laboratory amilasa

A Critical Evaluation of Laboratory Tests in AcutePancreatitisDhiraj Yadav, M.D., N. Agarwal, M.D., F.R.C.S., and C. S. Pitchumoni, M.D., M.A.C.G.Division of Gastroenterology and Department of Surgery, Our Lady of Mercy University Medical Center,New York Medical College, Bronx, New York

ABSTRACTAn ideal laboratory test in the evaluation of a patient withacute pancreatitis (AP) should, in addition to accuratelyestablishing the diagnosis of AP, provide early assessmentof its severity and identify the etiology. None of the testsavailable today meet all these criteria, and presently there isno biochemical test that can be considered the “gold stan-dard” for the diagnosis and assessment of severity of AP. Inthe diagnosis of AP, serum amylase and lipase remainimportant tests. Advantages of amylase estimation are itstechnical simplicity, easy availability, and high sensitivity.However, its greatest disadvantage is its low specificity. Anormal amylase would usually exclude the diagnosis of AP,with the exception of AP secondary to hyperlipidemia, acuteexacerbation of chronic pancreatitis, and when the estima-tion of amylase is delayed in the course of the disease. Themajor advantage of lipase is an increased sensitivity in acutealcoholic pancreatitis and in patients who initially present tothe emergency room days after the onset of the disease, aslipase remains elevated longer than amylase. Although onceconsidered to be specific for AP, nonspecific elevations oflipase have been reported in almost as many disorders asamylase, thus decreasing its specificity. Simultaneous esti-mation of amylase and lipase does not improve the accu-racy. Other enzymes for the diagnosis of AP—pancreaticisoamylase, immunoreactive trypsin, and elastase—aremore cumbersome and expensive and have no clear role inthe diagnosis of AP. No enzyme assay has a predictive rolein determining the severity or etiology of AP. Once thediagnosis of AP is established, daily measurements of en-zymes have no value in assessing the clinical progress of thepatient or ultimate prognosis and should be discouraged. Ahost of new serological and urinary markers have beeninvestigated in the last few years. Their main use is inpredicting the severity of AP. At present, serum C-reactiveprotein at 48 h is the best available laboratory marker ofseverity. Urinary trypsinogen activation peptides within12–24 h of onset of AP are able to predict the severity butare not widely available. Serum interleukins 6 and 8 seempromising but remain experimental. (Am J Gastroenterol2002;97:1309–1318. © 2002 by Am. Coll. of Gastroenter-ology)

INTRODUCTION

The diagnostic tests in the evaluation of a patient withsuspected acute pancreatitis (AP) must answer three ques-tions sufficiently early in the course of the disease: first andforemost, the tests should establish the diagnosis accurately,excluding other conditions that mimic AP with or withouthyperamylasemia. Second, at the earliest they should pro-vide an assessment of severity of AP to provide appropriatetreatment in the required setting: intensive care unitversusa regular floor or by an internist aloneversus a team ofmedical specialists that includes a surgeon, radiologist, andgastroenterologist trained in therapeutic endoscopy. Third,tests should also help in establishing the etiology for AP, soas to offer a definitive treatment such as cholecystectomyand/or to prevent recurrences of AP, as in the case of apatient with hyperlipidemic AP. The tests should be easilyavailable, cost-effective, and easily repeatable. Obviously,every patient should not undergo all tests.

This article, an extension of a number of recent reviewson the topic, including ours published in 1990 in this journal(1), critically evaluates the standard serum and urine testsand discusses the advances in laboratory technology thatmay have new diagnostic possibilities. In addition to ourprevious article on this topic (1), we have included relevantarticles after 1990, using a broad-based MEDLINE searchusing the termsacute pancreatitis, diagnosis, laboratorytests, and severity. Relevant articles as well as their refer-ences were reviewed.

SERUM TESTS FOR DIAGNOSIS OF AP

A host of serum enzymes such as amylase, lipase, trypsino-gen, elastase, phospholipase A2, ribonuclease, etc are avail-able to diagnose AP and/or to assess the severity, but ele-vated amylase levels continue to be the “gold standard”among the serum markers.

Serum AmylaseIn AP, serum amylase rises as a result of both increasedrelease and, to some extent, reduced catabolism. The mostfamiliar units of expression are the Somogyi unit (SU) andthe international unit (IU). One SU per 100 ml is equivalentto 1.85 IU/L. The normal values are 60–160 SU/100 ml or

THE AMERICAN JOURNAL OF GASTROENTEROLOGY Vol. 97, No. 6, 2002© 2002 by Am. Coll. of Gastroenterology ISSN 0002-9270/02/$22.00Published by Elsevier Science Inc. PII S0002-9270(02)04122-9

110–300 IU/L in serum and 35–260 SU/h or 65–480 IU/Lin urine (2).

SENSITIVITY. The sensitivity of any test is dependent onthe criteria used for diagnosis. If the test itself is taken as acriterion for diagnosis, as is often the case for serum amy-lase in AP, its sensitivity will be artificially raised to 100%(3). In one cohort (4), a 99.6% sensitivity for serum amylasereflected such subjective inclusion criteria. However, thesensitivity decreases to 81–95% when CT scan or ultra-sound is used to ascertain the clinical diagnosis of AP (5, 6).

Normoamylasemia is reported in 19–32% of patientswith AP (6, 7). Diagnosis of AP is questioned if the amylaselevel is normal. However, there are three major factors thatlower the sensitivity:

1. Time interval since onset of attack. A major confoundingfactor is the variable time elapsed between the onset ofthe symptoms and the first blood analysis in a givenpatient. Within 24 h of the onset of symptoms, all en-zymes are elevated; thus, serum amylase is as sensitive asserum lipase, pancreatic isoamylase (P-isoamylase), im-munoreactive trypsin, or elastase (8). Amylase is the firstone to return toward normal values, and as such, after thefirst hospital day it is the least sensitive of the enzymatictests for pancreatitis. Normalization can sometimes occurvery rapidly, indicating early resolution of the disease,increased urinary clearance, or, less frequently, extensivedestruction of the pancreas with cessation of pancreaticamylase production (6).

2. Alcoholic pancreatitis. Spechler et al. (7) noted thatpatients with acute alcoholic pancreatitis frequently(32%) have normal serum amylase levels. Furthermore,in patients with alcoholic AP, normoamylasemia wassignificantly associated with number of previous attacks(0.7 vs 0.4, p � 0.01), indicating a parenchyma that is nolonger able to produce sufficient amounts of enzymes(6).

3. Hypertriglyceridemia. Serum or urinary amylase levelsmay be normal in as many as 50% of patients withabdominal pain and hypertriglyceridemia, in whom aclinical diagnosis of AP is considered and when CTshows pancreatic inflammation (9). The hyperlipidemiainterferes with the amylase assay, leading to a spuriouslynormal result. A circulatory inhibitor of serum amylaseexists, as suggested by demonstration of elevated amy-lase by dilution of the serum (10), but is probably nottriglyceride itself (11). In all situations where AP issuspected with a normal serum amylase, triglycerideestimations must be done, and if lactescent plasma ispresent, serial dilution techniques should be employed tocircumvent false negative amylase results.

