tube insertion. Any patient who initially tolerated early postoperative feeding, but returned to the hospital within 2 weeks of discharge with emesis that required bowel rest and intravenous hydration/nutrition, was defined as a hospital readmission.The initial part of the cohort (41 patients) was designed to examine the feasibility and safety of early postoperative feeding after elective open colon resection. In the second phase of the study (159 patients), identifying the possible factors that might contribute to the inability to tolerate early postoperative feeding was emphasized. Finally, in the most recent phase of the study, 10 mg of metoclopramide, administered every 6 hours for the entire hospitalization, was incorporated into the early postoperative feeding protocol to determine if the addition of a gastrointestinal motility agent could increase the ability to tolerate early postoperative feeding.The factors analyzed for possible correlation with the patient's ability to tolerate the early postoperative feeding protocol included the following: sex, age (in years), operative time (in minutes), presence of comorbid illness, presence of associated procedures, type of colectomy, and use of metoclopramide therapy. Each factor was initially tested in a univariate logistic regression model with failure to tolerate the early postoperative feeding protocol as the outcome variable. In addition, a multivariate logistic regression model with all factors included was employed to test the significance of each factor while adjusting for all other factors. The statistical software package of S-Plus 4.5 (MathSoft Inc, Seattle, Wash) was used for all logistic regression models. The χ2and Fisher exact tests in StatXact 2.05 (CYTEL Software Corporation, Cambridge, Mass) were also used to initially test for correlations between failure to tolerate the protocol and various categorical factors.RESULTSThe early postoperative feeding protocol was used in 200 patients following elective open colon resection. The mean patient age was 63.1 years (age range, 25-88 years) for the 104 men and 96 women in the study. The procedures most commonly performed included right hemicolectomy (32.5%), low anterior resection (25%), sigmoid colectomy (14%), and left hemicolectomy (14%). Less commonly performed procedures included total abdominal colectomy or total proctocolectomy (5.5%), colostomy closure (4.5%), abdominoperineal resection (2.5%), and transverse colectomy (2%).Overall, 173 (86.5%) of the 200 patients tolerated the early postoperative feeding protocol. The mean hospital stay for these 173 patients was 3.4 days (range, 3-6 days), and 155 patients (89%) were discharged home by POD 4. Of the 27 patients who failed to tolerate early feeding, 16 patients (8.0%) experienced emesis and the inability to tolerate an early postoperative diet during the initial hospitalization. Five (31%) of these 16 patients required NG tube insertion. Their mean hospital stay was 7.5 days (range, 5-11 days); none were readmitted to the hospital. Eleven patients (5.5%) initially tolerated the early postoperative feeding protocol and were discharged home, but required later readmission due to emesis. Of these 11 patients, 6 (54.5%) required NG tube insertion. Three patients (1.5%) required 2 or more readmissions owing to persistent ileus. The mean duration of rehospitalization was 4.6 days (range, 1-9 days). The overall length of stay for these 11 patients, inclusive of all admissions, was 9.7 days (range, 4-13 days). Including the length of hospitalization time for readmissions, the mean hospital stay for all 200 patients was 4.2 days (range, 3-13 days).Five patients (2.5%) developed wound infections. One patient (0.5%) developed pancreatitis requiring readmission, and 1 patient (0.5%) developed a perineal wound dehiscence after an abdominoperineal resection. Two patients (1.0%) experienced postoperative hemorrhage: 1 patient after an abdominoperineal resection developed pelvic hemorrhage requiring a 5-U blood transfusion, and 1 patient required immediate reexploration and vessel ligation for hemoperitoneum following a total proctocolectomy. One patient developed an early postoperative small-bowel obstruction requiring reexploration. No instances were noted of anastomotic leak, abdominal abscess, or aspiration pneumonia.For the last 159 patients in the study, data were collected to identify the possible factors which might contribute to the inability to tolerate early postoperative feeding. In this group of patients, the mean operative time was 124 minutes (range, 65-305 minutes). Thirty-seven patients (23.2%) had at least 1 significant comorbid medical illness including the following: diabetes mellitus (17 patients), coronary artery disease/cardiomyopathy (16 patients), hypothyroidism (4 patients), end-stage renal disease (2 patients), and scleroderma (2 patients). Forty-three patients (27%) underwent 43 other surgical procedures simultaneously with colon resection including the following: adhesiolysis (21 patients), liver resection/wedge biopsy (10 patients), cholecystectomy (5 patients), enterectomy (2 patients), oophorectomy (2 patients), partial cystectomy (1patient), radical nephrectomy (1 patient), and radical retropubic prostatectomy (1 patient).When patients who tolerated early postoperative feeding were compared with those who did not, no significant differences were noted for age, comorbid medical illness, operative time, or additional surgical procedures. However, 18 (20.9%) of the 86 men failed early feeding, whereas only 5 (6.8%) of the 73 women did so (P=.01, χ2test). Additionally, 5 (45.5%) of the 11 patients undergoing total abdominal colectomy or total proctocolectomy failed to tolerate early postoperative feeding, compared with 13 (12.2%) of the 189 patients undergoing other types of colectomy (P=.01, Fisher exact test). Furthermore, using multivariate analysis, only 2 factors were consistently significant—sex and type of operation (total abdominal colectomy or total proctocolectomy vs other colectomy) (Table 1).Factors Associated With Feeding Intolerance*FactorEarly Feeding ProtocolPFailed (n = 23)Tolerated (n = 136)UnivariateMultivariateMean operative time, min123.3132.2.41.84Comorbid medical illness6 (26)31 (22.8).73.58Mean age, y60.063.3.26.46Associated operation7 (30)36 (26.4).69.97SexMale18/86 (20.9)68/86 (79.1).02†.001†Female5/73 (6.8)68/73 (93.2). . .. . .Type of colectomyTAC or TPC‡5/11 (45.5)6/11 (54.5).006†.007†Other18/148 (12.1)130/148 (87.8). . .. . .Use of metoclopramide therapyYes7/49 (14.2)42/49 (85.8).97.82No20/151 (13.2)131/151 (86.8). . .. . .*Values are expressed as number (percentage) or number/total population (percentage) unless otherwise indicated. Ellipses indicate not applicable.†Boldfaced values indicate statistically significant factor.‡TAC or TPC indicates total abdominal colectomy or total proctocolectomy.In the second phase of the study, 49 patients were administered metoclopramide as part of the early postoperative feeding protocol, to determine if a gastrointestinal motility agent could improve the ability to tolerate early postoperative feeding. Of these, 7 (14.2%) of the 49 patients failed to tolerate early postoperative feeding, compared with 20 (13.2%) of the 151 patients who failed early postoperative feeding and did not receive metoclopramide therapy. No significant difference was noted for the number of readmissions due to ileus (7.3% vs 6.1%), mean length of the hospitalization (3.77 vs 3.67 days), and the number of patients discharged home by POD 4 (both 87%) in patients who did not receive metoclopramide therapy vs those who did.COMMENTThe traditional approach to postoperative feeding after colon resection required waiting for resolution of postoperative ileus, as evidenced by the passage of flatus and/or bowel movement. It was believed that early postoperative feeding might actually worsen and prolong postoperative ileus, by creating a vicious cycle of increasing intraluminal secretion, distention, and progressive motor dysfunction. Most important of all was the possibility of anastomotic leak, which could result from increased intraluminal pressure producing anastomotic tension, ischemia, and eventual disruption.Based on the assumption that smaller incisions and decreased bowel manipulation would produce less ileus, laparoscopic and laparoscopic-assisted colectomy provided evidence that early postoperative feeding could be tolerated without a significant increase in complications. However, several recent trials have demonstrated the feasibility and safety of early postoperative feeding following open colon resection as well.In our 1996 study, we compared 41 patients who were managed by the early postoperative feeding protocol after elective colon surgery with 41 historical controls undergoing similar operations who were managed by traditional postoperative feeding criteria.Equivalent numbers of patients experienced postoperative ileus, yet the mean length of hospitalization was considerably shorter in the early postoperative feeding group (4.2 vs 6.7 days), and no differences were noted in postoperative complications.Following open colectomy, Binderow et alalso showed no difference in the duration of postoperative ileus between patients fed by an early postoperative feeding protocol vs traditional criteria, yet identified a trend toward decreased length of hospitalization (6.7 vs 8.0 days) in the early postoperative feeding group. Hawalsi et aldemonstrated that in patients undergoing elective "conventional" colectomy who were given an early postoperative feeding 24 hours after the operation, their hospital stay was reduced from 8 to 4 days. Reissman et alcompared 80 patients undergoing open colon resection who were managed by traditional postoperative feeding protocol with 80 patients undergoing similar open procedures but who were managed by an early postoperative feeding protocol. No significant differences were noted between the early postoperative and regular feeding groups for the rate of emesis, need for NG tube reinsertion, duration of ileus, or overall complications.In our study, no patient developed an anastomotic leak or intra-abdominal abscess, confirming our previous findings that early postoperative feeding following open colectomy does not predispose patients to an increased incidence of complications. Other studies of early postoperative feeding after open colon resection have also shown no significant increase in infection rates, anastomotic leak, or other complications.This protocol was designed to take advantage of well-recognized differences between the average time of return of gastric, small-bowel, and colonic motor activity following laparotomy. Because gastric and small-bowel motility normally return within 48 hours of operation, postoperative feeding was started then. Since colonic ileus generally lasts 3 to 5 days, solid food was given on POD 3. The proof of this protocol design is reflected in our results, with more than 80% of the patients discharged home by POD 4, and a mean initial hospital stay of only 3.8 days. Importantly, even when all readmissions were included in analyzing the overall length of hospitalization, the mean hospital stay was only 4.2 days—considerably shorter than when the traditional postoperative feeding criteria were used. More significantly, this hospital stay is shorter than that reported in more than 540 patients in multiple laparoscopic colectomy trials, in which the mean hospital stay was 6.7 days (range, 4.0-12.3 days).Although the results of our study were good, 13.5% of our patients were still unable to tolerate early postoperative feeding after elective open colon resection. In our study, we attempted to elicit the risk factors that might predispose patients to fail early postoperative feeding following open colon resection. If risk factors could be identified, perhaps appropriate alterations in management could be made to avoid feeding intolerance and, therefore, reduce the risk for potential complications, as well as costly hospital readmissions.The risk factors associated with failure to tolerate early postoperative feeding were male gender and total abdominal colectomy or total proctocolectomy. We are uncertain why male patients were significantly more likely to fail early feeding. Possible mechanisms include anatomical differences, physiological differences (eg, hormone-dependence of gastrointestinal motility), or social/cultural differences between male and female perceptions of recovery and health. If anything, anatomical differences between men and women (increased mean colon length in women) and physiological differences in gastric motility between men and women (slower gastric emptying in women), would suggest an opposite effect than was observed.Patients undergoing total abdominal colectomy or total proctocolectomy demonstrated significant inability to tolerate an early postoperative diet. The logical explanation for this finding is the amount of bowel manipulation, the length of incision, and time of exposure associated with these more extensive operations. However, time as an independent variable was not significant in our analysis, and the mean operative time for total abdominal colectomy or total proctocolectomy in this study was 158 minutes (range, 105-235 minutes), not significantly different from the mean operative time for the other colectomies (122 minutes). Additionally, other extensive procedures such as adhesiolysis, which are time-consuming and also require extensive bowel retraction and manipulation, did not seem to affect the ability to tolerate early postoperative feeding. It remains unclear why patients having total abdominal colectomy or total proctocolectomy are more intolerant of early postoperative feeding.We also hypothesized that the addition of metoclopramide therapy may speed the time to resolution of gastric ileus and allow more patients to tolerate an early postoperative diet. However, in the 49 patients receiving routine metoclopramide therapy, no difference was noted in the failure rates, length of hospitalization, or number of patients discharged home by POD 4. It seems as if the addition of metoclopramide therapy offers no advantage in the success of early postoperative feeding after elective open colectomy. In the future, however, further study with different motility agents, as well as the use of nonopiate analgesics, may be done to determine if they have any notable effect on the results of early postoperative feeding.To the best of our knowledge, this study represents the largest series of patients undergoing early postoperative feeding after open colon resection. We believe our data clearly demonstrate that early postoperative feeding and early hospital discharge following colon resection are not limited to laparoscopic techniques alone. The ideal method to compare laparoscopic vs open techniques, however, is a randomized, controlled trial that carefully incorporates identical means of postoperative feeding. Although the Phase 3 Prospective Randomized Trial Comparing Laparoscopic-Assisted Colectomy vs Open Colectomy, currently underway and accruing subjects, aims primarily at comparing survival between the 2 operative approaches, quality of life and cost-effectiveness are also important study objectives.It is important to stress that in any randomized trial examining length of hospital stay and/or costs of medical treatment following laparoscopic vs open procedures, patients must be given an identical feeding protocol to make accurate and appropriate comparisons between the 2 techniques.In patients undergoing elective open colon resection, early postoperative feeding is safe, with no evidence of increased morbidity. Early postoperative feeding is also highly effective in reducing hospital stay. For unknown reasons, men and patients undergoing total abdominal colectomy or total proctocolectomy are predisposed to feeding intolerance. Failure to tolerate early postoperative feeding does not result in increased complications or increase the overall length of hospitalization. We believe that early postoperative feeding can be safely and effectively performed in the vast majority of patients undergoing elective open colon resection.IPuenteJLSosaDSleemanUDesaiNTranakasRHartmannLaparoscopic assisted colorectal surgery.J Laparoendosc Surg.1994;4:1-7.WRPetersTLBartelsMinimally invasive colectomy: are the potential benefits realized?Dis Colon Rectum.1993;36:751-756.PADeanRWBeart JrHNelsonTDElftmannRTSchlinkertLaparoscopic-assisted segmental colectomy: early Mayo Clinic experience.Mayo Clin Proc.1994;69:834-840.JChoiTXO'ConnellSafe and effective early postoperative feeding and hospital discharge after open colon resection.Am Surgeon.1996;62:853-856.HOrtizPArmendarizCYarnozIs early postoperative feeding feasible in elective colon and rectal surgery?Int J Colorectal Dis.1996;11:119-121.PReissmanTATeohSMCohenEGWeissJJNoguerasSDWexnerIs early oral feeding safe after elective colorectal surgery: a prospective randomized trial.Ann