Role of Early Multisystem Organ Failure as Major Risk Factor for Pancreatic Infections and Death in Severe Acute Pancreatitis

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2006;4: Role of Early Multisystem Organ Failure as Major Risk Factor for Pancreatic Infections and Death in Severe Acute Pancreatitis BETTINA M. RAU,* ASTRID BOTHE, MARTINA KRON, and HANS G. BEGER *Department of General, Visceral, and Vascular Surgery, University of the Saarland, Homburg/Saar; and Department of General Surgery and Department of Biometry and Medical Documentation, University of Ulm, Ulm, Germany Background & Aims: Infection of necrosis is considered as principal determinant of outcome in necrotizing pancreatitis and as potential complication after operative treatment of sterile necrosis. In this report a new concept is proposed. Methods: Of 392 patients with necrotizing pancreatitis, 135 patients with operatively treated sterile necrosis were stratified into 3 postoperative entities: secondary pancreatic infections (PIN, group I), pancreatic contaminations (group II), and sterile courses (group III). Ninety-five patients with conservatively treated sterile necrosis (group IV) served as controls. Results: Secondary PIN developed in 64 (47%) patients and contaminations in 37 (27%) patients, whereas 34 (25%) patients remained sterile postoperatively. Secondary PIN and contaminations were both diagnosed after a median of 3 weeks after disease onset. Early/ preoperative multisystem organ failure (MODS) affecting >2 organs was more frequent in group I (35%) than in group II (5%), group III (12%), and group IV (7%) (P <.003); mortality rates were 38%, 3%, 21%, and 7%, respectively (P <.001). Multiple logistic regression identified early/preoperative MODS and extent of intrapancreatic necrosis as major risk factors to develop secondary PIN in operatively treated sterile necrosis. However, irrespective of operative or conservative treatment, only early onset MODS >2 organs proved to be the predominant risk factor for death. Conclusions: Early MODS and extended intrapancreatic necrosis are risk factors for secondary PIN after operative treatment of sterile necrosis. In contrast, the ultimate outcome strongly depends on early and high systemic illness, whereas local pathology and operative procedure seem to be less important. Infection of necrosis is considered as principal risk factor of morbidity and mortality in necrotizing pancreatitis. Ever since the first report on predisposing factors, natural course, and consequences of infected necrosis was published 20 years ago, 1 a wealth of clinical data have confirmed the detrimental role of this morphologic complication on outcome 2 4 and therapeutic decision making in acute pancreatitis Although treatment of necrotizing pancreatitis has become increasingly conservative during the past decade, 9,12 17 general agreement exists that documented infection of pancreatic necrosis in the presence of systemic signs of sepsis is the only accepted indication for operative intervention. 18,19 Even if the presence of intrapancreatic and extrapancreatic necrosis is substantial for the development of pancreatic infections, the majority of patients can be successfully managed by conservative means as long as the necrotic process remains sterile A limited extent of intrapancreatic necrosis and a low incidence of mostly transient organ failure are common characteristics found in this specific group of patients. 3,9,12,13 Very recently, a number of reports have underscored the role of early organ failure as an important prognostic factor in acute pancreatitis. Persisting organ failure beyond the first week after disease onset has been shown to be a reliable predictor of nonsurvival in severe acute pancreatitis On the other hand, persisting organ failure despite maximum intensive care support has been advocated as indication for surgery in sterile necrosis by some authors, 10,13 which is still the most controversial issue in the management of this disease. There is clearly no more doubt that operative treatment is not necessary in the majority of patients with sterile necrosis. However, a subgroup of patients with clinically severe disease not responding to conservative management exists in whom the appropriate treatment is still under discussion. 11,23 In this specific setting, opponents of operative intervention have raised substantial concern that surgery might convert the sterile process Abbreviations used in this paper: APACHE II, Acute Physiology and Chronic Health Evaluation II; CE-CT, contrast-enhanced computed tomography; FNA, fine-needle aspiration; MODS, multiorgan dysfunction syndrome; PIN, pancreatic infection; SIRS, systemic inflammatory response syndrome by the American Gastroenterological Association Institute /06/$32.00 doi: /j.cgh