SPECIFICITY. The greatest limitation of serum amylase isits lack of specificity. Besides AP, conditions that increaseserum amylase levels include many diseases and derange-ments of the biliary tract, liver, intestines, genitourinary

tract, lungs, breast, prostate, central nervous system, and, ofcourse, salivary glands (12). Abnormal serum levels alsooccur in the presence of metabolic disturbances such asrenal failure, liver dysfunction, diabetic ketoacidosis, hypo-perfusion, eating disorders like anorexia nervosa and bu-limia (13), and abdominal and nonabdominal trauma includ-ing head injury, as well as with the use of various drugs (2,14). Interestingly, hyperamylasemia, seen commonly inHIV-infected persons and in up to 40% of intoxicated al-coholics, often does not signify AP, as predominantly S-type amylase levels alone go up (15, 16). An importantobservation is that, after ERCP, hyperamylasemia is notedin 25–43% of patients but only 1–4% have evidence ofpancreatitis (17). Furthermore, persistent hyperamylasemiamay be a normal variant (18) and has been recently de-scribed as a benign abnormality in many members of certainfamilies (19) (Table 1).

Another critical limiting factor that significantly influ-ences the specificity is the issue of what signifies an abnor-mal value (3). More than 200 techniques of amylase deter-mination are described in the literature, creating a host of“normal” values that differ both quantitatively and in meth-ods of expression (2, 20). Further, different studies havearbitrarily used various cutoff values ranging from just theupper limit of normal (300 IU/L) (3) to more than threetimes the upper limit of normal (�1000 IU/L) (21). Nointernational reference method or cutoff value has yet beenadopted toward establishing a standardized tool. With acutoff value of 300 IU/L (upper limit of normal), totalamylase has a sensitivity of 91–100% but a specificity of71–98%. Increasing the cutoff value to 1000 IU/L increasesthe specificity to around 100% but decreases sensitivity to aslow as 61%. Ideally, sensitivity should be increased at the

Table 1. Conditions With Increased Amylase

Abdominal disordersPancreatic disorders: acute pancreatitis, chronic pancreatitis,pseudocysts, pancreatic trauma, pancreatic cancerNonpancreatic intra-abdominal conditions: perforated bowel,mesenteric infarction, intestinal obstruction, appendicitis,peritonits, abdominal aortic aneurysm, ruptured ectopicpregnancy, fallopian and ovarian cysts, salpingitis, hepatitis

Extra-abdominal conditionsSalivary diseases, renal failure, ketoacidosis, pneumonia,cerebral trauma, burns, anorexia nervosa, bulimia,nonabdominal surgery

MacroamylasemiaIdiopathic hyperamylasemia

Familial and nonfamilialDrug induced

Definite association: azathioprine, L-asparginase,sulfonamides, tetracycline, didanosine, methyldopa, estrogens,furosemide, pentamidine, 5-aminosalicylic acid compounds,valproic acid, salicylate, thiazide, calcium, vinca alkaloidsProbable association: glucocorticoids, nitrofurantoin,phenformin, rifampin, FK-506 (tacrolimus), metronidazole,6-mercaptopurine, procainamide, diphenoxylate,chlorthalidone, cimetidine, cytosine arabinoside, cisplatin,cyclosporin A

1310 Yadav et al. AJG – Vol. 97, No. 6, 2002

expense of specificity when the penalty associated withmissing the disease is high. On the other hand, specificityshould be increased relative to sensitivity when the costs orrisks associated with further diagnostic techniques are sub-stantial. Because AP can be serious, it is wiser to increasethe sensitivity at the expense of specificity.

SEVERITY OF AP. Virtually all investigators agree thatthe magnitude of increase in amylase activity does notcorrelate with the severity of the disease. It is even possiblefor patients with very severe necrotizing AP to have normalor low values of amylase, indicating an inverse relationshipbetween the amylase level and the severity of the disease(22, 23). Clinically, AP does not seem to behave differentlywhen serum amylase is normal or elevated (6, 24). Once thediagnosis of AP is established, daily measurements of serumamylase have little value in assessing the clinical progress ofthe patient or ultimate prognosis (25). However, persistenthyperamylasemia that does not return to normal within 5–10days has been shown to correlate with complications such aspseudocysts, necrosis, or abscess.

ETIOLOGY OF AP. In general, patients with biliary pan-creatitis have markedly higher initial serum amylase levelsthan those with alcoholic pancreatitis or pancreatitis fromother causes. Hiatt et al. (26) observed that only 11% ofpatients with biliary disease had initial serum amylase val-ues lower than 1000 IU/L, whereas only 6% with alcoholicpancreatitis had initial amylase values higher than 1000IU/L.

CLINICAL PRACTICE RECOMMENDATIONS. The ad-vantage of serum amylase estimation lies in its technicalsimplicity and ready availability in all hospitals. In contrast,its greatest disadvantage is its overall low specificity. Be-cause the presence of a raised serum amylase in the clinicalsetting of abdominal pain is not entirely specific, it is de-sirable to confirm it by imaging studies. A normal amylaselevel would nearly exclude the diagnosis, with the excep-tions of possible hyperlipidemic pancreatitis, acute exacer-bation of chronic pancreatitis, and delayed estimation in thecourse of the disease. Indeed, AP should not be dismissed inthe presence of an amylase level that is either normal or onlymildly elevated on initial evaluation if the clinical suspicionfor AP is high and it is prudent to seek additional tests.Serum amylase has no value in assessment of severity oretiology of AP.

LipaseLipase (triacylglycerol acylhydrolase) is mainly synthesizedand stored as granules in the pancreatic acinar cells. Morethan 99% is subsequently excreted in the ductal systems,and less than 1% diffuses via the lymphatics and capillariesinto the general circulation (27). The concentration gradientbetween pancreatic tissue and serum is over 20,000-fold.Other sources of lipase are the tongue, esophagus, stomach,duodenum, leukocytes, adipose tissue, lung, and breast milk.

However, the lipase concentration in the pancreas is 100-fold greater than in the liver, duodenum, and small bowels(28). In AP, increased permeability in the basal pole of theacinar cells accounts for the pronounced rise of the enzymein the serum. Usually, serum lipase increases within 4–8 hafter onset of symptoms, peaks at 24 h, and returns tonormal after 8–14 days (29). Lipase assay is fast, reliable,practical, and almost as sensitive as an amylase assay. Thecost of the lipase assay compares favorably with amylaseassays, and the technique can be available 24 h a day, 7 daysa week in most hospitals.

SENSITIVITY. The sensitivity of lipase ranges from 85%to 100% (3, 21), with some reporting it to be less sensitivethan serum amylase (3) and others believing it to be moresensitive than amylase (21, 30). The major advantage oflipase is an increased sensitivity in acute alcoholic pancre-atitis and with late clinical presentation, as lipase remainselevated longer than does amylase. Clavien et al. (6) foundthat in patients with AP who had normal amylase, more thantwo thirds had elevated lipase levels. Gumaste et al. haveshown in their study on patients with AP and nonpancreaticabdominal pain (31) that the sensitivity of lipase levels ofgreater than three times normal is much higher than amylaselevels (100% sensitivity and 99% specificity, vs 72% sen-sitivity and 99% specificity for amylase levels). In a recentstudy of ERCP-induced AP (32), mean lipase values werefour times higher than amylase levels 2 h after the procedurein those who developed AP.

