Jooa. D . D . VOT.UMS V I I I MARCH, 1941 NUMBIr 3

Carcinoma of the Head of the Pancreas: A Review of Forty Cases By

SAVERIO CHARLES FRANCO, M.D. BROOKLYN, N E W YORK

INTRODUCTION

O UR interes t in the subject of carcinoma of the head of the pancreas was aroused some years ago

because i t mimicked other abdominal disease and i t seemed to be a very obscure type of abdominal neo- plasm. Our g rea tes t surpr ise came in several cases of supposed duodenal ulcer in people in middle life. There was a s tory of ulcer pain, usually the first episode, and indirect evidence ef duodenal ulcer in tha t a de- formed and i r r i tab le duodenal cap was found on fluoroscopy and X-ray. Improvement did not occur and a f te r a period of one to two months jaundice and an abdominal mass appeared. There were other cases where the s tory suggested abdominal neoplasm but re- peated gas t ro- intes t inal X-rays were negative. Af t e r months of intractable abdominal discomfort and pain an exploratory laparotomy or the clinical course dis- closed carcinoma of the head of the pancreas.

We have followed the same cr i te r ia in classifying carcinoma of the head of the pancreas as Graham de- scribes in his work on disease of the liver and bi l iary passages.

Graham (1) states tha t "within an area having a radius of not more than 0.5 cms., a carcinoma may arise in anyone of the following s t ruc tures : (1) the ampulla of Vater, (2) the end of the common bile duct, (3) the end of the duct of Wirsung, (4) the glandular t issue at the head of the pancreas, (5) the duodenal mucous m e m b r a n e covering the b i l iary papil la."

Leven (2) in a review of 678 cases of p r imary carci- noma of the pancreas s tated tha t 56.3 per cent were in the head of the gland and 30.7 per cent were diffusely spread through the pancreas.

Our series comprises fo r ty cases, fourteen proven by autopsy and the remainder by operation. Of the cases operated upon twelve had biopsies. The remain- ing fourteen cases showed not only a mass in the head of the pancreas but also abdominal metastases. Eleven of these operated cases died in the hospital within a period of t ime vary ing from a few days to a month. One pat ient died at home within two months. Only two pat ients were discharged and we were unable to ob- tain a follow-up as the cases occurred in 1926. A follow-up on two cases, not included in this series, of mass in the head of the pancreas without abdominal metastases showed them to be well several years later, i l lus t ra t ing tha t at t imes a purely clinical diagnosis can be wrong. As determined by autopsy one ease arose in the ampulla of Vater (colloid carcinoma), one arose in the lower bile ducts and the remainder in the head of the pancreas. At operation only once

Note: These for ty cases were admitted to the Long Island College Hospi tal in the period f rom 1924 to 1939 inclusive.

Submitted July 10, 1940,

was it thought that the carcinoma originated at the ampulla of Vater.

INCIDENCE

The grea tes t number of cases occurred in the sixth decade, more than 50 per cent. The youngest pat ient was th i r ty -s ix years of age and the oldest was seventy- seven years. Grouped according to decades, the age incidence is as follows:

4th decade . . . . . . . . . . . . . . 3 cases 5th decade . . . . . . . . . . . . . . 4 cases 6th decade . . . . . . . . . . . . . . 22 cases 7th decade . . . . . . . . . . . . . . 8 cases 8th decade . . . . . . . . . . . . . . 3 cases

There were t w e n t y - t h r e e males and seventeen ~females. Only two negro pat ients were present in this group, one male and one female.

PAST HISTORY

Eight pat ients gave a past h is tory of a gas t r ic illness. Four cases had previous gall bladder disease with colic and two of these had a choIecystectomy per- formed with rel ief of symptoms. One pat ient had an at tack of jaundice about one year p r io r to the carci- noma and at autopsy a definite cholecysti t is was seen. Three pat ients had recurrent indigest ion; one of these had an ulcer syndrome.

