- Email: [email protected]
Obesity Increases the Severity of Acute Pancreatitis: Performance of APACHE-O Score and Correlation with the Inflammatory Response
Original Paper Received: June 15, 2005 Accepted after revision: November 23, 2005 Published online: April 19, 2006
Pancreatology 2006;6:279–285 DOI: 10.1159/000092689
Obesity Increases the Severity of Acute Pancreatitis: Performance of APACHE-O Score and Correlation with the Inflammatory Response Georgios I. Papachristou a Dionysios J. Papachristou a Haritha Avula a Adam Slivka a David C. Whitcomb a–c a Division of Gastroenterology, Hepatology and Nutrition, Department of Medicine, b Cell Biology and Physiology, and c Human Genetics, University of Pittsburgh, Pittsburgh, Pa., USA
Key Words Obesity Acute pancreatitis, obesity Acute pancreatitis, severity APACHE-O scoring system APACHE-II Inflammatory response
Abstract Background: Obese patients appear to be at risk for complications of acute pancreatitis (AP). APACHE-O score has been suggested to improve APACHE-II accuracy in predicting severe outcome in AP. Aims: To determine if APACHE-O adds any predictive value to APACHE-II score and to test the hypothesis that obese patients are at increased risk of severe AP (SAP) because of a more intense inflammatory response to pancreatic injury. Methods: 102 AP patients were prospectively studied. Using a body mass index (BMI) 130, 28% of the subjects were obese. Nineteen patients developed organ dysfunction and were classified as SAP. Receiver-operating curves for prediction of SAP were calculated using admission APACHE-II and APACHE-O scores. Binary logistic regression was performed to assess if obesity is a risk for SAP and to determine the clinical factors associated with severe disease. Serum levels of IL-6, MCP-1 and CRP as
© 2006 S. Karger AG, Basel and IAP 1424–3903/06/0064–0279$23.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com
Accessible online at: www.karger.com/pan
well as Ranson’s scores were compared between obese and non-obese patients. Results: Admission APACHE-O (area under the curve AUC 0.895) and APACHE-II (AUC 0.893) showed similar accuracy in predicting severe outcome. BMI was identified as a significant risk for SAP (OR 2.8, p = 0.048) and mortality (OR 11.2, p = 0.022). CRP levels were significantly higher in obese AP patients (p = 0.0001) as well as Ranson’s score (p = 0.021). IL-6 and MCP-1 levels were higher in obese patients but did not reach statistical significance. Conclusions: Obesity is an independent risk for SAP. Admission APACHE-O score is not more accurate than APACHE-II. Our study results suggest that obesity increases the severity of AP by amplifying the immune response to injury. Copyright © 2006 S. Karger AG, Basel and IAP
Introduction
Acute pancreatitis (AP) is an acute inflammatory condition of the pancreas with a widely variable outcome. Most patients recover without complication. However, approximately 20% develop failure of one or more organ systems. The severity of AP is a function of the extent of
David C. Whitcomb, MD, PhD Division of Gastroenterology, Hepatology and Nutrition UPMC Presbyterian, Mezzanine Level 2, C Wing 200 Lothrop Street, Pittsburgh, PA 15213 (USA) Tel. +1 412 648 9604, Fax +1 412 383 7236, E-Mail [email protected]
pancreas injured and the intensity of the inflammatory response. Early death is often linked to the systemic inflammatory response syndrome (SIRS) with cardiovascular collapse or acute respiratory distress syndrome, while late death is more often associated with infected pancreatic necrosis, sepsis and multisystem organ failure. AP is a complex clinical syndrome encompassing a variety of mechanisms of pancreatic injury, activation of multiple inflammatory cascades and endangering multiple organs systems in a highly heterogeneous patient population. Thus, clinicians and scientists have struggled with developing a simple test to accurately predict clinical outcome. The development and use of different definitions, critical measurements and classification systems presents the additional problems of comparing complications and outcome. Currently, the most widely used definition of severe AP (SAP) is based on the Atlanta Consensus Conference in 1992 [1]. In the Atlanta consensus, SAP is defined by the presence of complications that are either systemic (e.g. organ system dysfunction) or local (e.g. peripancreatic fluid collection, pseudocyst, abscess, pancreatic necrosis). However, the risk for local complications and systemic complications and the consequences of these complications reflect different mechanisms and different outcomes. Furthermore, improved outcomes relating to early resolution of organ failure suggest that the Atlanta definition will need revisiting. The most commonly used predictive multifactorial scoring systems are the Ranson’s score [2] and the Acute Physiology and Chronic Health Evaluation (APACHE) II [3]. APACHE-II can be calculated on admission, in contrast to Ranson’s score, which requires 48 h for full assessment. Furthermore, APACHE-II can be updated daily over the hospital course, allowing monitoring of disease progression and response to therapy. Components of both tests measure the end-organ failure to respond to an inflammatory challenge (SIRS) and therefore represent indirect markers of the inflammatory response (i.e. arteriolar pO2 !60 mm Hg, rise in BUN 15 mg/dl, fluid sequestration 16 liters in Ranson’s score). It has been clearly demonstrated that regardless of the cut-off selection, their overall accuracy does not exceed 80% [4]. Obesity has been shown to be an independent risk factor for severe outcome in patients with AP [5–10]. Recently, Martinez et al. [11] conducted a meta-analysis suggesting that obesity (body mass index (BMI) 130 kg/m2) increased the risk of developing AP (OR 2.6, 95% CI 1.5– 4.6), as well as the severity (OR 2.6, 95% CI 1.5–4.6), but failed to find a significant difference in mortality. Although the increased risk or frequency of AP could be
280
Pancreatology 2006;6:279–285
related to the higher prevalence of gallstones in the obese [12], obesity appeared to be directly related to both local (OR 4.3, 95% CI 2.4–7.9) and systemic (OR 2.0, 95% CI 1.1–4.6) complications in the meta-analysis. The mechanism by which obesity increases the severity of AP is unclear. However, the observation that, in humans, the severity of AP is directly related to the intensity of the immune response [4] leads to the hypothesis that obese patients are at risk of more severe complications of AP than non-obese patients because of a more intense inflammatory response to pancreatic injury. Obesity may affect the immune response to injury. Adipose tissue is an important source of cytokines and obesity can be considered a state of chronic inflammation. Obese humans have increased levels of several proinflammatory markers, called adipokines [13]. For example, serum tumor necrosis factor (TNF-) [14], C-reactive protein (CRP) [15], interleukin-6 (IL-6) [16] and monocyte chemotactic protein (MCP-1) [17] concentrations are elevated in the obese when compared to non-obese patients and they fall after reduction of adipose tissue depots. Recently, a modification of the APACHE-II scoring system was proposed that included a factor for obesity. The proposed ‘APACHE-O’ scale adds 1 point for BMI between 26 and 30 kg/m2 (overweight) and 2 points for index 130 kg/m2 (obese). Johnson et al. [18] used APACHE-O score in patients with AP and suggested that this system improves severity prediction. Herein we report a clinical study that investigates the relationship between obesity and SAP and tests the utility of the APACHE-O scoring system for predicting SAP. The aims of this study are: (1) to investigate prospectively whether APACHE-O adds any predictive value to admission APACHE-II score in a North American population, (2) to assess if a BMI 130 or other clinical factors recorded on admission represent a significant risk for organ system failure, and (3) to compare early and later inflammatory markers, including serum levels of IL-6, MCP-1 and CRP as well as Ranson’s scores between obese and non-obese AP patients.