2 1054 RAU ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 8 into an infected one with a concomitantly increased risk of morbidity and mortality, thus being harmful rather than beneficial. 9,14 Hence, no study has ever investigated the incidence and severity of secondary PIN in patients with sterile necrosis subjected to operative treatment. The present report addressed this topic, which, in addition, questions our current understanding of the pathophysiology and role of infection of necrosis in severe acute pancreatitis. Patients and Methods Patients A total of 1520 patients with acute pancreatitis were treated at the Department of General Surgery, University Hospital of Ulm, Germany, between May 1982 and May There were 1128 patients with interstitial edematous and 392 patients with computed tomography and/or intraoperatively proven necrotizing pancreatitis. In 107 patients treatment was nonoperative, 96 had sterile necrosis (mortality, 7.3%), and 11 patients had fine-needle aspiration (FNA) proven infection of necrosis (mortality, 9.1%). Operative treatment was performed in 285 patients with necrotizing pancreatitis, intraoperative bacteriology revealed primary infected necrosis in 140 (49%) patients, whereas necrosis was sterile in 145 (51%) patients. In the patient group with intraoperatively proven sterile necrosis, a complete postoperative bacteriologic follow-up was available in 135 patients who underwent further analysis. In the nonoperatively treated patient group 95 patients were analyzed. Data acquisition was retrospective from May 1982 until December 1992; since January 1993 patients were studied prospectively. Definition of Postoperative Infective Entities Microbial cultures were routinely obtained from indwelling lavage catheters in the lesser sac or any other retroperitoneal/abdominal compartment in all patients at least once per week until complete removal. Cultures from pancreatic/ peripancreatic necrosis or abscesses were taken during any reoperation. Whenever bacteria or fungi were isolated, the culture was considered positive. Depending on the postoperative course, all 135 patients with intraoperatively proven sterile necrosis were classified into 3 different entities: group I, secondary PIN were defined as bacteriologically proven intraabdominal infections that required relaparotomy, interventional drainage, or prolonged ( 40 days) continuous closed lavage of the lesser sac; group II, pancreatic contaminations were defined as patients with positive cultures from abdominal drains who did not require relaparotomy, intervention, or prolonged closed lavage; and group III, all patients in whom no positive cultures from abdominal drains were cultured throughout the whole postoperative course comprised the sterile group. In the conservatively treated group (group IV), sterile necrosis was defined by negative FNA (n 44) or an uneventful clinical course without clinical or laboratory evidence of pancreatic infections. Definition of Organ Failure and Systemic Complications Remote organ failure was defined as follows: (1) pulmonary failure: pao 2 60 mm Hg at room air or mechanical ventilation; (2) renal failure: creatinine 180 mol/l or hemofiltration/dialysis; and (3) shock: systolic blood pressure 80 mm Hg over 15 minutes or pressure support. Multiorgan dysfunction syndrome (MODS) was defined as the presence of 2 or more organ failures. Definitions of systemic complications were systemic inflammatory response syndrome (SIRS) as described previously 11 and sepsis as positive blood culture and/or positive FNA/ positive intraoperative bacteriology in the presence of SIRS. Definition of the Extent of Intrapancreatic and Extrapancreatic Necrosis The extent of intrapancreatic (pancreatic parenchymal) necrosis was based on preoperative contrast-enhanced computed tomography (CE-CT) in 92 patients, which corresponded well with the intraoperative findings. In 26 patients with inconclusive CT findings and in 17 patients without preoperative CE-CT, the extent of intrapancreatic necrosis was assessed intraoperatively; in 1 patient with contamination (group II) no detailed data were available. The compartments/ areas affected by extrapancreatic fatty tissue necrosis were based on intraoperative findings only. In 5 patients of group I, in 1 patient of group II, and in 3 patients of group III no detailed intraoperative data were available. All conservatively treated patients (group IV) had undergone CE-CT. Treatment Initial treatment was conservative in all patients and comprised maximum intensive care unit support. Prophylactic antibiotics were administered since Indications for necrosectomy in patients with sterile necrosis were persisting or deteriorating organ failure and/or abdominal complications, despite maximum intensive care unit treatment during a time period of up to 7 days in the absence of FNA-proven infection of necrosis. Otherwise, FNA-proven infection of intrapancreatic/extrapancreatic necrosis in the presence of SIRS was an indication for surgery. Surgical treatment and relaparotomies were performed as reported previously. 11,24 Statistics Statistical analysis was performed by using the Kruskal-Wallis test for comparison of continuous variables between subgroups. The 2 test was applied to compare proportions between groups. P values.05 were considered significant. Multiple logistic regression with forward selection was performed to define risk factors of secondary PIN and nonsurvival in the group of operatively treated patients with sterile necrosis (n 135) and in all patients with sterile necrosis, irrespec-