SYMPTOMATOLOGY

Thir ty- three cases had symptoms four months or less. Two had symptoms for as long as e ight months. More than 25 per cent had symptoms of a month or less on admission. Often an acutely developing jaundice would cause them to seek medical advice. The various periods of durat ion of symptoms may be classified as follows:

1 month - - 12 cases 5 months - - 3 cases 2 months - - 8 cases 6 months - - 1 case 3 months - - 7 cases 7 months - - 1 case 4 months - - 6 cases 8 months - - 2 cases

The most common symptoms were abdominal pain, jaundice, severe weight loss and a change in bowel habit. A few cases presented a painless jaundice. About 30 per cent of the cases complained of severe abdominal pain, marked weight loss, and constipation without the presence of jaundice. These were the most difficult to diagnose because the X-ray often failed to reveal any localizing lesion.

(1) Abdominal Pain. This occurred in th i r ty - fou r cases (85 per cent) . The most common type of pain was a cramp-like epigas t r ic pain (16 cases), tn three cases this pain was aggravated by food. In seven cases th is pain radia ted to the lumbar area. At t imes the pain radia ted down over the entire lower abdomen. A gall bladder type of pain with attacks of colic in the r igh t upper quadrant occurred in eight eases. Five of these radiated to the r igh t costovertebral area and two

65

66 AMERICAN JOURNAL OF DIGESTIVE DISEASES VOLUMS VIII NUMRER 3

Plate 1. Widening of the duodenal curve and oblitera- tion of the markings on the medial aspect of the first and second portions of the duodenum.

radia ted to the r ight scapula. F ive cases had an ulcer type of pain and one of these had pain rad ia t ing to the lower dorsal spine. Three cases complained only of a dull constant pain in the lumbar spine. But there were fourteen cases in which the pain radia ted to the back. A burning umbilical pain occurred in two cases and in one of these i t rad ia ted to the back. There was no definite pathological basis for the different types of pain. In those cases with the ulcer type of pain only one had ulceration of the duodenum. On the other hand there were four other cases of invasion of the duodenum or stomach which did not have ulcer type of pain. In the cases with the gall bladder type of pain only one had gall stones.

(2) Jaundice. This was present in twenty-eight cases (70 per cent) and three more cases became jaundiced a f te r admission. The jaundice was usually present a shorter t ime than the pain. Twenty-three cases were less than six weeks in duration, three cases were of three months duration, and two cases were of five months duration. Two pat ients showed a marked diminution in the jaundice af te r i ts onset. Five cases were associated with chills and fever. One of these had cholecystitis, two had gall stones, one had a lung abscess, and in one case nothing was found to account for the chills and fever.

(3) Severe Weight Loss. Marked loss in weight occurred in twenty- four cases (60 per cent) . This sometimes was present before the onset of pain or jaundice. I t varied f rom ten to eighty pounds, and most of the cases lost more than twenty pounds.

(4) Changes in Bowel Habit. A change in bowel habi t was noted in fourteen cases. Marked consti- pation occurred in nine cases and in conjunction with the abdominal pain it raised the suspicion of colon

neoplasm. Diar rhea was present in four cases and one case had both constipation and diarrhea.

(5) Reflex Gastric Symptoms. Nausea and vomit- ing occurred in fourteen cases. Two cases, in the ad- vanced stages, had bloody vomitus. Ten cases com- plained of anorexia in addition to the above mentioned symptoms.

(6) Fever. Six cases noted fever in addition to the other complaints. One case had an undetermined fever for eight months as an outs tanding complaint. This case also had an enlarged spleen and was t rea ted at another hospital with X-ray for Hodgkin 's disease. At autopsy a double carcinoma was found, one at the head of the pancreas and the other in the left ureter.

PHYSICAL E X A M I N A T I O N

(1) Temperature. A tempera ture r e a c t i o n w a s noted in 50 per cent of the cases and varied from 101 ~ to 103 ~ . In the ma jo r i ty no cause was found for the tempera ture reaction but in a few it was explained on local pathology in the b i l iary t rac t and pancreas such as gall stones, cholecystitis, cholangitis, pancreat ic ab- scesses and pancreat ic calculi.