Surg.1995;222:73-77.SRBinderowSMCohenSDWexnerJJNoguerasMust early postoperative oral intake be limited to laparoscopy?Dis Col Rectum.1994;37:584-589.AHawalsiDMSchroderLRLloydRFeatherstoneElective conventional colectomy in the era of laparoscopic surgery.Am Surgeon.1996;62:589-593.HOrtizPArmendarizCYarnozEarly postoperative feeding after elective colorectal surgery is not a benefit unique to laparoscopy-assisted procedures.Int J Colorectal Dis.1996;11:246-249.JWMilsomBBohmKAHammerhoferVFazioESteigerPElsonA prospective, randomized trial comparing laparoscopic versus conventional techniques in colorectal cancer: a preliminary report.J Am Coll Surg.1998;187:46-57.TMKhaliliPRFleshnerJRHiattColorectal cancer: comparison of laparoscopic with open approaches.Dis Col Rectum.1998;41:832-838.BPSaundersMFukumotoSHalliganWhy is colonoscopy more difficult in women?Gastrointest Endosc.1996;43:124-126.LCKnightHPParkmanKLBrownDelayed gastric emptying and decreased antral contractility in normal premenopausal women compared with men.Am J Gastroenterol.1997;92:968-975.FLDatzPEChristianJMooreGender-related differences in gastric emptying.J Nucl Med.1987;28:1204-1207.LStocchiHNelsonLaparoscopic colectomy for colon cancer: trial update.J Surg Oncol.1998;68:255-267.Bruce E. Stabile, MD, Torrance, Calif:This paper has a real potential impact for the practices of many surgeons. In this paper Drs Di Fronzo, Cymerman, and O'Connell have convincingly shown that diligent application of a clinical pathway involving early feeding can expedite patient recovery and early hospital discharge following open colon resection. Among their 200 consecutive patients, the mean total length of stay for all hospitalizations was only 4.2 days, with a range of 3 to 13 days. Only 5.5% of the patients required readmission after early discharge and their mean rehospitalization duration was only 4.6 days, with a range of 1 to 9 days. These are truly excellent results and underscore the cost effectiveness of their protocol. They were no doubt made possible in some part by the total absence of anastomotic leaks, intra-abdominal abscesses, and fatality. Nasogastric tubes were used only selectively in the 5.5% of the patients who experienced repeated emesis. Oral liquids were routinely begun on POD 2 and a regular diet on POD 3. Patient-controlled analgesia using morphine was employed until POD 3 when oral narcotics were substituted. The 86.5% overall success rate of the protocol is undeniable evidence that early postoperative feeding is safe and effective. The results also demonstrate that postoperative adynamic ileus is uncommonly so profound as to preclude early feeding and hospital discharge. Unless the patient is experiencing nausea and emesis, perhaps surgeons should keep their stethoscopes in their pockets and stop ceaselessly interrogating patients about their passage of flatus. Along these lines, I would like to ask Dr O'Connell whether he ignores clinical signs and symptoms other than repetitive emesis for purposes of NG tube insertion, or even maintaining or reinstituting NPO (nothing by mouth) status? Furthermore, are anorexic patients forced to eat on POD 2 or 3?I am gratified to see that the authors have corroborated earlier findings by demonstrating that early feeding does not exacerbate postoperative ileus and does not cause anastomotic leaks or other complications. I somewhat doubt that the finding of sex difference will stand up to the study of larger populations. That will be an interesting issue to pursue further. It is my belief that the more extensive mesenteric and, therefore, retroperitoneal dissection is responsible for the observed more prolonged ileus among patients undergoing total abdominal colectomy or total proctocolectomy compared with all other colon resections. The extent of retroperitoneal autonomic nerve disruption rather than bowel manipulation per se is probably the more important factor in the ileus induction. Would Dr O'Connell care to comment on this?The authors found no other correlations with failure of the early feeding protocol. I would like to ask why they did not consider PCA dosage consumed worthy of their study. Was any assessment of total narcotic dose or rate of consumption done in these patients? I believe this may be an important issue worthy of their future consideration.I have a few final questions. Were epidural catheters used in any patients? If they were, how did their use influence tolerance of the early postoperative feeding protocol? Do the authors have any anecdotal experience with prokinetic agents other than metoclopramide that show promise in this particular setting? Finally, how has the success of your early feeding and early discharge program with open colectomy influenced your opinions regarding the appropriateness and cost effectiveness of laparoscopic or laparoscopic-assisted colon resection?Theodore Schrock, MD, San Francisco, Calif:We have seen enough data in this and other papers to believe that this effect is genuine. We, too, have observed that patients who have had the entire colon resected are different. Bruce Stabile said it may be due to autonomic nerve disruption from extensive mesocolic dissection. I have thought it was loss of the main intestinal gas reservoir, namely, the colon, that is responsible for intolerance to small amounts of gas, but I am curious about the authors' view.The second question is about postdischarge follow-up. Was there a protocol involving personal calls from surgeons or nurses to these patients, or were they simply advised to call if they were having any problems?Thomas Russell, MD, San Francisco:We did a study similar to this a few years ago, and I agree completely with the authors' premise. Many surgeons are still using NG tubes after elective colon resections. To these surgeons early feeding is a foreign concept. To feed early a protocol needs to be developed and allow taking small amounts of fluid throughout the day as the patient desires. Hopefully improved prokinetic agents will further shorten the period of ileus.John Owings, MD, Sacramento, Calif:Patients' autonomic function is subject to suggestion by physicians. We, in a prospective study, previously demonstrated that ileus was similarly subject to perioperative suggestion by the physicians. My question then for the authors is what sort of suggestion were these patients given, namely, were the patients given clear liquids at POD 2 and told to take them and regular diet at POD 3 and told to take that, or were they simply offered up to clear liquids on POD 2 and up to regular diet on POD 3?Edward Phillips, MD, Los Angeles, Calif:I have a question regarding protocols for nausea or nausea prevention. Were protocols in place prior to nausea, or was an aggressive protocol in place to treat nausea with Zofran (ondansetron hydrochloride), droperidol, or Compazine (prochlorperazine maleate) in your patients?Ronald V. Maier, MD, Seattle, Wash:Since I agree with the authors' conclusions, it must be a good paper. Two questions. One of the important issues is how soon and how much the patients ambulate postoperatively. My impression is that men are more resistant than women. Did the authors notice if women ambulated earlier and whether that had a significant impact on their ability to eat sooner. The second is, as we are driven by medical economics and efficiency to force human physiology to fit a time clock, might the authors comment on what is an acceptable failure rate? A near 10% incidence of vomiting and a 5% readmission rate is not unreasonable, but should we start giving liquids on POD 1, send them home on POD 3 and accept an increased readmission rate and emesis rate? Do the authors have any thoughts on what are acceptable failure rates rather than waiting for objective evidence that the colon has returned to normal function?Laurence F. Yee, MD, San Francisco:I congratulate the authors on a well-designed, controlled study, and I agree with their conclusions. I have a question regarding how you decide when to advance these patients' diets. Are there certain criteria regarding the amount patients have to tolerate on POD 2 to have a regular diet by POD 3? Second, what is your discharge criteria? Do your patients have to pass flatus or have a bowel movement prior to discharge?Mark L. Welton, MD, San Francisco:We have been feeding our patients early now for 4 or 5 years, and Dr Schrock mentioned our lack of success. I agree with the authors' discussion about men not doing as well. I think it comes down to intelligence. Women are smarter than men, and when you tell a woman to take it easy, she will start out with a postsurgical diet (tea, toast, crackers, juice, English muffins), and they nibble at it, drink a little bit, and they do okay. If you tell a man to start out light, he has 3 cheeseburgers and a pizza and then vomits. I specifically recommend that they just go real slow, and it works best if the wife is there to control them.John E. Connolly, MD, Orange, Calif:Having had the pleasure of being a visiting professor in various parts of the world, particularly the United Kingdom, going back many years, I was always amazed and somewhat aghast to see that they did not use NG tubes and fed patients earlier than we. Thus, finally we have a well-conducted study that questions a long accepted habit at least in the United States. I predict that many nonbelievers will change their ways.Dr O'Connell:This is probably the first time that I had so much support in anything that I have ever presented, even at home, especially at home. But I think maybe I am preaching to the choir. I actually thought there would be a lot of objections since this is such a paradigm shift and a real change in our traditional methods. Surgeons are slow to change from their traditional methods. This study started with me watching people do laparoscopic colectomy a few years ago. They would look through the scope for 2 or 3 hours and, since these were usually laparoscopic assisted, they would make an incision, take the colon out, resect it, and anastomose it, drop it back in, and then feed them early. They said it was because of the laparoscopic technique. I really thought that this was Dumbo's magic feather, that they thought it was laparoscopic technique but they could always do it. They could always fly. They just didn't know it. They thought they needed the crutch of the laparoscopic technique. We can learn from these kinds of things just like we learn from laparoscopic herniorrhaphy, the use of the mesh. We don't have to use laparoscopy, but use of mesh is the way that herniorrhaphy is done increasingly in the world.I have answers to several of the questions. First of all, Dr Stabile, essentially we don't use an NG tube unless the patient vomits and vomits repeatedly which means 2 emeses. There are no other factors that really make us put in the NG tube routinely.The second question was, were the patients forced to eat, even if they were anorexic? Do we suggest to them beforehand? First of all these patients are prepared before they come into the hospital of what our protocol would be. So they already have an idea of when they are going to be fed and accept that, buy into it, and when they are going to be discharged, and buy into it. We don't force them to feed on the night of POD 2. A clear liquid diet is offered to them. I tell them that this is a test and if they feel sick, if they feel nauseated, don't drink it. They are not forced to. We are interested to see if they pass that test. Then the next day they go on to a regular diet. Again, the sex influence may come in. That was brought up. Women will abide by that and still have a lot of their tray left, but they say it went down pretty well. For men, you come in 5 minutes later, and it's all gone. They are ready to move on to the next step. Maybe that is part of the difference. The sex difference was really quite surprising to me. I thought it would be the time of surgery, comorbid illnesses such as diabetes, hypothyroidism, or associated operations that would add to the ileus, and what we really came up with was sex was one of the big factors. I was surprised by that. You would think that women actually would fail more often because various physiological studies have shown that they are slower to get over an ileus, they have a longer colon to body size, and they have delayed gastric emptying compared with men. So you would think physiologically they would fail. Why don't they fail? I think there are several factors. Maybe they go along with the protocol, they don't try to force the feeding as much as men. I think the other thing which we didn't look at, and several people brought it up, is the amount of morphine used. This is patient controlled, and men may not have the same pain tolerance as women. Women have already been through childbirth, many of them and so forth, and perhaps a little ileus and a little incision doesn't bother them a bit. I think they may use less morphine, but we have to examine that in further studies.As far as the total abdominal colectomy is concerned, I agree with Dr Stabile. I think this is because of the extensive mesenteric and retroperitoneal nerve manipulation. We have abandoned total proctocolectomy from the early feeding protocol. However, these patients also failed the traditional method. After total proctocolectomy, patients fail no matter whether you do early or late feeding. There were no epidurals used in this study. Epidurals were really introduced after the study was running and I didn't want to add another variable. But, again, maybe one future direction is to use epidurals rather than morphine. We have dropped metoclopramide; it didn't seem to be of any benefit. So our next variable that we recently added was cisapride.As far as laparoscopic and its cost effectiveness, what we are really saying is that any study that is done in the future really has to keep the variables of the feeding protocol the same, otherwise you are not really comparing open colectomy with laparoscopic. You are maybe comparing 2 different feeding protocols, traditional vs early feeding protocol.As far as Dr Schrock's question, we do not do routine calls to these patients. We don't want to suggest that they should be sick or vomiting at home. We just tell them to call us if there is a problem. Most don't. Five percent did come in because they did have nausea or emesis.It is important to have a protocol and stick with it and not vary it to have a proper protocol. If you feed them too early, you are going to fail. What we did, and I am not saying this is the ideal one, is to come up with this protocol because most studies have shown that after 24 hours the small intestine is back. After 48 hours the stomach is back, and then it takes 3 to 5 days for the colon to get over its ileus. So we didn't want to feed the patient before the stomach ileus was over. That's why we waited for 48 hours. Could you push it and get another 6 or 8 hours off or 12 hours? Maybe, but this was a rationale of the protocol and it was a physiologically thought out protocol. As far as Dr Phillips' question, we don't use anti-emetics in general, either in traditional or early feeding protocol as far as Zofran, Compazine, etc. I don't think anti-emetics should probably be used in most surgical patients. I don't think their vomiting is due to central mediated problems. That's fine for chemotherapy and sea sickness, but I don't think it is right for surgical patients.I do agree that early ambulation is very important. The patients are essentially up that evening of surgery and walking. Again, what is the acceptable rate of failure? I can't tell you what the acceptable rate is, but I have shown you that whether we are using the traditional method or the early feeding protocol method, about 12% of the patients fail. Some people get prolonged ileuses even though you wait forever for them to get over them. I think not using the NG tube and early feeding is actually a good method. There have been multiple studies that show that the intestines do better with solutions and food in them, etc. We used to question years ago when people would actually put jejunostomy tubes in and start feeding the patient immediately after the surgery. But this may be right. Enterocytes do better when there is food in contact with them, and it may actually help them get over the ileus.The discharge criteria: We do not wait for them to pass gas or have a bowel movement. Most people can do that at home like they have been doing for centuries and do not have to do that in the hospital. The idea that you have to keep the patient in the hospital hovering over them and asking them, have you passed gas, is really not necessary for the proper care of these patients.Presented at the 70th Annual Session of the Pacific Coast Surgical Association, San Jose del Cabo, Baja California Sur, February 14, 1999.We thank Stacey L. Stern, MS, for expert assistance with the statistical analysis.Reprints: Theodore X. O'Connell, MD, Department of Surgery, 4747 Sunset Blvd, Los Angeles, CA 90027 (e-mail: theodore.x.o'[email protected]).