Methods Patients The Severity of Acute Pancreatitis Study 1 (SAPS1) protocol was approved by the Institutional Review Board of the University of Pittsburgh Medical Center. Informed written consent was obtained from all the patients before their enrollment into the study. The diagnosis of AP was based on upper abdominal pain or ab-
Papachristou /Papachristou /Avula /Slivka / Whitcomb
dominal localizing signs and plasma amylase and/or lipase levels at least 3 times above the upper limit of normal. The time interval between the onset of symptoms and admission to the hospital was no more than 48 h. Patients were recruited within 24 h from the time of admission. Patients entering the study had blood samples obtained daily along with recording of clinical, physiological and biochemical measurements necessary for Ranson’s, APACHE-II and APACHEO score calculations. Patients were managed according to the standard clinical practice. The patient’s clinical records until discharge from hospital and in follow-up visits were reviewed by two investigators (G.I.P. and D.J.P.) and all local and systemic AP complications were recorded. The term severe AP (SAP) was reserved for remote organ dysfunction including cardiovascular, pulmonary and renal failure necessitating ICU admission or organ support (meaning systolic pressure !90 mm Hg or use of vasopressors, arterial pO2 !60 mm Hg at room air or mechanical ventilation and serum creatinine 12 mg/ dl after rehydration or initiation of hemodialysis) [1]. Pancreatic necrosis was assessed by contrast-enhancing abdominal CT scan, which was performed in most of the patients (77 out of 102; 76%). Hospital mortality was defined as death within the same hospitalization for AP. We assessed if BMI 130 or other factors recorded on admission are associated with organ system failure. Clinical factors identified on history-taking were assessed, including gender, race, age, BMI, active smoking, alcohol use, presence of comorbidites (diabetes mellitus, coronary artery disease, chronic obstructive pulmonary disease, chronic renal insufficiency and cirrhosis) and etiology of AP (gallstones, post-endoscopic retrograde cholangiopancreatography (post-ERCP) procedure, idiopathic and alcohol-associated). All the above clinical factors have common features that can be assessed and recorded upon admission. Serum Cytokine Assay We assessed the inflammatory response in AP by measuring IL-6, MCP-1, and CRP serum concentrations. Serum samples collected within 24 h of the time of admission were assayed for IL-6, MCP-1 and CRP levels, quantified using a fluorescence-based capture sandwich immunoassay based on Luminex Flowmetrix system (Luminex, Austin, Tex., USA) and run in the University of Pittsburgh Cancer Institute’s Luminex Core Facility. Samples were analyzed using the Bio-Plex suspension array system, which includes a fluorescent reader and a Bio-Plex Manager analytical software (Bio-Rad Laboratories, Hercules, Calif., USA) [19]. Statistical Analysis Sample size was measured using one-sided calculations with of 0.05 and power of 80%, based on an incidence of obesity (BMI 130 kg/m²) in one third of our population. Using the proposed odds ratio in the recent literature [11], a sample size of 95 AP patients (71 non-obese and 24 obese) allows sufficient power to detect a 2.5fold increase in the risk of severe disease in obese subjects. Receiver-operating curves for prediction of SAP were calculated using admission APACHE-II and APACHE-O scores. The area under the curve (AUC) was determined for comparing predictive accuracy. The diagnostic accuracy of individual clinical scores for severe outcome was assessed using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy.
Obesity Increases the Severity of Acute Pancreatitis
Table 1. Systemic and local complications in 102 AP patients
Systemic (n = 19)
n
Local (n = 23)
n
Respiratory failure Cardiovascular failure Renal failure Single organ failure Multisystem organ failure Death
15 8 10 10 9 5
Pseudocysts Infected pseudocysts Pancreatic necrosis Infected necrosis Pseudoaneurysms
13 2 11 3 2
Binary logistic regression analysis was performed to assess if BMI 130 or other clinical factors recorded on admission represent a significant risk for organ system failure. Serum protein levels and Ranson’s scores were compared between AP patients with a BMI 130 (obese) vs. BMI ^30 (non-obese) by using Mann-Whitney test. For all statistical comparisons a p value !0.05 was considered significant.
Results
Clinical Features of Patients We prospectively studied 102 consecutive patients with AP admitted to the University of Pittsburgh Medical Center between June 2003 and October 2004. Twenty-nine subjects (28%) were obese (BMI 130). There were 50 males and 52 females with a mean age of 50 years (range 15–90). Ninety-one were Caucasian (89%), 9 were African-American (9%) and 2 were Hispanic (2%). The mean BMI was 27.4 (range 13.6–42.8). Etiologic classification of AP included biliary in 36 patients (35%), postERCP procedure in 21 (20%), idiopathic in 18 (18%), alcohol-associated AP in 15 (15%), drug-induced in 5 (5%), hypertriglyceridemia in 4 (4%) and pancreatic masses in 3 (3%). Nineteen subjects developed organ failure and were classified as having SAP, of which 11 developed local complications including 9 patients with pancreatic necrosis. Etiologic classification in these 19 patients with SAP included biliary in 6 patients, post-ERCP procedure in 3, idiopathic in 3, alcohol-associated AP in 5, drug-induced in 1, and hypertriglyceridemia in 1. Overall, 23 of the 102 patients (23%) developed local complications including pancreatic pseudocysts, pancreatic necrosis and splenic or hepatic artery pseudoaneurysms. Systemic and local complications are presented in table 1.
Pancreatology 2006;6:279–285
281
35
Percent of subjects
1.00
Sensitivity
0.75
30 25 20 15
17% 14%
14% 8%
10 5 0
0.50
BMI ≥30 BMI <30
31%
1.4% Organ failure
Necrosis
Death
Fig. 2. Frequency (%) of organ failure, necrosis and death in obese
0.25
(BMI 130) vs. non-obese (BMI ^30) subjects with AP.
APACHE-O APACHE-II 0 0
0.25
0.50
0.75
1.00
1 – Specificity
Fig. 1. Receiver-operating curves for prediction of severity using admission APACHE-O score (—) and admission APACHE-II score (- - -). The areas under the curve (AUC) were 0.895 for APACHE-O and 0.893 for APACHE-II.