(2) Jaundice. Jaundice was present in twenty- eight cases on admission and was a later development in three more cases. Eus te rman and Wilbur (3) l ist th i r ty - th ree cases of carcinoma of the head of the pancreas without jaundice in a group of 403 cases of p r imary carcinoma of the pancreas.

(3) Enlarged Liver. An enlarged liver was found in twenty-seven cases. The enlargement varied from a short distance below the costal marg in to the um- bilicus. I t was usually hard and smooth. In two cases only was i t described as nodular.

(4) Enlarged Gall Bladder. An e n l a r g e d gall

Plate 2. Constriction of the middle of the second portion of the duodenum with dilatation of the first and second portions of the duodenum.

JOVR. D.D. FRANCO----CARCINOMA OF THE HEAD OF THE PANCREAS 67 MARCH, 1941

Plate 3. Constriction at the junction of the second and third portions of the duodenum with dilatation of the second portion. There is also evidence of pressure on the greater curvature of the stomach near the pylorus.

b ladder was present in sixteen cases. I t was described as cystic and ballotable.

(5) Palpable Mass. A palpable mass in the epi- gas t r ium other than liver or gall bladder was noted in five cases. I t was thought to be the or iginal tumor. Only one was described as pulsat ing with the aor ta and being movable. Autopsy showed this mass to be the tumor in the head of the pancreas. Eus te rman (4) states "contrary to the general ly accepted belief, about half of all such tumors (pancreat ic) were found to be somewhat movable."

(6) Palpable Spleen. A palpable spleen was de- scribed twice. One was enlarged two fingers below the costal margin and this was confirmed by operation. A n o t h e r - w a s enlarged one finger below the costal marg in and this was confirmed by autopsy. This spleen weighed 575 grams. Macroscopic infarcts were present and microscopically metastat ic carcinoma was seen.

(7) Other Physical Findings. Ascites was present in four cases. Leg edema was noted in two cases. A Virchow's node was found in one case and a rectal node above the prosta te in another case. In one case generalized purpura occurred even before the jaundice was noted.

LABORATORY F I N D I N G S

(1) Blood Studies. The average hemoglobin was close to 80 per cent. The two lowest readings, 50 per cent and 37 per cent, were in the two cases complicated by gas t r ic hemorrhage due to duodenal erosions. An- other case with erosion of the tumor into the stomach had 65 per cent hemoglobin. The average white cell count var ied f rom 7,000 to 10,000 with 70 per cent polymorphonuclear cells. Several elevated counts were associated with local suppurat ion such as small ab-

scess of the pancreas, suppurat ion in the duct of Wir- sung, and severe necrosis in the tumor. Some elevated counts were unexplained. The coagulation t ime was prolonged in two of eight cases but there were no bleeding phenomena. The bleeding t ime was prolonged in one of six cases and this was associated with purpur ic manifestat ions.

(2) Icterus. The icterus index was determined in twenty-nine cases. Seventeen cases (60 per cent) were below 50 and seven cases were below 20. Only six cases were over 100 and one case was above 200. The icterus index was rechecked pre-operat ively in seven- teen cases. There was a progressive increase in the icterus in nine cases. The icterus fluctuated in eight cases and sometimes dropped to a subicteric level. The cholesterol was always over 200 mgs. and increased with the icterus. The quali tat ive van den Berg test was only done a few times and it showed an immediate direct reaction in the presence of icterus. The stool was l ight in the presence of icterus. When the icterus index was over 25, bile appeared in the urine.

As is the case with obstruct ive jaundice, urobili- nogen was absent in the urine. However, short ly a f te r this group of cases was studied we had an opportuni ty to learn about another case of severe jaundice due to carcinoma of the head of the pancreas in which uro- bilinogen was present in the urine. In this case i t was felt tha t the numerous metastases to the liver had caused sufficient l iver damage to account for the uro- bilinogen in the urine. I t is conceivable tha t in a tong- s tanding b i l ia ry obstruction sufficient l iver damage could occur to allow urobilinogen to appear in the urine, but clinically this has not occurred. The per- s is tent absence of urobilinogen in the urine in the presence of icterus over a period of two weeks points

Plate 4. Displacement of the pyloric end of the stomach to the left. Displacement of the duodenum to the right with obliteration of the first portion and part of the second portion.