de Virgilio, Christian; Bui, Hao; Donayre, Carlos; Ephraim, Linda; Lewis, Roger J.; Elbassir, Magdi; Stabile, Bruce E.; White, Rodney
doi: 10.1001/archsurg.134.9.947pmid: 10487588
HypothesisAdverse cardiac event rates following endovascular abdominal aortic aneurysm (EAAA) and open abdominal aortic aneurysm (OAAA) repair are similar. We also hypothesized that the Eagle criteria (Q wave on electrocardiogram, diabetes, angina, congestive heart failure, age >70 years, and ventricular ectopy) are useful predictors of cardiac events in patients undergoing EAAA repair.DesignProspective (patients undergoing EAAA repair) and retrospective (patients undergoing OAAA repair).SettingPublic teaching and Veterans Affairs medical centers.PatientsEighty-three EAAA and 63 OAAA repairs.Main Outcome MeasuresMyocardial infarction, congestive heart failure, and cardiac death.ResultsPatients with EAAA were older (73 vs 68 years, P=.003). There were no differences in the mean number of Eagle criteria (1.2 vs 1.3), cardiac event rates (6% vs 4.8%), or mortalities (3.6% vs 4.8%). Within the EAAA group, congestive heart failure (P=.005) and Q wave on electrocardiogram (P=.006) were the only predictors of cardiac events.ConclusionsPatients undergoing OAAA and EAAA repair had similar cardiac event rates and mortality. In patients undergoing EAAA repair, history of congestive heart failure and Q wave on electrocardiogram were predictors of cardiac events.SINCE THE FIRST report by Parodi et al,interest in endovascular repair of abdominal aortic aneurysms (EAAA) has grown markedly. Multiple series have demonstrated the short-term effectiveness of EAAA repair.However, few data are available regarding the cardiac morbidity and mortality of EAAA repair as compared with standard open repair of abdominal aortic aneurysms (OAAA).The Eagle criteria (aged >70 years, diabetes, angina, history of congestive heart failure [CHF], Q wave on electrocardiogram [ECG], and ventricular arrhythmia requiring therapy) have been shown to be predictive of adverse cardiac events following elective vascular surgery.However, the utility of these criteria for predicting adverse cardiac events in patients undergoing EAAA repair has not been investigated.The purpose of this study was to compare cardiac morbidity and postoperative mortality for EAAA vs OAAA repair and to determine the predictive value of the Eagle criteria in patients undergoing EAAA repair.PATIENTS AND METHODSData for consecutive patients undergoing elective EAAA repair at Harbor-UCLA Medical Center (Torrance, Calif) and West Los Angeles Veterans Affairs Medical Center (Los Angeles, Calif) were prospectively collected from July 1995 to July 1998. Data for patients undergoing elective OAAA repair at these 2 centers during the same period were reviewed retrospectively. Most patients undergoing EAAA repair were outside referrals, whereas most patients undergoing OAAA repair were not. The reasons for outside referral varied. In most instances, referral was the result of patient preference for EAAA repair. In other cases, patients were referred by an outside physician for perceived increased risk for OAAA repair due to prior laparotomy, advanced age, multiple prior myocardial infarctions, poor pulmonary function, or low cardiac ejection fraction. Patients having an emergency operation, those with a suspected ruptured AAA, and those with pararenal, suprarenal, and thoracoabdominal aneurysms were excluded. Cardiac risk factors for each patient, including the Eagle criteria, were recorded. In 6 patients undergoing EAAA repair and in 1 patient undergoing OAAA repair, the presence of a Q wave on ECG could not be determined due to the presence of a pacemaker. Perioperative complications were defined as those occurring within 30 days of surgery or during the index hospital stay. Adverse postoperative cardiac events were defined as Q-wave and non–Q-wave myocardial infarction, CHF, ventricular tachycardia, unstable angina, cardiac arrest, and cardiac death. Myocardial infarction was defined as an elevation of the serum creatinine kinase MB isoenzyme to more than 5% and/or new Q wave on ECG or persistent changes in the ST-sT wave with or without chest pain for longer than 30 minutes. Congestive heart failure was defined as (1) symptoms or signs of pulmonary congestion (shortness of breath or rales); (2) symptoms or signs of new left or right ventricular failure (cardiomegaly, S3 heart sound, jugular venous distention, or peripheral edema); and (3) abnormal chest radiography findings (vascular redistribution and interstitial or alveolar edema). Ventricular tachycardia was defined as consecutive premature ventricular contractions lasting longer than 30 seconds, more than 30 beats, and resulting in hypotension to less than 90 mm Hg. Unstable angina was defined as typical precordial chest pain consistent with ischemia, lasting 30 minutes or longer, unresponsive to nitroglycerin and rest, or a crescendo pattern of angina occurring at a lower threshold or higher frequency. Cardiac arrest and cardiac death were defined as arrest or death from a dysrhythmia or CHF caused by a cardiac complication.Categorical variables were examined for associations by means of χ2or Fisher exact tests where appropriate. Continuous variables were compared with the use of the Wilcoxon rank sum test. P<.05 was considered statistically significant. No adjustment was made for multiple comparisons. Factors found to have a significant association on univariate analysis were entered in a multivariate logistic regression analysis. All statistical comparisons were performed with the SAS statistical software package (SAS Institute, Cary, NC).RESULTSThere were 146 patients who underwent AAA repair (83 EAAA and 63 OAAA). Ninety-two percent were male, with a mean age of 70.7 years (range, 49-88 years). The mean (SD) number of Eagle criteria per patient was 1.2 (1.1). The most frequent Eagle criterion was age older than 70 years (53%), followed by Q wave on ECG (27%), angina (17%), diabetes (13%), history of CHF (11%), and ventricular ectopy requiring therapy (6%). A comparison of clinical risk factors for patients undergoing EAAA and OAAA repair is listed in Table 1. Patients undergoing EAAA repair were significantly older than those undergoing OAAA repair (73 vs 68 years, P=.003), and were more likely to be older than 70 years. There was otherwise no difference in the Eagle risk factors.Table 1. Comparison of Patients Undergoing EAAA vs OAAA Repair*Risk FactorEAAA (n = 83)Open AAA (n = 63)PMean age, y7368.003Age >70 y52/83 (63)26/63 (41).01Q wave on ECG16/77 (21)21/62 (34).08Angina11/83 (13)14/63 (22).18Diabetes11/83 (13)8/63 (13)>.99CHF8/83 (10)8/63 (13).60Ventricular ectopy3/83 (4)6/63 (10).17Mean No. of Eagle criteria1.21.3.52≥2 Eagle criteria21/83 (25)20/63 (32).45*Data are presented as number/total number (percentage). EAAA indicates endovascular aortic abdominal aneurysm; OAAA, open abdominal aortic aneurysm; ECG, electrocardiogram; and CHF, congestive heart failure.Within the OAAA group, 33 (52%) had aortic tube grafts and 30 (48%) had aortoiliac bypasses for combined aortic and iliac aneurysms. All procedures were performed via a transabdominal approach using general anesthesia. The average length of hospital stay was 15.0 days. Within the EAAA group, 60 (72%) had aortic aneurysms alone and 23 (28%) had aortic and iliac aneurysms (P=.01 vs OAAA). All EAAA repairs were performed with a bifurcated modular endovascular prosthesis using general anesthesia in 79 patients (95%), local in 3 patients (4%), and epidural in 1 patient (1%). The average length of hospital stay was 5.7 days (P<.001 vs OAAA).Eight (5.5%) of the 146 patients suffered adverse cardiac events, including 4 patients with myocardial infarctions (3 Q wave and 1 non–Q wave), 3 patients with CHF alone, and 1 with ventricular tachycardia and CHF. Of the 83 patients undergoing EAAA repair, 5 (6%) had cardiac events, whereas 3 (4%) of the patients undergoing OAAA repair suffered a cardiac event (P>.99).Within the EAAA group, patients who suffered an adverse cardiac event were significantly older (82 vs 72 years, P=.007), and had a higher mean number of Eagle criteria per patient (3.2 vs 1.1, P<.001) than those without adverse cardiac events. On univariate analysis, Q wave on ECG (P=.006) and history of CHF (P=.005) were the only individual Eagle criteria that were predictive of adverse cardiac events (Table 2). The adverse cardiac event rate was 24% for patients with 2 or more Eagle criteria vs 0% for patients with 1 or no criteria (P=.001).Table 2. Univariate Analysis of Risk Factors for Adverse Cardiac Events in Patients Undergoing EAAA Repair*Risk FactorCardiac Event (n = 5)No Cardiac Event (n = 78)PAge >70 y5 (100)47 (60).15Q wave on ECG4 (80)12/72† (17).006Angina1 (20)10 (13).51Diabetes2 (40)9 (12).13History of CHF3 (60)5 (6).005Ventricular ectopy1 (20)2 (3).17≥2 Eagle criteria5 (100)16 (20).0007*Data are presented as number (percentage). EAAA indicates endovascular aortic abdominal aneurysm; ECG, electrocardiogram; and CHF, congestive heart failure.†Q-wave data were not available for 6 patients who have pacemakers.Within the OAAA group, patients with adverse cardiac events also had a significantly higher mean number of Eagle criteria per patient (3.3 vs 1.2, P=.01) but they were not significantly older (67 vs 72 years, P=.81) than those without adverse cardiac events. On univariate analysis, Q wave on ECG (P=.03), ventricular ectopy (P=.02), and history of CHF (P=.001) were the individual Eagle criteria that were predictive of adverse cardiac events. The adverse cardiac event rate was 15% for patients with 2 or more Eagle criteria vs 0% for patients with 1 or no criteria (P=.03).When we analyzed the patients undergoing EAAA or OAAA repair together, Q wave on ECG (P<.001), ventricular ectopy (P=.007), and a history of CHF (P<.001) were the Eagle criteria predictive of adverse cardiac events on univariate analysis, whereas diabetes (P=.07), angina, and age older than 70 years were not. No association was observed between EAAA repair and the occurrence of adverse cardiac events (odds ratio, 1.28; 95% confidence interval, 0.29-5.59). Because patients having EAAA repair were significantly older than those having OAAA repair, we performed an age-stratified analysis (age <70 vs ≥70 years) of the association between procedure type and adverse cardiac events using a Cochran-Mantel-Haenszel χ2test. This analysis found no association between procedure type and major cardiac events (odds ratio, 1.22; 95% confidence interval, 0.22-6.85). The cardiac event rate according to the number of Eagle criteria present is presented in Table 3. The cardiac event rate was 19.5% for patients with 2 or more Eagle criteria vs 0% for those with 1 or no criteria (P<.001). On multivariate logistic regression analysis with stepwise variable selection, only history of CHF (P=.002) and Q wave on ECG (P=.04) were predictive of adverse events.Table 3. Cardiac Event Rate by Number of Eagle CriteriaNo. of Eagle CriteriaEAAAOAAANo. of PatientsNo. of Cardiac EventsNo. of PatientsNo. of Cardiac Events02001301360290210213136251451105001160000Total835 (6%)633 (4.8%)There were a total of 6 (4.1%) postoperative deaths, of which 3 were EAAA and 3 were OAAA patients. Only 1 (0.7%) (in a patient undergoing EAAA repair) was of cardiac origin. The causes of the remaining noncardiac deaths were diffuse embolization with bowel ischemia (EAAA), gangrenous cholecystitis with sepsis (EAAA), and 1 patient each with a pulmonary complication, multisystem organ failure, and variceal bleeding from portal hypertension (all OAAA).COMMENTOur study demonstrated no difference in adverse cardiac event rate or postoperative mortality between EAAA and OAAA repair. Cardiac event rate was 6% for EAAA vs 5% for OAAA, whereas postoperative mortality was 4% for EAAA vs 5% for OAAA. Patients undergoing EAAA repair were significantly older. We therefore performed an age-stratified analysis that found no association between type of aortic abdominal aneurysm repair and adverse cardiac events. Additionally, there was no difference in the mean number of Eagle criteria per patient in the 2 groups, suggesting an equivalent cardiac risk. However, several limitations of the study should be addressed. First, data on OAAA repair were gathered retrospectively, whereas data on EAAA repair were prospective. Thus, some cardiac events in the former group may have been underdiagnosed. Second, many patients undergoing EAAA repair were referred from outside institutions, whereas patients undergoing OAAA repair were not. Many patients undergoing EAAA repair were referred because they were deemed to be high-risk candidates for OAAA repair, either because of advanced age or other comorbidities. These comorbidities, such as multiple prior myocardial infarctions, low cardiac ejection fraction, and pulmonary disease, may not be reflected accurately by the Eagle criteria.Most other studies on EAAA repair have focused on the technical complications related to the procedure, though some data on cardiac morbidity and mortality are available. May et alreported 8 cardiac complications (7%) in 108 patients who underwent attempted EAAA repair, with a 2.7% cardiac mortality. Moore et alnoted 1 myocardial infarction (2%) in 46 patients