Table 2. Sensitivity, specificity, positive predictive value (PPV),
negative predictive value (NPV) and overall accuracy of admission APACHE-O and APACHE-II for the prediction of SAP at a range of cut-off values Sensitivity Specificity PPV Admission APACHE-O >7 90 >8 84 >9 84 >10 74 Admission APACHE-II >7 90 >8 79 >9 74 >10 69
NPV
Accuracy
70 80 82 86
41 49 52 54
97 96 96 94
74 81 83 84
77 83 85 89
47 52 47 59
97 95 93 93
80 83 80 86
Comparison of Admission Scoring Systems in Predicting SAP Admission APACHE-O demonstrated similar accuracy (AUC 0.895) when compared to APACHE-II (AUC 0.893) (fig. 1) in predicting severe outcome. The sensitiv-
282
Pancreatology 2006;6:279–285
ity, specificity, predictive values and accuracy of admission APACHE scores at a range of cut-off values are seen in table 2. The selected admission APACHE-O cut-off value was 9 and showed a sensitivity of 84%, specificity 82%, PPV 52%, NPV 96% and accuracy 83%. Both admission APACHE-O and APACHE-II demonstrated an improved accuracy in predicting organ failure vs. predicting pancreatic necrosis (AUC 0.780 and 0.766, respectively). Identification of Clinical Factors Recorded at Admission Associated with Organ Failure Logistic regression analysis identified BMI 130 as significant risk factor for organ failure in AP (OR 2.8, p = 0.048, CI 1.1–7.9). The relative frequencies of pancreatic necrosis, organ failure and death according to BMI cutoff value of 30 are shown in figure 2. Thirty-one percent and 17% of the obese AP subjects developed organ failure and pancreatic necrosis respectively, whereas the frequencies were approximately half of these in the nonobese AP group (14% organ failure and 8% pancreatic necrosis). The other selected clinical factors recorded on admission (gender, race, age, smoking, alcohol use, presence of comorbidities and etiology of AP) were not associated with severe disease. There was a trend toward alcohol consumption and SAP (p = 0.06, OR = 2.9). Furthermore, BMI 130 is also a risk factor for mortality in AP (OR 11.2, p = 0.022, CI 1.05–572.9). Comparison of Levels of Inflammatory Markers between Obese and Non-Obese Patients Serum IL-6, MCP-1, and CRP levels in all AP patients, obese and non-obese subjects are presented in table 3. CRP levels were significantly higher in obese than non-
Papachristou /Papachristou /Avula /Slivka / Whitcomb
Table 3. Mean 8 SEM values of IL-6, MCP-1, CRP and Ranson’s score in all patients, obese, non-obese and p value when comparing values in obese vs. non-obese patients
IL-6, pg/ml MCP-1, pg/ml CRP, mg/dl Ranson’s score
All patients Obese
Non-obese
p value
5208217 6998214 10.981.2 1.6480.19
4848285 6298235 7.681.0 1.3480.19
0.368 0.325 0.0001 0.021
6128260 8708471 18.082.5 2.3880.45
obese patients with AP (p = 0.0001). IL-6 and MCP-1 levels were higher in obese subjects with a trend for statistical significance (p = 0.368 and 0.325, respectively). Ranson’s score was higher in obese subjects (p = 0.021).
Discussion
Lankisch and Schirren [5] were among the first to recognize that obesity was a risk factor for SAP. Since then about a dozen studies have confirmed and extended this observation [6–11, 18]. We prospectively collected and analyzed 102 subjects AP hospitalized at University of Pittsburgh, a tertiary care medical center. We found that a large fraction of our patients were obese (almost 30%) in contrast to previous studies from Europe [18], Mexico [8] and Asia [9] in which the percentage of obese subjects enrolled was either much less (2% in the Asian and 8% in the Mexican study) or not reported. This study was designed with adequate power and demonstrated a significant association between obesity and severity in AP. Using logistic regression analysis, we determined that obese AP patients are 2.8 times more likely to develop organ failure than non-obese subjects. We tested whether inclusion of obesity in a multifactorial scoring system improved predictive accuracy of the outcome of AP in our cohort, as has been proposed by Johnson et al. [18]. We found that this approach did not provide additional benefit. Adding a BMI score to APACHE-II did not improve the predictive accuracy on admission (APACHE-O AUC 0.895 vs. APACHE-II AUC 0.893). One possible reason for the limited utility of this system is that the only obese individuals that will be reclassified from mild to severe are those with a BMI of 26–30 with an exact APACHE-II score of 7, or those with a BMI of 130 with an APACHE-II score of 6 or 7. Therefore, use of APACHE-O at the above cut-offs could improve the sensitivity without overall loss of accuracy.
Obesity Increases the Severity of Acute Pancreatitis
Our logistic regression analysis suggests that all obese patients are at higher risk of systemic complications regardless of the cause of AP or amount of initial pancreatic injury. APACHE-II and APACHE-O scores should be included into a continuous rather than binary risk calculation. A primary question is why obese individuals are at higher risk for SAP. Mery et al. [10] investigated the type of fat and found that severity of AP was associated with android fat distribution (OR 9.23, 95% CI 1.67–51.07) and higher waist circumference (OR 13.41, 95% CI 2.43– 73.97) rather than the percent total body fat. They speculated that android fat distribution increased risk of severity since this type of obesity is associated with an ‘overactive’ immune response. Although obesity and inflammation is a broad subject, there is an important observation that may be relevant to the risk of AP. Takahashi et al. [20] used microarray technology in an attempt to identify the molecules explaining the relationship between obesity and vascular disorders by comparing mRNA expression of about 12,000 genes in white adipose tissue in normal vs. high fat diet-induced obesity in mice. MCP-1 is an early inflammatory cytokine and expression of MCP-1 mRNA was increased 7.2-fold in obese mice compared with normal mice, which was twice as high as any other gene. They also measured MCP-1 protein levels in adipocytes, which were found to be 5 times higher than in lean mice. MCP-1 levels in plasma were also increased in obese mice, and a strong correlation between plasma MCP-1 levels and body weight was identified. A similar relationship was found in humans. Troseid et al. [21] found that changes in plasma MCP-1 levels were significantly correlated to changes in visceral fat (r = 0.41, p = 0.02). Exercise resulted in reduction of the MCP-1 serum levels in obese subjects. MCP-1 has been shown to be an early marker of SAP in both animal models and clinical studies [22, 23]. Furthermore, our group has recently demonstrated that a genetic polymorphism (G allele at –2518 position) in the MCP-1 gene significantly increases the risk of severe disease in subjects with AP (OR 7.0, 95% CI 1.5–34) [24]. These observations suggest that obesity may increase the severity of inflammation through a fat-dependent (non-genetic) overexpression of MCP-1 (and other adipokines) leading to an amplification of the downstream immunological response to injury. However, additional ethnic and environmental differences may also be important since the effect of obesity on severity appeared to be worst in South Africa [6], moderate in Mexico [8] and least severe in Taiwan [9].
