critical review of diagnostic methods used in chronic ...factor for pancreatic cancer (2,3). once...

Critical review of diagnosticmethods used in chronic

pancreatic diseaseIVAN T BECK MD PHD FRCPC FACP FACG

THIS REVIEW DEALS WITH THEdiagnostic modalities used to in-vestigate chronic pancreatic diseaseand the methods employed to differen-tiate chronic pancreatitis from carci-noma of the pancreas. Otherconditions that lead to pancreatic in-sufficiency without pancreatitis or can-cer (such as cystic fibrosis, primarypancreatic atrophy of childhood, adultpancreatic lipomatosis or isolatedlipase-colipase deficiency, etc) will notbe discussed although these conditionsmay also lead to abnormal pancreaticfunction tests. The subject of the pres-ent paper was reviewed in more detailby the author two years ago (1). Thepresent paper is a more general over-view and a critical assessment of themethods used and an update on recentdevelopments.

The clinical features of chronic pan-creatitis and carcinoma are shown inFigures 1 and 2. The common denomi-nator of both diseases is replacement of

functioning parenchyma and the duct

system with nonfunctioning tissue: in-

flammatory tissue and fibrosis in

chronic pancreatitis, and tumour in

cancer of the pancreas. The clinical

symptoms, biochemical abnormalities

and anatomical distortions in both

conditions are due to ductal obstruc-

tion, diminished parenchymal or

ductal-cell function and the presence

of mass lesions. Except in the case of

functioning islet cell tumours (which

are not the subject of this review), dif-

ferentiation between the two condi-

tions may be very difficult (1-3). As

shown in Figures 3 and 4, cystic and

REVIEW

Queen’s University, Kingston, OntarioCorrespondence and reprints: Dr IT Beck, Division of Gastroenterology, Department of

Medicine, Hotel Dieu Hospital, Queen’s University, Kingston, Ontario K7L 5G2Received for publication April 5, 1994. Accepted June 13, 1994

IT BECK. Critical review of diagnostic methods used in chronic pancreatic disease. CanJ Gastroenterol 1995;9(1):51-60. This paper provides a balanced assessment of the vari-ous pancreatic function tests and imaging techniques used in the differential diagnosis ofchronic pancreatic disease. Function tests that study the digestive capacity of the pancreas(fat absorption of dietary lipids, fluorescein- or radiolabelled fats, bentiromide test, etc)have high specificity, but very low sensitivity. This is because 90% of pancreas has to bedestroyed before steatorrhea or creatorrhea occurs. Tests that directly measure pancreaticbicarbonate and protein secretion (secretin test, etc) are more accurate and may detectpancreatic dysfunction even before anatomical changes occur. Measurement of pancre-atic enzymes in serum or urine, or the decreased decline of serum amino acids during theirincorporation into pancreatic enzymes, are not sufficiently sensitive or specific to help di-agnose pancreatic disease. Sensitive and specific tumour markers are not yet available.Thus screening tests are not cost-effective – if they are negative, they do not exclude pan-creatic disease; and if positive, they have to be confirmed by more specific tests. Imagingtechniques are the most commonly used methods of investigation. The usefulness of ab-dominal survey films, barium studies, percutaneous transhepatic cholangiography, endo-scopic retrograde cholangiopancreatography (ERCP), ultrasonography, computedtomographic scan, magnetic resonance imaging and endoscopic ultrasonography is criti-cally reviewed. Most of the radiological methods can be combined with cytology or biopsy.Histology demonstrating malignancy establishes this diagnosis, but negative biopsies donot exclude malignant tumours. Presently only ERCP and endoscopic ultrasound can diag-nose cancers sufficiently early to allow for possible ‘curative’ surgery, and only endoscopicultrasound is capable to stage tumours for the assessment of resectability.

Key Words: Computed tomographic scan, Endoscopic retrograde cholangiopancreatography,Endoscopic ultrasonography, Imaging techniques, Pancreatic biopsies, Pancreatic function tests,Secretin test, Tumour markers

Analyse critique des méthodes diagnostiques utilisées dans lapancréatite chronique

RÉSUMÉ : Cet article présente une évaluation équilibrée des diverses épreuves de fonc-tion du pancréas et des techniques d’imagerie utilisées pour le diagnostic différentiel de lapancréatite chronique. Les épreuves de fonction qui portent sur la capacité digestive dupancréas (absorption des graisses d’origine alimentaire, fluorescéine ou graisses radio-marquées, épreuve au bentiromide, etc.) s’accompagnent d’un degré élevé de spécificité,mais d’une très faible sensibilité. Cela est dû au fait que 90 % du pancréas doit être détruitavant que ne survienne la stéatorrhée ou la créatorrhée. Les épreuves qui visent une me-sure directe des bicarbonates du pancréas et la sécrétion protéique (test de la sécrétine,

voir page suivante

CAN J GASTROENTEROL VOL 9 NO 1 JANUARY/FEBRUARY 1995 51

solid tumours can occur in chronic

pancreatitis as well as in benign and

malignant neoplasms (4). Further-

more, calcification can occur in

chronic pancreatitis and in several of

the benign or malignant tumours. In

addition, chronic pancreatitis is a risk

factor for pancreatic cancer (2,3).

Once the suspicion of pancreaticdisease has been raised on clinicalgrounds, investigation is warranted todetermine whether chronic pancreaticdisease is present, and if so, whetherthis is chronic pancreatitis or carci-noma. The methods employed can bedivided into two major categories. Thefirst consists of laboratory tests used to

demonstrate abnormal pancreaticfunction (pancreatic function tests)and immunoassays for tumour markersto differentiate between benign andmalignant disease. The second cate-gory consists of imaging and biopsytechniques used to establish the ana-tomical basis for abnormal function.Recent imaging techniques have re-placed, to a great extent, the pancreaticfunction tests, but evaluation and un-derstanding of the usefulness of indi-vidual tests is important in theinvestigation of patients with malab-sorption syndrome and in the differen-tiation of pancreatic maldigestion frommalabsorption due to other causes.