undergoing EAAA repair with no cardiac deaths.Baxendale et alhave demonstrated that EAAA repair causes less intraoperative hemodynamic and metabolic stress on the patient compared with OAAA repair. They noted a significant change in cardiac output, mean arterial pressure, and systemic vascular resistance related to aortic cross-clamping and lower limb reperfusion. With EAAA repair, there was only a transient rise in systemic vascular resistance with femoral artery clamping. Thus, one would anticipate a lower cardiac morbidity and mortality rate for EAAA repair. This expected reduction in cardiac morbidity was not noted in our study. This may be the result of a combination of factors. The EAAA repairs in our study were performed under general anesthesia, which itself is a major cardiac stress. Second, as mentioned earlier, patients undergoing EAAA repair may have had a greater cardiac risk than what was reflected by the Eagle criteria. We have since begun to selectively use local anesthesia for our high-risk patients undergoing EAAA repair. Whether this change in anesthetic approach will affect the cardiac event rate remains to be seen.The Eagle criteria have previously been demonstrated to be useful predictors of adverse cardiac events for patients undergoing major vascular surgery.Eagle et alreported a 3% adverse cardiac event rate in patients with no clinical criteria, a 15% event rate for those with 1 or 2 criteria, and a 50% event rate for those with 3 or more criteria. Our 2 previous studies of patients undergoing elective major vascular surgery confirmed the value of the Eagle criteria.We reported no cardiac morbidity or mortality for patients without a single criterion and a 7% cardiac event rate with 1 or more criteria (P<.05). To our knowledge, no one has investigated the utility of the Eagle criteria for patients undergoing EAAA repair. The present study demonstrated that Q wave on preoperative ECG and a history of CHF were predictors of postoperative cardiac morbidity for both OAAA and EAAA repair. Additionally, in patients with 2 or more Eagle criteria, the adverse cardiac event rate increased from 0% to 19.5%. The high cardiac morbidity associated with 2 or more criteria raises the question of the role of cardiac assessment in these patients. The present study did not attempt to answer the question of whether cardiac assessment is justified. Future studies are needed to address this important issue. Currently we are assessing whether the performance of EAAA with local or spinal anesthesia will reduce the perioperative cardiac morbidity in these high-risk patients.In summary, our study demonstrated no difference in adverse cardiac event rates or postoperative mortality rates between patients undergoing OAAA and EAAA repair. For patients with 2 or more Eagle criteria, the cardiac event rate rose to 19.5% compared with no cardiac morbidity for patients with only 1 criterion or no criteria. On multivariate analysis, Q wave on preoperative ECG and a history of CHF were predictive of cardiac events for both EAAA and OAAA repair. Future studies on cardiac risk assessment prior to AAA repair should focus on this subgroup of patients.JParodiJPalmazHBaroneTransfemoral intraluminal graft implantation for abdominal aortic aneurysms.Ann Vasc Surg.1991;5:491-499.WMooreRRutherfordfor the EVT InvestigatorsTransfemoral endovascular repair of abdominal aortic aneurysm: results of the North American EVT phase I trial.J Vasc Surg.1996;23:543-553.JParodiEndovascular repair of abdominal aortic aneurysms and other arterial lesions.J Vasc Surg.1995;21:549-557.CMialheCAmicabileJBecqueminEndovascular treatment of infrarenal abdominal aortic aneurysms by the Stentor system: preliminary results in 79 cases.J Vasc Surg.1997;26:199-209.TChuterBRisbergBHopkinsonClinical experience with a bifurcated endovascular graft for abdominal aneurysm repair.J Vasc Surg.1996;24:655-666.RWhiteCDonayreIWalotModular bifurcation endoprosthesis for treatment of abdominal aortic aneurysms.Ann Surg.1997;226:381-391.JMayGWhiteWYuConcurrent comparison of endoluminal versus open repair in the treatment of abdominal aortic aneurysms: analysis of 303 patients by the life table method.J Vasc Surg.1998;27:213-221.Cde VirgilioLKirbyRJLewisLimited utility of dipyridamole-thallium for predicting adverse cardiac events after vascular surgery.Vasc Surg.1998;32:275-279.Cde VirgilioSPakTArnellCardiac assessment prior to vascular surgery: is dipyridamole-sestamibi necessary?Ann Vasc Surg.1996;10:325-329.KEagleCColeyJNewellCombining clinical and thallium data optimizes preoperative assessment of cardiac risk before major vascular surgery.Ann Intern Med.1989;110:584-587.BBaxendaleDBakerAHutchinsonHaemodynamic and metabolic response to endovascular repair of infra-renal aortic aneurysms.Br J Anaesth.1996;77:581-585.Fred A. Weaver, MD, Los Angeles, Calif:This latest study analyzes the cardiac morbidity associated with open aneurysm repair vs endovascular stent graft replacement. Sixty-three open procedures were compared with 83 endovascular repairs. Development of myocardial infarction, congestive heart failure, ventricular tachycardia, unstable angina, cardiac arrest, or death served as primary end points. Eagle criteria were catalogued for all patients. Contrary to what has been touted as an advantage of the less-invasive endovascular repair, no significant difference in cardiac morbidity or mortality could be demonstrated between an open or an endovascular approach. The reasons for this are uncertain, and I would like the authors to comment or at least speculate on why they might think this would be so. It should also be noted that although the endovascular group was somewhat older, the Eagle criteria and thus the cardiac risk factors were equivalent. Another interesting point is that the cardiac-related morbidity was remarkably low in this particular series, with only 1 of the 6 deaths being directly related to cardiac morbidity. The rest of the deaths were due to noncardiac causes such as sepsis and respiratory and multiorgan system failure. The authors analyzed the patients' cardiac outcome by preoperative Eagle criteria using a univariate analysis and multivariate logistic regression analysis. Only Q wave on ECG or a history of congestive heart failure were found to be uniquely predictive of cardiac morbidity in the 146 aneurysm procedures. It was also found that patients with less than or equal to 1 Eagle criteria had no cardiac morbidity, whereas all cardiac morbidity was clustered in those individuals with 2 or more Eagle criteria.This is a very interesting paper and, although somewhat compromised by the retrospective nature of the open analysis vs the prospective nature of the endovascular analysis, it does challenge the perception of physicians, surgeons, and patients that endovascular aneurysm repair is inherently safer than an open approach.There are a few questions that I would like the authors to address. First, what kind of anesthetic was used for these procedures? Was the anesthesia used for the open procedure equivalent to the endovascular procedures? What kind of intraoperative hemodynamic monitoring was used? Was the monitoring for both procedures more or less the same?Second, how were the open procedures performed, retroperitoneal or transperitoneal? How many of the aortic aneurysms had iliac aneurysms as well, which, at least in my hands, makes for a more difficult open operation? Concerning the endovascular repair, were these all bifurcated prostheses or were tube stent graft prostheses used as well? You identified 2 specific Eagle criteria that were unique, highly predictive of cardiac events. Were all of the cardiac events that occurred clustered in patients with 1 or both of these criteria, making the rest of the Eagle criteria really irrelevant?Finally, you have studied and written a great deal on the cardiac evaluation of the patient undergoing vascular surgery. At present what are your recommendations? When do you obtain a noninvasive cardiac assessment such as persantine thallium or dobutamine echo? When should the patient have a cardiac cath?Samuel E. Wilson, MD, Orange, Calif:Why, after a less traumatic procedure, do the cardiac complications and morbidity occur with roughly the same incidence as after the open procedure?Dr Stabile:Our expectation with this study was that we would find that endovascular aneurysm repair would have a lower cardiac event rate, a lower cardiac death rate, and a lower overall death rate than open aneurysm repair. This clearly was not realized by the results of this study, but this is a preliminary study and we clearly have additional work to do and perhaps a more sophisticated analysis of risk factors will be required.The 2 treatment groups were very similar with respect to the Eagle criteria and, as Dr de Virgilio and the vascular group have shown in several prior studies, the Eagle criteria were again very highly predictive of cardiac events. The only significant difference amongst the Eagle criteria was that the endovascular treatment group was significantly older than the open treatment group. This was particularly true for the endovascular group of patients who had cardiac events. That group was fully 9 years older, with a mean age of 82 years, as compared with the remainder of the group, who suffered no cardiac events. The Eagle criteria were predictive in a very reliable way in that 2 or more criteria had a cardiac event rate of 20% whereas 1 or no criteria had 0 cardiac events. This corroborates a growing body of evidence suggesting this simple clinical assessment is highly valuable to the clinician in planning elective vascular operations.Dr Weaver pointed out the obvious limitation of the study, and that is comparing a prospectively evaluated endovascular group with a retrospectively evaluated open aneurysmectomy group. This is important because the majority of the endovascular patients were referred specifically because the referring physician had assessed the patient to be at high risk. The Eagle criteria unfortunately do not encompass all of the risk factors that are pertinent, such as low ejection fraction, multiple prior myocardial infarctions, recent severe cardiac symptoms, and noncardiac problems such as severe chronic obstructive pulmonary disease. Thus, we do not feel that the Eagle criteria give a full risk assessment. Potentially lethal adverse outcomes can occur, and we need to find assessment tools to evaluate those.These considerations may have obscured both higher risk in the endovascular group, as well as possibly more severe and unappreciated adverse outcomes in the open-treated group because of the retrospective nature of that evaluation.Dr Weaver asked that we specifically speculate on the lack of outcome differences. We need a better preoperative risk stratification system that takes into account cardiac issues not specifically encompassed by the Eagle criteria and also noncardiac potential morbidity.The anesthesia was general anesthesia for all patients. There was no difference in the anesthetic technique for the open vs the endovascular groups, and in general the same monitoring was used based on the same criteria. All open operations were transperitoneal. I do not have specific information, Dr Weaver, relating to the number of more complex aneurysms involving the iliac vessels.Early in the experience, straight tube grafts were exclusively used in the endovascular-treated group, but the vast majority of the patients now are being treated with a modular bifurcated device. We did find that a history of congestive heart failure and Q waves were by far the most predictive, and were the only independent predictors amongst the Eagle criteria. The others may well be irrelevant as suggested.Dr Wilson asked the causes of 5 of 6 deaths, as only 1 was a cardiac death. They were due to a variety of causes. The minority were technically related and occurred early in the experience with the endovascular devices. One patient had diffuse embolization to the bowel that resulted in sepsis and ultimately multiple organ failure. Another patient had acalculous cholecystitis that went on to gangrene, sepsis, and multiple organ failure. The other deaths were essentially anesthesia related. So there was a multiplicity of causes of death other than the 1 cardiac death. Again, I think this specifically points out the fact that we need a better and more comprehensive system for patient risk evaluation and only then will we be able to better select patients for endovascular vs open therapy.This study was supported in part by a grant from the Harbor-UCLA Research and Education Institute, Torrance, Calif (Dr de Virgilio).Presented at the 70th Annual Session of the Pacific Coast Surgical Association, San Jose del Cabo, Baja California Sur, February 14, 1999.We thank Roger J. Lewis, MD, PhD, for statistical analysis.Corresponding author: Christian de Virgilio, MD, Harbor-UCLA Medical Center, Division of Vascular Surgery, 1000 W Carson St, Box 25, Torrance, CA 90509.
doi: 10.1001/archsurg.134.9.951pmid: N/A
On the cameo page of the July issue of the ARCHIVES (1999;134:697), Lazar J. Greenfield, MD, was listed as Professor, Department of Surgery, University of Michigan Medical Center. His title is actually "Professor and Chair." The journal regrets the error.
