J Am Board Fam Pract: first published as 10.3122/15572625-12-1-68 on 1 January 1999. Downloaded from Spuriously Normal Amylase Levels in a Patient With Acute Pancreatitis Secondary to Hypertriglyceridemia

Kristen Okerberg, MD, and Makau Lee, MD, PhD

Acute pancreatitis is a disease that has varied clini­ function tests included total bilirubin 1.3 mg/dL cal manifestations often requiring clinicians to rely (normal, 0.2 - 1.2 mg/dL), aspartate aminotrans­ on laboratory tests, such as serum amylase and li­ ferase 25 lUlL (normal, 8 - 42 lUlL), alanine pase, as diagnostic aids. 1,2 Patients with acute pan­ aminotransferase 18 lUlL (normal, 3 - 36 lUlL), creatitis, however, can have normal amylase levels and alkaline phosphatase 60 lUlL (normal, 50 - as a result of hypertriglyceridemia. 3 The aims of 125 lUlL). Other pertinent laboratory values in­ this report are to describe a case of acute pancreati­ cluded a serum amylase of74 lUlL (normal, 40- tis secondary to hypertriglyceridemia, to discuss 125 lUlL), serum lipase of 106 lUlL (normal, 4 - the diagnostic dilemma caused by lactescent 24 lUlL), triglycerides of 9990 mg/dL (normal, plasma, and to review the therapeutic options for 10- 249 mg/dL), and a cholesterol level of 1255 this type of pancreatitis. mg/dL (normal, 150 - 199 mg/dL). A diagnosis of acute pancreatitis was made. Case Report The patient was treated with nothing by mouth, A 41-year-old woman was admitted for evaluation nasogastric suction, pain control with meperidine, of severe abdominal pain of 1 day's duration ac­ antiemetic therapy with promethazine, intrave­ companied by nausea, vomiting, and subjective nous insulin infusion, and parenteral hyperali­ fever. Her medical history was notable for diabetes mentation. On hospital day 2, a sonogram showed mellitus and hypercholesterolemia. The patient no evidence of gallbladder disease or dilatation of was a nonsmoker, and she denied any alcohol con­ bile ducts. Abdominal computed tomography sumption. Her only medication at admission was (CT) showed inflammatory changes surrounding subcutaneous insulin for her diabetes. the pancreas consistent with acute pancreatitis and At admission, the patient's temperature was a fatty liver. The patient's condition improved 100 oF, respiratory rate 18/min, pulse 89 beats per gradually during the next 9 days, and her serum li­ minute and blood pressure 124170 mmHg. She pase became normal by hospital day 10. She was http://www.jabfm.org/ had diffuse abdominal tenderness without guard­ gradually restarted on a low-fat, low-cholesterol ing or rebound tenderness, hypoactive bowel diet, and lipid-lowering therapy (ie, gemfibrozil sounds, and no hepatosplenomegaly. Rectally 600 mg twice-a-day) was begun. The patient was there were no masses, and stool was guaiac nega­ discharged after 14 days of hospitalization. At the tive. Findings on the remainder of the physical ex­ time of discharge, her serum triglyceride and cho-"

amination were unremarkable. Laboratory values lesterollevels were 303 mg/dL and 202 mg/dL, on 3 October 2021 by guest. Protected copyright. included a peripheral white blood cell count 10.8 x respectively. She was asymptomatic at 3 months' 103/mL (with a normal differential count), hemo­ follow-up. globin 12.9 g/dL, hematocrit 39 percent, and platelets 239,000/mL. Serum electrolytes were Discussion within normal limits with a blood urea nitrogen of Patients with acute pancreatitis typically complain 1.0 mg/dL, creatinine of 0.4 mg/dL, and calcium of upper quadrant abdominal pain that radiates to of 8.3 mg/dL (normal, 8.7 - 10.6 mg/dL). Liver the back, is relieved by sitting up, and is associated with nausea and vomiting. 1 Some, however, can have nonspecific complaints and varied findings on Submitted 26January 1998. 1 From the Deparonent of Medicine (KO, ML), University physical examination. Clinicians must thus rely on of Texas Health Science Center, San Antonio. Address reprint such laboratory tests as serum amylase and lipase requests to Makau Lee, MD, PhD, Division of Digestive Dis­ eases/Medicine, University of Mississippi Medical Center, to support the diagnosis. Because the serum amy­ 2500 North State St,Jackson, MS 39216-4504 lase assay tends to be less expensive and more read-