3 August 2006 EARLY MODS AND SEVERE ACUTE PANCREATITIS 1055 tive of the treatment (conservative, n 95; surgical, n 135). The influence of the following variables on the development of secondary PIN was analyzed in the surgical group: age, etiology of pancreatitis, early/preoperative pulmonary failure and mechanical ventilation, early/preoperative renal failure and dialysis/hemofiltration, early/preoperative shock and pressure support, early/preoperative MODS 2 organs, extent of intrapancreatic necrosis (30% vs 30% 50% vs 50%), extent of extrapancreatic necrosis (0 1 area vs 2 areas vs 2 areas), timing of surgery after onset of symptoms (1 week vs 2 weeks vs 3 weeks and later), overall incidence of mechanical ventilation, dialysis/hemofiltration, pressure support, and overall intensive care unit dependent MODS 2 organs. Secondary PIN and presence of bacteria were additional variables included to identify risk factors for death. In the total group of patients with sterile necrosis, the same factors as mentioned above including the kind of treatment (operative versus conservative) were included in the multiple logistic regression analysis. Extent of extrapancreatic necrosis was excluded from the analysis as a result of too many missing values (104 of 230 observations). Results Postoperative Infection Subgroups, Biologic Data, Etiology After necrosectomy for sterile necrosis, 64 patients (47%) developed secondary PIN (group I), 37 patients (27%) had contaminations (group II), and 34 patients (25%) remained sterile (group III) throughout the course of the disease. comprised 37 men and 27 women with a median age of 52 years (range, years); group II consisted of 26 men and 11 women with a median age of 48 years (range, years); and group III comprised 22 men and 12 women with a median age of 48 years (range, 7 80 years). Etiologic factors were alcohol in 26 (41%), gallstones in 18 (28%), and others in 20 (31%) patients of group I; these respective etiologies were found in 20 (54%), 6 (16%), and 11 (30%) patients of group II and in 19 (56%), 9 (27%), and 6 (18%) patients of group III. In group IV there were 65 men and 30 women with a median age of 47 years (range, years); etiology was alcohol in 52 (55%), gallstones in 21 (22%), and others in 22 (23%) patients. There was no difference concerning biologic data and etiology between all groups. Extent of Pancreatic Necrosis The extent of intrapancreatic and extrapancreatic necrosis is summarized in Table 1. Extended intrapancreatic necrosis 50% of the pancreas was found in almost half of the patients of group I. In contrast, limited necrosis of less than 30% was more often observed in patients of groups II and III. In the Table 1. Extent of Intrapancreatic and Extrapancreatic Necrosis in Operatively Treated Patients With Secondary PIN (I), Contaminations (II), Sterile Course (III), and in Nonoperatively Treated Patients With Sterile Necrosis (IV) I (n 34) V (n 95) Necrosis Intrapancreatic % a 20 (33%) 20 (57%) 22 (71%) 64 (67%) 30% 50% 14 (23%) 6 (17%) 4 (13%) 15 (16%) 50% b 27 (44%) 9 (26%) 5 (16%) 11 (12%) Extrapancreatic area c 25 (42%) 20 (56%) 25 (81%) na 2 areas 10 (17%) 6 (17%) 4 (13%) na 3 areas c 24 (41%) 10 (28%) 2 (6%) na na, not available. a P.005 between groups I III. b P.02 between groups I III. c P.005 between groups I III. conservatively treated group IV the extent of necrosis was comparable with groups II and III. In patients with secondary PIN (group I), the number of extrapancreatic compartments affected by fatty tissue necrosis was higher than in groups II and III. Incidence, Severity, and Persistence of Organ Failure Preoperative disease severity in terms of remote organ failure in all patient groups is shown in Table 2. Organ failure occurred early after onset of symptoms in all groups; the incidence and the severity of preoperative organ failures were higher in group I compared with patients of groups II and III. The lowest incidence of organ failure was present in group IV. Accordingly, preoperative disease severity in terms of Ranson scores and Acute Physiology and Chronic Health Evaluation II (APACHE II) scores within the first 3 days after hospital admission was highest in group I (Table 3). The overall incidence (preoperative and postoperative in operatively treated groups) and the onset of severe organ failure in all groups are shown in Table 4. The incidence of each severe organ failure was again higher and persisted during a longer time interval in group I than in the other groups. Timing of Admission and Treatment The time interval between onset of symptoms and admission to our institution was 2.1 days (median, range 0 35 days) in patients of group I, 2.4 days (median, range days) in patients of group II, 2.5 days (median, range days) in group III, and 2.0 days (median, range 0 38 days) in group IV, which was not