Abou-Zamzam, Jr, Ahmed M.; Moneta, Gregory L.; Edwards, James M.; Yeager, Richard A.; McConnell, Donald B.; Taylor, Jr, Lloyd M.; Porter, John M.
doi: 10.1001/archsurg.134.9.952pmid: 10487589
HypothesisExtrathoracic cervical grafts are safe and provide long-lasting stroke prevention in patients with disease not amenable to standard carotid bifurcation endarterectomy.DesignReview of a prospectively maintained vascular surgical registry.SettingCombined university and Department of Veterans Affairs vascular surgical service.ParticipantsPatients requiring surgery for carotid atherosclerotic occlusive disease not amenable to endarterectomy from January 1988 to March 1998.InterventionsCarotid interposition grafting, subclavian-carotid bypass, or carotid-carotid bypass.Main Outcome MeasuresPerioperative stroke and death, and life-table determination of freedom from stroke, stroke-free survival, and graft patency.ResultsSixty patients (mean age, 65.8 years; range, 36-83) underwent cervically based carotid grafting. All had greater than 70% stenosis or occlusion of the innominate, common carotid, or internal carotid arteries, and 30 (50%) had undergone at least 1 previous ipsilateral carotid endarterectomy. Indication for operation was stroke or transient ischemic attack in 46 (77%) and asymptomatic high-grade stenosis in 14 (23%). Operative procedures included 31 (52%) carotid interposition grafts, 18 (30%) subclavian-carotid grafts, and 11 (18%) carotid-carotid grafts. Mean follow-up was 29 months (range, 1-117 months). Perioperative stroke rate was 5% (3/60) all in symptomatic patients, and there were no perioperative deaths. By life-table analysis, freedom from stroke was 92% at 1 and 5 years. Stroke-free survival was 90% at 1 year and 61% at 5 years. Primary graft patency was 94% at 1 year and 84% at 5 years, with assisted primary patency of 90% at 5 years.ConclusionCervical carotid artery grafts for complicated or recurrent carotid atherosclerosis not amenable to endarterectomy are durable and provide excellent freedom from stroke with low perioperative morbidity and mortality.THE ROLE of carotid bifurcation endarterectomy (CEA) for symptomatic and asymptomatic atherosclerotic occlusive disease of the carotid bifurcation has been clarified over the past decade by the results of several randomized trials of CEA.The increase in performance of CEA that these studies have generated will likely lead to an increase in the prevalence of recurrent carotid artery stenosis.Recurrent carotid artery stenosis may not be amenable to standard techniques of endarterectomy so that other procedures such as carotid interposition grafting or carotid bypass may be required.Greater awareness of the role of extracranial carotid disease in stroke has also increased identification of patients with occlusive disease proximal to the carotid bifurcation. Because of the rarity of these lesions, their natural history has been extrapolated from carotid bifurcation disease. Patients with these proximal lesions may be treated with direct repair via a sternotomy or with a cervically based bypass to the carotid artery (subclavian-carotid or carotid-carotid bypass).While carotid interposition grafting and cervically based carotid bypass are theoretically attractive alternatives to direct reconstruction using a median sternotomy, the safety and long-term stroke prevention of such procedures is unclear. The purpose of this study was to review our experience with carotid interposition and bypass grafting in terms of operative complications and durability of stroke prophylaxis.PATIENTS AND METHODSA review of the vascular registry at Oregon Health Sciences University and the Portland Veterans Affairs Medical Center, Portland, Ore, was performed covering the period January 1988 through March 1998. All patients who underwent surgery for occlusive disease of the innominate or carotid arteries not amenable to standard CEA were included. This included patients who had disease considered by the surgeon at operation to be unsuitable for endarterectomy, as well as all disease that would require sternotomy for direct surgical repair.Patient demographics and atherosclerotic risk factors at the time of surgery (smoking, hypertension [need for antihypertensive medication], diabetes mellitus [oral hypoglycemic medications or insulin], and coronary artery disease [angina, abnormal electrocardiogram, congestive heart failure, coronary artery bypass graft or coronary angioplasty]) were documented. A specific history of cerebral vascular disease (stroke, transient ischemic attack, amaurosis fugax, or prior CEA) was recorded.The indication for operation was classified as either symptomatic (stroke, transient ischemic attack, amaurosis fugax, global ischemia) or asymptomatic (high-grade stenosis). Operative details reviewed included the conduit used for grafting, the origin and termination of the grafts, the use of intraoperative shunts, and total operative time. Perioperative morbidity was documented. Clinical follow-up routinely included an early postoperative visit with a cervical duplex ultrasound examination followed every 6 months by clinical visits to assess for freedom from stroke and repeated cervical duplex ultrasound examination to document graft patency and development of graft stenosis.Analysis of graft patency, freedom from stroke, and survival were performed by the life-table method. The influence of the type of operation performed (interposition or bypass), graft material (vein or prosthetic), and operative indication (symptomatic or asymptomatic) on perioperative stroke were determined by χ2analysis. Long-term graft patency and freedom from stroke of the various operations and conduits were compared by the log-rank test. Data were analyzed by a statistical software program (JMP; SAS Institute, Cary, NC). All differences were considered statistically significant at P<.05.RESULTSDuring the 10-year study period, 60 patients underwent carotid interposition grafting or cervically based carotid bypass for occlusive cerebrovascular disease. Mean age at operation was 65.8 years (range, 36-83 years). There were 40 men and 20 women with the typical atherosclerotic risk factors. Other patient demographics are detailed in Table 1. One half of the patients had undergone at least 1 previous ipsilateral CEA, and 1 patient had undergone previous carotid angioplasty and stenting. Indication for operation was cerebral ischemia in 46 patients (77%) and asymptomatic high-grade lesions (80%-99% stenosis) in 14 patients (23%) (Table 2).Table 1. Patient DemographicsCharacteristicNo. (%)SexMale40 (67)Female20 (33)Diabetes10 (17)Hypertension32 (53)Smoker (current or former)56 (93)Coronary artery disease32 (53)Hyperlipidemia14 (23)Previous stroke or transient ischemic attack45 (75)Previous carotid endarterectomyIpsilateral30 (50)Contralateral20 (33)Table 2. Operative DetailsCharacteristicVariableOperative indication, No. (%)Stroke8 (13)Transient ischemic attack23 (38)Amaurosis fugax12 (20)Global ischemia3 (5)Asymptomatic high-grade lesion14 (23)Procedure, No. (%)Carotid interposition31 (52)Vein grafts27Dacron4Bypass29 (48)Subclavian-carotid bypass18Dacron7Vein grafts11Carotid-carotid bypass11Dacron10Polytetrafluoroethylene1Operative time, mean (range), minAll procedures170 (65-270)Interposition165 (115-240)Bypass175 (65-270)Intraoperative shunt used, No. (%)29 (48)The 60 operations performed included 31 carotid interposition grafts and 29 cervically based bypass grafts. The types of operations are shown in Figure 1and listed in Table 2. Carotid interposition grafts were performed for recurrent carotid bifurcation disease not amenable to endarterectomy in 21 patients, a combination of atherosclerotic stenosis and small internal carotid aneurysms in 4 patients, and disease not amenable to first-time endarterectomy owing to radiation in 3 patients, and complex extensive plaque (requiring a very long endarterectomy or extruding through the adventitia) in 3 patients.Figure 1.Diagrammatic representation of types of operations performed. A, Right-left carotid-carotid (4 patients); B, left-right carotid-carotid (7 patients); C, left subclavian-carotid (13 patients); D, right subclavian-carotid (5 patients); E, left carotid interposition (17 patients); and F, right carotid interposition (14 patients). Typical atherosclerotic lesions are shown.Cervical bypass grafts were performed for proximal common carotid artery stenosis in 16 patients (including 5 occlusions), innominate stenosis in 4 patients (including 2 occlusions), and for combinations of common carotid and bifurcation stenosis in 9 patients. Cervical grafts were combined with recipient-side CEA in 8 patients and with donor-side CEA in 4 patients. Subclavian-carotid grafts were tunneled behind the sternocleidomastoid muscle. Carotid-carotid bypasses were all tunneled in a subplatysmal plane anterior to the trachea.Interposition grafts were performed with a venous conduit in 27 patients, and with Dacron in 4 patients. Bypass grafts were performed with Dacron in 17 patients, polytetrafluoroethylene in 1 patient, and a venous conduit in 11 patients. Average operative time was 170 minutes, with no difference between interposition and bypass grafting procedures (165 minutes vs 175 minutes; P=.34).Operative complications are listed in the tabulation below.ComplicationNo. (%) of PatientsStroke3 (5)Death0Transient ischemic attack2 (3)Superficial wound infection2 (3)Hematoma2 (3)Cranial nerve palsy (temporary)2 (3)Myocardial infarction1 (2)Arrhythmia1 (2)Urinary tract infection1 (2)There were 3 perioperative strokes for an operative stroke rate of 5%. All 3 strokes occurred in symptomatic patients. One patient underwent a venous left carotid interposition and awoke with a right hemiparesis. Upon reexploration, the graft was found to be thrombosed. This graft was revised and the patient had a minimal deficit at the time of hospital discharge. A second patient awoke with a neurologic deficit following a right carotid interposition. Upon reexploration, the graft was patent and intraoperative angiography revealed no technical problems. This patient was left with a moderate deficit at the time of discharge from the hospital. The third perioperative stroke occurred in a patient 3 days following a subclavian-carotid bypass. The bypass was patent on duplex scan and the patient recovered with minimal deficit.There was no statistical difference in rate of perioperative strokes in symptomatic vs asymptomatic patients (3/46 [6.5%] vs. 0/14 [0%]) (P=.78). Similarly, the type of operation (interposition vs bypass) did not influence perioperative stroke rate (2/31 [6.5%] vs 1/29 [3.5%]; P=.95). The type of conduit (vein vs prosthetic) also had no effect on early stroke rate (2/38 [5.3%] vs 1/22 [4.5%]; P=.63).Average follow-up was 29 months (range, 0-117 months). During follow-up there was 1 patient with late stroke in the ipsilateral cerebral distribution 11 months following a Dacron carotid interposition graft. The rates for 5-year freedom from stroke, survival, and stroke-free survival were 92%, 69%, and 61%, respectively (Figure 2).Figure 2.Life-table analysis. A, Freedom from stroke; B, survival; C, stroke-free survival; D, primary and assisted primary graft patency. Bars represent SE; dashed lines, beyond SE >10%.Primary graft patency was 84% at 5 years and assisted primary patency was 90% at 5 years (Figure 2). There were 2 patients with late graft occlusions. A carotid interposition vein graft occluded 2 months postoperatively and resulted in a transient ischemic attack. A Dacron subclavian-carotid bypass occluded 39 months postoperatively and was asymptomatic. There were 2 patients with late graft revisions—1 at 12 months for a stenosis at the distal end of an interposition vein graft that was causing transient ischemic attacks, and 1 at 13 months for an asymptomatic stenosis at the distal end of a Dacron carotid-carotid bypass.Life-table analysis revealed no difference in 5-year freedom from stroke when analyzed by the type of operation (interposition vs bypass, 93% vs 92%; P=.95) or the conduit used (vein vs prosthetic, 95% vs 89%; P=.30). There were also no differences in 5-year primary graft patency when interposition grafts were compared with bypass grafts (88% vs 83%; P=.54), and when vein grafts were compared with prosthetic grafts (90% vs 77%; P=.94).COMMENTSome patients with extracranial carotid artery occlusive disease cannot be treated with standard CEA. Depending on the location of the occlusive lesion, carotid interposition grafting and various configurations of cervically based carotid bypasses are conceptually attractive procedures for the treatment of such patients. However, there have been concerns regarding operative morbidity, optimal graft material, graft patency, and the long-term stroke prophylaxis associated with carotid grafting.Our data suggest that these operations can be performed with acceptable morbidity and mortality. There were no operative deaths in this series and the operative stroke rate was 5%, with all strokes occurring in symptomatic patients. These results compare favorably with those of other reports.Fry et alreported a combined perioperative stroke and death rate of 5% in a group of 20 subclavian-carotid bypasses. A similar perioperative stroke and death rate (4.3%-6.3%) was reported by Berguer et alin 2 recent series of cervical bypasses. Salam et aldescribed 31 patients undergoing carotid-carotid or subclavian-carotid bypass without a single perioperative stroke or death. Such cervically based procedures appear to offer lower morbidity and mortality than transthoracic repairs where the combined stroke and death rate has been reported in the range of 14% to 16%.In addition, both interposition grafting and cervically based carotid bypass appear quite durable, with an assisted primary patency in this series of 90% at 5 years. In our experience, only 2 grafts required revision to maintain patency and 2 other grafts occluded during follow-up with neither occlusion resulting in a stroke. The durability of these cervical grafts argues against the suggestion that cervical bypasses offer patency rates inferior to direct transthoracic repairs.Some authors have suggested that prosthetic extrathoracic carotid grafts may provide superior durability over venous conduits.Recently, Rockman et alreported that in redo carotid surgery, the use of a vein interposition led to an increased risk of recurrent stenosis or occlusion compared with prosthetic grafts.Their series, however, contained only 9 vein interpositions in 82 redo carotid operations, and had an overall stroke rate of 8.6% with interposition grafting. Other authors have demonstrated no difference or even superior results with vein grafts.Synn et alconcluded that vein is a better conduit than prosthetic material in longer bypasses from the subclavian artery to the carotid bifurcation. In our study, there appears to be no difference in the durability of prosthetic vs vein conduits in extrathoracic carotid grafting. Both conduits provided excellent long-term results. We conclude, therefore, that the choice of conduit is relatively unimportant in extrathoracic carotid grafting. While our preference is for autogenous conduits, we believe that a prosthetic conduit is superior to a poor-quality venous conduit. In addition, certain procedures such as carotid-carotid bypass appear best performed with a crimped or externally supported prosthetic conduit to avoid kinking in a curved configuration and we have found crimped 7-mm Dacron to be an excellent conduit in this setting.Martin et alreported 75% long-term freedom from stroke in a small series of patients with cervical grafts, raising concerns regarding poor long-term stroke prophylaxis with extrathoracic carotid grafting. Despite the advanced degree of extracranial carotid occlusive disease present in the patients in our series, life-table analysis found 92% of surviving patients remained stroke-free 5 years following cervical grafting. These results compare very favorably with stroke prophylaxis following CEA for both symptomatic