LABORATORY TESTS(TABLE 1)

Pancreatic function tests

Tests that assess digestive capacity ofthe pancreas (Table 2): Tests based onmeasuring the digestive capacity of thepancreas are less sensitive than thosethat directly measure pancreatic secre-tion. This is because the normal pan-creas has considerable reserve capacity,and steatorrhea and creatorrhea be-come biochemically detectable onlywhen lipase and trypsin secretion hasbeen reduced to less than 10% of nor-mal (5), and steatorrhea may not be de-tectable until 75% of the pancreas hasbeen resected. Thus, fecal fat studies areinsensitive in mild chronic pancreaticdisease.Fecal fat balance study: Steatorrheaoccurs due to maldigestion (pancreaticdisease) and to malabsorption (as a re-sult of many other conditions, includ-ing intestinal disease). Differentiationbetween the two can be achieved by themethod described below.Fecal fat balance study before andduring pancreatic replacement ther-apy: Steatorrhea due to pancreaticdisease is diagnosed if fat absorption im-proves with oral ingestion of pancreaticenzymes. Gastric acid may lower duo-denal pH below the level that is opti-mal for lipase activity, and thesimultaneous administration ofenzymes and a H2 receptor antagonistmay further improve fat absorption(6).

Colipase, a factor needed for appro-priate lipase activity, plays an impor-tant role in fat digestion, and steator-rhea may occur in congenital isolatedlipase-colipase deficiency (7).Screening tests for pancreatic maldi-gestion: These tests, which study fat,starch or peptide digestion, have beendeveloped to simplify the cumbersomefecal fat balance study. Most of themare less sensitive and specific than the72 h fecal fat excretion.Macroscopic and microscopic examination

of stool: Considering the low sensitiv-ity of the balanced chemical fecal fatdetermination, one would expect thatthe qualitative (microscopic) exami-nation of stool is devoid of sensitivity,

Figure 1) Clinical and biochemical features caused by the anatomical changes of chronic pancreati-tis. Reproduced with permission from reference 1

etc.) sont plus précises et peuvent déceler une dysfonction pancréatique avant même quene surviennent des changements anatomiques. La mesure des enzymes pancréatiques séri-ques ou urinaires ou l’atténuation de la baisse des amino-acides sériques durant leur incor-poration aux enzymes pancréatiques ne sont pas suffisamment sensibles ni précises pourcontribuer au diagnostic de la pancréatite. Les marqueurs tumoraux sensibles et spéci-fiques se font encore attendre. Donc, les épreuves de dépistage ne sont pas économiques; sielles sont négatives, elles permettent d’exclure une maladie pancréatique; si elles sontpositives, elles doivent encore être confirmées à l’aide d’autres tests plus spécifiques. Lestechniques d’imagerie sont les méthodes d’investigation les plus fréquemment utilisées.L’utilité des plaques simples de l’abdomen, des épreuves barytées, de la cholangiographietranshépatique percutanée, de la cholangiopancréatographie endoscopique rétrograde(CPER), l’échographie, la scintigraphie, l’imagerie par résonnance magnétique et l’écho-graphie endoscopique sont passées en revue. La plupart des méthodes radiologiques peu-vent être combinées à la cytologie ou à la biopsie. L’histologie permet d’établir undiagnostic de néoplasie le cas échéant, mais les biopsies négatives ne permettent pas d’ex-clure les tumeurs malignes. À l’heure actuelle, seule la CPER et l’échographie endoscopiquepeuvent diagnostiquer les cancers suffisamment tôt pour permettre une chirurgie curativepossible et seule l’échographie endoscopique peut déterminer si la chirurgie est faisable.

52 CAN J GASTROENTEROL VOL 9 NO 1 JANUARY/FEBRUARY 1995

BECK

and thus may lead to diagnostic confu-sion.Pancreolauryl test: This pancreatic func-tion test is based on the digestion of di-dodecanoic ester of fluorescein by pan-creatic lipase. Because 90% of lipase ac-tivity has to be destroyed before eventhe accurate 72 h fecal fat balance studybecomes abnormal, it is not astonishingthat the sensitivity of the pancreolauryltest (which is based on a single dose ofadministered fluorescein fat) is low (8).The sensitivity of the test was reportedin severe pancreatic disease to be be-tween 80 and 90% but it is much lowerin mild chronic pancreatitis (1).Bentiromide test: This test is based onthe digestion of benzoyl-L tyrosil-L-p-aminobenzoic acid (NBT-PABA).Because the tyrosil-p-aminobenzoicacid bond is specific for chymotrypsin,this synthetic peptide is digested onlyby chymotrypsin. The hydrolysis of thispeptide results in the release of p-amino-benzoic acid, which is absorbed and ex-creted in urine. As this test is based onintraduodenal digestion by chymotryp-sin, which is present in the pancreas ingreat excess, the test becomes positiveonly in patients with severe pancreaticinsufficiency (8,9).Other screening tests for pancreatic

maldigestion: The 14C triolein and fatty

Figure 3) Classification of benign tumours of the pancreas

Figure 4) Classification of malignant tumours of the pancreas

Figure 2) Chemical and biochemical features caused by the anatomical and hormonal alterations of cancer of the pancreas. Reproduced with permissionfrom reference 1

Review of diagnostic methods for chronic pancreatitis

acid absorption and breath tests (thefirst depends on duodenal lipase activ-ity and the second on intestinal absorp-tion) have been used to differentiatebetween maldigestion and malabsorp-tion. However, similar to the nonra-dioactive 13C-CO2 breath test for fat

and carbohydrate digestion, the 14C fatabsorption tests become positive onlyin patients with severe pancreatic in-sufficiency (1,2).