Pancreatology 2006;6:279–285
283
To test the hypothesis that obesity amplifies the inflammatory response to pancreatic injury through different adipokines, many of which have been found to be upregulated in obese individuals, we assessed the inflammatory response in AP in multiple ways. We measured the serum levels of three different inflammatory markers within 24 h of admission and also calculated the Ranson’s score. We demonstrated that obese patients with a BMI 130 had significantly higher CRP levels (later marker). Mean IL-6 and MCP-1 concentrations (early marker) were higher in obese patients than nonobese patients, but did not reach statistical significance. Ranson’s score, which is an indirect marker of the inflammatory response, was also significantly higher in obese subjects. The major limitation with respect to assessing these markers was the variability in time delay between the onset of pancreatitis and admission – which would have a large impact on accurately measuring the ‘early markers’ IL-6 and MCP-1. Thus, it is possible that in many of our patients, the levels of the early markers might have already started to decline at the time of the sample draw. Peak CRP levels occur at 48–72 h from the time of injury. CRP tends to maintain high levels for a longer period of time and therefore may be a more accurate marker of the inflammatory response in our patients. An alternate hypothesis is that obesity amplifies the in-
flammatory cascade at sites downstream to MCP-1 and IL-6 actions. The highly significant increase in CRP and Ranson’s scores in obese patients, however, supports the hypothesis that obesity amplifies the immune response. This hypothesis would need to be supported by further studies since the possibility that the obese patients suffer more severe pancreatic injury and have a proportionally increased inflammatory response cannot be excluded. It would be of interest to study these markers in obese patients suffering from other varieties of acute abdominal insult such as perforated bowel or diverticulitis. In summary, this study demonstrated that in populations with a large fraction of obese subjects, admission APACHE-O score does not improve the predictive accuracy of APACHE-II for severe outcome in AP. We confirmed that obesity constitutes an independent risk for organ failure in AP in our cohort. Our study results suggest that obesity increases the severity of AP by amplifying the immune response to injury.
Acknowledgement This work was supported by DK061451 (D.C.W.) and DK054709.
References 1 Bradley EL: A clinically based classification system for acute pancreatitis: summary of the Atlanta symposium. Arch Surg 1993;128:586– 590. 2 Ranson JHC, Rifkind KM, Roses DF, et al: Prognostic signs and the role of operative management in acute pancreatitis. Surg Gynaecol Obstet 1974;139:69–81. 3 Knaus WA, Draper EA, Wagner DP, et al: APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818–829. 4 Papachristou GI, Whitcomb DC: Predictors of severity and necrosis in acute pancreatitis. Gastroenterol Clin North Am 2004; 33: 871– 890. 5 Lankisch PG, Schirren CA: Increased body weight as a prognostic parameter for complications in the course of acute pancreatitis. Pancreas 1990;5:626–629. 6 Funnell IC, Bornman PC, Weakley SP, et al: Obesity: an important prognostic factor in acute pancreatitis. Br J Surg 1993; 80: 484– 486.
284
7 Martinez J, Sanchez-Paya J, Palazon JM, et al: Obesity: a prognostic factor of severity in acute pancreatitis. Pancreas 1999;19:15–20. 8 Suazo-Barahona J, Carmona-Sanchez R, Robles-Diaz G, Milke-Garcia P, Vargas-Vorackova F, Uscanga-Dominguez L, Pelaez-Luna M: Obesity: a risk factor for severe acute biliary and alcoholic pancreatitis. Am J Gastroenterol 1998;93:1324–1328. 9 Tsai CJ: Is obesity a significant prognostic factor in acute pancreatitis? Dig Dis Sci 1998;43: 2251–2254. 10 Mery CM, Rubio V, Duarte-Rojo A, SuazoBarahona J, Pelaez-Luna M, Milke P, RoblesDiaz G: Android fat distribution as predictor of severity in acute pancreatitis. Pancreatology 2002;2:543–549. 11 Martinez J, Sanchez-Paya J, Palazon JM, Suazo-Barahona J, Robles-Diaz G, Perez-Mateo M: Is obesity a risk factor in acute pancreatitis? A meta-analysis. Pancreatology 2004;4: 42–48.
Pancreatology 2006;6:279–285
12 Torgerson JS, Lindroos AK, Naslund I, Peltonen M: Gallstones, gallbladder disease, and pancreatitis: cross-sectional and two-year data from the Swedish Obese Subjects (SOS) and SOS reference studies. Am J Gastroenterol 2003;98:1032–1041. 13 Lyon CJ, Law RE, Hsueh WA: Minireview: adiposity, inflammation, and atherogenesis. Endocrinology 2003;144:2195–2200. 14 Dandona P, Weinstock R, Thusu K, AbdelRahman E, Aljada A, Wadden T: Tumor necrosis factor- in sera of obese patients: fall with weight loss. J Clin Endocrinol Metab 1998;83:2907–2910. 15 Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB: Elevated C-reactive protein levels in overweight and obese adults. JAMA 1999;282:2131–2135. 16 Khaodhiar L, Ling PR, Blackburn GL, Bistrian BR: Serum levels of interleukin-6 and C-reactive protein correlate with body mass index across the broad range of obesity. JPEN J Parenter Enteral Nutr 2004;28:410–415.
Papachristou /Papachristou /Avula /Slivka / Whitcomb
17 Christiansen T, Richelsen B, Bruun JM: Monocyte chemoattractant protein-1 is produced in isolated adipocytes, associated with adiposity and reduced after weight loss in morbid obese subjects. Int J Obes Relat Metab Disord 2005; 29:146–150. 18 Johnson CD, Toh SK, Campbell MJ: Combination of APACHE-II score and an obesity score (APACHE-O) for the prediction of severe acute pancreatitis. Pancreatology 2004;4:1–6. 19 Seideman J, Peritt D: A novel monoclonal antibody screening method using the Luminex100 microsphere system. J Immunol Methods 2002;267:165–171.
Obesity Increases the Severity of Acute Pancreatitis
20 Takahashi K, Mizuarai S, Araki H, Mashiko S, Ishihara A, Kanatani A, Itadani H, Kotani H: Adiposity elevates plasma MCP-1 levels leading to the increased CD11b-positive monocytes in mice. J Biol Chem 2003; 278: 46654– 46660. 21 Troseid M, Lappegard KT, Claudi T, Damas JK, Morkrid L, Brendberg R, Mollnes TE: Exercise reduces plasma levels of the chemokines MCP-1 and IL-8 in subjects with the metabolic syndrome. Eur Heart J 2004;25:349–355. 22 Grady T, Liang P, Ernst SA, Logsdon CD: Chemokine gene expression in rat pancreatic acinar cells is an early event associated with acute pancreatitis. Gastroenterology 1997; 113: 1966–1975.
23 Rau B, Baumgart K, Kruger CM, Schilling M, Beger HG: CC-chemokine activation in acute pancreatitis: enhanced release of monocyte chemoattractant protein-1 in patients with local and systemic complications. Intensive Care Med 2003;29:622–629. 24 Papachristou GI, Sass DA, Avula H, Lamb J, Lokshin A, Barmada MM, Slivka A, Whitcomb DC: The MCP-1 -2518 G allele acts as a potent disease modifier in acute pancreatitis. Clin Gastroenterol Hepatol 2005;3:475–481.