and asymptomatic carotid occlusive disease. Freedom from stroke at 5 years was 87% in the North American Symptomatic Carotid Endarterectomy Trialand 95% in the Asymptomatic Carotid Atherosclerosis Study.Although the need for extracranial carotid grafts is relatively infrequent, the increased numbers of CEAs performed along with greater physician awareness of the relationship between carotid occlusive disease and stroke will likely lead to an increase in the need for these procedures. Surgeons performing such procedures can be assured that they are offering these high-risk patients a safe, efficacious, and durable procedure.HJMBarnettDWTaylorMEliasziwBenefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis.N Engl J Med.1998;339:1415-1425.Executive Committee for the Asymptomatic Carotid Atherosclerosis StudyEndarterectomy for asymptomatic carotid artery stenosis.JAMA.1995;273:1421-1428.North American Symptomatic Carotid Endarterectomy Trial CollaboratorsBeneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis.N Engl J Med.1991;325:445-453.European Carotid Surgery Trialists' Collaborative GroupRandomised trial of endarterectomy for recently symptomatic carotid stenosis: final results of the MRC European Carotid Surgery Trial (ECST).Lancet.1998;351:1379-1387.JVTuELHannanGMAndersonThe fall and rise of carotid endarterectomy in the United States and Canada.N Engl J Med.1998;339:1441-1447.RBerguerMDMoraschRAKlineTransthoracic repair of innominate and common carotid artery disease: immediate and long-term outcome for 100 consecutive surgical reconstructions.J Vasc Surg.1998;27:34-42.TMSullivanSubclavian-carotid bypass to an "isolated" carotid bifurcation: a retrospective analysis.Ann Vasc Surg.1996;10:283-289.TASalamRBSmith IIIABLumsdenExtrathoracic bypass procedures for proximal common carotid artery lesions.Am J Surg.1993;166:163-167.WRFryJDMartinGPClagettWJFryExtrathoracic carotid reconstruction: the subclavian-carotid artery bypass.J Vasc Surg.1992;15:83-89.WSMooreExtra-anatomic bypass for revascularization of occlusive lesions involving the branches of the aortic arch.J Vasc Surg.1985;2:230-232.RSMartinWHEdwardsJLMulherinWHEdwards JrSurgical treatment of common carotid artery occlusion.Am J Surg.1993;165:302-306.AYSynnRTAChalmersWJSharpJJHoballahTFKresowikJDCorsonIs there a conduit of preference for a bypass between the carotid and subclavian arteries?Am J Surg.1993;166:157-162.RBerguerMDMoraschRAKlineAKazmersMSFriedlandCervical reconstruction of the supra-aortic trunks: a 16-year experience.J Vasc Surg.1999;29:239-248.MAMansourSSKangWHBakerCarotid endarterectomy for recurrent stenosis.J Vasc Surg.1997;25:877-883.DRosenthalJPArchie JrMHAvilaSecondary recurrent carotid stenosis.J Vasc Surg.1996;24:424-429.RBerguerJAGonzalezRevascularization by the retropharyngeal route for extensive disease of the extracranial arteries.J Vasc Surg.1994;19:217-225.DPVogtNRHertzerPJO'HaraEGBevenBrachio-cephalic arterial reconstruction.Am J Surg.1982;196:541-552.CBRockmanTSRilesRLandisRedo carotid surgery: an analysis of materials and configurations used in carotid reoperations and their influence on perioperative stroke and subsequent recurrent stenosis.J Vasc Surg.1999;29:72-81.Jeffrey L. Ballard, MD, Loma Linda, Calif:Dr Abou-Zamzam and colleagues from the Oregon Health Sciences Center have presented a 10-year review of 60 patients who underwent carotid interposition grafting or cervically based carotid artery bypass for occlusive disease of the innominate and carotid arteries. None of these vessels were amenable to standard endarterectomy techniques. In this group of typical vascular surgery patients, there were 31 carotid interposition grafts and 29 cervically based bypass grafts. Surgical results were excellent as one might expect from this group of highly talented vascular surgeons. Data were well analyzed and thoroughly support their conclusions.There were no operative deaths in the series and the operative stroke rate was 5%. All of these strokes occurred in symptomatic patients. Life-table analysis demonstrated that the postoperative stroke risk was not affected by the type of operation or the conduit used during the procedure. Additionally, there were no significant differences in 5-year cumulative graft patency when interposition grafts were compared to bypass grafts or when vein grafts were compared to prosthetic grafts.Based upon these excellent results, the authors conclude that these extrathoracic arterial grafts performed for arterial occlusive disease not amenable to endarterectomy are safe, efficacious, and durable. Furthermore, they conclude that the choice of conduit is relatively unimportant for these cervically based carotid grafting procedures.At Loma Linda University Medical Center, our results over the past 3 years parallel those of the Oregon group. We have performed 23 similar procedures with 1 perioperative stroke and no operative deaths. Most of the grafts were PTFE. Graft patency rate and stroke-free survival are similar to the results just presented.Just to ensure that Dr Abou-Zamzam does not waltz off the podium unscathed, I have a few questions. (1) In the manuscript, it is stated that some of the carotid interposition grafts were performed for "recurrent carotid bifurcation disease not amenable to endarterectomy." What criteria did you use to come to this conclusion? Was this an intraoperative assessment or was there something specific about your preoperative evaluation that led you to this conclusion prior to surgical intervention? Also, how did you know that the innominate artery was not amenable to endarterectomy? (2) One carotid interposition vein graft occluded 2 months postoperatively. Was this event possibly due to a technical error, and, if so, what is your intraoperative protocol for ensuring that there are no technical defects in these bypass grafts? That is, do you perform routine intraoperative duplex ultrasound or angiography following these vascular reconstructions? (3) What outpatient protocol do you use to follow these grafts after the initial procedure? (4) Does your group have any experience with ministernotomy? This technique utilizes a limited skin incision extending from the sternal notch down to the Angle of Louis. The manubrium and upper sternum are divided in the midline down to the third intercostal space and then the sternum is transected transversely, creating an upside down "T" incision. This facilitates direct exposure of the upper pericardium, which can then be divided vertically so that the base of the heart and arch vessels can be readily exposed. We have recently performed innominate and common carotid artery endarterectomy, patch angioplasty, and bypass with encouraging preliminary results. Are you familiar with this modification of median sternotomy, and when might you favor this direct approach for some innominate or other great vessel disease? (5) My last question revolves around your choice of conduit. I have favored synthetic conduits for these reconstructions because of the larger size of the subclavian and carotid arteries compared to a greater saphenous vein conduit. If you had a cervically based bypass procedure to do tomorrow, what would be your conduit of choice?Clifford W. Deveney, MD, Portland, Ore:The authors have concluded that there is no relationship between previous stroke and stroke in this study. However, their numbers are relatively small and their stroke rate is so small (5%) that I think they would need a larger group to have the power to detect such a difference.Stephen N. Etheredge, MD, Oakland, Calif:I rise to speak against the saphenous vein as a conduit of choice except in those cases where you are dealing with an infective process. Two significant problems are the inability to back flush from the brain down because of the valves that are present in the vein itself and the second one being that those same valve rings in a high-flow situation increase your risk of hyperplasia at those locations and may lead to further stenosis. If the prosthetic material is equally good, I would suggest that you stay away from the vein unless you have to use it for infective reasons.Cornelius Olcott, MD, Stanford, Calif:There has been increasing interest in endovascular balloon dilatation for innominate lesions. What is your experience with endovascular techniques for brachiocephalic lesions?Dr Moneta:Dr Ballard asked about the criteria for determining recurrent bifurcation lesions not amenable to endarterectomy. All of our patients having carotid operations for recurrent disease undergo angiography. If the lesion is a long lesion, or if this is a second recurrence following 2 previous bifurcation endarterectomies, or if it is a recurrence following radiation treatment, we will tend to treat those patients with grafting rather than attempt to try to do another endarterectomy. He also asked about criteria for not considering endarterectomy of the innominate artery. Ours are not much different from anyone else's. If there is extensive calcification of the aortic arch, we won't do an innominate endarterectomy because of dangers associated with clamping the arch. Also, if the orifices of the innominate and the left common carotid artery are close together such that placement of a clamp on the arch will compromise both lumens, we will not attempt endarterectomy of the innominate artery.Also, if a patient has had a previous sternotomy, we don't do endarterectomies on the innominate just because of the risk and difficulty of opening a previous sternotomy. And of course there are patients whom we judge to be too high a medical risk for that procedure.Our intraoperative assessment of patients following extrathoracic cervical grafting is quite simple. In virtually all cases, only a continuous-wave Doppler examination is used. Angiography is rarely and selectively performed. We do not do routine intraoperative duplex scanning or angiography for any type of vascular reconstruction on our vascular surgical service.The follow-up protocol for evaluation of these patients is a duplex scan shortly postoperatively, then every 3 months for the first year for patients with venous conduits, and then every 6 months indefinitely. Prosthetic conduits are followed every 6 months indefinitely.I don't have any experience with the mini sternotomy that Dr Ballard advocated. We have used a full sternotomy for direct reconstruction of the arch vessels.About the conduit that is used, we prefer a prosthetic conduit for carotid to carotid bypasses because of the curve that is inherent in those grafts. I prefer a vein graft when the distal anastomosis is to the internal carotid artery. I find it is easier to sew a vein to the internal carotid artery and the size matches nicely. If the 2 ends are to the common carotid and subclavian vessels, then a prosthetic conduit is a nice size match and a technically easier procedure.Dr Deveney indicated we need more numbers to really know the stroke rate. More numbers always help. We will keep track.Steve Etheredge doesn't like saphenous vein for these procedures. We do under the circumstances previously discussed. Our data suggest both vein and prosthetic work well. We have done a few innominate artery angioplasties. We feel these patients have to be without the contraindications to operation in case there is a problem.Presented at the 70th Annual Session of the Pacific Coast Surgical Association, San Jose del Cabo, Baja California Sur, February 14, 1999.Reprints: Gregory L. Moneta, MD, Oregon Health Sciences University, Division of Vascular Surgery, OP-11, 3181 SW Sam Jackson Park Rd, Portland, OR 97201 (e-mail: [email protected]).
Caruso, Daniel M.; Battistella, Felix D.; Owings, John T.; Lee, Steven L.; Samaco, Rodney C.
doi: 10.1001/archsurg.134.9.958pmid: 10487590
HypothesisPerihepatic packs used to control hemorrhage after liver injury increase the risk of complications and this risk increases the longer packs are left in place.DesignRetrospective case series.SettingUniversity level I trauma center.PatientsConsecutive patients with hepatic injury.Main Outcome MeasuresLiver-related complications (biliary leak and abscess), rebleeding, and mortality.ResultsOne hundred twenty-nine of 804 patients with liver injuries were treated with perihepatic packing. Of the 69 who survived more than 24 hours, 75% lived to hospital discharge. Mortality rates were 14% and 30% in patients with and without liver complications, respectively (P=.23). Liver complication rates were similar (P=.83) when packs were removed within 36 hours (early [33%]) or between 36 and 72 hours (late [29%]) after they were placed; the rebleeding rate was greater in the early group (21% vs 4%; P<.001).ConclusionsLiver complications associated with perihepatic packing did not affect survival. Removing liver packs 36 to 72 hours after placement reduced the risk of rebleeding without increasing the risk of liver-related complications.MASSIVE LIVER injury with uncontrollable hemorrhage is still one of the most challenging problems faced by trauma surgeons. In the early 20th century, intra-abdominal (perihepatic) packing became the "mainstay of therapy" for uncontrollable liver hemorrhage. Unfortunately, infection rates and associated morbidity were high.At the end of World War II, Madding et alreported a lower infection rate and a decrease in overall morbidity and mortality rates using liver resection and drainage for major hepatic injury. Perihepatic packing was felt to have no place in the practice of surgery.Although some surgeons continue to condemn the use of perihepatic packing owing to the high incidence of intra-abdominal abscess and sepsis, others tout its usefulness in severe liver injury, especially when patients develop intraoperative coagulopathy, hypothermia, and acidosis.Proponents of perihepatic packing have demonstrated that better antibiotics, improved surgical techniques, and "planned" reoperations have resulted in better survival rates and fewer infectious complications with perihepatic packing than with hepatic resection in the acute postinjury setting.We performed this study to reinforce the concept that perihepatic packing is a valid option in the treatment of patients with uncontrolled hemorrhage due to a major liver injury. We also wanted to determine if the timing of pack removal affected the rate of rebleeding and the incidence of postoperative liver-related complications such as biloma, bile leak, and intra-abdominal abscess.PATIENTS AND METHODSWe performed a retrospective case study of 804 consecutive patients with liver injuries identified at celiotomy who had been operated on by the trauma service at the University of California, Davis, Medical Center from June 1, 1988, to January 1, 1997. Patients younger than 15 years and those who were declared dead on arrival were excluded from the study. We identified patients with liver injuries by reviewing the operative records of patients who had an exploratory celiotomy. We collected clinical information from both the Trauma Registry and from each patient's medical record.We recorded the following information: age, sex, mechanism of injury (blunt vs penetrating), vital signs in the emergency department, operative procedure, intraoperative findings, length of operation, estimated blood loss, amount of blood transfused in the first 24 hours after injury, length of intensive care unit (ICU) and hospital stays, associated injuries, liver complications (such as biloma, persistent bile leak, and hepatic or subphrenic abscess), time to pack removal, and incidence of rebleeding. Complications and rebleeding were defined as follows. Bilomaindicated an intra-abdominal collection of bile that required percutaneous or operative drainage; persistent bile leakindicated a bile fistula that drained longer than 10 days; intra-abdominalor liver abscessindicated a culture-positive right subphrenic or hepatic fluid collection that required drainage; and rebleedingindicated a liver hemorrhage that required replacement of perihepatic packs. Overall severity of injury was measured using the Injury Severity Score (ISS)and predicted survival