Vitamin B12 absorption depends onsplitting the R protein from cobalaminby pancreatic enzymes. The Schillingtest may be abnormal in some patientswith severe pancreatic insufficiency;vitamin B12 blood levels usually remainnormal (10).Other tests of malabsorption: Clini-cally significant malabsorption of pro-teins and fat-soluble vitamins occursonly rarely in chronic pancreatic dis-ease. Thus, serum albumin is usuallynormal, and carotene (for vitamin A),serum calcium, phosphate and alkalinephosphatase (for vitamin D) andprothrombin time (for vitamin K) areusually normal. However, plasma vita-

min E levels and the vitamin E:totalplasma lipid ratio are abnormally low in75% of patients with chronic alcoholicpancreatitis and in 91% of those withsteatorrhea. Although it has been sug-gested that this determination may be apractical means of detecting steator-rhea in patients with alcoholic chronicpancreatitis (11), fecal fat balance stillremains the standard for abnormal fatexcretion. Low folate levels, if present,are usually related to excessive alcoholconsumption rather than pancreatic in-sufficiency.

Differentiation of malabsorptionfrom pancreatic maldigestion can beachieved by tests that depend on intes-tinal absorption rather than digestion.Thus, the xylose absorption test andlactose tolerance tests are normal inpancreatic insufficiency.Tests of pancreatic exocrine secretoryand synthetic function (Table 3):Tests in this category are based on thesecretory and synthetic function of thepancreas. Assessment of this functioncan be done either by directly deter-mining the composition of pancreaticsecretion obtained by peroral intuba-tion, or by deducing secretory functionsfrom the concentration of enzymes ortheir substrate in blood, urine or stool.Tests requiring duodenal intubation:Stimulation of pancreatic secretionscan be achieved either by parenterallyadministered hormones or by stimulat-ing the pancreatic secretions with in-traduodenally administered food.Direct stimulation of the pancreas with

parenterally administered hormones: Thefirst changes of chronic pancreatitis oc-cur in the ductal cells. Becausesecretin-stimulated bicarbonate secre-tion depends on ductal function, themost accurate method to assess pancre-atic function is the intraduodenalmeasurement of the secretin-stimulated pancreatic bicarbonate se-cretion. Based on over 5000 cases thesensitivity of the tests was 90% and thespecificity was 94%. False positive re-sults occurred in patients with va-gotomy, gastric surgery, diabetes andinaccurate tube placement (12). ‘Falsepositive’ results may not necessarily befalsely positive. Abnormal pancreaticfunction after vagotomy or gastric sur-

TABLE 2Tests designed to investigate the digestive capability of the pancreas

Fecal fat excretion (does not differentiate between pancreatic and other types of

malabsorption)

Fecal fat excretion before and during pancreatic replacement therapy

Screening tests

Macroscopic and microscopic examination of stool

Starch digestion test for amylase (hydrogen,14

C-CO2 breath test)131

I-triolein and fatty acid digestion and absorption test (blood and urine)14

C-triolein and fatty acid digestion and absorption test (14

C-CO2 breath test)

1,3 distearyl-2 carboxyl-13

C octanoyl glyceryl test for digestion by lipase (13

C-CO2breath test)

Fluorescein dilauryl test for digestion by lipase

N-benzoyl-L-tyrosyl-L-p-aminobenzoic acid (NBT-PABA) urinary excretion test

Other tests of malabsorption that may be positive in chronic pancreatitis

Low serum protein

Malabsorption of fat-soluble vitamins

Vitamin A: Low serum carotene

Vitamin D: Low serum calcium, elevated alkaline phosphatase

Vitamin K: Prolonged prothrombin time

B12 absorption (Schilling test and double-labelled Schilling test)

Modified from reference 1

TABLE 3Tests designed to study pancreatic synthetic and exocrine secretory function

Tests requiring duodenal intubation (‘invasive’)

Direct stimulation with secretin, cholecystokinin, caerulein, bombesin

Indirect stimulation: Lundh test meal, fatty acid or amino acid test meals

Tests not requiring duodenal intubation (‘noninvasive’)

Fecal enzyme determination: chymotrypsin

Serum or plasma levels (with or without provocative testing) of:

Substances required in synthetic activity: amino acids

Substances synthesized: pancreatic amylase, trypsinogen, lipase, pancreatic

polypeptide


TABLE 1Laboratory tests used to investigatechronic pancreatic disease

Pancreatic function tests designed to

assess:

Digestive capacity of the pancreas

Pancreatic synthetic and exocrine

secretory function

Pancreatic endocrine function

Tests designed to differentiate between

chronic pancreatitis and cancer



BECK

gery may relate to abnormal neuroen-docrine control of pancreatic secre-tions. Evidence is accumulating that ahigh proportion of insulin-dependentdiabetics may have exocrine pancreaticdysfunction, including ductular abnor-malities (13). Hypersecretion was ob-served in patients with cirrhosis. Falsenegatives were mainly due to inaccu-rate tube placement, which can beavoided by careful fluoroscopy (12).The secretin test becomes positive be-fore any other function test, and an ab-normal secretin test precedes the earli-est structural abnormalities that can bedemonstrated by endoscopic retrogradepancreatography (14).

The hormones cholecystokinin(CCK), caerulein and bombesin stimu-late enzyme secretion by alveolar cells.Diminished hormone stimulated in-traduodenal enzyme concentration oc-curs early in pancreatic insufficiency.These tests, however, have not been aswell standardized as the secretin test(12).