4 1056 RAU ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 8 Table 2. Early/Preoperative Organ Failure in Patients With Secondary PIN (I), Contaminations (II), and Sterile Course (III) Before Surgical Treatment of Sterile Necrosis and in Nonoperatively Treated Patients With Sterile Necrosis (IV) I (n 34) V (n 95) n (%) n (%) n (%) n (%) Pulmonary failure a 52 (81) 25 (68) 17 (50) 47 (49) Onset on day of AP, median (range) 3 (1 15) 3 (1 25) 3 (1 8) 4 (1 26) Mechanical ventilation a 42 (65) 15 (41) 11 (32) 14 (15) Renal failure b 29 (46) 8 (22) 6 (18) 12 (13) Onset on day of AP, median (range) 3 (1 10) 2.5 (1 4) 3.5 (1 8) 2 (1 17) Dialysis/hemofiltration b 14 (22) 1 (3) 2 (6) 5 (5) Shock c 31 (49) 8 (22) 8 (24) 15 (16) Onset on day of AP, median (range) 3 (1 45) 3.5 (1 26) 3 (1 8) 4.5 (1 22) Pressure support d 28 (44) 7 (19) 7 (21) 13 (14) MODS e 36 (56) 11 (30) 10 (29) 20 (21) MODS 2 organs 14 (22) 9 (24) 6 (18) 13 (14) MODS 3 organs 22 (35) 2 (5) 4 (12) 7 (7) AP, acute pancreatitis. a P.006 between groups I III. b P.008 between groups I III. c P.007 between groups I III. d P.02 between groups I III. e P.008 between groups I III. different. Necrosectomy was performed 5.4 days (median, range 1 44 days) after disease onset in group I, 6.0 days (median, range days) in group II, and 7.1 days (median, range days) in group III, with no differences between the groups. Preoperative antibiotics were administered in 78% (50/64) of the patients in group I for 2 days (median, range 1 10 days), in 62% (23/37) of the patients in group II for 2 days (median, range 1 7 days), and in 59% (20/34) of the patients in group III for 2 days (median, range 1 9 days), which was not different as well. Table 3. Staging Scores After Hospital Admission in Patients With Secondary PIN (I), Contaminations (II), and Sterile Course (III) Before Surgical Treatment of Sterile Necrosis and in Nonoperatively Treated Sterile Necrosis (IV) I (n 34) V (n 95) Ranson a 6 (2 10) 4 (1 9) 3 (0 9) 3 (0 11) APACHE II 24h b 14 (1 28) 10 (0 20) 7 (3 22) 8 (3 22) APACHE II 48h c 11 (0 27) 8 (0 15) 8 (2 15) 7 (2 15) APACHE II 72h d 11 (2 21) 9 (0 16) 8 (1 16) 7 (1 16) NOTE. Data are presented as medians and ranges. a P.0001 between groups I III. b P.001 between groups I III. c P.005 between groups I III. d P.002 between groups I III. Microbiologic Findings The time interval between disease onset and occurrence of infection/positive intra-abdominal bacteriology and the duration of antibiotic treatment up to this point in patients with primary infected necrosis (n 140), secondary PIN (group I), and contaminations (group II) are shown in Table 5. All infectious entities were diagnosed in median at the end of the third week after disease onset. After surgery for sterile necrosis, secondary PIN occurred after 16 days (median, range 2 72 days) and contaminations after 13 days (median, range 3 34 days), which was not different. Microbiologic findings in patients with secondary PIN (group I) and contaminations (group II) are shown in Table 6 and did not show a difference. Mortality Mortality rate significantly differed among the different groups. Only 1 of 37 patients died in group II (3%), whereas 24 of 64 patients (38%) in group I and 7 of 34 (21%) patients in group III did not survive (P.003). Among conservatively treated patients with sterile necrosis (group IV), 7 of 95 (7%) patients died. Causes of death in group I were septic MODS in 23 and hypovolemic shock caused by major bleeding in 1 patient occurring 44 days (median, range 5 118) days after disease onset. In group II the only death observed was the consequence of septic MODS as a result of severe pulmonary infection and acute respiratory distress syndrome on day 35 of pancreatitis. In group III all patients died of non-septic (toxic) MODS on day 10 (median, range 3 26 days) after onset of symptoms. In group IV causes of death were non-septic (toxic) MODS in 3 patients, pulmonary embolism in 2 patients, septic MODS as a result

5 August 2006 EARLY MODS AND SEVERE ACUTE PANCREATITIS 1057 Table 4. Overall Incidence, Onset, and Persistence of Severe Organ Failure in Operatively Treated Patients With Secondary PIN (I), Contaminations (II), Sterile Course (III), and in Nonoperatively Treated Sterile Necrosis (IV) I (n 34) V (n 95) n (%) n (%) n (%) n (%) Mechanical ventilation a 60 (94) 28 (76) 18 (53) 14 (15) Start on day of AP b 3 (1 45) 3 (1 26) 4 (1 22) 6 (2 85) Persistence/days b,c 26 (1 173) 8 (1 35) 4 (1 43) 7 (1 47) Dialysis/hemofiltration a 26 (41) 2 (5) 4 (12) 5 (5) Start on day of AP b 5.5 (1 58) 10.5 (3 18) 4 (2 7) 4 (2 25) Persistence/days b 16 (1 86) 44 (17 70) 2 (2 12) 2 (1 24) Pressure support a 53 (84) 14 (38) 11 (32) 15 (16) Start on day of AP b 4 (1 49) 4 (1 26) 4 (1 21) 5 (2 85) Persistence/days b,d 10 (1 44) 6 (1 21) 4 (1 