SPECIFICITY. Lipase elevation is not specific to AP, al-though it may be slightly better than amylase. Apple andassociates (30) have shown that lipase activity is four timesgreater than amylase activity in the pancreas. Second, pan-creatic tissue in chronic pancreatitis demonstrates a substan-tial decline in both amylase and lipase activity, with amylaseactivity showing a greater decrease compared to lipase (91%vs 26%). In conditions of extrapancreatic injury, lipase isalso elevated. Mumps, types I and IV hyperlipoproteine-mias, peptic ulcer, acute cholecystitis, extrahepatic biliaryobstruction, liver diseases, small bowel obstruction, intesti-nal infarction, perforated bowel, acute renal failure, fractureof bone, crush injury, fat embolism, and the postcholecys-tectomy syndrome are some examples (7, 33). In our studyof patients with diabetic ketoacidosis (34) we found non-specific elevation of lipase to occur more frequently thanamylase elevation. Recently, inflammatory bowel diseaseand familial pancreatic hyperenzymemia have been in-cluded among causes of lipase elevation (19, 35).

The reference point determined in different laboratoriesmay differ even for identical methods (27, 36). Further,different authors have arbitrarily used different cutoff val-ues. Steinberg et al. (3) have shown that the upper limit ofnormal itself provided the best cutoff value. Keim andassociates (5) suggest that 2-fold elevated lipase valuesshould be used as the cutoff, whereas Gumaste et al. (31)

1311AJG – June, 2002 Evaluation of Laboratory Tests in Acute Pancreatitis

advocate a cutoff level of three times normal. They observedthat when lipase elevation is not due to AP, the elevation isusually less than three times normal. However, Frank andGottlieb (29) have reported patients with lipase greater thanthree times normal secondary to renal insufficiency, malig-nant tumors, cholecystitis, esophagitis, and hypertriglyceri-demia. Other major disadvantages of lipase assay are a)presence of as many as four fractions of lipase in serum ofpatients with pancreatitis, b) the macroforms or macroli-pasemia contribute to hyperlipasemia, and c) technical dif-ficulties are more with lipase assays than amylase assays(27, 36, 37).

CLINICAL PRACTICE RECOMMENDATIONS. Themajor advantage of lipase is an increased sensitivity in acutealcoholic pancreatitis and late clinical presentation, as lipaseremains elevated longer than amylase. Its specificity may beslightly better than amylase; however, it is increasinglybeing recognized that nonspecific elevations of lipase can beseen in as many disorders as amylase. Serum lipase also hasno value in assessment of severity or etiology of AP.

Amylase, Lipase, or BothControversy exists whether amylase and lipase should beused alone or in combination to avoid overlooking patientswith AP. Opinions vary on the preferential test (38). Al-though amylase continues to be the screening test for AP (3,6, 39), a number of studies have challenged the primarydiagnostic role of serum amylase, and a case has been madefor the use of serum lipase instead (31, 40, 41).

Within 24 h after onset of symptoms, both amylase andlipase values have high sensitivity and specificity, withlipase having a slightly higher diagnostic value. Amylaseappears to be a better test in gallstone pancreatitis and lipasefrom alcoholic and other causes. The differences in perfor-mance of the two tests, though small, are definite (42).Simultaneous evaluation of amylase and lipase does notimprove the accuracy (5, 43).

Other Tests of Limited or No ValueNormal circulating amylase consists of P-isoamylase (40%of total amylase) and a salivary-type isoamylase (60%) (18).In AP, P-isoamylase is expected to rise; hence the estima-tion of this fraction is theoretically attractive. Importantly,some nonpancreatic abdominal emergencies such as acutebiliary tract disease, perforated duodenal ulcer, intestinalobstruction, infarction, and ruptured abdominal aortic aneu-rysm are also associated with an increase in P-isoamylase.Thus, a major group of differential diagnoses is not elimi-nated (44), and therefore, measurement of isoenzymes in theserum has been largely abandoned (25).

Macroamylases are large molecules of amylase whereabnormal proteins—IgA, IgG, or IgM—produce a largemolecular weight complex that cannot be cleared throughthe kidneys (45). It occurs in 0.1% of the population (46)and in up to 2.7% of hospitalized patients (47). In a largemajority of individuals macroamylasemia is not clinically

significant. However, in an asymptomatic individual withelevated serum amylase, which causes concern, an estima-tion of urine amylase that shows low levels will settle theissue.

Immunoreactive trypsinogen (IRT) has a sensitivity of97–100%, a specificity of 83%, and a positive predictivevalue of 46–74% (3, 8). However, the levels are also highin malignant neoplasms of the pancreas, diabetes mellitus,chronic renal failure, hypercalcemia, hypertriglyceridemia,cirrhosis of the liver, chronic pancreatitis, and extrahepaticobstructive jaundice (48). Therefore, many believe that IRThelps to confirm the pancreatic origin of a raised serumamylase, but does not much improve the diagnostic accu-racy in patients with suspected AP with normal or onlymildly elevated amylase. IRT is a more difficult test toperform and requires 24 h to complete.

Elastase-1, a proteolytic enzyme liberated in the course ofAP, has a specific elastolytic action that is responsible fordigestion of blood vessel walls and vascular complications.By radiommunoassay, one can demonstrate elevated elas-tase levels in the serum in all cases of AP. The test, however,lacks specificity, as it is elevated in two thirds of pancreaticcancer patients and less frequently in chronic pancreatitis(17). Also, serum elastase levels did not correlate withdisease severity or development of complications. It is alsounable to discriminate between alcohol and gallstone pan-creatitis (8). Elastase confers no benefit as a diagnostic testnor does it provide any prognostic information. The onlystrength of IRT and elastase assays is that they remainelevated for 7–10 days after the onset of AP, and elastase isthe more sensitive of the two. Their determination may beindicated in patients who present late and in whom thediagnosis of pancreatitis is in doubt.

CLINICAL PRACTICE RECOMMENDATIONS. Pres-ently, there is no role of isoamylases, IRT, macroamylases,and elastase estimations in the routine management of pa-tients with AP.

SERUM MARKERS OF SEVERITY OF AP

HematocritRecently, hemoconcentration has been identified to be astrong risk factor and early marker for necrotizing pancre-atitis and organ failure (49, 50). An admission Hct of �47and a failure of admission Hct to decrease at 24 h representa strong risk factor for the development of pancreatic ne-crosis. Baillargeon et al. (49) compared 32 patients withnecrotizing pancreatitis to an equal number of patients withmild pancreatitis. At 24 h, 81% (26/32) of patients meteither of the criteria (admission HCT � 47 or failure of HCTto decrease), compared to 12.5% (4/32) of those with mildAP (p � 0.01). The sensitivity and specificity using thesecriteria on admission were 34% and 91% and, at 24 h, 81and 88%, respectively.