Other ‘invasive’ tests requiring duo-denal intubation have been proposed.For instance, the above-mentionedhormones increase secretion of calciumand lactoferrin, and abnormally highconcentrations can be measured induodenal juice of patients with earlychronic pancreatitis. However, duode-nal lactoferrin levels are also elevatedin patients with duodenitis (1,2). Apancreatic function test was developedincorporating 75Se methionine intopancreatic proteins. Increased 75Semethionine labelled protein excretioninto the duodenal juice has been re-ported to occur in early chronic pan-creatitis, but this observation has notyet been confirmed (1,2).Indirect stimulation tests: The indirecttest described by Lundh (15) measuresthe secretory response to an intraduo-denally administered liquid test meal.Diminished enzyme secretion in re-sponse to the test meal occurs in pan-creatic insufficiency. However, indirectpancreatic secretory stimulation de-pends on the release of secretin and CCKfrom the duodenal mucosa, which maybe impaired in patients with duodenalmucosal abnormalities. False positiveresults have been reported in patients

with celiac disease. Tests based on duo-denal perfusion with fatty acids oramino acids have the same limitations.Tests that do not require duodenal in-tubation: Because ‘invasive tests’ aredifficult to perform and some patientsresist prolonged nasoduodenal intuba-tion, investigators have attempted todevelop sensitive and specific tests thatcan be performed without intubation.Fecal chymotrypsin: The overall prote-olytic activity in feces reflects the ac-tivities of a mixture of bacterial, pan-creatic and other peptidases. Therefore,a method using specific substances forfecal chymotrypsin activity is more spe-cific. Unfortunately, fecal chymotryp-sin becomes abnormal only in ad-vanced pancreatic disease (1,2).Serum or plasma amino acids: Duringamino acid incorporation into newlysynthesized pancreatic protein, a fall inplasma amino acid levels occurs aftersecretin and pancreozymin stimulation.Slowing of the rate of plasma aminoacid decrease may suggest exocrinepancreatic insufficiency. Unfortu-nately, overlaps occur between normalpatients and those with pancreatic dis-eases (1,16).Serum and urine levels of pancreatic en-

zymes: Serum and urinary isoamylaseand lipase are useful in the diagnosis ofacute pancreatitis. However, these en-zymes are of no help in the diagnosis ofchronic pancreatitis or pancreatic neo-plasm, except during an acute relapse orthe presence of pseudocysts where theseenzymes may remain elevated duringthe active presence of a pseudocyst.Tests assessing endocrine pancreaticfunction: Frank diabetes does not oc-cur in chronic pancreatic disease untilapproximately 90% of the islets havebeen destroyed. Therefore, demonstra-tion of carbohydrate intolerance is nothelpful for the early diagnosis of pancre-atic disease. Both insulin and glucoselevels vary considerably in patientswith chronic pancreatitis, and theirlevels do not provide useful informa-tion (2).

Tumour markers for differentiatingcancer from chronic pancreatitis

Many attempts have been made todevelop diagnostic techniques that do

not require laparotomy or biopsy. Se-rum and duodenal tumour markers ini-tially appeared the most promising;however, many of the proposed tests,such as carcinoembryonic antigen(CEA), pancreatic oncofetal antigenand ribonuclease, are not sufficientlysensitive to provide a differential diag-nosis (17).

More promising are the carbohy-drate antigens, especially CA 19-9, themarker that has the closest associationwith pancreatic cancer (17,18). It hasbeen reported that this carbohydrateantigen has a sensitivity of 78% in re-sectable and 91% in unresectable pan-creatic cancer. The specificity of thetest is 92%. Unfortunately, the test hasa relatively low sensitivity in early de-tectable lesions, so it cannot be used asa screening test. The antigen was origi-nally extracted from a colon cancer, ispositive, and is expressed in cancers ofthe colon (19) and other gastrointesti-nal carcinomas. The monoclonal anti-body is expensive, and the test does notappear to be cost-effective at present.

METHODS TO DETERMINETHE ANATOMICAL BASIS FOR

ABNORMAL FUNCTION

Imaging techniques (Table 4)

Indirect radiological methods: Con-ventional x-ray techniques cannot dif-ferentiate between the densities ofpancreatic and peripancreatic tissues.Therefore, except when the pancreas iscalcified, the abdominal survey filmdoes not show the pancreas. Attemptsto visualize pancreatic anatomy directlyby isotope scanning techniques (75Semethionine) do not provide imagesthat could accurately differentiate nor-mal from diseased pancreas. Therefore,before the existance of ultrasound andcomputed tomography (CT) scan, di-rect visualization of the pancreas wasnot possible. Up to the late 1970s, ra-diological diagnosis of pancreatic dis-ease depended entirely on ‘indirectmethods’. These were based on detec-tion of distortion, compression or in-vasion by benign or malignant massesof hollow organs (stomach, duodenum,colon), ducts (pancreatic or common


duct) or blood vessels filled with con-trast material.Barium meal and barium enema: Bar-ium studies of the stomach and duode-num may demonstrate deformities ofthe posterior wall of the stomach, andof the bulbar, postbulbar and peripapil-lary regions of the duodenum. Differen-tiation between cancer and chronicpancreatitis is difficult. Occasionally,large pseudocysts or tumours may com-press or invade the colon, resulting inabnormalities of the barium enema.Angiography: Ultrasonography and CTscanning have, for all practical pur-poses, replaced angiography in the diag-nosis of pancreatic disease.Percutaneous transhepatic cholangi-ography: Chronic pancreatic diseasemay distort or obstruct the transpancre-atic portion of the common bile duct.Injection of dye into the intrahepaticduct system via the percutaneous routeprovides excellent visualization of thecommon duct. The method is relativelysafe, the complication rate due to sepsis,bile leak and bleeding is around 3%,and the mortality is less than 0.2%. Un-til the introduction of endoscopic ret-rograde cholangiopancreatography(ERCP), percutaneous transhepaticcholangiography (PTC) was the methodof choice for investigating patients withobstructive jaundice (20). Since ERCP isnow becoming increasingly available inprimary and secondary care hospitals,the use of PTC has been relegated to pa-tients in whom ERCP has failed. Clearly,

the advantage of ERCP over PTC is thatERCP provides the opportunity fortherapeutic intervention, with sphinc-terotomy or stenting in the same sit-ting. Furthermore, for pancreatic dis-ease, ERCP is superior because it allowsfor direct visualization not only of thecommon bile duct, but also of the pan-creatic duct.ERCP: ERCP plays a principal role inthe overall investigation of pancreaticdisease. The main indications for ERCPare to: differentiate between chronicpancreatitis and carcinoma in patientswith abnormal CT scan or ultrasound;differentiate between hepatobiliary andpancreatic diseases in patients with ob-structive jaundice; assess ductal anat-omy preoperatively in patients withchronic pancreatitis; prove or disprovethe presence of pancreatic disease in pa-tients in whom, based on history andfunction tests, there is a high degree ofsuspicion of pancreatic disease butwhose ultrasound and CT scans are nor-mal (21).