36) 4 (1 18) MODS a 53 (83) 14 (38) 11 (32) 13 (14) MODS 2 organs 28 (44) 12 (32) 7 (21) 8 (8) MODS 3 organs 25 (39) 2 (5) 4 (12) 5 (5) AP, acute pancreatitis. a P.0001 between groups I III. b Median, range. c P.001 between groups I III. d P.05 between groups I III. of pulmonary infection and acute respiratory distress syndrome in 1 patient, and fulminant bleeding in another patient. Patients died 7 days (median, range days) after onset of symptoms in this group. Risk Factors for Secondary Infections and Death Multiple regression analysis identified only early/ preoperative MODS and extended intrapancreatic necrosis 30% as risk factors for the development of secondary PIN in operatively treated patients with sterile necrosis (Table 7). The same analysis performed in all patients with sterile necrosis irrespective of treatment Table 5. Occurrence of PIN After Onset of Symptoms and Antibiotic Treatment in Patients With Primary Infected Necrosis, Secondary PIN (), and Contaminations () Symtom onset, positive bacteriology (days) Duration of antibiotic treatment (days) until diagnosis of secondary PIN/ contamination Primary infected necrosis (n 135) a 20 (1 207) 23 (4 76) 21 (9 156) b 12 (2 40) 9 (1 23) NOTE. Data are presented as medians and ranges. a Previous necrosectomy at referring hospital without further intervention at our department in 5 of 140 patients with primary infected necrosis. b Precise duration not assessable due to high number of pretreated referrals in whom length of antibiotic treatment was unknown. revealed 3 major types of severe organ failure but no morphologic factors such as extent of necrosis as risk factors for the development of secondary PIN (Table 8). Interestingly, neither the timing of operative treatment after onset of symptoms nor the operative procedure itself was found to be a risk factor for the development of secondary PIN or death. Among operatively treated patients with sterile necrosis, multiple logistic regression identified early/preoper- Table 6. Microbiologic Findings in Secondary PIN () and Contaminations () n (%) n (%) Cultures available 63 a 25 b Monomicrobial 24 (38) 14 (56) Polymicrobial 39 (62) 11 (44) Gram-negative only 8 (13) 2 (8) Gram-positive only 34 (54) 18 (72) Both 17 (27) 4 (16) Candida only 3 (5) 1 (4) Total findings of germs Escherichia coli 8 (13) 0 Enterococcus 22 (34) 7 (28) Staphylococcus aureus 10 (16) 1 (4) Klebsiella 5 (8) 0 Staphylococcus epidermidis 33 (53) 17 (68) Pseudomonas 7 (11) 2 (8) Candida 17 (27) 4 (16) a Secondary PIN: in 1 patient with macroscopically infected necrosis, no bacteria/fungi could be cultured. b In 12 patients no complete differentiation of positive cultures was performed.

6 1058 RAU ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 8 Table 7. Multiple Logistic Regression With Forward Selection: Risk Factors for Development of Secondary PIN and for Nonsurvival After Operative Treatment of Sterile Necrosis (n 135) Risk factor Odds ratio 95% Confidence interval P value Outcome: secondary PIN Early/preoperative MODS or 1 organ organs organs Extent of intrapancreatic.0312 necrosis 30% % 50% % Outcome: nonsurvival Etiology.0021 Alcoholic 1.00 Biliary Others Early/preoperative MODS or 1 organ organs organs Overall dialysis/.0185 hemofiltration Yes addition, infection of pancreatic necrosis is believed to influence organ failure and survival in severe acute pancreatitis, 1 4,9,10,25 27 and intervention is necessary to overcome the detrimental consequences. 1,5 12 However, little attention has been paid to the temporal sequence and severity between organ failure and the development of local complications, especially of infected necrosis. 28 The present analysis shows for the first time that severe organ failure within the first week after onset of symptoms before any kind of intervention is closely linked to clinically relevant pancreatic infections, which occurred in median 2 weeks later. Similar observations have been reported by Büchler et al, 9 Götzinger et al, 10 and Bosscha et al 7 in their series of patients who presented with a comparable high incidence of MODS before the ultimate intraoperative diagnosis of infected necrosis. What are the potential underlying pathomechanisms for this hypothesis? Clinical and experimental studies have shown that macrocirculatory 29 and microcirculatory derangements 30 are a frequent and early feature in severe acute pancreatitis. An impaired mircrocirculation not only promotes the progression of pancreatic damage but also affects a variety of other organs including the intestine, ultimately resulting in MODS. In severe acute ative MODS 2 organs, overall need for dialysis/hemofiltration, and a nonalcoholic etiology as risk factors for death (Table 7). Local parameters of severity such as extent of intrapancreatic or extrapancreatic necrosis, the presence of bacteria or secondary PIN, and timing of surgery had no such influence on mortality. Similar risk factors were found in the total series of patients with sterile necrosis (Table 