68 AMERICAN JOURNAL OF DIGESTIVE DISEASES VOLUME VIII NUMBER 3

to an obstructive type of jaundice. A severe hepatitis such as o c c u r s in cincophen poisoning, arsenical poisoning, and even in Weil's disease can be associated with temporary obstruction of bile canaliculi and pre- vent the appearance of urobilinogen in the urine, but this does not last more than seven to ten days as the patient recovers and urobilinogen appears in the urine or the patient dies of severe liver damage. The test for urobilinogen in the urine can be done according to the method of White (5) who has devised a simple and practical quantitative procedure.

There have appeared in the literature several re- ports of procedures to differentiate the icterus due to biliary calculi from that due to carcinoma. These, Sparkman (6) and O'Hage (7), are based on the quantitative determination of urobilinogen in the feces and urine. Though the figures vary depending on the method used, all agree that the urobilinogen in the stool and urine in the presence of carcinoma of the head of the pancreas is absent or is much less than in the case of biliary calculi. O'Hage (7) using the Watson method reported that urobilinogen in the feces is more than 10 mgms. in the presence of icterus due to stone and less than 10 mgms. in the cases of carcinoma of the pancreas and biliary passages. This test proved accurate in more than 90 per cent of cases of biliary obstruction studied by O'Hage.

(3) Pancreatic Ferments. In this group no study was made of the pancreatic ferments. Diamond (8) and his co-workers have published the results of study on the pancreatic secretion following the stimulation of pancreatic juice by the injection of secretin. Their test is of value in recognizing mechanical obstruction of the pancreatic duct alone or in conjunction with the common duct and could be used to differentiate biliary calculi from carcinoma of the head of the pancreas.

(4) Other Laboratory Procedures. Glycosuria was found three times in urine tests on thirty-six cases. The blood sugar was elevated in 30 per cent of the cases, ranging from 128 mgs. to 180 mgs. The blood chemistries were otherwise normal. Only a few tests of liver function were done. These were the sugar tolerance and galactose tolerance and they were all normal. An achlorhydria was present in six out of twenty-three cases tested following the injection of histamine.

X-RAY STUDY

(1) Gastro-Intestinal Series. G a s t r o - i n t e s t i n a l series were studied in twenty-six cases and abnormali- ties were described in sixteen cases as listed in Table I and shown in part in the accompanying plates.

TABLE I

Gastro-intestinal X-ray findings in carcinoma of the head of the pancreas

G. I. Series--26 cases Normal--l O cases Abnormal~16 cases (1)---4 cases had greater curvature change near the

pylorus with duodenal changes. (a ) - -py lorus cupped out, 2nd and 3rd portions

of the duodenum not seen. ( b ) - - g r e a t e r c u r v a t u r e defect near the

pylorus, redundency of 2nd and 3rd portions of duodenum.

( c ) - - g r e a t e r c u r v a t u r e defect near the pylorus, 1st and 2nd portions of duo- denum dilated. (Plate I I I )

(d)- -pylor ic area is poorly filled, 2nd and 3rd portions of duodenum incompletely filled and displaced u p w a r d and l a t e r a l l y . (Plate IV)

(2)--Duodenal changes alone 11 cases. (a)---displacement--5 cases - -4 displaced to

right, 1 to left. (b)---dilatation of the 1st portion of duo-

denum and constriction of some part of the 2nd port ion--4 cases. Plate II)

(c)---obliteration o f m a r k i n g s on medial border of 1st and 2nd portion of duo- denum--2 cases.

1 case--2nd portion displaced to right. 1 case---2nd and 3rd portions rounded out. (Plate I)

(3 ) - -Py lorospasm--1 case. (4)--Duodenal cap pulled to gall bladder by ad-

h e s i o n s - 1 case. In fourteen cases there were real changes in the

duodenum such as displacement, dilatation and obli- teration of markings on the medial border. In four of these cases there were in addition defects on the greater curvature near the pylorus. We feel, however, that these changes are associated with an advanced lesion. Feldman (9) has described as an early change in enlargement of the head of the pancreas a filling defect, shaped like an inverted three, in the ampullary portion of the duodenum. Negative X-ray studies by no means rule out pathology in the head of the pancreas.