Pancreatology 2006;6:279–285
285
Pancreatology 2006;6:279–285 DOI: 10.1159/000092689
Obesity Increases the Severity of Acute Pancreatitis: Performance of APACHE-O Score and Correlation with the Inflammatory Response Georgios I. Papachristou a Dionysios J. Papachristou a Haritha Avula a Adam Slivka a David C. Whitcomb a–c a Division of Gastroenterology, Hepatology and Nutrition, Department of Medicine, b Cell Biology and Physiology, and c Human Genetics, University of Pittsburgh, Pittsburgh, Pa., USA
Key Words Obesity Acute pancreatitis, obesity Acute pancreatitis, severity APACHE-O scoring system APACHE-II Inflammatory response
Abstract Background: Obese patients appear to be at risk for complications of acute pancreatitis (AP). APACHE-O score has been suggested to improve APACHE-II accuracy in predicting severe outcome in AP. Aims: To determine if APACHE-O adds any predictive value to APACHE-II score and to test the hypothesis that obese patients are at increased risk of severe AP (SAP) because of a more intense inflammatory response to pancreatic injury. Methods: 102 AP patients were prospectively studied. Using a body mass index (BMI) 130, 28% of the subjects were obese. Nineteen patients developed organ dysfunction and were classified as SAP. Receiver-operating curves for prediction of SAP were calculated using admission APACHE-II and APACHE-O scores. Binary logistic regression was performed to assess if obesity is a risk for SAP and to determine the clinical factors associated with severe disease. Serum levels of IL-6, MCP-1 and CRP as
© 2006 S. Karger AG, Basel and IAP 1424–3903/06/0064–0279$23.50/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com
Accessible online at: www.karger.com/pan
well as Ranson’s scores were compared between obese and non-obese patients. Results: Admission APACHE-O (area under the curve AUC 0.895) and APACHE-II (AUC 0.893) showed similar accuracy in predicting severe outcome. BMI was identified as a significant risk for SAP (OR 2.8, p = 0.048) and mortality (OR 11.2, p = 0.022). CRP levels were significantly higher in obese AP patients (p = 0.0001) as well as Ranson’s score (p = 0.021). IL-6 and MCP-1 levels were higher in obese patients but did not reach statistical significance. Conclusions: Obesity is an independent risk for SAP. Admission APACHE-O score is not more accurate than APACHE-II. Our study results suggest that obesity increases the severity of AP by amplifying the immune response to injury. Copyright © 2006 S. Karger AG, Basel and IAP
Introduction
Acute pancreatitis (AP) is an acute inflammatory condition of the pancreas with a widely variable outcome. Most patients recover without complication. However, approximately 20% develop failure of one or more organ systems. The severity of AP is a function of the extent of
David C. Whitcomb, MD, PhD Division of Gastroenterology, Hepatology and Nutrition UPMC Presbyterian, Mezzanine Level 2, C Wing 200 Lothrop Street, Pittsburgh, PA 15213 (USA) Tel. +1 412 648 9604, Fax +1 412 383 7236, E-Mail [email protected]
pancreas injured and the intensity of the inflammatory response. Early death is often linked to the systemic inflammatory response syndrome (SIRS) with cardiovascular collapse or acute respiratory distress syndrome, while late death is more often associated with infected pancreatic necrosis, sepsis and multisystem organ failure. AP is a complex clinical syndrome encompassing a variety of mechanisms of pancreatic injury, activation of multiple inflammatory cascades and endangering multiple organs systems in a highly heterogeneous patient population. Thus, clinicians and scientists have struggled with developing a simple test to accurately predict clinical outcome. The development and use of different definitions, critical measurements and classification systems presents the additional problems of comparing complications and outcome. Currently, the most widely used definition of severe AP (SAP) is based on the Atlanta Consensus Conference in 1992 [1]. In the Atlanta consensus, SAP is defined by the presence of complications that are either systemic (e.g. organ system dysfunction) or local (e.g. peripancreatic fluid collection, pseudocyst, abscess, pancreatic necrosis). However, the risk for local complications and systemic complications and the consequences of these complications reflect different mechanisms and different outcomes. Furthermore, improved outcomes relating to early resolution of organ failure suggest that the Atlanta definition will need revisiting. The most commonly used predictive multifactorial scoring systems are the Ranson’s score [2] and the Acute Physiology and Chronic Health Evaluation (APACHE) II [3]. APACHE-II can be calculated on admission, in contrast to Ranson’s score, which requires 48 h for full assessment. Furthermore, APACHE-II can be updated daily over the hospital course, allowing monitoring of disease progression and response to therapy. Components of both tests measure the end-organ failure to respond to an inflammatory challenge (SIRS) and therefore represent indirect markers of the inflammatory response (i.e. arteriolar pO2 !60 mm Hg, rise in BUN 15 mg/dl, fluid sequestration 16 liters in Ranson’s score). It has been clearly demonstrated that regardless of the cut-off selection, their overall accuracy does not exceed 80% [4]. Obesity has been shown to be an independent risk factor for severe outcome in patients with AP [5–10]. Recently, Martinez et al. [11] conducted a meta-analysis suggesting that obesity (body mass index (BMI) 130 kg/m2) increased the risk of developing AP (OR 2.6, 95% CI 1.5– 4.6), as well as the severity (OR 2.6, 95% CI 1.5–4.6), but failed to find a significant difference in mortality. Although the increased risk or frequency of AP could be
280
Pancreatology 2006;6:279–285
related to the higher prevalence of gallstones in the obese [12], obesity appeared to be directly related to both local (OR 4.3, 95% CI 2.4–7.9) and systemic (OR 2.0, 95% CI 1.1–4.6) complications in the meta-analysis. The mechanism by which obesity increases the severity of AP is unclear. However, the observation that, in humans, the severity of AP is directly related to the intensity of the immune response [4] leads to the hypothesis that obese patients are at risk of more severe complications of AP than non-obese patients because of a more intense inflammatory response to pancreatic injury. Obesity may affect the immune response to injury. Adipose tissue is an important source of cytokines and obesity can be considered a state of chronic inflammation. Obese humans have increased levels of several proinflammatory markers, called adipokines [13]. For example, serum tumor necrosis factor (TNF-) [14], C-reactive protein (CRP) [15], interleukin-6 (IL-6) [16] and monocyte chemotactic protein (MCP-1) [17] concentrations are elevated in the obese when compared to non-obese patients and they fall after reduction of adipose tissue depots. Recently, a modification of the APACHE-II scoring system was proposed that included a factor for obesity. The proposed ‘APACHE-O’ scale adds 1 point for BMI between 26 and 30 kg/m2 (overweight) and 2 points for index 130 kg/m2 (obese). Johnson et al. [18] used APACHE-O score in patients with AP and suggested that this system improves severity prediction. Herein we report a clinical study that investigates the relationship between obesity and SAP and tests the utility of the APACHE-O scoring system for predicting SAP. The aims of this study are: (1) to investigate prospectively whether APACHE-O adds any predictive value to admission APACHE-II score in a North American population, (2) to assess if a BMI 130 or other clinical factors recorded on admission represent a significant risk for organ system failure, and (3) to compare early and later inflammatory markers, including serum levels of IL-6, MCP-1 and CRP as well as Ranson’s scores between obese and non-obese AP patients.