C-Reactive Protein (CRP)Serum CRP is an acute phase reactant that is elevated inseveral inflammatory conditions and serves as a nonspecificmarker for inflammation. CRP levels peak on the 3rd or 4thday, and values of �150 mg/L when done 48 h after theonset of symptoms are now accepted as a proven predictorof severity of AP (42). In a study by Wilson et al. (51), peakCRP levels of �210 mg/L were able to differentiate severeAP from the milder form with a sensitivity of 83–84% anda specificity of 74–85%. Another recent study (52) hasshown CRP to be superior to interleukin 1B (IL-1B), IL-8,and tumor necrosis factor � (TNF-�) and equivalent to IL-6in predicting severe pancreatitis on day 2. CRP has alsobeen reported to have an overall accuracy of 93% in detect-ing pancreatic necrosis. Serum CRP therefore is consideredto be the gold standard for predicting severity of AP (51–55).

CRP is widely available, easy to measure, and cheap toperform. The major drawback of CRP is that it takes 48–72h to peak, a delay similar to other methods used for severityassessment in AP.

Polymorphonuclear ElastaseIn severe AP, neutrophils accumulate in the pancreas, pro-ducing lysosomal proteases—mainly elastase—and a majorfactor for pancreatic necrosis. It also causes activation ofcomplement, kinins, and fibrinolytic systems, leading tomultiple organ system failure. Obviously, estimation ofPMN elastase is not a diagnostic test for AP but mayindicate severity. Uhl et al. (56) have shown that elastaselevels differentiate edematous from necrotic pancreatitis.The PMN elastase level is comparable to CRP in predictingnecrosis. Its advantage over CRP is that its peak levels arereached on day 1 of onset and the levels fall rapidly inpatients with edematous pancreatitis, compared to CRP val-ues, which remain elevated.

Pancreatitis-Associated Protein (PAP)PAP, an acute phase protein, is secreted from pancreaticacinar cells (57, 58), especially in AP. It induces extensivebacterial aggregation and thus may play a role in the pre-vention of bacterial infection in AP. PAP levels have beencorrelated with severity of pancreatitis in rats (59) andhumans (60), suggesting a prognostic role.

Phospholipase A2 and RibonucleasePhospholipase A2 and ribonuclease are elevated in AP butnot in healthy individuals. Phospholipase A2 is produced inthe pancreas and also by neutrophil activation. Several stud-ies have also shown it to be a good early marker of severepancreatitis (61–64). In a recent study by Mayer et al. (65),levels of secretory synovial-type phospholipase were signif-icantly higher in patients with infected necrosis than thosewith sterile necrosis, and levels of �300 ng/ml on 2 suc-cessive days within the first 4 days predicted infected ne-crosis with a high sensitivity and specificity. Warshaw andLee (66) have reported a relationship between serum levels

of pancreatic ribonuclease and the need for operative treat-ment of pancreatic necrosis or abscesses. They found that,among 24 patients with normal ribonuclease levels, only onerequired surgical treatment for abscesses. In contrast, 11 of13 patients with elevated ribonuclease required surgicalintervention. These relationships need further evaluation ina large group of patients. The assays are cumbersome andnot currently available for clinical use.

InterleukinsThe activation of inflammatory cells that release cytokinesplays an important role in the pathogenesis of the disease.Various studies have demonstrated that IL-6 and IL-8 peakwithin the first 24 h after onset of symptoms and are sig-nificantly higher in patients with severe AP (67–72). Astudy of serum markers in ERCP-induced AP showed thatthe earliest peak was of serum IL-8, 12 h after the procedure,followed by IL-6’s peak at 24 h and CRP’s peak at 72 h.Chen and associates (52) have observed that, when com-pared to serum TNF-�, IL-1B, IL-8, and CRP, IL-6 is thebest early predictor (day 1 after admission) of severe AP.Using a cutoff value of �400 pg/ml, the sensitivity, speci-ficity, and accuracy were 89%, 87%, and 88%, respectively,on day 1. Furthermore, patients with fatal outcomes showedthe most markedly elevated IL-6 concentrations (2–5 timesthe mean values in severe pancreatitis) on days 1 and 2, andthey remained elevated on day 7. The disadvantage is thatthe routine determination of IL-6 is not yet widely available.A rapid dipstick method for estimation of IL-8 is also underinvestigation (72).

In contrast, IL-10 reduces the inflammatory response inexperimental pancreatitis. In humans, Pezzelli and associ-ates (68) have also observed higher levels of IL-10 in thesera of patients with mild disease.

Tumor Necrosis FactorThe prognostic significance of serum TNF in AP has notbeen established, as its release is variable and phasic. de-Beaux et al. (73) have shown that the concentration ofsoluble TNF receptors is able to differentiate mild, severe,and fatal attacks of AP. Banks et al. (74) observed nosignificant difference between mild and severe AP, whereasChen and associates (52) found significantly elevated levelsin severe AP on days 1–3 but not on days 4 and 7. The roleof TNF assay as a prognostic marker remains unclear.

Other Serum MarkersPreliminary studies indicate serum procalcitonin to be avaluable marker for the prediction of infected pancreaticnecrosis as well as septic multiorgan failure (75, 76). Sev-eral other markers have been evaluated recently to assess theseverity of AP. These include plasma soluble intercellularadhesion molecule 1 (76, 77), serum levels of extracellularmatrix (78), serum levels of the activation peptide of car-boxypeptidase B (79), serum amyloid A (80), and serumtrypsinogen-2 and trypsinogen-2-�-1-antitrypsin complex(81). The use of these tests is currently restricted to research


settings, and further studies defining their clinical impor-tance in assessment of severity of AP are awaited.

Clinical Practice Recommendations for Serum Tests ofSeverityAlthough a host of newer serum markers hold promise forthe future, they are still experimental, used in researchsettings, and not widely available. Serum CRP is the bestavailable serum marker presently to assess the severity ofAP. A cutoff level of �150 mg/L is now accepted as aproven predictor of severity.

CRITERIA UTILIZING MULTIPLE LABORATORY TESTS

Blood urea nitrogen with glucose levels have been shown tobe helpful in identifying patients with severe AP (82). Themost commonly used criteria for predicting severity areRanson’s criteria (83), which include 11 signs with prog-nostic significance. The mortality is related to the number ofthese signs present: 0.9% with less than three positive prog-nostic signs, 16% with three or four, 40% with five or six,and 100% with more than six signs (84). A modification ofRanson’s signs as suggested by Osborne et al. (85) is usedin the United Kingdom. They have excluded Hct, basedeficit, and fluid sequestration, but included serum albuminof �3.2 g/dl as an important criterion of severity.

Acute Physiology and Chronic Health Evaluation(APACHE-II) uses the worst values of 12 physiologicalmeasurements, age, and previous health status to provide ageneral measure of severity of disease (86). The physiolog-ical variables considered are temperature, mean arterialpressure, heart rate, respiratory rate, arterial oxygenation,arterial pH, serum sodium, serum potassium, serum creati-nine, Hct, white blood cell count, and Glasgow Coma Scale.An APACHE-II score of �8 indicates severe AP. Theadvantages of APACHE-II over other prognostic criteria areobjective determination of AP within hours of admissionand the ability to be recalculated daily so as to follow thecourse of the disease and response to therapy.

