Lipase/amylase ratio. A new index that distinguishes acute episodes of alcoholic from nonalcoholic acute pancreatitis

Because of observations that patients with acute episodes of alcoholic pancreatitis had high serum lipase levels whereas patients with gall stone pancreatitis had high serum amylase levels, a prospective study was undertaken to determine whether the ratio of serum lipase to serum amylase, a newly computed ratio, would discriminate between acute episodes of alcoholic and nonalcoholic pancreatitis. In phase one, 30 consecutive patients with acute pancreatitis were entered into the study and divided into groups A and B. Patients with renal failure were excluded from the study. Group A consisted of 20 patients in whom the etiology of pancreatitis was alcohol. Group B consisted of 10 patients whose pancreatitis was nonalcoholic in etiology (predominantly gallstones). Serum lipase values in group A ranged 492 to 25,706 U/L (median, 3433 U/L) and in group B from 711 to 31,153 U/L (median, 1260 U/L). These differences were not significant statistically. Serum amylase values in group A ranged from 104 to 2985 U/L (median, 331 U/L) and in group B from 423 to 13,000 (median, 1187 U/L). Although these figures were statistically different (P < 0.005), there was a considerable degree of overlap in the values between the two groups. The lipase/amylase ratio calculated from the blood sample obtained at presentation appeared to be a promising discriminatory index. The lipase/amylase ratio was calculated by using the amylase and lipase levels expressed as multiples of the upper limit of normal in each case. The lipase/amylase ratios in the alcoholic group ranged from 2.2 to 14.8, whereas the lipase/amylase ratio in nonalcoholic pancreatitis ranged from 0.31 to 1.93. These differences were statistically significant (P < 0.005). A lipase/amylase ratio of > 2 was indicative of an alcoholic etiology, and a ratio of < 2 suggested that the pancreatitis was nonalcoholic in nature. In phase two, this lipase/amylase ratio of 2 was applied prospectively to an unselected population of 21 consecutive patients with acute pancreatitis. Thirteen patients had a lipase/amylase ratio of > 2; in 11 of them, the etiology of the pancreatitis was alcohol. Eight patients had a lipase/amylase ratio of < 2; of them, only 1 patient had an alcoholic etiology for the pancreatitis. These differences were statistically significant (P < 0.05). The sensitivity of a lipase/amylase ratio of > 2 in detecting acute alcoholic pancreatitis was 91%, the specificity was 78%, and the positive predictive value was 85%.