Methods Patients The Severity of Acute Pancreatitis Study 1 (SAPS1) protocol was approved by the Institutional Review Board of the University of Pittsburgh Medical Center. Informed written consent was obtained from all the patients before their enrollment into the study. The diagnosis of AP was based on upper abdominal pain or ab-
Papachristou /Papachristou /Avula /Slivka / Whitcomb
dominal localizing signs and plasma amylase and/or lipase levels at least 3 times above the upper limit of normal. The time interval between the onset of symptoms and admission to the hospital was no more than 48 h. Patients were recruited within 24 h from the time of admission. Patients entering the study had blood samples obtained daily along with recording of clinical, physiological and biochemical measurements necessary for Ranson’s, APACHE-II and APACHEO score calculations. Patients were managed according to the standard clinical practice. The patient’s clinical records until discharge from hospital and in follow-up visits were reviewed by two investigators (G.I.P. and D.J.P.) and all local and systemic AP complications were recorded. The term severe AP (SAP) was reserved for remote organ dysfunction including cardiovascular, pulmonary and renal failure necessitating ICU admission or organ support (meaning systolic pressure !90 mm Hg or use of vasopressors, arterial pO2 !60 mm Hg at room air or mechanical ventilation and serum creatinine 12 mg/ dl after rehydration or initiation of hemodialysis) [1]. Pancreatic necrosis was assessed by contrast-enhancing abdominal CT scan, which was performed in most of the patients (77 out of 102; 76%). Hospital mortality was defined as death within the same hospitalization for AP. We assessed if BMI 130 or other factors recorded on admission are associated with organ system failure. Clinical factors identified on history-taking were assessed, including gender, race, age, BMI, active smoking, alcohol use, presence of comorbidites (diabetes mellitus, coronary artery disease, chronic obstructive pulmonary disease, chronic renal insufficiency and cirrhosis) and etiology of AP (gallstones, post-endoscopic retrograde cholangiopancreatography (post-ERCP) procedure, idiopathic and alcohol-associated). All the above clinical factors have common features that can be assessed and recorded upon admission. Serum Cytokine Assay We assessed the inflammatory response in AP by measuring IL-6, MCP-1, and CRP serum concentrations. Serum samples collected within 24 h of the time of admission were assayed for IL-6, MCP-1 and CRP levels, quantified using a fluorescence-based capture sandwich immunoassay based on Luminex Flowmetrix system (Luminex, Austin, Tex., USA) and run in the University of Pittsburgh Cancer Institute’s Luminex Core Facility. Samples were analyzed using the Bio-Plex suspension array system, which includes a fluorescent reader and a Bio-Plex Manager analytical software (Bio-Rad Laboratories, Hercules, Calif., USA) [19]. Statistical Analysis Sample size was measured using one-sided calculations with of 0.05 and power of 80%, based on an incidence of obesity (BMI 130 kg/m²) in one third of our population. Using the proposed odds ratio in the recent literature [11], a sample size of 95 AP patients (71 non-obese and 24 obese) allows sufficient power to detect a 2.5fold increase in the risk of severe disease in obese subjects. Receiver-operating curves for prediction of SAP were calculated using admission APACHE-II and APACHE-O scores. The area under the curve (AUC) was determined for comparing predictive accuracy. The diagnostic accuracy of individual clinical scores for severe outcome was assessed using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy.
Obesity Increases the Severity of Acute Pancreatitis
Table 1. Systemic and local complications in 102 AP patients
Systemic (n = 19)
n
Local (n = 23)
n
Respiratory failure Cardiovascular failure Renal failure Single organ failure Multisystem organ failure Death
15 8 10 10 9 5
Pseudocysts Infected pseudocysts Pancreatic necrosis Infected necrosis Pseudoaneurysms
13 2 11 3 2
Binary logistic regression analysis was performed to assess if BMI 130 or other clinical factors recorded on admission represent a significant risk for organ system failure. Serum protein levels and Ranson’s scores were compared between AP patients with a BMI 130 (obese) vs. BMI ^30 (non-obese) by using Mann-Whitney test. For all statistical comparisons a p value !0.05 was considered significant.
Results
Clinical Features of Patients We prospectively studied 102 consecutive patients with AP admitted to the University of Pittsburgh Medical Center between June 2003 and October 2004. Twenty-nine subjects (28%) were obese (BMI 130). There were 50 males and 52 females with a mean age of 50 years (range 15–90). Ninety-one were Caucasian (89%), 9 were African-American (9%) and 2 were Hispanic (2%). The mean BMI was 27.4 (range 13.6–42.8). Etiologic classification of AP included biliary in 36 patients (35%), postERCP procedure in 21 (20%), idiopathic in 18 (18%), alcohol-associated AP in 15 (15%), drug-induced in 5 (5%), hypertriglyceridemia in 4 (4%) and pancreatic masses in 3 (3%). Nineteen subjects developed organ failure and were classified as having SAP, of which 11 developed local complications including 9 patients with pancreatic necrosis. Etiologic classification in these 19 patients with SAP included biliary in 6 patients, post-ERCP procedure in 3, idiopathic in 3, alcohol-associated AP in 5, drug-induced in 1, and hypertriglyceridemia in 1. Overall, 23 of the 102 patients (23%) developed local complications including pancreatic pseudocysts, pancreatic necrosis and splenic or hepatic artery pseudoaneurysms. Systemic and local complications are presented in table 1.
Pancreatology 2006;6:279–285
281
35
Percent of subjects
1.00
Sensitivity
0.75
30 25 20 15
17% 14%
14% 8%
10 5 0
0.50
BMI ≥30 BMI <30
31%
1.4% Organ failure
Necrosis
Death
Fig. 2. Frequency (%) of organ failure, necrosis and death in obese
0.25
(BMI 130) vs. non-obese (BMI ^30) subjects with AP.
APACHE-O APACHE-II 0 0
0.25
0.50
0.75
1.00
1 – Specificity
Fig. 1. Receiver-operating curves for prediction of severity using admission APACHE-O score (—) and admission APACHE-II score (- - -). The areas under the curve (AUC) were 0.895 for APACHE-O and 0.893 for APACHE-II.
