TREATMENT OF CARCINOMA OF THE AMPULLA OF VATER

TREATMENT OF CARCINOMA OF THE AMPULLA OF VATERALLEN 0. WHIPPLE, M.D., WiLLIAM BARCLAY PARSONS, M.D.,

AND CLINTON R. MULLINS, M.D.NEW YORK, N. Y

FROM THE DEPARMENT OF SURGERY, COLUMBIA UNIVESITY

A DEFINITE advance has been made in the last five or six years in thesurgery of the pancreas, notably in the brilliant cures of hyperinsulinism bythe removal of adenomata and more recently by the excision of a large partof the pancreas itself. These cures have been reported by American surgeonswith no mortality in the series thus far published.

Because of a technique used in successful operations on a number ofadenomata, cyst-adenomata and benign cysts of the pancreas, we becameinterested again in attacking the problem of malignancy of the pancreas andperi-ampullar region. The fact that a large part of the pancreas can besafely excised, and that in animals the main ducts can be ligated as theyenter the duodenum without permanent damage to their well being, suggesteda new approach to the problem. This was impressed upon us after an un-successful attempt to remove a carcinoma of the ampulla of Vater by thetransduodenal method.

Involved in the previous attempts at radical removal of carcinoma of thepapilla of Vater and the head of the pancreas are certain factors which havecompromised its success and made it such a hazardous procedure as to beprohibitive in the minds of even the ablest surgeons. The first of these wasthe mistaken belief that the flow of pancreatic juice is essential to life, whichled surgeons to attempt to reestablish this flow into the duodenum or jejunumby implanting the resected head of the pancreas or the cut end of the ductinto the upper intestine. The activation of the pancreatic ferments by duo-denal contents compromised any type of anastomosis in the human subject,especially around the posterior aspect of the duodenum devoid of peritoneum.Experimentally the operation of pancreato-enterostomy has been done suc-cessfully in dogs by Sauve,' Coffey,2 and Sweet.3 However, in dogs thepancreas is mobile and covered by peritoneum, quite in contrast to man.There are four case reports in the literature (Halsted, Hirschel, Kausch,Tenani4) in which this operation of reimplantation of the head of the pancreasor the duct was successfully carried out after a resection of the carcinoma ofthe ampulla. Two others reported by Koerte and Mayo-Robson died fol-lowing this procedure. Our patient operated upon in this manner died fromleakage around the anastomosis. We can form no idea of the number of un-published cases in which it has been tried and resulted fatally, but it hasseemed to us, because of the digestive action of the pancreatic ferments, thatthe hazard of this operation is too great to be advocated.

The second factor was the attempt to carry out the excision of these763

WHIPPLE, PARSONS AND MULLINS Annals of SurgeryOctober. 1935

tumors in oiie stage, wlhatever the nmethod used. The victimiis of these tumorsare as a rule deeply jaundliced, have a hemorrlhagic diathesis, are depleted,undernourished, astlieniic, and(I lhave severe liver damlage. The majority ofthese patients cannot survive such a major operation until the associatedconditions have been relieved. This factor has been recognized in recentyears, and a preliminary short-circuiting operation to relieve jaundice hasbeen carried out.

As a result of mistakes which were made in operating on our first twocases we arrived at a technic for radical removal for carcinoma of the papillaor the peri-ampullar region embodying the following principles:

i-That after resecting the descending limb of the duodenum with thepancreas wide of the growth no attempt should be made to reestablish thecontinuity of the duodenum or of the pancreas with the intestine.

tX ~Hepatic\

Common $DJGastroduodenal

FIG. I.-Normal anatomy of biliary tract and vessels in the operative field.

2 That the operation be a two stage one with definite steps in each stageand with the following technic:

(a) Spinal pantocaine anesthesia.(b) A right rectus or epigastric midline incision for the first stage, with

the following steps:(i) Posterior gastro-enterostomy.(2) Ligation and section of the common duct below the cystic duct after

determining the patency of the cystic duct, and leaving a long black silk liga-ture as an indicator on the lower stump of the sectioned common duct. Thegreat difficulty in finding the unsectioned duct in our first case and in divid-ing it at a sufficiently high level impelled us to divide the duct at the firstprocedure and to mark it with a clear signal.

(3) A cholecystgastrostomy to the anterior surface of the stomach wellaway from the pylorus, using an anastomotic opening at least 2 cm. in diame-ter, in order to avoid subsequent stenosis and cholangitis.

3-Three to four weeks after the first stage the second procedure is carried764

Volume 102 CARCINOMA OF AMPULLA OF VATERNumber 4

out through a transverse incision above the umbilicus through both recti, ifnecessary, under spinal anesthesia. The steps in this second stage are asfollows:

FIRST OPERATION

SECOND OPERATION END STAGE

FIG. 2.-Consecutive steps in the two stages of the operation.

(a) Ligation of the pancreaticoduodenlal aind gastroduodenial arteries.(b) Resection of the descending portion of the duodenum with inversion

of the upper and lower ends, and a V-shaped excision of the head of the pan-

765

Gastroduo- -denal A.

RefroperitonealArea to be drained.

WHIPPLE, PARSONS AND MULLINS Annals of Su1g9°y

creas wide of the growth together with the common duct, using the silk liga-tures as a guide to the lower cut end of the duct.