(2) Gall Bladder X-ray. Gall bladder X-ray, ac- complished mainly by the use of intravenous dye was done in nine cases. Three of these were normal and six were abnormal. The most common abnormality was poor concentration of the dye with some dilatation of the gall bladder and failure to empty.

(3) Barium Enema. Barium enema study was done in two cases only and one of these showed a constric- tion at the hepatic flexure which at operation was found to be due to metastatic extension.

OPERATION Twelve cases had only an exploratory operation.

Fur ther operative procedure was u n d e r t a k e n in twenty-two cases and consisted usually of anastomosis of the gall bladder to the stomach or duodenum and sometimes jejunum. A gastro-enterostomy was often done in addition to the anastomosis between the gall bladder and the stomach or duodenum. In one case where only anastomosis of the gall bladder to the first part of the duodenum was done, there occurred ob- struction in the second portion of the duodenum due to the tumor growth.

The gall bladder was described in thirty-one cases either at operation or autopsy. A dilated gall bladder was found in twenty-two cases and in two cases the dilatation was due to hydrops as a result of the ob- struction of the cystic duct by tumor. A small con- tracted gall bladder was described four times and a normal gall bladder was found in five cases.

The cases without jaundice, which numbered twelve on admission, had operative findings similar to the

Joua. D . D . F R A N C O - - - C A R C I N O M A OF T H E H E A D OF T H E P A N C R E A S 69 MARCH, 1941

jaundiced cases so that it was impossible to satis- factorily explain the lack of icterus.

PATHOLOGY The type of carcinoma was determined by autopsy

and biopsy in twenty-six cases. Twenty-three had adenocarcinoma, two were colloid carcinoma, and one was a squamous type. McGee (10) reported on epi- theliod carcinoma of the pancreas several years ago. Pancreatic fibrosis was found in seven of the fourteen autopsied cases. Invasion of the duodenum occurred in five cases, though in only two was there actual erosion into the duodenum. Invasion of the stomach occurred in one case with the appearance of a pos- terior wall prepyloric ulcer.

DURATION From the onset of symptoms to the death of the

patient the average duration was two and a half months. This is shorter than most series and is proba- bly due to the high incidence of operative intervention. The shortest duration clinically was one month, a few lasted six to eight months, and one case lasted a year.

D I F F E R E N T I A L DIAGNOSIS In differential diagnosis one has to consider choleli-

thiasis, hepatitis and malignant disease of the liver. (1) Cholelithiasis. Chill and f e v e r a r e m o r e

common. The jaundice is more intermittent. Enlarged gall bladder is not so apt to be present. Tumor mass is not felt. The urobilinogen content of the feces is higher. The gastro-intestinal series do not show the above described changes in the region of the duo- denum. The presence of gall stones does not rule out concomitant malignancy. Often the differential diag- nosis can only be made by operative procedure.

(2) Hepatitis. (a) Cirrhosis of Liver. There may be a clinical story of overindulgence in alcohol. There is a more progressive dyspepsia. Physical examination reveals "spider angiomata," splenomegaly, prominent abdominal veins, ascites and often the liver is not felt. There is persistent urobilinogen in the urine. The total protein of the blood is diminished and there is a reversal of the albumin-globulin ratio. There may be diminution in the liver function as determined by the various tests. Gastro-intestinal X-ray series will reveal the esophageal varices.

(b) Hepatitis due to cincophen, arsenic and carbon tetrachloride. A severe hepatitis of this type may temporarily mimic the obstructive jaundice due to carcinoma of the head of the pancreas even though the icterus is considerably higher than in carcinoma. The main difficulty is that urobilinogen may be absent from the urine in a severe hepatitis for seven to ten days but it returns with the improvement of the patient. There is evidence of disturbed renal function

with nitrogen retention. The liver function tests are abnormal. A history of the etiological agent is helpful.