LABORATORY TESTS TO PREDICT THE ETIOLOGY OF AP

The height of amylase or lipase does not predict the etiologyof AP. A reported benefit of estimating serum lipase is thatit might suggest the etiology of AP. Gumaste et al. (87)found that a lipase/amylase ratio of �2 was suggestive ofacute alcoholic pancreatitis. Several other studies have triedto determine the utility value of the lipase/amylase ratio(88–91). Although all studies did not agree that a lipase/amylase ratio is a precise indicator, most reveal a definitetrend—values of �3 are highly indicative of acute alcoholicpancreatitis, whereas low values (�1–2) are more sugges-tive of gallstone pancreatitis (90). Controversy still existswith regard to the utility of the ratio in clinical practice. Arecent practice guideline article (25) did not recommend it.Table 2 summarizes the laboratory tests in AP.

Several studies have looked at the utility of liver functiontests in predicting gallstone pancreatitis (92–94). Accordingto a recent meta-analysis (95), an ALT level of �150 U/L(approximately a 3-fold elevation) has a positive predictivevalue of 95% in diagnosing acute gallstone pancreatitis.However, the sensitivity is only 50%. Bilirubin and ALPwere not found to be as useful as ALT (95).

URINE TESTS

DiagnosisAlthough routine use of urinary amylase is not done widelyto evaluate a patient with AP, recently several studies havereported the use of urinary dipstick tests for screening casesof AP in the emergency room (96, 97). A dipstick test fordetecting pancreatic amylase in urine by an immunochro-matography principle using the monoclonal antibodies spe-cific for pancreatic amylase in initial studies has shownpromising results (97). The test has a high specificity of 97%and is likely to become useful in the emergency roomsetting. Other tests have reported the clinical utility of theurine trypsinogen-2 dipstick test in AP for screening pa-tients with suspected AP as well as predicting the severity(98, 99). In a study of 525 consecutive patients presenting tothe emergency room with abdominal pain, the sensitivityand specificity of diagnosing AP were 96% and 92%, re-spectively. All nine patients with severe AP were detectedby the dipstick (98).

SeverityTrypsinogen activation peptides (TAPs) are the highly con-served tetra-1-aspartyl-1-lysyl amino terminal peptides re-

Table 2. Laboratory Tests in AP

Tests for diagnosisSerum tests

1. Amylase2. Isoamylases3. Macroamylases4. Lipase5. Immunoreactive trypsin6. Elastase

Urine tests1. Amylase2. Lipase

Fluid amylaseTests for severity

Serum tests1. CRP2. Interleukins (IL-6, IL-8, etc)3. Polymorphonuclear elastase4. PAP5. Phospholipase6. Procalcitonin7. TNF-�8. Ribonuclease9. Methalbumin

Urine testsTAPs


leased during the activation of trypsinogen to trypsin.Though normally produced in the intestinal lumen afteractivation of trypsinogen by enterokinase, they are not ab-sorbed, because of degradation by enteric oligopeptidases(100–103). In AP, premature intrapancreatic activation oftrypsinogen leads to the release of TAPs into blood, lym-phatics, and peripancreatic tissue, leading to increased con-centrations in body fluids. TAPs are liberated within the firstfew hours of the onset of AP, and they peak within 12–24h of admission.

Experimental studies done in animals and humans havefound estimation of TAPs to be helpful as a prognosticmarker in differentiating severe from mild AP. Gudgeon etal. in their study (103) found urinary TAP estimation to bemuch better than serum CRP levels. When a urinary con-centration of �2 nmol/L was used as a cutoff on admission,the sensitivity and specificity for predicting severe AP were80% and 90%. When the highest concentration of TAPs inthe first 24 h was used, the sensitivity and specificity ofurinary TAP levels of �10 ng/ml were 100% and 85% inpredicting severe AP (104).

Neoptolemus and associates (105) in a prospective studycompared levels of urinary TAPs with serum CRP and thethree currently used scoring systems in 172 patients with AP(35 with severe disease) and 74 controls. Urinary TAPconcentrations differed significantly between mild and se-vere AP at 24 h and 48 h after onset of symptoms and alsoat 24 h and 48 h after admission. CRP concentrations dif-fered significantly at 48 h but not at 24 h. TAPs at 24 h (�34mmol/L) were similar to an APACHE-II score of �8 (0–12h after admission) with respect to sensitivity (58% vs 58%),specificity (73% vs 76%), positive predictive value (39% vs40%), negative predictive value (86% vs 87%), and accu-racy (70% vs 72%) (Table 3).

CLINICAL PRACTICE RECOMMENDATIONS. The as-say of TAPs for determining severity of AP is appealing asa single marker that is able to provide accurate severityprediction within 24 h after onset of symptoms. Theseencouraging results should be considered preliminary, andfurther studies are needed to establish its role in the evalu-ation of AP. Presently, TAP assays are not widely available.

ASCITIC FLUID ANALYSIS

Early appearance of ascites is seen in over 60% of cases ofsevere AP (106). This peritoneal exudate is rich in activatedlipolytic and proteolytic enzymes, vasoactive substances,and several other proinflammatory mediators. Ascites mayplay a role in the transfer of toxic mediators into the sys-temic circulation or may be a reflection of the locoregionalnecrotizing process (106). A peritoneal tap can providecorroborative evidence of AP by the presence of a highamylase concentration, especially if sterile fluid is aspirated(106). The biochemical composition of peritoneal fluid dur-ing AP, however, reveals no major differences between theprincipal etiological categories of AP (107, 108).

McMahon et al. (109) have shown that volume and colorof peritoneal fluid are indicators of the severity of an attackof AP. Severe AP is indicated in the presence of one or moreof the following: greater than 20 ml of free peritoneal fluid,dark-colored free intraperitoneal fluid, and lavage fluiddarker than a pale straw color obtained after peritoneallavage with a liter of normal saline. However, peritoneallavage is an invasive procedure that is not applicable forpatients with mild disease and contraindicated in patientswith previous scars, obesity, coagulopathy, or difficulty incatheterization. It is associated with visceral puncture in0.8% of patients. It is also a poor system for accurategrading of patients with biliary pancreatitis (108).

CLINICAL PRACTICE RECOMMENDATIONS. Diag-nostic peritoneal lavage is an invasive test and a poor systemfor grading the severity of AP. It has never been acceptedwidely into clinical practice.

RESULTS AND DISCUSSION

It is clear that there is no biochemical test that can beconsidered to be a gold standard for the diagnosis or assess-ment of severity of AP. Amylase and lipase remain impor-tant tests in the diagnosis of AP. Lipase, which was initiallythought to be more specific than amylase, has recently beenshown to be almost as nonspecific as amylase. The use ofother tests like P-isoamylase, IRT, elastase, urinary amylase

Table 3. Comparison of Laboratory Tests for AP

Amylase Lipase Serum CRP Urinary TAPs Interleukins (IL-6, IL-8)

Sensitivity 67–100%* 82–100% Test for severity Test for severity Test for severitySpecificity 85–98% 82–100%† Test for severity Test for severity Test for severityPrediction of severity None None Yes

(�150 mg/L at48 h)

Yes(peak within 12–24 h)

Yes(peak within 12–24 h)

Comments Cheap, widelyavailable

Cheap, widelyavailable

Cheap, widelyavailable, bestavailable labtest for severity

Not widely available,expensive

Experimental

* Poor sensitivity in hyperlipidemic AP, acute or chronic AP due to alcohol, and delayed estimation.† Better than amylase. Lipase is increasingly being recognized as nonspecific.


and lipase, and fluid amylase has no clear role in the eval-uation of patients with AP. Urinary TAPs within 12–24 hand serum CRP at 48 h are now considered by many to begood markers for predicting severity of AP. A host of newserological markers have been investigated in the last fewyears to predict the severity of AP early. Some of them showpromise but have yet to prove their superiority.