68 JABFP Jan.-Feb.1999 Vol. 12 No. I

.... J Am Board Fam Pract: first published as 10.3122/15572625-12-1-68 on 1 January 1999. Downloaded from ily available than lipase assay in most hospitals, enzyme elevations relative to the course of the dis­ most clinicians will obtain amylase measurements ease in most studies, and (3) the lack of standard­ first in cases of suspected pancreatitis.I ,2 Recogniz­ ized enzyme detection methods.2 The sensitivity ing factors that reduce the sensitivity and speci­ and specificity of amylase and lipase are reported ficity of serum amylase and lipase can help prevent to vary considerably depending on the detection unnecessary invasive diagnostic procedures and al­ method used, ranging from 70 to 100 percent and low for appropriate treatment. One factor that can 33 to 89 percent for serum amylase, and from 74 to lead to spuriously normal amylase values is hyper­ 100 percent and 34 to 100 percent for serum li­ triglyceridemia.3 Specific treatment to reduce pase, respectively.2 WIthin these ranges, most re­ serum triglycerides can hasten clinical recovery cent studies document higher sensitivity and speci­ during the acute episode and allow risk factor ficity for lipase.I ,2 Because elevated serum lipase is modification.4,5 not masked by hypertriglyceridemia, lipase may be Patients with acute pancreatitis and concomitant used to improve diagnostic sensitivity in cases of hypertriglyceridemia most commonly report a his­ suspected pancreatitis with normal amylase values. tory of poorly controlled diabetes, alcoholism, pre­ Clavien et al8 have shown that the overall diagnos­ disposing pharmacologic agents (such as estrogens, tic sensitivity for acute pancreatitis improved from diuretics, or glucocorticoids),2 or dietary risk fac­ 81 to 94 percent when serum lipase was measured tors.6 The incidence of hyperlipidemia in patients in addition to amylase. As illustrated by this case with acute pancreatitis has been reported in the report as well, serum lipase levels should be ob­ range of 12 to 38 percent,7 which encompasses pa­ tained in patients suspected to have acute pancre­ tients with transient, mild lipid abnormalities that atitis despite normal amylase levels. would not precipitate acute pancreatitis. Our re­ Treatment of acute pancreatitis generally in­ port, however, refers specifically to the 4 to 19 per­ cludes intravenous fluid replacement, pain man­ cent of patients with acute pancreatitis who have el­ agement, restoration of fluid and electrolyte bal­ evated concentrations of particulate triglycerides in ance, and observation either in the intensive care their plasma.7,8 In this subset of patients, hyper­ unit or on the medical unit depending on initial triglyceridemia can act as both a precipitating cause clinical assessment.ll Patients who fail to improve and a continuing risk factor for pancreatitis.5 within 72 hours or who deteriorate rapidly require Hypertriglyceridemia compounds the diagnosis dynamic contrast CT to rule out necrotizing pan­ of acute pancreatitis by causing spuriously normal creatitis.ll Symptomatic patients with necrotizing amylase levels.3 Plasma triglyceride levels higher pancreatitis should undergo CT-guided fine-nee­ than 500 mg/dL interfere with in vitro determina­ dle aspiration to rule out infection and to deter­ http://www.jabfm.org/ tion of the actual amylase level by preventing the mine whether surgical intervention is required. 1,1 I calorimetric reading of the assay end point.2 Serial Overall, although pancreatitis caused by hyper­ dilutions of the patient's sample with the assay triglyceridemia has the same prognosis as other buffer to reduce interference of light transmission causes of the acute episode,6 early recognition and by hyperlipidemic serum can reveal an abnormal treatment have been shown to hasten clinical re­