Table 2. Sensitivity, specificity, positive predictive value (PPV),
negative predictive value (NPV) and overall accuracy of admission APACHE-O and APACHE-II for the prediction of SAP at a range of cut-off values Sensitivity Specificity PPV Admission APACHE-O >7 90 >8 84 >9 84 >10 74 Admission APACHE-II >7 90 >8 79 >9 74 >10 69
NPV
Accuracy
70 80 82 86
41 49 52 54
97 96 96 94
74 81 83 84
77 83 85 89
47 52 47 59
97 95 93 93
80 83 80 86
Comparison of Admission Scoring Systems in Predicting SAP Admission APACHE-O demonstrated similar accuracy (AUC 0.895) when compared to APACHE-II (AUC 0.893) (fig. 1) in predicting severe outcome. The sensitiv-
282
Pancreatology 2006;6:279–285
ity, specificity, predictive values and accuracy of admission APACHE scores at a range of cut-off values are seen in table 2. The selected admission APACHE-O cut-off value was 9 and showed a sensitivity of 84%, specificity 82%, PPV 52%, NPV 96% and accuracy 83%. Both admission APACHE-O and APACHE-II demonstrated an improved accuracy in predicting organ failure vs. predicting pancreatic necrosis (AUC 0.780 and 0.766, respectively). Identification of Clinical Factors Recorded at Admission Associated with Organ Failure Logistic regression analysis identified BMI 130 as significant risk factor for organ failure in AP (OR 2.8, p = 0.048, CI 1.1–7.9). The relative frequencies of pancreatic necrosis, organ failure and death according to BMI cutoff value of 30 are shown in figure 2. Thirty-one percent and 17% of the obese AP subjects developed organ failure and pancreatic necrosis respectively, whereas the frequencies were approximately half of these in the nonobese AP group (14% organ failure and 8% pancreatic necrosis). The other selected clinical factors recorded on admission (gender, race, age, smoking, alcohol use, presence of comorbidities and etiology of AP) were not associated with severe disease. There was a trend toward alcohol consumption and SAP (p = 0.06, OR = 2.9). Furthermore, BMI 130 is also a risk factor for mortality in AP (OR 11.2, p = 0.022, CI 1.05–572.9). Comparison of Levels of Inflammatory Markers between Obese and Non-Obese Patients Serum IL-6, MCP-1, and CRP levels in all AP patients, obese and non-obese subjects are presented in table 3. CRP levels were significantly higher in obese than non-
Papachristou /Papachristou /Avula /Slivka / Whitcomb
Table 3. Mean 8 SEM values of IL-6, MCP-1, CRP and Ranson’s score in all patients, obese, non-obese and p value when comparing values in obese vs. non-obese patients
IL-6, pg/ml MCP-1, pg/ml CRP, mg/dl Ranson’s score
All patients Obese
Non-obese
p value
5208217 6998214 10.981.2 1.6480.19
4848285 6298235 7.681.0 1.3480.19
0.368 0.325 0.0001 0.021
6128260 8708471 18.082.5 2.3880.45
obese patients with AP (p = 0.0001). IL-6 and MCP-1 levels were higher in obese subjects with a trend for statistical significance (p = 0.368 and 0.325, respectively). Ranson’s score was higher in obese subjects (p = 0.021).
Discussion
Lankisch and Schirren [5] were among the first to recognize that obesity was a risk factor for SAP. Since then about a dozen studies have confirmed and extended this observation [6–11, 18]. We prospectively collected and analyzed 102 subjects AP hospitalized at University of Pittsburgh, a tertiary care medical center. We found that a large fraction of our patients were obese (almost 30%) in contrast to previous studies from Europe [18], Mexico [8] and Asia [9] in which the percentage of obese subjects enrolled was either much less (2% in the Asian and 8% in the Mexican study) or not reported. This study was designed with adequate power and demonstrated a significant association between obesity and severity in AP. Using logistic regression analysis, we determined that obese AP patients are 2.8 times more likely to develop organ failure than non-obese subjects. We tested whether inclusion of obesity in a multifactorial scoring system improved predictive accuracy of the outcome of AP in our cohort, as has been proposed by Johnson et al. [18]. We found that this approach did not provide additional benefit. Adding a BMI score to APACHE-II did not improve the predictive accuracy on admission (APACHE-O AUC 0.895 vs. APACHE-II AUC 0.893). One possible reason for the limited utility of this system is that the only obese individuals that will be reclassified from mild to severe are those with a BMI of 26–30 with an exact APACHE-II score of 7, or those with a BMI of 130 with an APACHE-II score of 6 or 7. Therefore, use of APACHE-O at the above cut-offs could improve the sensitivity without overall loss of accuracy.
Obesity Increases the Severity of Acute Pancreatitis
Our logistic regression analysis suggests that all obese patients are at higher risk of systemic complications regardless of the cause of AP or amount of initial pancreatic injury. APACHE-II and APACHE-O scores should be included into a continuous rather than binary risk calculation. A primary question is why obese individuals are at higher risk for SAP. Mery et al. [10] investigated the type of fat and found that severity of AP was associated with android fat distribution (OR 9.23, 95% CI 1.67–51.07) and higher waist circumference (OR 13.41, 95% CI 2.43– 73.97) rather than the percent total body fat. They speculated that android fat distribution increased risk of severity since this type of obesity is associated with an ‘overactive’ immune response. Although obesity and inflammation is a broad subject, there is an important observation that may be relevant to the risk of AP. Takahashi et al. [20] used microarray technology in an attempt to identify the molecules explaining the relationship between obesity and vascular disorders by comparing mRNA expression of about 12,000 genes in white adipose tissue in normal vs. high fat diet-induced obesity in mice. MCP-1 is an early inflammatory cytokine and expression of MCP-1 mRNA was increased 7.2-fold in obese mice compared with normal mice, which was twice as high as any other gene. They also measured MCP-1 protein levels in adipocytes, which were found to be 5 times higher than in lean mice. MCP-1 levels in plasma were also increased in obese mice, and a strong correlation between plasma MCP-1 levels and body weight was identified. A similar relationship was found in humans. Troseid et al. [21] found that changes in plasma MCP-1 levels were significantly correlated to changes in visceral fat (r = 0.41, p = 0.02). Exercise resulted in reduction of the MCP-1 serum levels in obese subjects. MCP-1 has been shown to be an early marker of SAP in both animal models and clinical studies [22, 23]. Furthermore, our group has recently demonstrated that a genetic polymorphism (G allele at –2518 position) in the MCP-1 gene significantly increases the risk of severe disease in subjects with AP (OR 7.0, 95% CI 1.5–34) [24]. These observations suggest that obesity may increase the severity of inflammation through a fat-dependent (non-genetic) overexpression of MCP-1 (and other adipokines) leading to an amplification of the downstream immunological response to injury. However, additional ethnic and environmental differences may also be important since the effect of obesity on severity appeared to be worst in South Africa [6], moderate in Mexico [8] and least severe in Taiwan [9].
Pancreatology 2006;6:279–285
283
To test the hypothesis that obesity amplifies the inflammatory response to pancreatic injury through different adipokines, many of which have been found to be upregulated in obese individuals, we assessed the inflammatory response in AP in multiple ways. We measured the serum levels of three different inflammatory markers within 24 h of admission and also calculated the Ranson’s score. We demonstrated that obese patients with a BMI 130 had significantly higher CRP levels (later marker). Mean IL-6 and MCP-1 concentrations (early marker) were higher in obese patients than nonobese patients, but did not reach statistical significance. Ranson’s score, which is an indirect marker of the inflammatory response, was also significantly higher in obese subjects. The major limitation with respect to assessing these markers was the variability in time delay between the onset of pancreatitis and admission – which would have a large impact on accurately measuring the ‘early markers’ IL-6 and MCP-1. Thus, it is possible that in many of our patients, the levels of the early markers might have already started to decline at the time of the sample draw. Peak CRP levels occur at 48–72 h from the time of injury. CRP tends to maintain high levels for a longer period of time and therefore may be a more accurate marker of the inflammatory response in our patients. An alternate hypothesis is that obesity amplifies the in-
flammatory cascade at sites downstream to MCP-1 and IL-6 actions. The highly significant increase in CRP and Ranson’s scores in obese patients, however, supports the hypothesis that obesity amplifies the immune response. This hypothesis would need to be supported by further studies since the possibility that the obese patients suffer more severe pancreatic injury and have a proportionally increased inflammatory response cannot be excluded. It would be of interest to study these markers in obese patients suffering from other varieties of acute abdominal insult such as perforated bowel or diverticulitis. In summary, this study demonstrated that in populations with a large fraction of obese subjects, admission APACHE-O score does not improve the predictive accuracy of APACHE-II for severe outcome in AP. We confirmed that obesity constitutes an independent risk for organ failure in AP in our cohort. Our study results suggest that obesity increases the severity of AP by amplifying the immune response to injury.
Acknowledgement This work was supported by DK061451 (D.C.W.) and DK054709.
References 1 Bradley EL: A clinically based classification system for acute pancreatitis: summary of the Atlanta symposium. Arch Surg 1993;128:586– 590. 2 Ranson JHC, Rifkind KM, Roses DF, et al: Prognostic signs and the role of operative management in acute pancreatitis. Surg Gynaecol Obstet 1974;139:69–81. 3 Knaus WA, Draper EA, Wagner DP, et al: APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818–829. 4 Papachristou GI, Whitcomb DC: Predictors of severity and necrosis in acute pancreatitis. Gastroenterol Clin North Am 2004; 33: 871– 890. 5 Lankisch PG, Schirren CA: Increased body weight as a prognostic parameter for complications in the course of acute pancreatitis. Pancreas 1990;5:626–629. 6 Funnell IC, Bornman PC, Weakley SP, et al: Obesity: an important prognostic factor in acute pancreatitis. Br J Surg 1993; 80: 484– 486.
284
7 Martinez J, Sanchez-Paya J, Palazon JM, et al: Obesity: a prognostic factor of severity in acute pancreatitis. Pancreas 1999;19:15–20. 8 Suazo-Barahona J, Carmona-Sanchez R, Robles-Diaz G, Milke-Garcia P, Vargas-Vorackova F, Uscanga-Dominguez L, Pelaez-Luna M: Obesity: a risk factor for severe acute biliary and alcoholic pancreatitis. Am J Gastroenterol 1998;93:1324–1328. 9 Tsai CJ: Is obesity a significant prognostic factor in acute pancreatitis? Dig Dis Sci 1998;43: 2251–2254. 10 Mery CM, Rubio V, Duarte-Rojo A, SuazoBarahona J, Pelaez-Luna M, Milke P, RoblesDiaz G: Android fat distribution as predictor of severity in acute pancreatitis. Pancreatology 2002;2:543–549. 11 Martinez J, Sanchez-Paya J, Palazon JM, Suazo-Barahona J, Robles-Diaz G, Perez-Mateo M: Is obesity a risk factor in acute pancreatitis? A meta-analysis. Pancreatology 2004;4: 42–48.
Pancreatology 2006;6:279–285
12 Torgerson JS, Lindroos AK, Naslund I, Peltonen M: Gallstones, gallbladder disease, and pancreatitis: cross-sectional and two-year data from the Swedish Obese Subjects (SOS) and SOS reference studies. Am J Gastroenterol 2003;98:1032–1041. 13 Lyon CJ, Law RE, Hsueh WA: Minireview: adiposity, inflammation, and atherogenesis. Endocrinology 2003;144:2195–2200. 14 Dandona P, Weinstock R, Thusu K, AbdelRahman E, Aljada A, Wadden T: Tumor necrosis factor- in sera of obese patients: fall with weight loss. J Clin Endocrinol Metab 1998;83:2907–2910. 15 Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB: Elevated C-reactive protein levels in overweight and obese adults. JAMA 1999;282:2131–2135. 16 Khaodhiar L, Ling PR, Blackburn GL, Bistrian BR: Serum levels of interleukin-6 and C-reactive protein correlate with body mass index across the broad range of obesity. JPEN J Parenter Enteral Nutr 2004;28:410–415.
Papachristou /Papachristou /Avula /Slivka / Whitcomb
17 Christiansen T, Richelsen B, Bruun JM: Monocyte chemoattractant protein-1 is produced in isolated adipocytes, associated with adiposity and reduced after weight loss in morbid obese subjects. Int J Obes Relat Metab Disord 2005; 29:146–150. 18 Johnson CD, Toh SK, Campbell MJ: Combination of APACHE-II score and an obesity score (APACHE-O) for the prediction of severe acute pancreatitis. Pancreatology 2004;4:1–6. 19 Seideman J, Peritt D: A novel monoclonal antibody screening method using the Luminex100 microsphere system. J Immunol Methods 2002;267:165–171.
Obesity Increases the Severity of Acute Pancreatitis
20 Takahashi K, Mizuarai S, Araki H, Mashiko S, Ishihara A, Kanatani A, Itadani H, Kotani H: Adiposity elevates plasma MCP-1 levels leading to the increased CD11b-positive monocytes in mice. J Biol Chem 2003; 278: 46654– 46660. 21 Troseid M, Lappegard KT, Claudi T, Damas JK, Morkrid L, Brendberg R, Mollnes TE: Exercise reduces plasma levels of the chemokines MCP-1 and IL-8 in subjects with the metabolic syndrome. Eur Heart J 2004;25:349–355. 22 Grady T, Liang P, Ernst SA, Logsdon CD: Chemokine gene expression in rat pancreatic acinar cells is an early event associated with acute pancreatitis. Gastroenterology 1997; 113: 1966–1975.
23 Rau B, Baumgart K, Kruger CM, Schilling M, Beger HG: CC-chemokine activation in acute pancreatitis: enhanced release of monocyte chemoattractant protein-1 in patients with local and systemic complications. Intensive Care Med 2003;29:622–629. 24 Papachristou GI, Sass DA, Avula H, Lamb J, Lokshin A, Barmada MM, Slivka A, Whitcomb DC: The MCP-1 -2518 G allele acts as a potent disease modifier in acute pancreatitis. Clin Gastroenterol Hepatol 2005;3:475–481.
Pancreatology 2006;6:279–285
285