Reprint requests and correspondence: C. S. Pitchumoni, M.D.,M.A.C.G., M.P.H., Professor of Medicine and Preventive & Com-munity Medicine, New York Medical College, Director, Depart-ment of Medicine & Chief of GI, Our Lady of Mercy MedicalCenter, 600 East 233rd Street, Bronx, NY 10466.

Received Feb. 6, 2001; accepted Jan. 15, 2002.

REFERENCES

1. Agarwal N, Pitchumoni CS, Sivaprasad AV. Evaluating testsfor acute pancreatitis. Am J Gastroenterol 1990;85:356–66.

2. Salt WB, Schenker S. Amylase—its clinical significance: Areview of the literature. Medicine 1976;55:269–89.

3. Steinberg WM, Goldstein SS, Davis ND, et al. Diagnosticassays in acute pancreatitis. Ann Intern Med 1985;102:576–80.

4. Jacobs ML, Daggett WM, Civetta JM, et al. Acute pancre-atitis. Analysis of factors influencing survival. Ann Surg1977;185:43–51.

5. Keim V, Teich N, Fiedler F, et al. A comparison of lipase andamylase in the diagnosis of acute pancreatitis in patients withabdominal pain. Pancreas 1998;16:45–9.

6. Clavien PA, Robert J, Meyer P, et al. Acute pancreatitis andnormoamylasemia. Not an uncommon combination. AnnSurg 1989;210:614–20.

7. Spechler SJ, Dalton JW, Robbins AH, et al. Prevalence ofnormal serum amylase levels in patients with acute alcoholicpancreatitis. Dig Dis Sci 1983;28:865–9.

8. Ventrucci M, Pezzilli A, Naldoni P, et al. Serum pancreaticenzyme behavior during the course of acute pancreatitis.Pancreas 1987;2:506–9.

9. Toskes PP. Hyperlipidemic pancreatitis. Gastroenterol ClinNorth Am 1990;19:783–91.

10. Warshaw AL, Bellini CA, Lesser PB. Inhibition of serum andurine amylase activity in pancreatitis with hyperlipemia. AnnSurg 1975;182:72.

11. Mishkin S, Bates J, O’Hashi J, et al. Possible mechanisms ofnormal amylase activity in hyperlipemic pancreatitis. CMAJ1976;115:1016–9.

12. Webber J, Fromm D. Assessment, diagnosis, and initial treat-ment. In: Howard J, Yasuo I, Ihse I, Prinz R, eds. Surgicaldiseases of the pancreas, 3rd ed. Baltimore: Williams &Wilkins 1998:207–27.

13. Humphries CC, Adams LJ, Eckfeldt JH, et al. Hyperamy-lasemia in patients with eating disorders. Ann Intern Med1987;106:50–2.

14. Lott JA. The value of clinical laboratory studies in acutepancreatitis. Arch Pathol Lab Med 1991;115:325–7.

15. Berk JE, Fridhandler L, Webb SF. Does hyperamylasemia inthe drunken alcoholic signify pancreatitis? Am J Gastroen-terol 1979;71:557–62.

16. Dutta SK, Douglass W, Smalls UA, et al. Prevalence andnature of hyperamylasemia in acute alcoholism. Dig Dis Sci1981;26:136–41.

17. Clavien PA, Burgan S, Moossa AR. Serum enzymes and the

other laboratory tests in acute pancreatitis. Br J Surg 1989;76:1234–43.

18. Warshaw AL, Hawboldt MM. Puzzling persistent hypramy-lasemia, probably neither pancreatic nor pathologic. Am JSurg 1988;155:453–6.

19. Gullo L. Familial pancreatic hyperenzymemia. Pancreas2000;20:158–60.

20. Berk JE, Fridhandler J. Hyperamylasemia. Interpretation andnewer approaches to evaluation. Adv Intern Med 1980;26:253–64.

21. Thomson HJ, Obekpa PO, Smith AN, et al. Diagnosis ofacute pancreatitis: A proposed sequence of biochemical in-vestigation. Scand J Gastroenterol 1987;22:719–24.

22. Winslet M, Hall C, London NJ, et al. Relation of diagnosticserum amylase to aetiology and severity of acute pancreatitis.Gut 1992;33:982–6.

23. Lankisch PG, Burchard-Recker IS, Lehrich D. Underestima-tion of acute pancreatitis: Patients with only a small increasein amylase/lipase levels can also have or develop severeacute pancreatitis. Gut 1999;44:542–4.

24. Williamson RCN. Early assesssment of severity in acutepancreatitis. Gut 1984;25:1331–9.

25. Banks PA. Practice guidelines in acute pancreatitis. Am JGastroenterol 1997;92:377–86.

26. Hiatt JR, Calabria RP, Passaro E, et al. The amylase profile:A discriminant in biliary and pancreatic disease. Am J Surg1987;154:490–2.

27. Tietz NW, Shuey DF. Lipase in serum—the elusive enzyme:An overview. Clin Chem 1993;39:746–56.

28. Panteghini M. Lipase. Clin Chem News 1991;17(2):6–7 (re-view).

29. Frank B, Gottlieb K. Amylase normal, lipase elevated: Is itpancreatitis? A case series and review of the literature. Am JGastroenterol 1999;94:463–9.

30. Apple F, Benson P, Preese L, et al. Lipase and pancreaticamylase activities in tissues and in patients with hyperamy-lasemia. Am J Clin Pathol 1991;96:610–4.

31. Gumaste VV, Roditis N, Mehta D, et al. Serum lipase levelsin non-pancreatic abdominal pain versus acute pancreatitis.Am J Gastroenterol 1993;88:2051–5.

32. Gottelieb K, Sherman S, Lehman GA, et al. Early recognitionof post-ERCP pancreatitis by clinical assessment and serumpancreatic enzymes. Am J Gastroenterol 1996;91:1553–7.

33. Lott JA, Speicher CE, Nemesanszky E. Is serum amylase anobsolete test in the diagnosis of acute pancreatitis? ArchPathol Lab Med 1985;109:314–5 (editiorial).

34. Yadav D, Nair S, Norkus EP, et al. Non-specific hyperamy-lasemia and hyperlipasemia in diabetic ketoacidosis: Inci-dence and correlation with biochemical abnormalities. Am JGastroenterol 2000;95:3123–8.

35. Heikius B, Niemela S, Lehtola J, et al. Elevated pancreaticenzymes in inflammatory bowel disease are associated withextensive disease. Am J Gastroenterol 1999;94:1062–9.

36. Tietz NW. Support of the diagnosis of pancreatitis by enzymetest—old problems, new techniques. Clin Chim Acta 1997;257:85–98.