was determined using the TRISSmethodology and the 24-hour ICU Score.Hypotension in the emergency department was defined as a blood pressure of 90 mm Hg or less. Liver injuries were graded retrospectively based on the description found in the operative report using the scale developed by the Organ Injury Scaling Committee of the American Association for the Surgery of Trauma.The decision to use perihepatic packs was made by the attending trauma surgeon on a case-by-case basis. In general, we opt to pack liver injuries in patients with hepatic hemorrhage that persists after arterial bleeders have been controlled by direct suture ligation. Gauze laparotomy sponges were used to pack liver injuries in all patients treated with perihepatic packing. Perihepatic packs were removed once the patient's hypothermia, acidosis, and coagulopathy had resolved and the attending surgeon felt the patient was ready to return to the operating room. All patients had their packs removed between 24 and 72 hours after injury. If uncontrollable hemorrhage was encountered at the time of pack removal, perihepatic packs were replaced and removed at a subsequent operation.Statistical analysis was performed using the SYSTAT 8.0 software (SPSS Inc, Chicago, Ill). Continuous variables were compared using 2-sample ttests, and discrete variables were compared using either the Fisher exact or Yates corrected χ2tests, as appropriate.RESULTSLIVER INJURY COHORTDuring the 9.5-year study period, 804 patients with liver injuries had a celiotomy. Perihepatic packs were used in 129 patients (16%). Most patients had suffered a blunt injury (68% and 59% in patients with and without packs, respectively; P=.05). When we compared patients whose liver injuries were managed without packs with those who had their liver injuries packed, we found the packed group had a higher mean ± SD ISS (29 ± 15 vs 21 ± 14; P<.001), were more likely to be hypotensive on presentation (52% vs 23%; P<.001), had a higher-grade liver injury (4.2 ± 0.8 vs 1.7 ± 0.8; P<.001), had a larger intraoperative blood loss (4.1 ± 4.1 vs 1.0 ± 2.4 L; P<.001), required more blood transfusions in the 24 hours after injury (16 ± 14 vs 6 ± 4 packed red blood cell units; P<.001), and had a higher mortality rate (60% vs 15%; P<.001). The distribution of the liver injury grades in patients who had their liver injuries packed and those who did not is found in Table 1. We were unable to determine the liver injury grade in 39 of the 804 patients.Table 1. Distribution of Liver Injury Grades for Patients With Liver Injuries Identified During Celiotomy for Patients Managed With Perihepatic Packs vs Those Managed Without Perihepatic PacksLiver Injury GradeNo. (%) of PatientsPacked (n = 123)Not Packed (n = 642)I0 (0)310 (48.3)II0 (0)251 (39.1)III24 (19.5)72 (11.1)IV50 (40.7)6 (1)V48 (39)1 (0.2)VI1 (0.8)2 (0.3)PATIENTS WITH GRADE III-VI LIVER INJURIESWe identified a grade III-VI liver injury in 204 of the 804 patients; perihepatic packs were used in 123 of these patients. Again, patients who had their injuries packed were more severely injured than patients whose injuries were managed without packs. Patients with perihepatic packs had a higher ISS (29 ± 15 vs 22 ± 13; P<.001), were more likely to be hypotensive on presentation (52% vs 28%; P=.002), had a larger intraoperative blood loss (4.1 ± 4.2 vs 2.4 ± 4.5 L; P<.001), and had a more severe liver injury (mean grades of 4.2 ± 0.8 vs 3.2 ± 0.6, respectively; P<.001). Eighty percent of the patients with perihepatic packs had grade IV or V injuries compared with the patients without packs, 89% of whom had a grade III liver injury (Table 1). Patients who had their injuries packed and the group managed without packs were similar with respect to age (35 ± 14 vs 35 ± 15 years, respectively; P=.92), sex (72% vs 79% male, respectively; P=.36), and mechanism of injury (69% vs 56% blunt, respectively; P=.07). Although length of hospital stay was similar for the 2 groups, outcome with respect to mortality favored patients who were managed without liver packs (Table 2).Table 2. Liver-Related Complication Rates, Length of Hospital Stay, and Mortality Rates in Patients With Grade III Through VI Liver Injuries for Patients Managed With Perihepatic Packs vs Those Managed Without Perihepatic PacksPacked (n = 123)Not Packed (n = 81)PLiver-related complications, No. (%)18 (15)5 (6).10Mean ± SD length of hospital stay, d14.1 ± 23.115.4 ± 12.9.62Mortality, No. (%)76 (62)17 (21)<.001LIVER PACKING GROUPPerihepatic packs were used in 129 (16%) of the 804 patients; 35 of the 129 died in the operating room, an additional 25 patients died within 24 hours after injury, and 14 died later in their hospital course. Overall survival was 43% compared with a TRISS-predicted survival rate of 31%. Intraoperative deaths and deaths occurring within the first 24 hours after injury were due to irreversible disseminated intravascular coagulation or to severe head injury. These patients were excluded from further analysis.Sixty-nine patients survived the initial operation and lived longer than 24 hours after injury. The predicted survival rates based on the TRISS method and the 24-hour ICU Score were 47% and 56%, respectively. The actual survival rate for these 69 patients was 75%.In analyzing the 69 patients who survived at least 24 hours after injury, we found survivors and nonsurvivors to have similar characteristics (age, ISS, sex, mechanism of injury, grade of liver injury) with the exception that nonsurvivors were more likely to be hypotensive on presentation (53% vs 21%; P=.03) and required more blood transfusions in the first 24 hours after injury (20 ± 12 vs 12 ± 11 packed red blood cell units; P=.008). The liver complication rates were similar in the 2 groups, with 19 of 52 survivors and 3 of 17 nonsurvivors developing an intra-abdominal abscess or biliary leak (P=.25).Twenty-two patients developed liver-related complications. In comparing patients who developed liver-related complications with those who did not, we found their ISS was similar (25 ± 10 vs 27 ± 14, respectively; P=.53), their liver injury grade score was similar (4.1 ± 0.7 vs 3.9 ± 0.8, respectively; P=.56), and their intraoperative blood loss was similar (3.3 ± 3.4 vs 3.1 ± 3.0 L, respectively; P=.81). Patients who developed liver-related complications had longer hospital courses than those who did not (45 ± 33 vs 17 ± 16 days, respectively; P=.001); however, mortality rates were similar in the 2 groups (14% vs 30%, respectively; P=.23).TIMING OF PERIHEPATIC PACKING REMOVALAll patients had their packs removed between 24 and 72 hours after they were placed. Infection rates and rebleeding complications were similar when patients who had their packs removed within 36 hours were compared with those who had their packs removed between 36 and 72 hours after the initial packing. We hypothesized that leaving the packs in longer would increase the incidence of liver-related complications. However, when we compared patients who had their packs removed within 36 hours after they were placed with those who had their packs removed 36 to 72 hours later, we found the liver-related complication rates to be similar (Table 3).Table 3. Liver-Related Complication Rates, Rebleeding Rates, Length of Hospital and ICU Stays, and Mortality Rates for Patients Who Had Their Perihepatic Packs Removed Within 36 Hours (Early) vs Those Who Had Their Packs Removed 36 to 72 Hours (Late) After Injury*Packing RemovalEarly (n = 39)Late (n = 24)PLiver-related complications13 (33)7 (29).83Rebleeding requiring repacking8 (21)1 (4)<.001Mean ± SD length of ICU stay, d21 ± 2921 ± 15.99Mean ± SD length of hospital stay, d27 ± 3126 ± 19.97Mortality7 (18)7 (29).11*Data are presented as number (percentage) of patients unless otherwise indicated. ICU indicates intensive care unit.In comparing the 39 patients who had their packs removed within 36 hours (early) with those who had their packs removed 36 to 72 hours after they were placed (late), we found the 2 groups to be similar in age (31 ± 12 vs 33 ± 10 years, respectively; P=.47), mechanism of injury (56% vs 83% blunt injury, respectively; P=.15), presenting hypotensive (23% vs 33%, respectively; P=.20), ISS (25 ± 10 vs 25 ± 14, respectively; P=.96), liver injury grade (3.9 ± 0.7 vs 4.0 ± 0.8, respectively; P=.78), intraoperative blood loss (3.0 ± 2.3 vs 3.6 ± 4.3 L, respectively; P=.45), and amount of blood transfused in the first 24 hours after injury (14 ± 13 vs 13 ± 15 packed red blood cell units, respectively; P=.99).Despite the similarity in the number of liver-related complications in the 2 groups, the rate of recurrent liver hemorrhage requiring repacking was markedly higher in the group of patients who had their packs removed early (Table 3). Morbidity as measured by length of stay (hospital and ICU) and mortality rates were similar in the 2 groups (Table 3). Liver packs were successfully removed 2 days after they were replaced in patients who required repacking due to recurrent liver hemorrhage.COMMENTEven with the move to nonoperative management of liver injuries,we still encounter a subset of patients who require operative intervention to control liver hemorrhage. In fact, exsanguination from major liver injury is the most common cause of intraoperative death.Some patients with liver injuries will have continued hemorrhage even after direct ligation of bleeding vessels. Many of these patients have a coagulopathy and are acidotic and hypothermic. These patients benefit from a "damage control" strategy and from restoration of homeostasis in an ICU. Perihepatic packing can be used to tamponade and control hepatic hemorrhage and abbreviate the operation.Our goal was to determine if perihepatic packs increase the risk of intra-abdominal abscesses and bile leaks after major liver injuries, and to determine if the risk of liver-related complications increases the longer packs are left in place.Our findings support but do not prove the contention that perihepatic packing is a useful strategy in managing major liver injuries. Overall survival for the group of patients whose liver injuries were packed, including those who died in the operating room while an attempt was being made to pack their injuries, compares favorably with the TRISS-predicted survival for these patients. The difference between predicted survival (calculated using both the TRISS and 24-hour ICU Score methods) and actual survival rates was even more dramatic when we excluded patients who died in the operating room or the ICU within 24 hours after injury. Although comparisons of actual mortality rates with predicted mortality rates (TRISS or 24-hour ICU Score) cannot be used to establish the validity of a treatment strategy, it is reassuring to know that patients who had their liver injuries packed did not have a higher than predicted mortality.The mortality rates for both patients who had their injuries packed and those who did not were high. Their mortality rates are consistent with the severity of injury encountered in this select group of patients. By excluding patients who had liver injuries managed nonoperatively, we eliminated patients with less severe injuries and many of those with severe yet isolated liver injuries.To identify a more equivalent group of patients to compare with the group of patients who had their liver injuries packed, we narrowed our comparison to include only patients with a liver injury grade of III or more. Not surprisingly, patients who had their liver injuries packed were more severely injured and therefore had a higher mortality rate; however, we were surprised to find that the liver-related complication rates were similar in patients managed with or without liver packs. This probably represents a statistical aberrancy because many patients in the group treated with perihepatic packing die early after injury and do not have the opportunity to develop complications. Thus, the findings suggest that patients managed with perihepatic packing have a higher rate of liver-related complications; however, because patients managed with liver packs had more severe injuries and a worse physiologic status on presentation, it is not clear whether the higher complication rate should be attributed to the packs, the more severe injuries, or a combination of the two.Patients who developed liver-related complications had a longer hospital course; however, patients who developed liver complications did not have a higher mortality rate than those who did not develop complications.Many surgeons recommend that liver packs be removed as quickly as possible to reduce the rate of infectious complications.Our findings support previously made recommendations that the timing of pack removal is not as critical as reversing the detrimental effects of severe hemorrhagic shock and that the packs should be removed after "hemodynamic stability" is achieved.Based on our findings that the incidence of rebleeding was significantly greater in patients whose packs were removed within 36 hours, we recommend that the packs be removed between 36 and 72 hours after injury. When packs were removed more than 36 hours after injury, the rebleeding rate was low and the liver complication rate was similar to that of patients who had their packs removed within 36 hours of injury.CONCLUSIONSPerihepatic packing and "damage control" laparotomy can be lifesaving in patients who require urgent laparotomy because of persistent hypotension due to a major liver injury. Perihepatic packing seems to be associated with an increased risk of liver-related complications such as subphrenic or hepatic abscesses, bilomas, and bile fistulas; however, in our experience these complications were not overwhelming and were not associated with an increase in mortality. We recommend using perihepatic packing in patients with major liver injuries, especially in patients who are acidotic, coagulopathic, and hypothermic. The packs should be removed 36 to 72 hours after being placed.AJWaltThe mythology of liver trauma, or Babel revisited.Am J Surg.1978;135:12-18.JSaifiJBFortuneLGracaDMShahBenefits of intra-abdominal pack placement for the management of nonmechanical hemorrhage.Arch Surg.1990;125:119-22.GFMaddingWHPenistonLiver hemostasis.Surg Gynecol Obstet.1957;104:417-24.GFMaddingKBLawrencePAKennedyWounds of the liver.In: Forsee JH, ed. Forward Surgery of the Severely Wounded, Second Auxiliary Surgical Group. Washington, DC: US Army; 1945:303-326.AJWaltThe surgical management of hepatic trauma and its complications.Ann R Col Surg Engl.1969;45:319-339.RRIvaturyMNallathambiYGunduzRConstableMRohmanWMStahlLiver packing for uncontrolled hemorrhage: a reappraisal.J Trauma.1986;26:744-753.JASvobodaETPeterCVDangSNParksJHEllysonSevere liver trauma in the face of coagulopathy: a case for temporary packing and early reexploration.Am J Surg.1982;144:711-721.RHCarmonaDZPeckRCLim JrThe role of packing and planned reoperation in severe hepatic trauma.J Trauma.1984;24:779-784.DVFelicianoKLMattoxJMBurchCGBitondoGLJordan JrPacking for control of hepatic hemorrhage.J Trauma.1986;26:738-743.DVFelicianoKLMattoxGLJordan JrJMBurchCGBitondoPACruseManagement of 1000 consecutive cases of hepatic trauma (1979-1984).Ann Surg.1986;204:438-445.THCogbillEEMooreGJJurkovichDVFelicianoJAMorrisPMuchaSevere hepatic trauma: a multi-center experience with 1,335 liver injuries.J Trauma.1988;28:1433-1438.JLCueHGCryerFBMillerJDRichardsonHCPolk JrPacking and planned reexploration for hepatic and retroperitoneal hemorrhage: critical refinements of a useful technique.J Trauma.1990;30:1007-1013.SPBakerBO'NeillThe injury severity score: an update.J Trauma.1976;6:882-885.CRBoydMATolsonWSCopesEvaluating trauma care, the TRISS method: Trauma Score and the Injury Severity Score.J Trauma.1987;27:370-378.MJVassarCLWilkersonPJDuranCAPerryJWHolcroftComparison of APACHE II, TRISS, and a proposed 24-hour ICU point system for prediction of outcome in ICU trauma patients.J Trauma.1992;32:490-499.EEMooreTHCogbillGJJurkovichSRShackfordMAMalangoniHRChampionOrgan injury scaling: spleen and liver (1994 revision).J Trauma.1995;38:323-324.EHCarrilloAPlatzFBMillerJDRichardsonHCPolk