8). Again, morphologic factors such as extent of necrosis, presence of bacteria/secondary PIN, timing of operative treatment, and the operative procedure itself were not found to be risk factors for nonsurvival. Discussion There are 2 major findings of the present study that question our general understanding of the prognostic role and the underlying pathomechanisms of pancreatic infections. First, early MODS seems to be an important pathomechanism of subsequent pancreatic infections and is a major risk factor of death. Second, bacterial invasion of pancreatic necrosis is of less prognostic importance than previously believed. According to our current knowledge, the development of infected necrosis/pancreatic infections is a function of the extent of intrapancreatic and extrapancreatic necrosis and the time elapsed after onset of symptoms. 1,9 11 In Table 8. Multiple Logistic Regression With Forward Selection: Risk Factors for Development of Secondary PIN and for Nonsurvival in Operatively and Conservatively Treated Patients With Sterile Necrosis (n 230) Risk factor Odds ratio 95% Confidence interval P value Outcome: secondary PIN Overall mechanical.0038 ventilation Yes Overall dialysis/.0054 hemofiltration Yes Overall pressure support.0002 Yes Outcome: nonsurvival Etiology.0061 Alcoholic 1.00 Biliary Others Early/preoperative MODS or 1 organ organs organs Overall pressure support.0028 Yes

7 August 2006 EARLY MODS AND SEVERE ACUTE PANCREATITIS 1059 pancreatitis, failure of the gut barrier function with enhanced bacterial translocation is an early pathophysiologic event. 31,32 Bacterial translocation in turn is considered as the most important source of infected necrosis. Although bacterial translocation occurs early, infection of pancreatic necrosis becomes clinically manifest much later, usually within the second and third weeks of the disease. As shown in Tables 2 and 4, MODS occurred within days after disease onset; however, bacteria/fungi were isolated on median day 20 in any of the infected groups. Moreover, besides the presence of MODS, the duration of this complication seems to be another important aspect. Patients with secondary PIN group suffered from persistent MODS far beyond the first week after symptom onset, whereas patients in all other groups did not. In addition, early MODS might trigger additional mechanisms that render bacterial translocation into clinically manifest infections. Besides an overt proinflammatory response, recent clinical studies suggest that an impaired immune response is an early feature of patients with severe acute pancreatitis. Depending on the presence and severity of organ failure, an early and dramatic decrease of HLA-DR expression on circulating monocytes reflecting immunoparalysis was found and was associated with septic complications and poor outcome in the subsequent course. 33,34 This hypothesis could further explain the fact that, despite the positive bacteriologic status of necrosis but in the absence of severe organ failure, patients of the contaminated group probably had an immune response still sufficient enough to deal with the local problem of bacteria. Early MODS therefore seems to be an underlying pathomechanism promoting the development of clinically relevant pancreatic infections, just as it is a serious consequence, if infection once is manifest and uncontrolled during the later stages of the disease. Moreover, in the absence of infection, early MODS involving all 3 major organ systems was more than twice as high in group III than in contaminated patients, which further explains the higher mortality rate of 20% in this group. The present series provides the first analysis of the incidence and risk factors of PIN after surgical treatment of sterile necrosis. The results of our analysis show that operative treatment in patients with sterile necrosis at the end of the first week after symptom onset is no risk factor for the occurrence of pancreatic infections, and it does not increase mortality. Secondary PIN occurred in 47% of all operatively treated patients with sterile necrosis, which is identical to the incidence of primary infected necrosis in our overall surgical series of patients with necrotizing pancreatitis. 11 In addition, the diagnosis of secondary PIN and the occurrence of bacteria/fungi in the contaminated group were both observed around the end of the third week after disease onset, which is again similar to that in patients with primary infected necrosis. The comparison of patients with secondary PIN and contaminations raises, however, further questions about the general role of microbial invasion, infected necrosis, the precise definition of PIN, and its therapeutic consequences. Indeed, from a surgeon s standpoint, one can argue that the difference between the 2 groups might be due to the fact that patients of the contaminated group were drained effectively, whereas those of the secondary PIN group were not. On the other hand, patients of both groups had documented bacterial/fungal invasion of pancreatic necrosis with similar microbiologic spectrum, and all had SIRS (data not shown). Although no further intervention was performed, outcome in the contaminated group was extremely good. This observation is in accordance with recent clinical studies. Patients with FNA-proven infection of necrosis do not necessarily require immediate intervention; some of them do not even require any kind of intervention at all, 35 which is in contrast to the present treatment algorithms. 