amylase value that was previously masked by the covery.5,12 In addition to insulin, which promotes on 3 October 2021 by guest. Protected copyright. lactescent plasma.3 Other studies investigating the tissue uptake of lipids, specific treatment modali­ potential role of circulating amylase inhibitors ties can include fresh frozen plasma 13 and plasma­ have not been well substantiated.9,lo Alternatively, pheresis.S,I2 Plasmapheresis has been shown to patients with hypertriglyceridemia can also have a cause dramatic clinical improvement in patients more insidious course of disease progression than with acute pancreatitis and to prevent disease re­ other patients with acute pancreatitis, resulting in currences.5 The benefits of these treatments might normalization of amylase levels by the time they relate to the elimination of excessive plasma pro­ seek medical help.8 teases associated with acute pancreatitis, either by Determining the sensitivity and specificity of direct removal by plasmapheresisl2 or replenish­ serum amylase and lipase assays is difficult because ment of protease inhibitors, such as alpharmacro­ of (1) the lack of randomized, controlled studies in­ globulin, by donor plasma. 13 In the recovery phase, volving patients with confirmed diagnosis of pan­ treatment should also include elimination of sec­ creatitis, (2) the failure to document the timing of ondary causes of hypertriglyceridemia, dietary

Normal Amylase Levels With Pancreatitis 69 J Am Board Fam Pract: first published as 10.3122/15572625-12-1-68 on 1 January 1999. Downloaded from

modification with pharmacologic reduction of ical assessment of hyperlipidemic pancreatitis. Am] triglycerides, if necessary, and avoidance of alcohol GastroenteroI1995;90:2134-9. consumption. 14 7. BrunzellJD, Schrott HG. The interaction offamil­ In summary, hypertriglyceridemia can lead to ial and secondary causes of hypertriglyceridemia: role in pancreatitis. Trans Assoc Am Physicians the erroneous exclusion of acute pancreatitis as a 1973;86:245-54. diagnostic possibility because hypertriglyceridemia 8. Clavien PA, RobertJ, Meyer P, Borst F, Hauser H, can cause spuriously normal amylase levels. Recog­ Herrmann F, et al. Acute pancreatitis and nor­ nizing this subset of patients with acute pancreati­ moamylasemia. Not an uncommon combination. tis can lead to more efficacious acute treatment and Ann Surg 1989;210:614-20. long-term risk factor modification. 9. Mishkin S, Bates], O'HashiJ, Schneider P, Snider­ man AD, Wolf RO. Possible mechanisms of normal References amylase activity in hyperlipemic pancreatitis. Can Med Assoc J 1976;115: 1016-9. 1. Ranson ]H. Diagnostic standards for acute pancre­ 10. Kessler ]1, Miller M, Barza 0, Mishkin S. Hyper­ atitis. World] Surg 1997;21:136-42. lipemia in acute pancreatitis. Metabolic studies in a 2. Wong EC, Butch AW, Rosenblum]L. The clinical patient and demonstration of abnormal lipoprotein­ chemistry laboratory and acute pancreatitis. Clin triglyceride complexes resistant to the action of Chem 1993;39:234-43. lipoprotein lipase. AmJ Med 1967;42:968-76. 3. Fallat RW, Vester ]W, Glueck CJ. Suppression of 11. Tenner S, Banks PA. Acute pancreatitis: nonsurgical amylase activity by hypertriglyceridemia. JAMA management. WorldJ Surg 1997;21:143-8. 1973;225:1331-4. 12. Larvin M, Lansdown MR, McMahon M], Chalmers 4. Chait A, Brunzell]D. Chylomicronemia syndrome. AG, Turney]H, Brown AM. Plasmapheresis: a ra­ Adv Intern Med 1992;37:249-73. tional treatment for fulminant acute pancreatitis? 5. Piolot A, Nadler F, Cavallero E, Coquard ]L, J aco­ BM] 1988;297:593-4. tot B. Prevention of recurrent acute pancreatitis in 13. Leese T, Holliday M, Heath 0, Hall AW, Bell PRo patients with severe hypertriglyceridemia: value of Multicenter clinical trial of low volume fresh frozen regular plasmapheresis. Pancreas 1996; 13 :96-9. plasma therapy in acute pancreatitis. Br ] Surg 6. Fortson MR, Freedman SN, Webster PD 3rd. Clin- 1987;74:907-11. http://www.jabfm.org/ on 3 October 2021 by guest. Protected copyright.

70 ]ABFP ]an.-Feb.1999 Vol. 12 No.1