C-reactive Protein-its Role In Evaluating The Severity Of Hyperlipidemic Acute Pancreatitis

Objectives: To promote the level of diagnosis and treatment of hyperlipidemicacute pancreatitis by analyzing clinical features of hyperlipidemic acute pancreatitis(HLAP) and non hyperlipidemic acute pancreatitis (NHLAP), especially the role ofC reactive protein (CRP) in predicting the severity of acute pancreatitis.Methods: A total of1073episodes of patients with acute pancreatitis between July2009and June2013from two3A grade hospitals were retrospectively studied. Comparedthe clinical characteristics between hyperlipidemic acute pancreatitis andnon hyperlipidemic acute pancreatitis in age, gender, hospital stay, the recurrence rate andconcomitant diseases (diabetes, hypertension and fatty liver), severity, local complicationsand serum C reactive protein concentrations. Receiver operating characteristic curve (ROCcurve) analysis was used to choose the best cut off values for diagnostic C reactive proteinin patients with hyperlipidemic acute pancreatitis and non hyperlipidemic acutepancreatitis. Multiple linear regression analysis was used for risk factors of C reactiveprotein in patients with hyperlipidemic acute pancreatitis.Results: A total of1073episodes of acute pancreatitis (51.74±17.55years of age)were studied. There were208episodes of hyperlipidemic acute pancreatitis and865episodes of non hyperlipidemic acute pancreatitis. There were significant differencesbetween the patients with hyperlipidemic acute pancreatitis and non hyperlipidemic acutepancreatitis with respect to age, gender, length of hospital stay, local complications,pseudocyst formation, concomitant diseases (DM, hypertension, and fatty liver), theincidences of sever acute pancreatitis (Balthazar D, E), and the incidences of mild acutepancreatitis (MAP) and moderately severe acute pancreatitis (MSAP) according to the 2012revision of the Atlanta classification (P<0.05). There were no significant differencesin the duration of abdominal pain (P=0.809), admission delay (P=0.970), and the incidenceof sever acute pancreatitis according to the2012revision of the Atlanta classification(P=0.266) between the patients with hyperlipidemic acute pancreatitis andnon hyperlipidemic acute pancreatitis. The serum C reactive protein concentrations inpatients with hyperlipidemic acute pancreatitis were significantly higher than theC reactive protein levels in the patients with non hyperlipidemic acute pancreatitis(P<0.05). The serum C reactive protein concentrations of hyperlipidemic acute pancreatitispatients on days2and3were much higher than non hyperlipidemic acute pancreatitispatients, despite of the severity of acute pancreatitis defined by Balthazar’s criteria or bythe2012revision of the Atlanta classification. When severe acute pancreatitis was definedby Balthazar’s criteria (D and E), the highest Youden’s index of hyperlipidemic acutepancreatitis patients occurred on day2when the C reactive protein cut off value was162.75mg/L(sensitivity81%, specificity84%), but the highest Youden’s index ofnon hyperlipidemic acute pancreatitis patients was on day3when the C reactive proteincut off value was142.50mg/L(sensitivity69%, specificity85%). When severe acutepancreatitis was defined by the2012revision of the Atlanta classification, the highestYouden’s index of hyperlipidemic acute pancreatitis patients occurred on day2when theC reactive protein cut off value was172.95mg/L(sensitivity100%, specificity72%), andthe highest Youden’s index of non hyperlipidemic acute pancreatitis patients was on day3when the C reactive protein cut off value was197.00mg/L(sensitivity60%, specificity89%). Higher serum C reactive protein levels in hyperlipidemic acute pancreatitis caseswere correlated with higher incidences of diabetes and fatty liver, and lower incidences infemales, elevated very low density lipoprotein (VLDL) levels, and lower high densitylipoprotein (HDL) levels.Conclusions: Hyperlipidemic acute pancreatitis has characteristics of highermorbidity, lower age, higher rate of recurrence, local complications, and concomitantdiseases (DM, and fatty liver). The serum C reactive protein concentrations in patients with hyperlipidemic acute pancreatitis were significantly higher than the C reactive proteinlevels in the patients with non hyperlipidemic acute pancreatitis. An appropriate C reactiveprotein cut off value in predicting severe acute pancreatitis in hyperlipidemic acutepancreatitis patients was different from non hyperlipidemic acute pancreatitis patients.Higher serum C reactive protein levels in hyperlipidemic acute pancreatitis cases werecorrelated with higher incidences of diabetes and fatty liver, and lower incidences infemales, elevated very low density lipoprotein (VLDL) levels, and lower high densitylipoprotein (HDL) levels.