18,19 Corresponding observations have been reported by other authors 36 and were also found in a recent prospective international multicenter trial in severe acute pancreatitis. 37 Even if we cannot rule out that bacterial invasion might have been a consequence of necrosectomy, it still did not influence outcome as long as early MODS was absent. These findings emphasize the need for a revised definition of infection and better tools beyond FNA for the differentiation of both infectious entities. 37 Our findings do not support the assumption that operative treatment, if carried out at the end of the first week after symptom onset, is harmful in patients with sterile necrosis. On the other hand, our results are no plea to support this concept because we still have to face the question of whether it is beneficial. This is clearly an issue our study cannot answer, and so far, no study ever published has been able to do so. Dismissing the fact that meaningful prospective controlled data are still absent, the answer to this question is most likely no in the majority of these patients within the first and second weeks of the disease, 23 but what thereafter? An even more challenging hypothesis might be deduced from our results. The current concept of local morphologic factors such as extent and infection of necrosis as principal risk factors of disease severity and death needs to be reevaluated. It definitely has to be kept

8 1060 RAU ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 8 in mind that our analysis comprises a selected series of patients with initially sterile necrosis only. This selection bias might be one reason for the predominance of the systemic complications and organ failure. However, a number of other recent reports have shown that relevant factors of disease severity and determinants of outcome in acute pancreatitis are clearly related to the individual systemic host response in terms of MODS ,38 These observations almost necessarily lead to questioning our current therapeutic targets in acute pancreatitis. In focusing on well-established morphologic complications such as extent and infection of necrosis, novel treatment protocols including antibiotic prophylaxis as well as tissue-sparing open or minimally invasive surgical techniques failed to lower the still high mortality rates in the specific subset of patients with severe systemic disease. 7 12,19 Considering the importance of MODS in the present series, it is no surprise that targeting local morphologic factors is not the clue to our problem. On the basis of revised and improved classification systems, any further progress in the diagnosis and management of acute pancreatitis requires reassessment and evaluation of both established and future therapeutic approaches in the light of this new pathophysiologic and prognostic concept. In summary, early MODS and extent of intrapancreatic necrosis are key risk factors for the development of clinically relevant PIN after operative treatment of sterile necrosis. However, nonsurvival in necrotizing pancreatitis initially classified as sterile does not seem to depend on the subsequent microbial invasion of pancreatic necrosis or on the surgical procedure per se. Our results suggest that early pathophysiologic events reflected by MODS render the host susceptible to major complications and mortality. New concepts and definitions emphasizing the predominance of early systemic over local morphologic complications in acute pancreatitis are in demand. References 1. Beger HG, Bittner R, Block S, et al. Bacterial contamination of pancreatic necrosis: a prospective clinical study. Gastroenterology 1986;49: Isenmann R, Rau B, Beger HG. Bacterial infection and extent of necrosis are determinants of organ failure in patients with acute necrotizing pancreatitis. Br J Surg 1999;86: Le Mee J, Paye F, Sauvanet A, et al. Incidence and reversibility of organ failure in the course of sterile or infected necrotizing pancreatitis. Arch Surg 2001;136: Garg PK, Madan K, Pande GK, et al. Association of extent and infection of pancreatic necrosis with organ failure and death in acute necrotizing pancreatitis. Clin Gastroenterol Hepatol 2005; 3: Widdison AL, Karanjia ND. Pancreatic infection complicating acute pancreatitis. Br J Surg 1993;80: Farkas G, Marton J, Mandi Y, et al. Surgical strategy and management of infected pancreatic necrosis. Br J Surg 1996;83: Bosscha K, Hulstaert PF, Hennipman A, et al. Fulminant acute pancreatitis and infected necrosis: results of open