37. Tetraut GA. Lipase activity in serum measued with Ektachemis often increased in non-pancreatic disorders. Clin Chem1991;37:47–51.

38. Chase WC, Barker DE, Russel WL, et al. Serum amylase andlipase in the evaluation of acute abdominal pain. Am Surg1996;62:1028–34.

39. Moossa AR. Diagnostic tests and procedures in acute pan-creatitis. N Engl J Med 1984;311:639–43.

40. Kolars JC, Ellis CJ, Levitt MD. Comparison of serum amy-lase, pancreatic isoamylase and lipase in patients with hy-peramylasemia. Dig Dis Sci 1984;29:289–93.


41. Ventrucci M, Pezzilli A, Naldoni P, et al. A rapid assay forserum immunoreactive lipase as a screening test for acutepancreatitis. Pancreas 1986;4:320–3.

42. Dervenis C, Johnson CD, Bassi C, et al. Diagnosis, objectiveassessment of severity, and management of acute pancreati-tis. Santorini Consensus Conference. Int J Pancreatol 1999;25:195–210.

43. Werner H, Steinberg WM, Pauley C. Strategic use of indi-vidual and combined enzyme indicators for acute pancteatitisanalyzed by reciever operator characteristics. Clin Chem1989;35:967–71.

44. Piper-Bigelow C, Strocchi A, Levitt MD. Where does serumamylase come from and where does it go? Gastroenterol ClinNorth Am 1990;19:793–810.

45. Berk JE, Kizu H, Wilding P, et al. Macroamylasemia: A newrecognized cause for elevated serum amylase activity. N EnglJ Med 1967;277:941–6.

46. Imrie CW, King J, Henderson AR. Macroamylasemia—sur-vey of prevalence in a mixed population. N Engl J Med1972;287:931.

47. Boyle CE, Fraser CG. Macroamylasemia: How common isit? Br J Med 1985;291:1389.

48. Malvano A, Marchisio M, Massaglia A, et al. Radioimmu-noassay of trypsin-like substance in human serum. Scand JGastroenterol Suppl 1980;62:3–10.

49. Baillargeon JD, Ramagopal V, Tenner SM, et al. Hemocon-centration as an early risk factor for necrotizing pancreatitis.Am J Gastroenterol 1998;93:2130–4.

50. Brown A, Orav J, Banks PA. Hemoconcentration is an earlymarker for organ failure and necrotizing pancreatitis. Pan-creas 2000;20:367–72.

51. Wilson C, Heads A, Shenkin A, et al. C-reactive protein,antiproteases, and complement factors as objective markersof severity in acute pancreatitis. Br J Surg 1989;76:177–81.

52. Chen CC, Wang SS, Lee FY, et al. Proinflammatory cyto-kines in early assessment of the prognosis of acute pancre-atitis. Am J Gastroenterol 1999;94:213–8.

53. Mayer AD, McMahon MJ, Bowen M, et al. C-reactiveprotein: An aid to assessment and monitoring of acute pan-creatitis. J Clin Pathol 1984;37:207–11.

54. Puolakkainen P, Valtonen V, Paananen A, et al. C-reactiveprotein (CRP) and serum phospholipase A2 in the assessmentof the severity of acute pancreatitis. Gut 1987;28:764–71.

55. Uhl W, Buchler M, Malfertheiner P, et al. PMN-elastase incomparison with CRP, anti-proteases, and LDH as indicatorsof necrosis in human acute pancreatitis. Pancreas 1991;6:253–9.

56. Uhl W, Buchler M, Malfertheiner P, et al. PMN-elastase: Anew serum marker for the staging of acute pancreatitis.Digestion 1989;43:176–7.

57. Iovanna J, Orelle B, Keim V, et al. Messenger RNA sequenceand expression of rat pancreatitis associated protein is over-expressed during acute experimental pancreatitis. J BiolChem 1991;226:24664–9.

58. Keim V, Iovanna JL, Dagorn JC. The acute phase reaction ofthe exocrine pancreas: Gene expression and synthesis ofpancreatitis associated protein. Digestion 1994;35:65–72.

59. Keim V, Willemer S, Iovanna JL, et al. Rat pancreatitis-associated protein is expressed in relation to severity ofexperimental pancreatitis. Pancreas 1994;9:606–12.

60. Iovanna JL, Kein V, Nordback I, et al. Serum levels ofpancreatitis-associated protein as indicators of the course ofacute pancreatitis. Gastroenterology 1994;106:728–34.

61. Schroder T, Kivilaakso E, Kinnunen PK, et al. Serum phos-pholipase-A2 in human acute pancreatitis. Scand J Gastro-enterol 1980;15:633–6.

62. Nevalainen TJ. The role of phospolipase A in acute pancre-atitis. Scand J Gastroenterol 1980;15:641–50.

63. Buchler M, Malfertheiner P, Schadlich M, et al. Role ofphospholipase-A2 in human acute pancreatitis. Gastroenter-ology 1989;97:1521–6.

64. Bird NC, Goodman AJ, Johnson AG. Serum phospholipaseA2 activity in acute pancreatitis: An early guide to severity.Br J Surg 1989;76:731–2.

65. Mayer J, Rau B, Grewe M, et al. Secretory phospholipase A2in patients with infected pancreatic necrosis in acute pancre-atitis. Pancreas 1998;17:272–7.

66. Warshaw AL, Lee KH. Serum ribonuclease elevation andpancreatic necrosis in acute pancreatitis. Surgery 1979;86:227–34.

67. Gross V, Andreesen R, Leser HG, et al. Interleukin-8 andneutrophil activation in acute pancreatitis. Eur J Clin Invest1992;22:200–3.

68. Pezzelli R, Belli P, Miniero R, et al. Serum interleukin-6,interleukin-8 and alpha-2 microglobulin in early assessmentof severity of acute pancreatitis. Comparison with C-reactiveprotein. Dig Dis Sci 1995;40:2341–8.

69. Leser HG, Gross H, Scheibenbogen C, et al. Elevation ofinterleukin-6 concentration precedes acute phase responseand reflects severity in acute pancreatitis. Gastroenterology1991;101:782–5.

70. Viedma JA, Perez-Mateo M, Dominquez JE, et al. Role ofinterleukin-6 in acute pancreatitis: Comparison with C-reac-tive protein and phospholipase A. Gut 1992;33:1264–7.

71. Heath DI, Cruickshank A, Gudgeon M, et al. Role of inter-leukin-6 in mediating the acute phase protein response andpotential as an early means of severity assessment in acutepancreatitis. Gut 1993;34:41–5.

72. Rau B, Steinbach G, Mayer J, et al. The role of interleukin-8in the severity assessment of septic complication in necro-tizing pancreatitis. Digestion 1997;58(suppl 2):11.

73. deBeaux AC, Goldie AS, Ross JA, et al. Serum concentrationof inflammatory mediators related to organ failure with acutepancreatitis. Br J Surg 1996;83:349–53.

74. Banks RE, Evans SW, Alexander D, et al. Is fatal pancreatitisa consequence of excessive leukocyte stimulation? The roleof tumor necrosis factor alpha. Cytokine 1991;3:6–12.