(c) Catarrhal Jaundice. A prolonged episode of catarrhal jaundice is sometimes difficult to differenti- ate from carcinoma of the head of the pancreas. But the presence of urobilinogen in the urine, the dimi- nution in liver function, and the absence of X-ray findings should offer adequate differential diagnosis.

(3) Malignant disease of the liver. This is more commonly due to metastatic malignancy t h a n to pr imary malignancy. The location of the primary lesion by clinical measures and X-ray is important. When the pr imary lesion is in the stomach, especially on the greater curvature, the diagnosis may be difficult but study of the gastric contents and careful X-ray examination should make the distinction. Pr imary carcinoma of the gall bladder occurs usually in an older group of people, there is a greater incidence of cholelithiasis and the liver is very nodular.

DISCUSSION

Carcinoma of the head of the pancreas should be considered in cases where malignancy is suspected but the gastro-intestinal X-ray is negative or dubious. A change in the bowel status in the presence of abdomi- nal pain can be a lead. Weight loss is often marked. Jaundice is not always present. An enlarged liver and an enlarged gall bladder are the most common abdomi- nal findings. An achlorhydria is present in about 25 per cent of cases. The icterus when present is usually not of the severe type and it can be intermittent as well as progressive. The presence of sugar in the urine or an elevated blood sugar may be a clue in an obscure case. X-ray findings are present in about half the cases but these probably occur in well-advanced cases. Feldman (9) has described as an early change in carcinoma of the head of the pancreas a filling de- fect, shaped like an inverted three, in the ampullary portion of the duodenum. Studies of the pancreatic juice by the method of Diamond (8) should prove helpful. Operative procedures have been mainly pallia- tive---relieving the biliary obstruction and at times the intestinal obstruction. A very few cases of suc- cessful resection of the growth have been mentioned in the literature (11, 12).

SUMMARY

Forty cases of carcinoma of the head of the pancreas have been reviewed in an attempt to picture a clinical story that would lead to earlier diagnosis. Our hope in this disease rests on earlier diagnosis with perhaps more courage in earlier exploratory laparotomy in the suspected case. This would allow a greater percentage of resections ir~ an attempt to cure the disease.

REFERENCES 1. Graham, E. A, : Surg. Diag. W. B. Saunders Co., 3:471, 1930. 2. Levin, N. L . : P r imary Carcinoma of the Pancreas. Am. J .

Cancer, 18:852, Aug., 1933. 3. Eusterman, G. B. and Wilbur, D. L. : P r i m a r y Malignant

Neoplasm of the Pancreas. South. Msd. J. , 26:875, Oct., 1933. 4. Eusterman, G. B. and Balfour, D. C.: The Stomach and Duo-

denum. W. B. Saunders Co., 121, 1936. 5, White, F . : Galactose Tolerance and Urobilinogcn Tests in the

Differential Diagnosis of Painless Jaundice. Am. J . Dig. Dis. and Nutr i t . , 4:315, July, 1937.

G. Sparkman, R.: Studies of Urobilin. Arch. In t . Med., 63:872, May, 1933.

7. O'Hage, J . : Pr imary Carcinoma of the Pancreas. Minn. Med., 22:298, May, 1939.

8. Diamond, J . S. e t al: The Use of Secretin as a Clinical Test of Pancreatic Function. Am. J . Dig. DIS., 6:366, Aug., 1939.

9. Feldman, M. : The Roentgen Diagnosis of Early Enlargement of the Head of the Pancreas. Am. J . Dig. Dis., 6:237, June, 1939.

10. McGee, L. C.: Epitheliod Carcinoma of the Pancreas with Duo- denal Hemorrhage. Am. J. Clln. Path., 6:371, July, 1936.

11. Whiplale, A. O.: Surgical Trea tment of Carcinoma of Ampunary Region and I-Iead of Pancreas. Am. J. Surg., 40:260, 199S.

12. Crile, G., J r . : Successful Resection of Head of Pancreas for Carcinoma: Report of Case. Cleveland clin. Quart., 5:250, 1938.