(c) A ligation of the cut end of the duct of Wirsung-and the duct ofSantorini, if present-and the suturing of the two cut surfaces with inter-rupted fine silk.

(d) Drainage of the bed of the resected duodenum with cigarette drain.4-Silk technic for both the first and second stages of the operation, using

the finest silk for all but the large arteries.We report the three cases subjected to radical removal of carcinoma of

the papilla of Vater. The literature on the operative treatment of carcinomaof the peri-ampullar region of the duodenum was completely summarized byCohen and Colp4 in I927. Shorter reviews have also been published byFulde5 and Lauwers.8 Cohen and Colp were able to collect 59 cases treatedby radical operation-one was not excised, but had radium applied locally.Of these, 24 died following operation; I3 died of recurrence; one died ninemonths later of cholangitis; and 2I or 35.6 per cent survived the operationand lived for varying periods, Kelly's case being well after eight and a halfyears, and Koerte's case for 22 years.

TABLE I

Cases Reported from 1927 to 1935

Year Author No. Age Sex Diagnosis Operation ResultCases

1925 Homans I ? ? Carcinoma of Transduodenal removal with cautery Well afterampulla "several years"

1927 Fulde I 46 M. Carcinoma of Transduodenalremovalwithreimplan- Well after 2 yrs.ampulla tation ofcommon and pancreatic ducts

1927 Clar I ? ? Carcinoma of Transduodenalremovalwithreimplan- Well at end of(Pamperl) ampulla tation ofcommon and pancreatic ducts 5 yrs.

1928 Busch I 53 M. Carcinoma of Transduodenalremovalwithreimplan- Well after I yr.ampulla tation of common and pancreatic ducts

1928 Del Valle§ I 42 M. Carcinoma of Transduodenal removal, choledochot- Operative re-ampulla omy, cholecystectomy and appendec- covery

tomy1929 Klinkert (7) 2 53 M. Carcinoma of ist stage: Cholecystjejunostomy. Died after 3

ampulla 2nd stage: Transduodenal removal and mos. with livergastro-enterostomy metastases

Pollet's case (8)# ? ? Carcinoma of Ist stage: Cholecystjejunostomy. 2nd Operative re-ampulla stage: Transduodenal removal with recovery

implantation of common and pancre-atic ducts

2930 Bengolca I 37 F. Carcinoma of Excision through choledochotomy Reoperation 3ampulla opening and drainage of common duct mos. later for

jaundice. Me-tasis in locallumph node

2932 Walters I 5o M. Carcinoma of Transduodenal removal of papilla and Well after 2ampulla reimplantation of common duct into mos.

duodenum2932 JU(l(l I 38 M. Carcinoma of Transduodenal removal 6 wks. after Recurrence.

ampulla cholecystostomy and choledochotomy Lived over 21yrs.

I932 Pemberton* I 44 M. Carcinoma of Transduodenal removal, choledocho- Recurrence.ampulla dtiodenostomy, and reimplantation of Lived 2y yrs.

pancreatic duct into duodenum

766


TABLE I (Continued)

Year Author No. Age Sex Diagnosis Operation Result

1933 Potter I 57 F. Carcinoma of Transduodenalremovalwithreimplan- Operative re-ampulla tation of common and pancreatic ducts covery.

into duodenumI933 Cabott I 35 M. Carcinoma of Resection of lower common duct with Well for 8 yrs.

lower com- choledochoduodenostomymon duct

1933 Collert I 64 M. Carcinoma of Resection of lower end of common Postoperativelower com- duct deathmon duct

1933 Lauwers 2 51 M. Carcinoma of Transduodenal removal with cautery. Well after 3 yrs.ampulla Cholecystjejunostomy and I0 mos.

52 M. Carcinoma of Transduodenal removal with cautery. Well after 9ampulla Cholecystjejunostomy mos.

1934 Santero 2 50 M. Carcinoma of Transduodenal removal Both recurredampulla soon

72 F. Carcinoma of Transduodenal removalampulla

1935 Whipple I 6o F. Carcinoma of ist stage: Choledochoduodenostomy Postoperativeampulla and cholecystostomy. 2nd stage: Re- death from

section of ampulla and adjacent pan- duodenal leak-creas with pancreatoduodenostomy age

1935 Parsons I 53 M. Carcinoma of Ist stage: Cholecystgastrostomy. 2nd Died in 8 mos.ampulla stage: Resection ofduodenum andhead of cholangitis

of pancreas with closure of pancreaticstump. End-to-end suture of duode-num. 3rd op. gastro-enterostomy

1935 Whipple I 49 M. Carcinoma of ist stage: Gastro-enterostomy, liga- Well after 3ampulla tion and division of common duct, and mos.

cholecyst-gastrostomy. 2nd stage: Re-section of duodenum and head of pan-creas with closure of pancreatic stump

1935 Janest I ? M. Carcinoma of Ist stage: Cholecystgastrostomy. 2nd Postoperativeampulla stage: (3 wks. later) Resection of du- death from

odenum and adjacentpancreas, closure pneumonia.of pancreatic stump, ligation of com-mon duct, and gastro-enterostomy

* Reported by Walters.t Reported by Potter.t Personal communication.Schofield's case not included-treated by radium (I3).

§ Reported by Llambias, et at.# Reported by Klinkert.

TABLE II

All Reported Cases

Operation No. One Died Lived Two Died LivedCases Stage Stage

Transduodenal excision. . 65 60 22 38 5 0 5Resection of duodenum.. 6 3 I 2 3 I 2Retroduodenal excision. . 3 3 I 2Resection ofcommon duct 3 2 I I I 0 IResection of duodenumand pancreas with clo-sureofpancreaticstump 3 3 I 2

Total ... 80 68 26 43 12 2 10

(38%) (i6.6%)767

WHIPPLE, PARSONS AND MULLINS AnnalsofSurgerY

The reported cases since 1927 are listed in Table I, giving a total of 22cases including the three here )resented. In this group there were threepostoperative deaths, and six are known to have had recurrence of the tumor.Thus out of a total of 8o cases, 28 or 35 per cent died an operative death, and20 others, or 24.6 per cent, are known to have had recurrence. It is probablethat many of the remaining 34 cases also died from cancer, since only ashort follow up is available in the majority.

TABLE III

Presbyterian Hospital Cases with Autopsy

Case Age Sex Autopsy Diagnosis Metastases

9747 82 M. Carcinoma of ampullar region (probably common ductorigin) .o...................

9850 65 M. Carcinoma of terminal common duct with invasion ofpancreas and metastases to liver ................. +

10489 65 F. Carcinoma of ampulla (probably common duct)..... 0IOIII 55 M. Carcinoma of terminal common duct with extension to

pancreas and microscopic metastasis to a local lymphnode......................................... +

IO699 67 M. Carcinoma of terminal common duct .............. oIO839 56 F. Carcinoma of peri-ampullar region of duodenum.... oIIOII 60 M. Carcinoma of peri-ampullar region of duodenum.... oTotal-M. = 5.

F. =2.With metastases = 2.Without metastases = 5.

CASE REPORTSCASE I.-W. S. Unit History 70844. C. C.-Jaundice of ten weeks' duration. Fam11ily

I-Iistory.-Irrelevant. Past History.-Partial thyroidectomy for hyperthyroidism fouryears before. Patient had suffered from symptoms of epigastric fulness for seven yearsand a duodenal diverticulum had been demonstrated by roentgenogram.

Present Illness.-For ten weeks patient has noted progressive painless jaundice withanorexia, dark urine, occasional abdominal cramps and loss of ten pounds in weight.

Physical Examination.-A white female 6o years of age, deeply jaundiced. Tempera-ture I01.4°; pulse 75; respirations, 20; blood pressure I30/85. The liver is palpable 4 cm.below the right costal margin and beneath this an indefinite mass which seems to begallbladder. The physical examination is essentially negative otherwise.

Laboratory Data: Hemoglobin, 77 per cent; red blood cells, 3,670,ooo; white bloodcells, 8,350,000 (64 per cent neutrophils). Sedimentation rate, 70 Mm. in one hour.Blood amylase, 13.4. Duodenal drainage-bile stained fluid obtained, showing no choles-terol crystals and containing pancreatic ferments. Roentgenogram of the abdomen showedno calcified shadows. Urine, bile stained. Serum bilirubin, 6.8 mg. per IOO cc.

First Operation.-March i6, I934. Choledochoduodenostomy and cholecystostomy.Pathology.-The gallbladder was distended and a small mass in the region of the ampullaof Vater could be felt. Procedure.-As described. Course.-The patient did fairly wellfollowing operation, and the jaundice cleared. The drainage tract closed in the thirdweek.

Second Operation.-May 7, 1934. Resection of ampulla of Vater and pancreato-enterostomy. Resection of the duodenal wall in the ampullar region with part of the

768


adjacent panicreas and suture of pancreas into the duodenal wall defect. The distal com-mon duct was not well vistualized. Course.-The patient went rapidly down hill and diedabout 30 hours after operationi.

Pathologic Rcport.-Adenocarcinooma of the ampulla of Vater, apparently arisingfrom the duodenal mucosa. The ampullar opening was completely closed off. Autopsy.-Showed acute localized peritonitis with fat necrosis in the region of resection, with leakagefrom the lower end of the common duct. There were no metastases.

CASE II.-E. W. Unit history 422730. C. C.-Severe itching of skin and painlessjaundice, of two months' duration. Family History.-Grandfather died of carcinoma ofthe stomach. Past History.-Usual childhood diseases. Malaria many years ago.'Blood poisoning" nine years ago following infected left elbow. Gonorrhoea ten years agowith treatment.

Present Illness.-First admission of a 53 year old Nova Scotian boat builder foritching and jaundice, two months' duration. Nausea, no vomiting. Appetite very poor,

FIG. 3 -Case I.-Low power photomicrograph showing carcinoma of ampulla of Vaterwith stenosis of the lumen.~

unable to eat fatty foods. Clay colored stools. Loss of 20 pounds in. the past threemonths.

Physical Examinationt-Temperature, 100.20 ; pulse, 98; respirations, i8; bloodpressure, 104/6o. Poorly developed and emaciated old man, who is intensely jaundiced.Eyes-sclerae jaundiced. Lungs clear. Abdomen-distended liver edge two fingersdown. No spasm or tenderness.

Laboratory Data.-Blood amylase, 5.8. Serum bilirubin, I0.7 mg. per ioo cc. Wasser-mann, negative.

First Operation.-Cholecystgastrostomy. July i8, 194 Pathology.-A moderatenumber of adhesions were present in the right upper quadrant. The gallbladder wasthickened, pale, markedly distended, containing 300 cc. of caf6 au lait bile. No stoneswere made out in the gallbladder or common duct. The head of the pancreas presented afirm enlargement, somewhat nodular in shape, firm in consistency, lying on the deep sur-face of the organ. It was too deep to warrant the removal of a specimen. No metastaseswere seen in the liver. Procedure.-15 cm. right upper rectus incision. Silk technic.

769

WHIPPLE, PARSONS AND MULLINS Annals of Surgery

Procedure as described. Course.-Uneventful. Sutures out seventh day, retention suturesout twelfth day. Wound healed by primary union. Up fifteenth day. DischargedAugust 14, feeling very well, to return in about a week for second operation.

Second Operation.-Partial pancreatectomy. Partial duodenectomy. Duodenoduo-denostomy. August 2I, I934. Pathology.-The stomach was adherent to the undersurface of the liver, so that the cholecystgastrostomy could not be seen. The commonduct was enlarged, being a full centimeter in diameter. There was a hard mass at theregion of the papilla of Vater which apparently also involved the neighboring pancreas.At least, this area of the pancreas was thickened. There was a slight amount of freefluid in the upper peritoneal cavity. Procedure.-Transverse incision across upper ab-domen. Procedure as described. Patient was transfused at the end of the operation.

Pathologic Report.-Diagnosis: carcinoma of papilla of Vater.Course.-Persistent vomiting after this operation, approximately 2,000 cc. a day, with

nothing by mouth, was a serious complication. An obstruction at the site of the du-

M~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~k

FiG. 4.-Case I.-Higher magnification showing the fairly well differentiated glands inthe tumor.

odenoduodenostomy was thought to be the cause of this, so one week later a thirdoperation was done.

Third Operation.-Anterior gastro-enterostomy. Entero-enterostomy. August 29,I935. Pathology.-There was no free fluid. No evidence of fat necrosis. Area of pre-vious operation not inspected. Procedure.-Fifteen cm. left upper rectus incision. Pro-cedure as described. Silk technic. Course.-Patient had a very good recovery consideringthe severity of his previous operations. All wounds healed well. No leakage or drainageof pancreatic ferments. Following the pancreatectomy his blood amylase rose to 72 andthen gradually fell to around 30. Blood sugar has been normal. Studies of his proteinand fat digestion since operation show about an 8o per cent protein digestion and about50 per cent fat.

Patient continued well until about March 4, when he began to have abdominal dis-comfort associated with head cold. jaundice, light colored stools, dark urine, vomiting,for the past week. No itching. Chills and fever. He was readmitted March 25, I935.

770


Physical Examination.-Skin dry and warm. Eyes-sclerae jaundiced. Lungs.clear.Heart-sounds of good quality, no murmurs. Abdomen-incisional scars with herniation incentral portion of epigastrium. Liver is felt 3 to 4 cm. below costal margin and is dif-fusely tender abdominally and on rib percussion. Abdomen soft and non-tender elsewhere,but moderately distended. Temperature, IO4.60; pulse, IOO; respirations, 24; bloodpressure, I40/85.

Laboratory Data.-Blood amylase, II.4. Serum bilirubin, 8.2 mg. per Ioo cc. Bloodsugar, 1.30 Gm/L. Blood C02, 49CC. per cent. Blood calcium, 7.4 mg. per ioo cc. Bloodphosphorus, 1.7. Blood culture positive for B. lactis aerogenes. Course.-Patient ran aspiking temperature, Ioo°-Io4°, and gradually became weaker. He succumbed on Aprili8, I935, three and one-half weeks after admission and approximately eight months afterthe radical operation.

Autopsy.-The peritoneal cavity contained about 500 cc. of yellowish fluid. Therewere firm adhesions between the stomach and the under surface of the liver. The spleen

-VAFIG. S.-Case II.-Low magnification indicating complete replacement of lumen of ampulla

of Vater by tumor tissue.

was enlarged, weighing 320 Gm., and was of uniform red color. The liver weighed ap-proximately 3,IwO Gm., and contained numerous yellowish nodules varying in size from afew millimeters to i,2 cm. across. On section these contained purulent exudate. Thehepatic ducts were dilated and the gallbladder also, the latter containing numerous softyellowish stones and some sand-like material. The common duct was dilated, and at itslower end where it had been ligated there was a small amount of firm whitish tissuewhich microscopically showed carcinoma. There were no other areas of carcinomafound. The cholecystgastrostomy opening was contracted so that it admitted only asmall probe. The pancreas was I2 cm. in length, with grayish-yellow lobules and abscesscavities scattered throughout the parenchyma. There were no connections between thepancreas and the intestinal tract. The pancreatic duct was moderately dilated. The re-mainder of the examination, including the chest cavity and the thyroid, showed nothingexcept multiple abscesses in the lungs and kidneys, the result of the septicemia. Micro-scopically the liver showed evidences of the marked infection with many abscesses. Theportal areas showed some increase of connective tissue with inflammatory cells. There

771


was no fatty degeneration noted. The pancreas showed, in addition to the multiple ab-scesses, moderate fibrosis and atrophy of the acinar tissue.

It was the pathologist's opinion that the sequence leading to death was stenosis of thecholecystgastrostomy opening, with bile stasis, infection, septicemia and death.

CASE III.-M. J. Unit History 440080. C. C.-Itching and jaundice, and clay coloredstools of two and one-half months' duration. Family History.-Negative. Past History.-Negative except for malaria i8 years ago, Gonorrhoea 20 years ago. Gas and belchingp.c. for past 20 years.

Presentt Illness.-For past two and one-half months patient has had epigastric pain,colicky, and nausea and anorexia. No vomiting. This was followed two weeks later

... ._ ,.77

FIG. 6.-Case III.-Posterior view of gross specimen consisting of duodenum and headof pancreas. Note dilated common and pancreatic ducts.

by jaundice, clay colored stools and dark urine. No pain since the first attack, whichlasted one-half hour. Had previously had belching and distress after meals for 20 years.Loss of i8 pounds in three months.

Physical Exaniination.-A well developed colored man Of 49 with a greenish tingeto his skin, and markedly icteric sclerae. Heart not enlarged, sounds regular, good qual-ity. Lungs negative. Abdomeni-liver edge is felt 6 cm. below costal margin. No signsof fluid. Examinationi otherwise niegative. Temperature, 98.80; pulse, 70; respirationis,20; blood pressure, II0/70.

Laboratory Data.-Hemoglobin, 76; red blood cells, 4,000,000; white blood cells,7,400; polymorphonuclears, 66. Bleeding and clotting time, normal. Stool negative for

772


bile or blood. Serum bilirubin, IO.9. Wassermann-Cholesterol 3+, Alcohol, negative.Repeat Wassermann negative. Duodenal drainage-no bile obtained. Sedimentationrate 65, defibrinated 3, Takata slightly positive. Urine-bile positive.

Patient was discharged to decide on operation and returned two days later, January24, I935. Blood N.P.N.-.2o Gm./L Serum bilirubin, 8.o mg. per ioo cc. Roenltgenogramof abdomen negative for stones.

First Operation.-January 25, 1935. First stage procedure for excision of duodenumand head of pancreas for carcinoma, consisting of gastro-enterostomy, cholecystgas-trostomy, and ligation of the common duct below the cystic duct.

Pathology.-The findings in this patient promised unusually well for removal of whatwas considered to be a carcinoma of the papilla. The common duct obstruction wasthought to be due to a small tumor mass about two centimeter in diameter situated in theregion of the papilla. The common duct was markedly dilated, as was the gallbladder,but no stones were made out in either the gallbladder or common duct. There was ananomaly of the cystic duct which complicated the procedure somewhat, inasmuch as liga-tion of the common duct had to be carried out at a lower level than was desired becauseof the long cystic duct emptying into the common duct in the retroduodenal portion on thecommon duct. There were no enlarged nodes made out. No masses palpable in the liver.The liver was somewhat enlarged due to the common duct obstruction.

Procedure.-A reversed L-shaped incision was made in the right upper quadrant.Procedure as described. Silk technic. Course.-Little or no reaction following opera-tion. Jaundice subsided. In four days had bile in stools and bile disappeared from urinea few days later. Serum bilirubin dropped slowly from eight milligrams before operation,and was still showing a trace on the twenty-fifth day. On the thirteenth day resectionwas done.

Second Operation.-February 7, 1935. Resection of duodenum and head of pancreas.Spinal anesthesia. Procedure.-Transverse incision above umbilicus. Adhesions wererather numerous, but duodenum could be mobilized and turned forward to the left.After this the lower end of the common duct was isolated with some difficulty and theduodenum was resected, the point of section being just distal to the pylorus above andproximal to the superior mesenteric vessels below. A wedge-shaped portion of the headof the pancreas was then removed, so that the specimen included the ampulla of Vaterand the tumor. The duodenal ends were then inverted and the pancreatic stump closedwith interrupted silk sutures.

Course.-Little reaction following second operation, but a serous discharge on thethird day which on analysis showed pancreatic ferment. Blood amylase was 40 on theday of operation and 44 the following day-remained below 20 after this. Fasting bloodsugar rose to I.28 on the second- day, and was under IOO after this. The discharge per-sisted in moderate amounts up to the eighth day, following which there was a smallamount, apparently from the subcutaneous tissues at the site of the drain. Up on thetwelfth. Discharged on eighteenth day with a small draining wound.

Pathologic Report.-Gross.-The specimen consists of a portion of the head of thepancreas, the duodenum, and the common bile duct and pancreatic duct, removed in asecond stage operation for carcinoma of the papilla of Vater. In the fixed stage the duo-denum measures nine centimeters long on its convex surface. The serosal coat is roughand somewhat shrunken, owing to having been torn by retractors. A mass can be pal-pated through the wall in the region of the papilla.

On opening the duodenum there is a small ulceration about 6 Mm. in diameter,which is irregula-r in shape and occupies the site of the opening of the papilla into thelumen of the duodenum. No opening, however, can be made out. The tissue beneaththis is indurated and fixed to the wall.

The pancreatic tissue removed with the specimen measures 5 by 3 by 2.5 cm. and isattached to the concave surface of the duodenum in its upper portion. The common bileduct and the pancreatic duct can be seen coursing through the upper portion of the pan-

773


creatic tissue. The common bile duct, where it has been cut through, measures about7 Mm. in diameter. The pancreatic duct measures 4 Mm. in diameter. A probe insertedinito each of these in the fresh specimeni passed to the papilla but not through any openinginto the duodenum.

On section the two ducts are seeni to course side by side towards the region of thepapilla, and just below the papilla there is a firm, grayish-white, ovoid mass measuring1.3 by .8 by i cm. which appears to have not only involved the papilla of Vater but hascompletely destroyed the lower end of the pancreatic duct and bulges in toward the lumenof the lower end of the common duct, invading the wall of both of these and blocking theopening completely.

Situated on the upper anterior inner wall of the duodenum, 6 Mm. from the upperline of resection, is a small papillary shaped nodule measuring 3 Mm. in diameter and

FIG. 7.-Higher magnification of tumor from Case III, showing the car-cinoma cells.

raised above the mucosal surface i Mm. This has two tiny openings or depressionswithin it. A section of this, however, reveals no duct-like structure under it or any con-nection with the pancreas which might suggest an accessory pancreatic duct.

On the posterior aspect of the duodenum and pancreas six lymph nodes were found,the largest measuring 8 by 5 by 4 Mm. All of these are soft except one which is sit-uated immediately behind the tumor and firmly adherent to it, and is harder and filledwith white tissue, and appears to have been involved by direct extension.

Microscopic.-A section has been taken longitudinally through the common duct,pancreatic duct, ampulla, and includes a portion of the pancreas. The common ducthistologically appears patent, but is obstructed by the tumor mass which involves theentire lower end of the pancreatic duct and adjacent duodenum where it joins the loweraspect of the papilla of Vater. The tumor is a carcinoma made up of irregular, infil-trating glands and strands of small hyperchromatic epithelial cells which show mitoses onan average of i per high power field. These cells secrete large amount-s of mucin in

774


some areas. Both sides of the lower end of the common duct at the ampulla are involvedwith the tumor tissue. The surface of the tumor at the ampulla is ulcerated and coveredwith inflammatory cell debris. The tumor does not extend upwards along either thecommon or pancreatic ducts. No direct extension can be seen through the duodenal wallinto the pancreas.

A section through the small nodule at the upper line of resection of the duodenumshows this to be composed of an adenomatous mass of duodenal mucous glands which areperfectly regular in shape and only show a small amount of epithelial hyperplasia. A fewdeep crypts can be seen presenting on the lumen side, but no real ducts, which mightbe considered an accessory pancreatic duct.

Sections of all the lymph nodes show one to be definitely involved with a smallmetastasis. In addition, the firm white node adherent to the deep surface of the tumorshows carcinoma bordering it, but not actually invading, and extending completely throughthe adjacent issue. This would indicate that some tumor was left behind at the time ofoperation. The prognosis, owing to this last finding and owing to the metastasis, mustbe considered unfavorable. Diagnosis.-Carcinoma of ampulla of Vater with metastasesto retroperitoneal lymph nodes.

Second Admnssion, March 25, I935. Dicharged April I3, I935. The patient was welluntil March 25, I935, about two months after operation, when he was seized with suddensevere epigastric pain which became steadily worse, and was accompanied by a bilestained vomiting. He was readmitted, at which time his temperature was normal. Hehad upper abdominal spasm and tenderness. White- blood cells, 21,900; polymorpho-nuclears, 92 per cent; blood amylase, 72. Examination of the stools at this admission forfat absorption showed that he was utilizing 88 to go per cent of his fat intake. Bloodcalcium and phosphorus were normal.

Third Operation.-March 25, I935. Exploratory celiotomy. The patient was ex-plored under spinal anesthesia. In the region of the right upper quadrant there was asmall pocket containing 20 to 30 cc. of yellowish fluid, the culture of which showed B. coliand non-hemolytic streptococcus. The tissues in this area were very edematous, and al-though no fat necrosis was present, it was felt that the patient had a mild acute pan-creatitis. This area was drained, and the patient made an uneventful recovery.

Course.-The fistula drained what apparently was pancreatic juice (since it containedall the ferments), for about two weeks, and then closed. The patient was discharged ingood condition April I3, and has remained well since that time.

DISCUSSION

The operation proposed in this paper has certain advantages over thosepreviously described.

(i) It carries out the principle of cancer surgery in excising, en bloc,tissues wide of the growth. In this particular condition this principle is ap-plicable because extension of the growth from the papilla has been found tobe either into the neighboring pancreas or up along the lining of the com-mon duct.

(2) The hazard of surgery is reduced by a two stage procedure, insuringthe relief of jaundice and improvement of nutrition by the first stage andexcising the lesion radically when the risk to the patient is minimized. Inour first case of radical removal, duodenal obstruction appeared later becauseof the drag set up against the superior mesenteric vessels. We thereforehave inserted a gastro-enterostomy as part of the first stage.

There are certain theoretical objections to the removal of a large part ofthe duodenum and permanent obstruction of the pancreatic ducts.

775

WHIPPLE, PARSONS AND MULLINS Annals of SurgeryOctober, 1935

Hershey and Soskin9 and Berg and Zucker'0 noted fatty degeneration ofthe liver in animals with ligation of pancreatic ducts, and with total pan-createctomy.

Sweet3 described a peculiar translucence and atrophy of the spleen andthyroid in the dog after obstruction of the external pancreatic secretion. Inour patient coming to autopsy, because of a cholangitis eight months afterthe radical operation, none of the above changes was noted in liver, thyroid,or spleen.

The objection to this procedure on the basis of disturbed fat digestiondue to lack of pancreatic ferments was considered, but much to our surprisethe ultilization of fat in both of our patients several weeks after the radicalprocedure was remarkably good, 85 to go per cent of the fat intake was ab-sorbed, as shown by stool examinations. Furthermore, both patients gainedand maintained their normal weight three and seven months after operation.

That there may be a compensatory secretion of fat splitting ferment inthe upper intestine is suggested, but requires further experimental studybefore any positive statement can be made.A certain amount of atrophy of the acinar tissue of the pancreas has al-

ready taken place in these patients before they come to operation, due to theobstruction of the pancreatic duct by the neoplasm. This probably accountsfor the small amount of leakage of pancreatic juice after this operation.

REFERENCESSauve, L.: Des Pancreatectomies. Rev. de Chir., vol. 37, p. II3, I908.Coffey, R. C.: Pancreato-Enterostomy and Pancreatectomy. Am. Surg., vol. 50, p.

1238, I 909.Sweet, J. E.: The Surgery of the Pancreas. J. B. Lippincott Co., Philadelphia, Pa.,

I9I5.

Cohen, I., and Colp, R.: Cancer of the Peri-ampullary Region of the Duodenum.Surg., Gynec., and Obst., vol. 45, p. 332, I927.

Fulde, E.: Zentralbl. f. Chir., vol. 54, p. I48i, I927.Lauwers, E.: J. de Chir., vol. 42, p. 833, I933.Klinkert: Nederlandsch Tijdschrift voor Geneesk., vol. 732a, p. 443. Sept. 2I, 1929.Klinkert: Geneesk-gids., No. 2, vol. 7, p. 789, 1929.

'Hershey, J. M., and Soskin, S.: Am. Jour. Physiol., vol. 98, p. 74, 1931.Berg, B. N., and Zucker, T. F.: Proc. Soc. Exp. Biol. and Med., vol. 29, p. 68, 1931.

DIscUSSION.-DR. DAMON A. B. PFEIFFER (Philadelphia, Pa.) .-I havelittle in the way of personal experience to add. It is very fitting, however, thatthis Society should recognize the importance of this contribution because itis quite apparent that a new and brilliant chapter is being added to the surgeryof this condition.

None of us is too young to be aware of the timidity with which surgeonsin general have approached any radical interference on the pancreas. It istrue that in I884 Billroth, that great pioneer and innovator, reported com-plete removal of the pancreas with success from an immediate operativestandpoint, but his lead found few followers. A few years later, Mikulicz,then possessing the greatest authority in the surgical world, pointed out theenormous increase in mortality entailed in operations or removal of thestomach when the pancreas itself was in any way included.

Gradually, however, as a result of isolated experiences, it became apparent776

V'olume 102 CARCINOMA OF AMPULLA OF VATERNuimber 4

that some surgery, at least, could be done upon the pancreas, that limited re-sections could be made safely, and Finney, in this country, was one of thefirst to draw our attention to the feasibility of such procedures. He reportedbefore this Association in I9IO a case of resection of the body of the pancreasfor benign tumor. The two divided ends were reunited and the result wasentirely satisfactory.

Following this, sporadic cases of limited resection for various indicationsappeared in the literature and surgeons in general began to lose their fear ofoperative procedures on this organ.

The dangers of interfering with the head, however, deterred almost every-body from attempting radical intervention in that location. It is true Desjar-dins, in I897, devised a very ingenious operation that was satisfactory me-chanically for the removal of the head of the pancreas. It involved threeseparate anastomoses, one to restore continuity of the gastro-intestinal tract,one to reintroduce the bile into the intestine, and another to preserve the ex-ternal secretion of the pancreas, which was at that time thought to be essentialto life.

Naturally, such procedures are hardly feasible in cases as met clinicallyand Finney, in his paper, sounded a very pessimistic note as to the possibilityof ever being able to carry it out. However, this new contribution showsthat by proper combination of procedures it is entirely feasible to resect theampulla, which means resection of the head of the pancreas. By a little ex-tension of our imagination Inow we can see ourselves removing early car-cinomata from the head of the pancreas if we are fortunate enough to getand recognize them in time.

The difficulties of recognition of this condition in the early operable stageare great. The success of this procedure depends apparently upon recognitionof two things, first of all, the importance of separating the operation into twostages. That important military maxim of "divide and conquer" has been ap-l)lied successfully now in many situations and here apparently it gets a newapplication. It necessitates, however, an absolute diagram before embarkingu1)on the plan.

The second important thing to appreciate is that the external secretion ofthe pancreas is not essential to digestion and preservation of life. We mighthave deduced this from experiment and proved it to be true. As a matter offact, I think pathologists were aware of it before clinicians. I recall in mlycourse of pathology at Johns Hopkins, that Opie had somewhere obtaineda specimen of a pancreas at the autopsy table, which consisted of nothingmore than islands of lIangerhans imbedded in fibrous tissue. All the glandu-lar elements of external secretion had disappeared, but the patient had livedfor a considerable time. It remained for later work to show it is possible toavail ourselves of this fact clinically.

The difficulty of diagnosis to which I refer may be illustrated by a casewhich I had only recently, a man whose history and condition were in everyway typical of stone in the common duct. At operation he had a small thickgallbladder containing stones and a hugely dilated common duct. I opene(dhis common duct and found no stone, no sand, no indication of any foreignmaterial in the duct.

\Vith the greatest difficulty, I finally succeeded in introducing a probe intothe (luodeniuimi, (lilated the opening and felt I was (lealing with stricture ofthe papilla of Vater, put in a T tube, and stopped. He made a good recoveryand bile at first passed inlto his intestinie. After removal of the tube, however,his internal damage persisted and bile finally ceased to flow into his duodenum.It was apparent he had complete obstruction in his common duct. I felt pos-

777

WHIPPLE, PARSONS AND MULLINS Annals of Surgery

sibly I had overlooked a stone. I again explored his common duct, which atthat time was big enough so I could introduce my finger into the distal por-tion of the duct, and came to a dead end. I could feel it perfectly well. Therewas no suggestion of tumor formation. There was no stone, but simply adead end at the intestinal extremity. I made a choledochoduodenostomy andthe patient promptly got well and has remained so.

I had never met with a similar case, but Ellsworth Eliot referred me to areport by Herferth some years before in which he had found at autopsy whathe called prolapse and intussusception of the common duct. It was an entirelynew entity to me, but it seems to fit this case. In this condition the end ofthe common duct had bulged into the duodenum, reversed itself and mechani-cally completed the obstruction of the strictured papilla.

My admiration for the surgeons who carried out this very successful op-eration is not only for the ingenious, brilliant operative procedure which theyhave devised, but also for their boldness in making this a positive diagnosis inthis early stage.

DR. EVARTS A. GRAHAM (St. Louis, Mo.) .-I wish to compliment veryhighly Doctors Whipple, Parsons and Mullins for their brilliant work. It isanother example of the step-like progress which has to be made before afinal goal is reached.

I wish to discuss for a moment the question of the importance of the pan-creatic juice in digestion.

There was no evidence so far as I know that in a rapidly growing youngsubject it is possible to carry on a normal development without much pan-creatic juice until suddenly an emergency which confronted me approximatelytwo years ago seems to have established the fact. I operated on a baby thatwas less than two years of age because of intractible hypoglycemia whichretarded it greatly, both mentally and physically. We hoped to find a pan-creatic adenoma at operation, but the pancreas looked perfectly normal. Inview of the previous partial resections of the pancreas which had been madeby Judd and Finney among others, in the absence of finding a definite ade-noma, I decided that it might be worth while to make as nearly a total resec-tion of the pancreas in this youngster as was possible, without jeopardizingthe common bile duct.

The procedure was carried out successfully; the case was reported inSeptember, I934. One month ago I had a letter from this little girl's mother,which stated that the child had developed in a perfect manner physically, andthat there seemed to have been no difficulty, whatever, with her digestion.At operation I performed a subtotal removal of the pancreas, leaving only alittle strip of tissue along the common bile duct to protect it.

The operation, therefore, described by Doctors Whipple, Parsons, andMullins, I feel is based on what seems to me to be thoroughly sound physio-logic evidence of the fact that, so far as the secretions of the pancreas areconcerned, it is a perfectly safe procedure to carry this operation out, espe-cially in view of the fact that if a rapidly growing child can apparently getalong perfectly well without the external secretion of the pancreas, certainlyan adult ought to be able to do so. There is another point, however, whichdeserves some comment. What assurance have we that a patient can goalong indefinitely with a cholecystogastrostomy or a cholecystoduodenostomywithout developing a severe infection of the liver? The evidence with whichI am familiar indicates that almost invariably the patient succumbs frommultiple abscesses in the liver if the anastomosis remains open. It might bepossible, however, to devise an anastomosis which would prevent the ascend-

778

Volume 102Number 4 CARCINOMA OF AMPULLA OF VATER

ing infection, in a manner perhaps analogous to the Coffey operation oftransplantatioin of a ureter.

DR. WILLIAM PARSONS (New York).-Dr. Eliasoil is about to putblislhan article in which lhe discusses the question of cholecystogastrostomy. In ourpatient who died, we feel beyond a doubt his septicemia was the same typethat so many cholecystogastrostomies will show later on.

Doctor Eliason quotes Bernard in an article showing that about 25 percent of a rather large series of cholecystogastrostomies, performed for avariety of conditions, were followed by cholangitis. In this particular case,although we tried to make a large opening, the gallbladder was so enormousI think we did not estimate the degree to which it was going to contract lateron. At postmortem the cholecystogastrostomy opening admitted only a probeand it was quite obvious that this was the cause of the had result.

779

Documents

TREATMENT OF CARCINOMA OF THE AMPULLA OF VATER