The major recent advances are inthe therapeutic applications of thistechnique. The classic indication forpapillotomy is the presence of commonbile duct stones. Recently, sphincter ofOddi spasm has been recognized as apossible cause of recurrent abdominalpain and pancreatitis. It has been sug-gested that if sphincter of Oddi motilitystudy is abnormal, symptoms may be re-lieved or relapsing pancreatitis allevi-ated by sphincterotomy. If a

malignancy causes jaundice withsymptoms, palliative treatment withstenting is safer and less invasive thansurgical bypass procedures. Newerstents allow for less frequent cloggingof the insert (21). However, the use-fulness of pancreatic stone removaland dilation of the duct remains to beproven.

As some of the features of chronicpancreatitis and pancreatic cancer aresimilar, differential diagnosis betweenthe two may not always be easy. Themost important diagnostic features arethe changes that occur in the duct sys-tem. The pancreatic duct of normalsubjects is regular, with terminal ta-pering. Multiple irregularities of themajor and secondary pancreatic ductsare usually diagnostic of chronic pan-

creatitis, and dilated ducts due to par-tially obstructive tumour are sharplydelineated and have a smooth contour.However, as these smoothly dilatedducts may take on a ‘bead-like’ appear-ance, the ductal changes caused bycancer or chronic pancreatitis maysometimes be difficult to distinguish.The diagnosis becomes easier if theductal changes of chronic pancreatitisoccur in the absence of a localized ob-struction. According to several studies,the accuracy of ERCP in differentiatingbenign from malignant changes is be-tween 62 and 92%. The higher diag-nostic accuracy (92%) reported in1987 than in previous years may be re-lated to improved technique and moreaccurate interpretation of radiologicalimages (22).Direct visualization of the pancreas –Abdominal survey film: This examina-tion is useful as an indirect method todiagnose acute pancreatitis. The usualfindings are localized ileus (sentinelloop) or the cut-off sign of the colon.Routine films cannot distinguish thepancreas from surrounding soft tissues.However, once calcification occurs, di-rect visualization becomes possible.Calcium may be localized in a singlearea or extend throughout the organ.As to the differential diagnostic valueof calcification, calcium deposits occurmost frequently in chronic pancreatitis;however, they may also be present inthe walls of pseudocysts, cystic neo-

TABLE 5Ultrasonic and computerized tomo-graphic scan of chronic pancreatitis (CP)and cancer

CP Cancer

Mass + +++

Inhomogeneity + +

Cystic lesions +++ +

Calcification + +

Abnormal common duct + +

Abnormal pancreatic duct

Smooth dilation + +++

Beaded dilation + +++

Irregular dilation +++ +

+Rare; ++Common; +++Frequent

TABLE 4Imaging techniques

Indirect

Barium meal

Barium enema

Angiography

Percutaneous transhepatic

cholangiography

Endoscopic retrograde

cholangiopancreatography

Direct

Abdominal survey film (if calcified)

Ultrasonography

Computerized tomography (CT) scan

CT scan with dynamic scanning

Magnetic resonance imaging

Endoscopic ultrasonography


BECK

plasms and in the parenchyma of carci-nomas.Ultrasonography: Echographic diag-nosis of both chronic pancreatitis andpancreatic carcinoma is dependent onchanges in size and contour of thegland, echoreflectivity of the paren-chyma and alteration of ductal anat-omy (Table 5). Unfortunately, theshape, size and position of the organ

may differ from one normal individual

to another, and echoreflectivity in-

creases with age. In chronic pancreati-

tis, changes in gland size are not always

diagnostic because this condition may

cause pancreatic atrophy or enlarge-

ment. Alterations of contour may be

difficult, because the age-dependent in-

creases in echogenicity of the gland

may render delineation of the pancreas

from peripancreatic fibrous tissue diffi-

cult. A solid pancreatic mass of a tu-

mour cannot always be differentiated

from that of a focal mass caused by in-

flammation. Also, cystic tumours may

be difficult to differentiate from pseu-

docysts.

Interpretation of ultrasonographicchanges in the ducts are subject to thesame limitations were described for theductal characteristics observed in ERCP.In a retrospective analysis of 27 pa-tients with proven chronic pancreati-tis, similar ductal abnormalities werefound in patients with or without pan-creatic insufficiency, and in 13% of pa-tients with chronic pancreatitis theultrasound was normal (23).Computerized axial tomography:Computerized axial tomography (CAT)is one of the most accurate direct meth-ods to assess the pancreas. However, forthe same reasons as described for ultra-sound, except for the classic examplesof each disease, the differential diagno-sis of chronic pancreatitis from carci-noma may be difficult and not alwayspossible (Table 5). Mass lesions may becaused by inflammation or tumour, cys-

tic changes may be due to pseudocysts

or cystic neoplasia, and changes in duc-

tal anatomy in chronic pancreatitis

may resemble those seen in carcinoma

of the pancreas (24).