management of the abdomen and planned reoperations. J Am Coll Surg 1998;187: Takeda K, Matsuno S, Sunamura M, et al. Surgical aspects and management of acute necrotizing pancreatitis: recent results of a cooperative national survey in Japan. Pancreas 1998;16: Büchler MW, Gloor B, Müller CA, et al. Acute necrotizing pancreatitis: treatment strategy according to the status of infection. Ann Surg 2000;232: Gotzinger P, Sautner T, Kriwanek S, et al. Surgical treatment for severe acute pancreatitis: extent and surgical control of necrosis determine outcome. World J Surg 2002;26: Rau B, Bothe A, Beger HG. Surgical treatment of necrotizing pancreatitis by necrosectomy and closed lavage: changing patient characteristics and outcome in a 19-year single center series. Surgery 2005;138: Bradley EL III, Allen K. A prospective longitudinal study of observation versus surgical intervention in the management of necrotizing pancreatitis. Am J Surg 1991;161: Rau B, Pralle U, Uhl W, et al. Management of sterile necrosis in instances of severe acute pancreatitis. J Am Coll Surg 1995; 181: Bradley EL III. Surgical debridement is rarely necessary in sterile pancreatic necrosis. Pancreas 1996;13: Uomo G, Visconti M, Manes G, et al. Nonsurgical treatment of acute necrotizing pancreatitis. Pancreas 1996;12: Aultman DF, Bilton BD, Zibari GB, et al. Nonoperative therapy for acute necrotizing pancreatitis. Am Surg 1997;63: Ashley SW, Perez A, Pierce EA, et al. Necrotizing pancreatitis: contemporary analysis of 99 consecutive cases. Ann Surg 2001; 234: Nathens AB, Curtis JR, Beale RJ, et al. Management of the critically ill patient with severe acute pancreatitis. Crit Care Med 2004;32: Working Party of the British Society of Gastroenterology, Associations of Surgeons of Great Britain and Ireland, Pancreatic Society of Great Britain and Ireland, et al. UK guidelines for the management of acute pancreatitis. Gut 2005;54(Suppl 3): Isenmann R, Rau B, Beger HG. Early severe acute pancreatitis: characteristics of a new subgroup. Pancreas 2001;22: Buter A, Imrie CW, Carter CR, et al. Dynamic nature of early organ dysfunction determines outcome in acute pancreatitis. Br J Surg 2002;89: Johnson CD, Abu-Hilal M. Persistent organ failure during the first week as a marker of fatal outcome in acute pancreatitis. Gut 2004;53: Fernandez-del Castillo C, Rattner DW, Makary MA, et al. Debridement and closed packing for the treatment of necrotizing pancreatitis. Ann Surg 1998;228: Beger HG, Buchler M, Bittner R, et al. Necrosectomy and postoperative local lavage in necrotizing pancreatitis. Br J Surg 1988; 75: Perez A, Whang EE, Brooks DC, et al. Is severity of necrotizing pancreatitis increased in extended necrosis and infected necrosis? Pancreas 2002;25: Lankisch PG, Pflichthofer D, Lehnick D. No strict correlation between necrosis and organ failure in acute pancreatitis. Pancreas 2000;20: Tenner S, Sica G, Hughes M, et al. Relationship of necrosis to organ failure in severe acute pancreatitis. Gastroenterology 1997;113:

9 August 2006 EARLY MODS AND SEVERE ACUTE PANCREATITIS Gong ZY, Tang YQ. Onset of time of complications in patients with severe acute pancreatitis receiving nonoperative therapy. HBPD Int 2002;1: Beger HG, Bittner R, Buchler M, et al. Hemodynamic data pattern in patients with acute pancreatitis. Gastroenterology 1986;90: Vollmar B, Menger MD. Microcirculatory dysfunction in acute pancreatitis: a new concept of pathogenesis involving vasomotion-associated arteriolar constriction and dilation. Pancreatology 2003;3: Ammori BJ. Role of the gut in the course of severe acute pancreatitis. Pancreas 2003;26: Ammori BJ, Leeder PC, King R, et al. Early increase in intestinal permeability in patients with severe acute pancreatitis: correlation with endotoxemia, organ failure and mortality. J Gastrointest Surg 1999;3: Yu WK, Li WQ, Li N, et al. Mononuclear histocompatibility leukocyte antigen-dr expression in the early phase of acute pancreatitis. Pancreatology 2004;4: Satoh A, Miura T, Satoh K, et al. Human leukocyte antigen-dr expression on peripheral monocytes as a predictive marker of sepsis during acute pancreatitis. Pancreas 2002;25: Rünzi M, Niebel W, Goebell H, et al. Severe acute pancreatitis: nonsurgical treatment of infected necrosis. Pancreas 2005;30: Ramesh H, Prakash K, Lekha V, et al. Are some cases of infected pancreatic necrosis treatable without intervention? Dig Surg 2003;20: Rau B, Kemppainen EA, Gumb A, et al. Clinical value of Procalcitonin (PCT) in predicting infectious complications and overall prognosis in severe acute pancreatitis: a prospective international multicenter study (abstr). Pancreatology 2004;4: Halonen K, Leppäniemi AK, Puolakkainen PA, et al. Severe acute pancreatitis: prognostic factors in 270 consecutive patients. Pancreas 2000;21: Address requests for reprints to: Priv-Doz Dr med Bettina Rau, Department of General, Visceral, and Vascular Surgery, University of the Saarland, Kirrberger Strasse, Homburg/Saar, Germany. bettina.rau@unikinikum-saarland.de; fax: