Objectives: To define the pro- and anti-inflammatory cytokine response during acute severe pancreatitis and to evaluate its predictive value on hospital mortality.
Design: Prospective, multicenter study.
Setting: Nine multidisciplinary intensive care units (ICUs).
Patients: Fifty patients with a diagnosis of acute pancreatitis who were admitted to the ICUs during a 14-month period were prospectively enrolled.
Measurements and main results: Plasma concentrations of tumor necrosis factor (TNF)-alpha interleukin (IL)-1beta, IL-6, IL-10, IL-1 receptor antagonist (IL-1ra) were determined at the inclusion and during the ICU stay at Days 1, 3, 8, and 15. The patient population was analyzed by age, gender, previous health status, preexisting organ dysfunction, and type of acute pancreatitis. Physiologic variables were measured at inclusion and during ICU stay to calculate the new Simplified Acute Physiology Score II, the Acute Physiology and Chronic Health Evaluation II (APACHE II) score, and the number of organ system failures. Prognostic factors were determined by univariate methods and stepwise logistic regression analysis. Fifty patients were included, among whom 34 at the time of the ICU admission. Preinclusion symptom history was < or = 48 hrs in 78% of the patients. Eleven patients (22%) died during their hospital stay. At inclusion, 46 of 50 patients had elevated IL-6 serum levels (1512 +/- 635 pg/mL; normal value < 10 pg/mL), 36% of the patients had raised TNF-alpha concentrations, and all patients had an anti-inflammatory response (IL-10, 92 +/- 15 pg/mL [normal value < 10 pg/mL]; and/or IL-1ra, 7271 +/- 2530 pg/mL [normal value < 200 mg/mL]). During the follow-up period, pro- and anti-inflammatory cytokines remained elevated in at least 75% of the population. Positive correlations were found between inclusion pro- (IL-6) and anti-inflammatory cytokine concentrations at Day 1 (IL-10, IL-1ra; p < .0001) and between cytokines levels and the Simplified Acute Physiology Score II. While hospital mortality was linked to six factors in univariate analysis (age, cirrhosis, delay between hospitalization and ICU admission, severity of illness, and IL-10 and IL-6 plasma levels) when using stepwise logistic regression, only severity scoring indexes were predictive of death.
Conclusions: During acute severe pancreatitis, the pro- and anti-inflammatory cytokine response occurred early and persisted in the systemic circulation for several days. Although associated with the patient's severity at inclusion and outcome, cytokine plasma concentrations were unable to predict death accurately in individual patients. If confirmed, these results should be taken into consideration when selecting patients who are apt to benefit from new therapies aimed at modifying the immune inflammatory response.