Dynamic scanning (25) employsrapid intravenous injection of 150 mLof iodinated contrast material. Starting

immediately after the intravenousinjection of the bolus, cuts are madeat 1 cm intervals with a rapid scan se-quence of less than 2 s. This allows as-sessment of the relationship of a massto the surrounding vasculature, and de-tects invasion of arteries or veins, in-farctions and perfusion defects. Acorrect diagnosis of pancreatic cancerhas been reported in 91% of 174 pa-tients with a frequency of false positiveand false negative results of 8 and 1%,respectively. Unfortunately, the accu-racy of this method was studied in pa-tients with advanced cancer, and onlysix of the 174 patients had resectabledisease. Thus, even this improvementin diagnosis using CAT scanning does notallow for the early diagnosis of resecta-ble cancer (25). Of special diagnosticinterest are benign and malignant cys-tic lesions of the pancreas (Figures 3,4);their x-ray characteristics are reviewedin Table 6. Combined with ERCP, someof these tumours are easily recognizableand may have characteristic appear-ance such as classic pseudocyst ofchronic pancreatitis (Figure 5), thesunburst calcification of the frequently

benign serous cystadenomas (Figure 6)and the large thick fluid-filled cysts of

mucinous cystadeno carcinomas (Fig-ure 7).

Magnetic resonance imaging: Exami-nation of the pancreas with magneticresonance imaging (MRI) has been lim-ited due to distortion by respiratory mo-tion and difficulty in distinguishingbetween the bowel and pancreas. Thehead and body can be seen reasonablywell, but the tail is often not visualized.At present, the diagnostic capability ofthe CT scan is superior to that of MRI(26).

A recent report indicates that MRIcan noninvasively visualize the biliarytract by subjecting the images obtainedin the axial plane to a computer gener-ated projection of the cholangiogramin a coronal plane. With today’s tech-niques, the images are not as clear aswith invasive visualization of the ducts,and several episodes of 17 to 20 s breathholding are necessary to obtain an im-age (27).Endoscopic ultrasonography: Endo-scopic ultrasonography involves the at-tachment of an ultrasound transducerto the endoscope. The scope is intro-duced into the duodenum or into theposterior wall of the stomach. Under di-rect endoscopic observation an echo-gram of the pancreas is created, which iswell seen because it appears adjacent tothe endoscopic image. Echograms ofthe relation of the pancreas to the bili-

TABLE 6Cystic tumours

Ultrasound/computed to-

mographic scan

Endoscopic retrograde

cholangiopancreatography

Pseudo cyst

Simple –Single

–Homogeneous

–Communicates (70%)

Complicated –Hemorrhage

–Multiloculated

–May calcify

Serous cystadenoma

(microcystic

cystaden)

–Lobulated, large

–Multiple small cysts

–’Sunburst’ calcification

–Does not communicate

–Draped

Mucinous cystadeno CA

(macrocystic

cystaden CA)

–Large cysts

–Thick fluid

–Dystrophic calcification

–Does not communicate

–Draped

–Obstructed

Mucinous ductal ectasia –Tumour in duct

–Cystic dilation of duct

–Communicates

–Thick fluid, draped


Papillary cystic tumour –Tumour in duct

–Obstruction

–May cause acute

pancreatitis

–Obstruction

–Draped


–Possible pseudocyst

Based on reference 4


ary system, portal vein, arteries and theaorta can be obtained. Differentiationof malignant and benign tumours of thepancreas is possible for masses of 30 mmdiameter, and even as small as 20 mm. Itis somewhat less sensitive for tumours ofless than 20 mm diameter (28,29).

It is unlikely that this examinationwill ever become the primary methodto detect small and fully resectable can-cers, because it is never used beforesome other technique (ERCP, CT scan,etc) has already raised the suspicion of

a tumour. However, by now endoscopicultrasonography has proven to be thebest method to assess peripancreatic lo-cal invasion and help to decide preop-eratively on resectability of the tumour(28,29).

Methods to establishhistopathological diagnosis

The many methods that have beenused to provide specimens for histologi-cal diagnosis are shown in Table 7.Every method has certain specific limi-

tations. Cytology from duodenal juice

has a very low yield of cells. Cytology

obtained by cannulation of the pancre-

atic duct during ERCP is cumbersome.A cytology brush or the recently devel-oped biopsy forceps can be introducedeasily into the head of the pancreasduring ERCP. However, obtainingspecimens from other parts of the glandmay be difficult and often impossible(30).

The major disadvantage of the ul-trasound or CT scan thin needle biopsyis that it may be difficult to differenti-ate tumour mass from pericancerousedema on the scan. If the latter is biop-sied a false negative cytology is ob-tained (31). Reports have also indi-cated that occasionally the tumour mayspread along the needle tract (32). Themajor disadvantage is that these biop-sies are carried out only once the tu-mour is large enough to be seen on ul-trasound or CT scan, and in mostinstances, by then it is too late for cura-tive removal of the malignancy. Thus,

Figure 7) Computed tomographic scan of mucinous cystadeno carci-noma. Note the multiloculated fluid-filled large cysts

Top left Figure 5) Computed tomographic scan of two large pseudocystsof the pancreas. Note the sharply delineated outline of the cysts

Bottom left Figure 6) Computed tomographic scan of a serous cystaden-oma. Note the ‘sunburst’ calcification within the cyst

TABLE 7Methods to obtain material for histology

Cytology from duodenal juice

Cytology from pancreas obtained during endoscopic retrograde

cholangiopancreatography by collection during secretin stimulation brushing

Cytology by ultrasonography or computed tomographic scan guided thin needle biopsy

Cytology obtained during endoscopic ultrasonography

Biopsy obtained during endoscopic retrograde cholangiopancreatography

Biopsy obtained during surgery


BECK

in most instances these biopsies areonly used to confirm by histology thepresence of unresectable cancer. Directcytology is now possible during endo-scopic ultrasonography (33). The as-piration needle is inserted directly intothe small tumour. Although data froma prospective study are not yet avail-able, this method may surpass the accu-racy of other types of biopsies.

All these methods have a commondefect: they exhibit low sensitivity buthigh specificity. In other words, if nocancer cells are found, malignancy isnot excluded with certainty, but thepresence of cancer cells provides an un-questionable diagnosis of malignancy.The major limitation is that many ofthese tumours are desmogenic and can-cer cells may be difficult to find. Thisalso holds, not only for aspiration cy-tology, but also for surgical biopsies(34).