75. Rau B, Steinbach G, Gansauge F, et al. The potential role ofprocalcitonin and interleukin-8 in the prediction of infectednecrosis in acute pancreatitis. Gut 1997;41:832–40.

76. Mandi Y, Farkas G, Takacs T, et al. Diagnostic relevance ofprocalcitonin, IL-6, and sICAM-1 in the prediction of in-fected necrosis in acute pancreatitis. Int J Pancreatol 2000;28:43–51.

77. Kaufmann P, Demel U, Tilz GP, et al. Time course of plasmaintercellular adhesion molecule-1 (sICAM-1) is related toseverity of acute pancreatitis. Hepatogastroenterology 1999;46:2565–71.

78. Lohr M, Hummel F, Martus P, et al. Serum levels of extra-cellular matrix in acute pancreatitis. Hepatogastroenterology1999;46:3263–70.

79. Appelros S, Petersson U, Johnson C, et al. Activation peptideof carboxypeptidase C and anionic trypsinogen as early pre-dictors of the severity of acute pancreatitis. Br J Surg 2001;88:216–21.

80. Rau B, Steinbach G, Baumgart K, et al. Serum amyloid Aversus C-reactive protein in acute pancreatitis: Clinical valueof an alternative acute-phase reactant. Crit Care Med 2000;28:736–42.

81. Hedstrom J, Kemppainen E, Andersen J, et al. A comparisonof serum trypsinogen and trypsin-2-alpha-1-antitrypsin com-plex with lipase and amylase in the diagnosis and assessment


of severity in the early phase of acute pancreatitis. Am JGastroenterol 2001;96:424–30.

82. Fan ST, Choi TK, Lai ECS, et al. Prediction of severity ofacute pancreatitis: An alternative approach. Gut 1989;30:1591–5.

83. Ranson JH, Pasternack BS. Statistical methods for quantify-ing the severity of clinical acute pancreatitis. J Surg Res1977;22:79–91.

84. Ranson JH. Acute pancreatitis. Curr Probl Surg 1979;16:1–84.

85. Osborne DH, Imrie CW, Carter DC. Biliary surgery in thesame admission for gallstone acute pancreatitis. Br J Surg1981;68:758–61.

86. Knaus WA, Draper EA, Wagner DP, et al. APACHE II: Aseverity of disease classification system. Crit Care Med 1985;13:818–29.

87. Gumaste VV, Dave PB, Weissman D, et al. Lipase/amylaseratio. A new index that distinguishes acute episodes of alco-holic from non-alcoholic acute pancreatitis. Gastroenterology1991;101:1361–6.

88. Sadowski DC, Todd JK, Sutherland LR. Biochemical modelsas early predictors of the etiology of acute pancreatitis. DigDis Sci 1993;38:637–43.

89. Moster SG, Herbsmann D, Kniaz JL, et al. Use of lipase:amylase (LA) ratio in distinguishing alcoholic versus gall-stone causes of acute pancreatitis. Am J Gastroenterol 1993;88:1536 (abstract).

90. Tenner SM, Steinberg W. The admission serum lipase:amylase ratio differentiates alcoholic from non-alcoholicacute pancreatitis. Am J Gastroenterol 1992;87:1755–8.

91. Lankisch PG, Petersen M. Lipase/amylase ratio: Not helpfulin the early etiological differentiation of acute pancreatitis. ZGastroenterol 1994;32:8–11.

92. Wang SS, Lin XZ, Tsai YT, et al. Clinical significance ofultrasonography, computed tomography and biochemicaltests in the rapid diagnosis of gallstone-related pancreatitis. Aprospective study. Pancreas 1988;3:153–8.

93. Scholmerich J, Gross V, Johannesson T, et al. Detection ofbiliary origin of acute pancreatitis: Comparison of laboratorytests, ultrasound, computed tomography, and ERCP. Dig DisSci 1989;34:830–3.

94. Ros E, Navarro S, Bru C, et al. Occult microlithiasis inidiopathic acute pancreatitis: Prevention of relapses by cho-lecystectomy or ursodeoxycholic acid. Gastroenterology1991;101:1701–9.

95. Tenner S, Dubner H, Steinberg W. Predicting gallstone pan-creatitis with laboratory parameters: A meta-analysis. Am JGastroenterol 1994;89:1863–6.

96. Burkitt DS. The rapignost-amylase test in acute pancreatitis.Br J Surg 1987;74:1063.

97. Hedstrom J, Svens E, Kenkimaki P. Evaluation of new uri-nary amylase test strip in the diagnosis of acute pancreatitis.Scand J Clin Lab Invest 1998;58:611–6.

98. Kylanpaa-Back ML, Kemppainen E, Puolakkainen P, et al.Reliable screening for acute pancreatitis with rapid urinetrypsinogen-2 test strip. Br J Surg 2000;87:49–52.

99. Pezzilli R, Morselli-Labate AM, d’Alessandro A, et al. Time-course and clinical value of the urine trypsinogen-2 dipsticktest in acute pancreatitis. Eur J Gastroenterol Hepatol 2001;13:269–74.

100. Guy O, Lombardo D, Bartelt DC, et al. Two human trypsi-nogens. Purification, molecular properties, and n-terminalsequence. Biochemistry 1978;17:1669–75.

101. Hurley PR, Cook A, Jehanli A, et al. Development of radio-immunoassays for free tetra-L-aspartyl-L-lysine trypsinogenactivation peptides (TAP). J Immunol Methods 1988;111:195–203.

102. Schmidt J, Fernandez-del Castillo C, Rattner DW, et al.Trypsinogen activation peptides in experimental ratpancreatitis: Prognostic implications and histopathologic cor-relates. Gastroenterology 1992;103:1009–16.

103. Gudgeon AM, Heath DI, Hurley P, et al. Trypsinogen acti-vation peptides assay in the early prediction of severity ofacute pancreatitis. Lancet 1990;335:4–8.

104. Tenner S, Fernandez-del Castillo C, Warshaw A, et al. Uri-nary trypsinogen activation peptide (TAP) predicts severityin patients with acute pancreatitis. Int J Pancreatol 1997;21:105–10.

105. Neoptolemus JP, Kemppainen EA, Mayer JM, et al. Earlyprediction of severity in acute pancreatitis by urinarytrypsinogen activation peptides: A multicenter study. Lancet2000;355:1955–60.

106. Dugernier T, Laterre PF, Reynaert MS. Astices fluid insevere acute pancreatitis: From pathophysiology to therapy.Acta Gastroenterol Belg 2000;63:264–8.

107. Dubick MA, Mayer D, Majumdar APN, et al. Biochemicalstudies in peritoneal fluid from patients with acutepancreatitis: Relationship to etiology. Dig Dis Sci 1987;32:305–12.

108. Corefield AP, Cooper MJ, Williamson RCN, et al. Predictionof severity in acute pancreatitis: Prospective comparison ofthree prognostic indices. Lancet 1985;2:403–7.

109. McMahon MJ, Playforth MJ, Pickford IR. A comparativestudy of methods for the prediction of severity of attacks ofacute pancreatitis. Br J Surg 1980;67:22–5.


laboratory amilasa

Documents