JrNon-operative management of blunt hepatic trauma.Br J Surg.1998;85:461-468.RRutledgeJPHuntCWLentzA statewide, population-based time-series analysis of the increasing frequency of nonoperative management of abdominal solid organ injury.Ann Surg.1995;222:311-326.DBHoytEMBulgerMMKnudsonDeath in the operating room: an analysis of a multi-center experience.J Trauma.1994;37:426-432.JRGarrisonJDRichardsonASHilakosPredicting the need to pack early for severe intra-abdominal hemorrhage.J Trauma.1996;40:923-929.EDegiannisRDLevyGCVelmahosTMokoenaADaponteRSaadiaGunshot injuries of the liver: the Baragwanath experience.Surgery.1995;117:359-364.FMenegauxPLangioisJPChigotSevere blunt trauma of the liver: study of mortality factors.J Trauma.1993;35:865-869.RHCarmonaRCLim JrGCClarkMorbidity and mortality in hepatic trauma: a 5-year study.Am J Surg.1983;144:88-94.VBaracco-GandolfoOVidarteVBaracco-MillerMdel CastilloProlonged closed liver packing in severe hepatic trauma: experience with 36 patients.J Trauma.1986;26:754-756.Ronald V. Maier, MD, Seattle, Wash:This is an important paper by Dr Caruso and his colleagues dealing with a very important problem and a very frustrating problem. We have improved outcomes from severe injury, particularly multiple system injury, dramatically over the last 3 to 4 decades. However, exsanguination from severe hepatic injury continues to frustrate us and it is one of the very few injuries that still has a high mortality due to uncontrollable exsanguination. As stated by the authors, we have now come full cycle in our approach to this injury. At the turn of the century, packing was adopted as lifesaving but had an excessive list of complications, primarily infectious, because the packs were left in place for 1 to 2 weeks to ensure hemostasis; other approaches such as advanced ligature techniques were not available to the surgeon at that time. As surgical techniques improved, perioperative support advanced, and blood banking with component therapy for coagulopathy and other issues developed, the approaches developed for elective hepatic resection were then adopted for the treatment of the injured liver. Major hepatic resections became the norm for the injured liver. But as trauma systems improved and patients with increasingly severe liver injury survived to reach the OR, the surgeon is faced with injuries that could not be definitively treated as in the elective setting. There was not excellent hemostasis, and patients with massive blood loss or other systemic injuries would rapidly became hypothermic, coagulopathic, and subsequently acidotic and die of their ongoing blood loss. Recently, identification of the patient with massive liver injury at high risk for hypothermia and coagulopathy early in the procedure has enabled us to control the predominantly venous bleeding by packing and to use full ICU support and subsequent "early" removal to prevent complications that had been identified in the past. While unproven by any prospective randomized trial, those who frequently encounter these challenging cases rapidly adopted this "new" technique of early packing, which of course was not new but just a reiteration of a century-old approach. The focus is to avoid entering the vicious spiral that leads to the death of these patients.It is unfortunate that we are using a technique that has never been studied in a prospective, randomized trial with matched cohorts to prove that it is the most effective approach. Similarly, in this study, this same assumption based on no hard data was undertaken by these authors with no attempt to test whether the packing is beneficial compared with definitive repair or resection. As I mentioned, this is unfortunate, since looking at the numbers, one might not be impressed. The mortality in the packed group was 60% compared with 7% in the nonpacking group. If one were to present that to the local HMO, one would quickly see which approach would be preferred. Even when the authors limited the analysis to grade III-VI severity injuries, the mortality was 3 times higher in the packed group in these similarly injured livers. Of course, as we frequently argue in such issues, packing must have been done in the more severely injured patients. The authors have somewhat documented this by using injury stratification such as ISS, blood loss, and hypotension. Yet we are still missing a true matched cohort to prove we have made the right assumption. The authors state they did this study to reinforce the concept of packing as a valid concept, but do they have any truly matched cohort data to prove the concept is correct?For those patients who survived at least 24 hours, the mortality was only 25%. I looked at a similar 5-year population at Harborview Medical Center in Seattle, Wash, and we had very similar results with a 22% mortality in those who survived 24 hours. But, an important issue is what of those who died during the first 24 hours? Those who died of irreversible DIC? Early packing is meant to avoid irreversible DIC. How many of the 60 patients they excluded who died early died of DIC? Could DIC have been prevented and could the patients have been saved if packing had occurred earlier in the operation? Was there too much delay during the operation due to intraoperative attempts to definitively treat the injuries? Do the authors have any data on the timing of packing during the case? I constantly reinforce to residents that they need to make the diagnosis, identify the severity of injuries, and start packing within 5 to 10 minutes of opening the abdomen to be most effective. Do the authors have any data as to whether packing earlier in the course of the operation was more effective than starting later when the patient was already becoming coagulopathic, particularly since they had a significant number of patients die of their irreversible DIC postoperatively and bleeding intraoperatively?However, the main goal of the study was to determine if timing of pack removal affected the rate of rebleeding and the incidence of postoperative complications such as biloma, bile leak, and abscesses. The authors demonstrate that rebleeding occurred more frequently, 21% vs 4% in the early removal group. This is what one would expect. The overall importance of complications was confirmed in that while there was no effect on mortality, there was a significant increase in length of stay, 45 days with complications vs 17 days without. This avoidance of complications is important. They noted that there was no difference in complications between packs removed early (<36 hours) than late (36-72 hours). This removal at 36 to 72 hours appears to be safest with no increase in complications. Unfortunately the numbers of patients become very small in this final analysis, 39 patients vs 24. While I believe the results, the risk of a β error is significant and I wondered if the authors could comment on this potential problem.In addition, because of our fear of reproducing the complications of late pack removal, we have adopted a very conservative approach in these patients to the point of almost rushing the patient back emergently to the OR to remove the packs. When one says early, does one mean less than 12 hours or 24 hours? Does one mean a reasonable 36 to 72 hours as the authors propose? The "36- to 72-hour group" I would argue is not an extended removal but still early removal. Thus the authors in this study have aided us greatly in making a rational timing decision for reoperation. What is not addressed is what is the safe limit at the far end? We at Harborview have noted that there is a noninfectious systemic toxicity or systemic inflammatory response that appears to occur after 3 to 4 days in response to these retained sponges as foreign bodies in the peritoneal cavity. Even though they are not infected, the patients appear to be at risk for this systemic toxicity and damage to other organs such as the lung and liver itself. Do the authors have any data on how long is safe? Should we go longer than 3 days to further ensure hemostasis and sealing of the liver parenchyma? When do infectious and other complications begin? When do we fall into the trap of truly reinventing the wheel and leaving them in too long?Steven N. Parks, MD, Fresno, Calif:How do you determine the time to bring the patient back to remove the packs? We have tried to correct the coagulopathy, be sure there is no acidosis, and make sure there is no more hypothermia before we bring them back to remove the packs. I have tried to remove the packs at 36 to 48 hours, not going as long as 72, if possible. One of the reasons for taking the packs out earlier is because they smell bad at 72 hours, but I don't disagree that they should be left in until the appropriate time. Would you tell us what the appropriate time is?Demetrios Demetriades, MD, Los Angeles, Calif:Most importantly this paper agrees with my biases and principles. I do believe that the greatest advancement in liver injuries in the last 10 years is the very early packing. In my opinion a good experienced trauma surgeon will do a quick evaluation and within minutes should pack and come out. At our center the patient will go from the OR to the angiography suite. Do the authors have any experience with that?Charles Scudamore, MD, Vancouver, British Columbia:I was interested in the authors' experience with their bile fistulae. We have been able to avoid operating on significant bile fistulae just by sphincterotomy and/or stenting the bile duct. They all close spontaneously. I wondered if the authors have had a similar experience.John Mayberry, MD, Portland, Ore:It is important that we keep discussing this even though it is well established in our practices. I wonder if the authors stratified their patients according to the Abdominal Trauma Index or other abdominal injury such as bowel injury in regards to the development of abscess and other complications. What were the causes of death?Thomas Berne, MD, Los Angeles:I have 2 questions. When you said "rebleed," it sounded like you meant that the rebleeding occurred as you were removing the pack. Or was it just "rebleeding" between at some other time? The other question is did hollow viscus injury affect the timing of the removal of packs?Wilton A. Doane, MD, Santa Barbara, Calif:According to the record, mortality was higher in patients who did not develop complications. That seems a little puzzling.Dr Battistella:As usual, Dr Maier, you hit the nail on the head with all of your comments. I will take your questions a little bit out of order because it will allow me to address some of the issues simultaneously.First of all you pointed out correctly that we did not show that the perihepatic packing technique is any better than any other technique with respect to controlling hemorrhage. That was not the goal of our study. We were as disturbed as you are by the fact that we still have a high rate of persistent uncontrolled bleeding, both in the operating room and postoperatively, with this technique. We had 35 patients who died in the operating room. All of those died of uncontrolled DIC. Many of the patients presented moribund. Some presented with no blood pressure, a few had ER thoracotomies, etc, etc. Many were the type of patient who would be very difficult to save under the best of circumstances. But, nonetheless, we are not convinced that perihepatic packing is the end-all treatment for all liver injuries and, in fact, it may be that some combination of packing and hepatic artery ligation or, as Dr Demetriades pointed out, maybe packing and prompt angio to embolize specific bleeders may be the best treatment option in patients who do not respond to liver packing.As we looked at the patients with respect to the decision to pack them, all patients who were packed had the decision made early in their course. The total operative time for all patients regardless of how you analyze them, whether they were packed and survived or were packed and developed complications, was about 90 minutes. That represents the time the patient spent in the operating room (from entering and leaving the operating room). I can't tell you how quickly the decision to pack the patient was made, but I can tell you that the total amount of time the patients spent in the operating room was fairly short.The second issue that you raised is whether our study suffers from a β error with respect to the timing of pack removal. All I can tell you is that our series is a fairly large series as compared with others that are reported in the literature. The point I am trying to make is that liver injuries that require packing are rare and in order to get an answer to that question that won't be subject to a β error we will require a prospective multicenter trial. Over a 10-year period we were able to accumulate an experience with 69 patients in an institution that is fairly busy. If you look at some of the other series, the Feliciano series, the Cogbill series, etc, etc, their numbers are pretty similar. Some of those are multicenter series. Therefore, in order to avoid a type II error, we have to organize a much larger patient base.Dr Parks, you asked about the timing of pack removal. It goes without saying and I apologize if we didn't emphasize this point, but aggressive efforts are made to resolve the patient's hypothermia, acidosis, and coagulopathy before removing the packs. Then the next question is when the patient is looking good and the hypothermia, acidosis, and coagulopathy are resolved, do you need to wait any longer? In some of our patients, we obviously didn't and the conclusion that we draw from our experience is that you probably should until 36 to 72 hours after the packs were placed to allow the clot to mature.Dr Berne asked the question about the rebleeding and how we defined that. Rebleeding was defined as those patients who had to be repacked at the time that an attempt was made to remove the packs because of recurrent uncontrollable hemorrhage.Dr Demetriades asked about the use of angiography. I addressed that question in part already, but let me go on to say that we have not taken full advantage of our interventional radiologists when dealing with severe liver injury. We utilize angio in those patients who have evidence of continued bleeding after they are packed. We do not routinely angio all patients who have been packed. Maybe an approach that utilizes angio in all patients with liver injuries that require packing might avoid delays in those patients who do have continued arterial bleeding. So we take your suggestion not lightheartedly.Dr Scudamore asked about bile fistulae. We have a limited experience with stenting the common bile duct in patients with persistent bile leaks. I do think that it makes a difference. We didn't address that in this paper, but I can tell you from my experience with anecdotal cases that patients who had prolonged bile fistulae resolved very quickly once their duct was stented. I do think that stenting the common bile duct in those patients is effective in resolving the bile leak.Dr Mayberry asked if we stratify patients according to their Abdominal Trauma Index score. We did not and we probably should have. I think that also addresses Dr Berne's question regarding concomitant bowel injuries. We did not specifically address the associated hollow viscus injury, nor did we specifically address the Abdominal Trauma Index.Dr Doane, you asked about the fact that we had a higher mortality rate in the group of patients who did not develop complications. I think this has a simple explanation. As you recall, most of the deaths occur early in the process and those patients really don't have an opportunity to develop complications. There is no survival advantage if you develop a complication.Presented at the 70th Annual Session of the Pacific Coast Surgical Association, San Jose del Cabo, Baja California Sur, February 15, 1999.Reprints: Felix D. Battistella, MD, Department of Surgery, University of California, Davis, Health System, 4301 X St, Sacramento, CA 95817-2214 (e-mail: [email protected]).
Showing 1 to 10 of 24 Articles