CLINICAL APPROACH TODIAGNOSIS

There are three questions that needto be answered during investigation:first, whether chronic pancreatic dis-ease is present; second, whether this isbenign or malignant; and finally,whether a malignant neoplasm is stillcuratively resectable. Before the intro-duction of ultrasound and CT scan, pa-tients were referred to gastroenterolo-gists because of abdominal pain,diarrhea, weight loss or jaundice. Thegastroenterologist’s job was to diagnosethese complex problems by means oflaboratory investigation, indirect imag-ing techniques and, since the 1970s, byERCP. This has changed with the intro-duction of both echography and CTscanning.

Presently, patients are referred be-cause of suspected pancreatic diseasedetected as an incidental finding on ul-trasound or CT scan, frequently orderedfor reasons unrelated to suspected pan-creatic disease. These patients do notneed pancreatic investigation to provethe presence of pancreatic disease and,in contrast to past practice, will not un-dergo function tests but will proceed di-rectly to ERCP.

There are, however, patients whoare still referred for investigation of

symptoms suggestive of pancreatic dis-ease without previous work-up. Inthese patients, the first examination or-dered is the abdominal ultrasound. Ifthis is negative, a CT scan is carried out.If this is also normal and if pancreaticdisease is strongly suspected, an ERCPfollows. If this is also negative, in thepresence of a strong clinical suspicion,a secretin test is carried out.

Some clinicians still use pancreaticfunction tests and even recent researchpapers have used them (11,16,35). Thisis why these tests were reviewed in thefirst section of the paper, even thoughthere may be little indication for theirutilization. Specifically, because oftheir poor predictive value, presentlyavailable ‘screening tests’ are not cost-effective. For all practical purposesthere are only two pancreatic functiontests that are still of use: the 72 h fecalfat balance before and during pancre-atic enzyme administration and the se-cretin test. The first is used in thecourse of investigations of patientswith steatorrhea. Because the secretintest is more sensitive than any of theimaging methods (14), it is used wherechronic pancreatitis is suspected butthe ERCP is still normal.

The methods used to achieve thesecond objective, ie, to differentiatebetween chronic pancreatitis and car-cinoma, is reviewed in the algorithmshown in Figure 8. If a mass is found on

ultrasound, a guided fine-needle biopsy

is carried out. A positive biopsy for a tu-

mour is diagnostic, but a negative one

does not exclude malignancy. If the ul-

trasound is negative, a dynamic CT

scan may demonstrate the lesion and aCT guided biopsy may provide a diagno-sis for malignancy. If, however, the bi-opsy yields no malignant cells, an ERCPis carried out to assess ductal anatomy.The role of endoscopic ultrasonogra-phy combined with endoscopic ultra-sound guided cytology needs to befurther evaluated. The main diagnosticrole of this test will be to assess the na-ture of tumours or cysts that are toosmall for regular echography or CT scanguided fine-needle biopsy.

The third objective of investigationis to establish whether a malignancy isresectable in patients in whom distantmetastases have been excluded. In thepast, angiography used to be employedto assess vascular invasion. As endo-scopic ultrasound becomes more andmore established, this will become themethod of choice because it can assessmost accurately size, vascular invasionand lymph node involvement.

In spite of the major advances inmethods of investigation, mortality dueto cancer of the pancreas has not di-minished during the past 20 years (36).This is because cancers of the pancreascan grow to considerable size beforethey cause symptoms. Once symptoms

Figure 8) Algorithm to differentiate carcinoma from chronic pancreatitis. CT Computed tomogra-phy; ERCP Endoscopic retrograde cholangiopancreatography; US Ultrasonography

are present, it is still difficult to differ-entiate chronic pancreatitis from can-cer unless the cytology is positive.None of the tests can diagnose cancersless than 2 cm in diameter, except inthe rare instance where a small tumourobstructs the papilla. Those tests thatcan diagnose small and resectable tu-mours, ie, ERCP and endoscopic echo-graphy, are done only once a suspicionof pancreatic disease has arisen on thebasis of less sensitive investigations.Even if tumours are discovered early,surgery is difficult and has considerablemortality, and radiotherapy andchemotherapy remain ineffective. Un-til a very sensitive and accurate screen-ing blood test that can detect verysmall cancers has been developed anduntil improvement in therapy has beenachieved, the high mortality due tocancer of the pancreas will remain un-altered.

ACKNOWLEDGEMENTS: I am indebted toDr Paul G O’Brien, Assistant Professor, and DrKhan Nguyen, Associate Professor of Diagnos-tic Radiology, Queen’s University, for provid-ing the CT scans for Figures 5, 6 and 7. I am alsograteful to Wendy Gorski for the conscien-tious typing of this manuscript.

REFERENCES1. Beck IT. Diagnosis of chronic pancreatic

disease. In: Thomson ABR, Shaffer E,eds. Modern Concepts inGastroenterology. New York: PlenumMedical Book Company, 1992:3:179-221.

2. Bank B. Chronic pancreatitis: Clinicalfeatures and medical management.Am J Gastroenterol 1986;81:153-67.

3. Warshaw A, Castillo C. Pancreaticcarcinoma. N Engl J Med1992;326:455-65.

4. Warshaw A, Compton C, LewandrowskiK, et al. Cystic tumors of the pancreas.Ann Surg 1990;212:432-43.

5. DiMagno EP, Go VLW, SummerskillWHJ. Relations between pancreaticenzyme outputs and malabsorption insevere pancreatic insufficiency. N Engl JMed 1973;222:813-5.

6. Regan PT, Malagelada JR, DiMagno EP,et al. Comparative effects of antacids,cimetidine and enteric coating on thetherapeutic response to oral enzymes insevere pancreatic insufficiency. N Engl JMed 1977;297:854-8.

7. Ghisan FK, Moran JR, Durie PR, et al.Isolated congenital lipase-colipase

deficiency. Gastroenterology1984;86:1580-2.

8. Lankisch PG, Schreiber A, Otto J,et al. Pancreolauryl test: Evaluation of atubeless pancreatic function test incomparison with other indirect and directtests for exocrine pancreatic function.Dig Dis Sci 1983;28:490-3.

9. Cavallini G, Piubello W, Brocco G.Reliability of the Bz-Ty-PABA and thepancreolauryl test in the assessment ofexocrine pancreatic function. Digestion1983;27:129-37.

10. Brugge WR, Goff JS, Allen NC.Development of a dual label Schillingtest for pancreatic exocrine functionbased on the differential absorption ofcobalamin bound to intrinsic factor andR protein. Gastroenterology1980;78:937-49.

11. Marotta F, Labadarios D, Frazer L,Girdwood A, Marks IN. Fat solublevitamin concentration in chronicalcohol-induced pancreatitis.Relationship with steatorrhea.Dig Dis Sci 1994;39:993-8.

12. Dreiling DA. Investigation of pancreaticfunction. In: Beck IT, Sinclair DG, eds.Exocrine Pancreas. London: Churchill,1971:154-66.

13. Nakanishi K, Kobayashi T, Miyashita H,et al. Exocrine pancreatic ductograms ininsulin-dependent diabetes mellitus. AmJ Gastroenterol 1994;89:762-6.

14. Waye JD, Adler M, Dreiling DA. Thepancreas: A correlation of function andstructure. Am J Gastroenterology1978;69:176-81.

15. Lundh G. Pancreatic exocrine function inneoplastic and inflammatory disease: asimple and reliable new test.Gastroenterology 1962;42:275-80.

16. Maringhini A, Nelson DK, Jones JD,DiMagno EP. Is the plasma amino acidconsumption test an accurate test ofexocrine pancreatic insufficiency.Gastroenterology 1994;106:488-93.

17. Benini L, Cavallini G, Zordan D.A clinical evaluation of monoclonal (CA19-9, CA50, CA 12-5) and polyclonal(CEA, TPA) antibody-defined antigensfor the diagnosis of pancreatic cancer.Pancreas 1988;3:61-6.

18. Yamamura H, Tatsuta M, Ifo M.Effectiveness of discriminant analysis ofserum Ca-19-9 and elastase 1 in thediagnosis of pancreatic carcinoma.Pancreas 1989;4:401-5.

19. Shimono R, Mori M, Akazawa K, et al.Immunohistochemical expression ofcarbohydrate antigen 19-9 in colorectalcarcinoma. Am J Gastroenterol1994;89:101-5.

20. Harbin WP, Mueller PR, Ferrucci JT Jr.Transhepatic cholangiography:Complications and use patterns of thefine needle technique: a

multi-institutional survey. Radiology1980;135:15-22.

21. Cotton BP, William CB. PracticalGastrointestinal Endoscopy, 3rd edn.Boston: Blackwell Scientific Publications,1990:85-156.

22. Deltenre M, De Reuck M, De Koster E.Endoscopic retrogradecholangiopancreatography in thediagnosis of 247 pancreatic malignancies.A reappraisal. Acta Gastroenterolog Belg1987;50:102-11.

23. Alpern MB, Sandler MA, Kellman GM.Chronic pancreatitis: Ultrasonic features.Radiology 1985;155:215-9.

24. Karasawa E, Goldberg HI, Moss AA. CTpancreatogram in carcinoma of thepancreas and chronic pancreatitis.Radiology 1983;148:489-93.

25. Freeny PC, Marks WM, Ryan JA.Pancreatic ductal adenocarcinoma:diagnosis and staging with dynamic CT.Radiology 1988;166:125-33.

26. Stark DD. The liver, pancreas and spleen.In: Higgins CB, Hricak H, eds. MagneticResonance Imaging of the Body. NewYork: Raven Press, 1987:347-72.

27. Ishizaki Y, Wakayama T, Okada Y,et al. Magnetic resonancecholangiography for evaluation ofobstructive jaundice. Am J Gastroenterol1993;88:2072-7.

28. Ysauda K, Mukai H, Fujimoto S.The diagnosis of pancreatic cancer byendoscopic ultrasonography. GastrointestEndosc 1988;34:1-8.

29. Mukai H, Nakajima M, Yasuda K,et al. Evaluation of endoscopicultrasonography in the pre-operativestaging of carcinoma of the ampulla ofVater and common bile duct.Gastrointest Endosc 1992;38:676-83.

30. Kubota Y, Takaoka M, Tani K, et al.Endoscopic transpapillary biopsy fordiagnosis of patients withpancreaticobiliary ductal strictures. Am JGastroenterol 1993;88:1700-4.

31. Holm HH, Torp-Pedersen S, Larsen T.Percutaneous fine needle biopsy. ClinGastroenterol 1985;14:423-49.

32. Smith FP, MacDonald JS, Schein PS, etal. Continuous seeding of pancreaticcancer by skinny-needle aspirationbiopsy. Arch Intern Med 1980;140:855.

33. Tio TL, Sie L, Tytgat GNJ.Endosonography and cytology indiagnosing and staging pancreatic bodyand tail carcinoma. Dig Dis Sci1993;38:59-64.

34. Moossa AR, Altorki N. Pancreatic biopsy.Surg Clin North Am 1983;63:1205-14.

35. Weber P, Seibold F, Jenss H. Acutepancreatitis in Crohn’s disease. J ClinGastroenterol 1993;17:286-91.

36. Gaudet L, Lee J. National CancerInstitute of Canada: Canadian CancerStatistics 1993. Toronto: CanadianCancer Society, 1993.


BECK

Submit your manuscripts athttp://www.hindawi.com

Stem CellsInternational

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

Immunology ResearchHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinson’s Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttp://www.hindawi.com

critical review of diagnostic methods used in chronic ...factor for pancreatic cancer (2,3). once...

Documents