Vesico-intestinal fistula—Actual and incipient

VESICOdNTESTINAL

FISTULA-ACTUAL

AND INCIPIENT

EARLY DIAGNOSIS AND TREATMENT JOSEPH A.

LAZARUS, M.D. AND MORRIS

S. MARKS,

M.D.

NEW YORK, NEW YORK

T

HE subject of vesico-intestina1 fistula has been frequentIy discussed in medica Iiterature from a variety of angIes. It shaI1 be our purpose to treat this subject from the uroIogica1 aspect, particuIarIy with the view of emphasizing its earIy recognition in uroIogic diagnosis, and suggesting a few additiona sideIights in treatment. FistuIous communications between the bIadder and the various segments of the intestina1 tract are not uncommon, but stiI1 suffIcientIy infrequent to warrant comment. Higgins, for exampIe, in 1936 was abIe to coIIect 583 cases, incIuding thirtyfive from the CIeveIand Clinic. KeIIogg, in 1938, reported 592 cases. Yet BaIch, discussing Higgins’ paper, stated that he was onIy abIe to find nineteen cases in going through the fiIes of the record rooms of six IndianapoIis hospitaIs having a tota of I ,738 beds. TYPES OF VESICO-INTESTINAL FISTULA Higgins’ grouping of cases is both simple and practica1. He groups them as: (I) .traumatic, (2) atraumatic or inflammatory, (3) tumors and (4) congenita1. For our purpose we have found it somewhat more convenient to group the cases under two main heads, nameIy, (I) fistuIas arising primariIy from the bIadder and (2) fistuIas arising primarily from the intestine. AIthough this CIassification faiIs to take into account the group of traumatic cases as we11 as the smaI1 number of congenita1 cases, it is more pertinent to the text of this communication to Iimit our remarks to the groups of inflammatory and neopIastic Iesions, since it is with these cases that the uroIogist is most apt to dea1.

FistuIas may be actua1 or incipient. ActuaI or fuIIy deveIoped fistuIas may occur in two forms: (a) direct communications between the bIadder and bower through short straight sinuses, and (6) Iong narrow tortuous sinuses connecting both viscera. The former is usuaIIy seen in cases due to carcinoma and tubercuIosis, whiIe the Iatter occurs most frequentIy in cases resuIting from infections. MaIes are more frequently affected than femaIes in the proportion of 3 to I (PascaI). The reason for this discrepancy has been attributed to the intervention of the uterus between the bIadder and the intestine, and to the prevaIence of diverticuIitis of the sigmoid in the male. The greatest incidence is between the ages of fifty and sixty. A carefu1 anaIysis of the cases coIIected by Higgins, KeIIogg and BaIch seems to indicate that inflammatory Iesions of the bowe1 constitute the underIying cause of the f%tuIas in the majority of cases (51 per cent), whiIe tumors of the bowe1 were the underIying cause in 21.3 per cent. The bIadder was the primary site of the lesion in I 1.6 per cent of cases. Higgins’ statistics seem to indicate that vesical caIcuIi constituted the most frequent cause of fistula formation in the bIadder group (39.3 per cent), whiIe carcinoma ranked second with 30 per cent. VesicaI diverticuIa were found in 24.2 per cent of the cases. An anaIysis of the inflammatory intestina group showed diverticulitis to have been the primary cause in 65.8 per cent, tubercuIosis was found in 17.8 per cent and appendicitis was present fourteen times (9.5 per cent) (Higgins). In KeIIogg’s series appendicitis was present in only 4 per cent of cases, whiIe the incidence of

526

NEW SERIES VOL. LIX, No. 3

Lazarus,

Marks-FistuIa

appendicitis in BaIch’s group was 2.5 per cent. In Higgins’ series the most frequent neopIastic Iesion of the bowe1 was carcinoma of the rectum (50.8 per cent), as compared with I 1.8 per cent in BaIch’s of the sigmoid was group. Carcinoma present in 45.9 per cent of the cases in Higgins’ coIIection. As regards the Iocation of the fistuIa, the rectum and bIadder were invoIved in 43.5 per cent (Higgins); 47.4 per cent (BaIch); and in onIy 16 per cent (KeIIogg). A fistuIous communication between bIadder and sigmoid was present in 23.8 per cent (Higgins), 26.3 per cent (BaIch) and 63 per cent (KeIIogg). The incidence of iIeovesical fistuIa was 4. I per cent (Higgins), 15.7 per cent (BaIch) and 4 per cent (KeIIogg). A communication between the bladder and the appendix was found fifteen times (3.4 per cent) in Higgins’ series; twice (10.6 per cent) by BaIch and in 4 per cent of the cases coIIected by noted eighty-two inKeIIogg. Higgins stances of vesicocoIonic fistuIas in a tota of 441 coIIected cases (18.6 per cent). SYMPTOMS

Actual Fistulas. The passage of gas or stoo1 through the urethra constitutes unequivoca1 evidence of a vesico-intestina1 fistuIa. AIthough the passage of urine from the rectum is aIso pathognomonic of this condition, this finding is not as readily ascertained as the former. It is, of course, obvious that coIIatera1 symptoms are present which in turn depend upon the segment of bowe1 with which the bIadder communicates. If the initia1 Iesion happens to be a carcinoma of the rectum which has subsequentIy broken into the bIadder, the presenting symptoms wiI1 be those characteristic of recta1 carcinoma, nameIy, constipation aIternating with diarrhea, bIood and mucus in stooIs and Ioss of weight. The same holds true when the initia1 Iesion is diverticuIitis of the sigmoid, in which event the outstanding symptoms are pain in the Ieft Iower quadrant of the

American Journal

ofSurgery

527

abdomen, with bIood, pus and mucus in the stoo1. Symptoms referabIe to the bIadder depend upon the degree of involvement present. With the actua1 passage of intestina contents into the vesica1 Iumen, acute cystitis invariabIy ensues, with urinary frequency, dysuria and urgency as outstanding symptoms. The urine contains bIood, pus and intestina1 contents. WhiIe the passage of gas in the urine is extremeIy suggestive of a vesico-intestinal comit is occasionaIIy found in munication, diabetic patients with infected urines. ShouId an ascending infection occur, the picture is that of an acute pyeIonephritis with chiIIs, fever and Iumbar pain. When the Iesion is primariIy in the bIadder, vesica1 symptoms naturaIIy predominate, and in fact wiI1 most frequentIy compIeteIy mask the intestina1 symptoms. Diarrhea with or without tenesmus is not an uncommon symptom, particuIarIy when an ampIe ffow of urine into the bowe1 is present. Incipient (Threatened) Fist&as. In the event the bIadder is the primary seat of the Iesion, it is extremeIy diffIcuIt to outIine a syndrome which is at a11 suggestive of a threatened communication with an adjacent segment of bowe1, since the bIadder Iesion must be extensive before it can Iead to a vesico-intestina1 fistuIa. We11 defined uroIogica1 symptoms are present Iong before threatening perforation occurs, so that from the uroIogica1 standpoint, one can hardIy suspect a threatened break-through to be detectabIe by changes in the genitourinary symptoms. Suspicion might be directed to this possibihty, however, by the sudden onset of intestina1 symptoms in a patient suffering from a serious vesica1 Iesion. When the Iesion originates in a segment of bowe1, an incipient or threatened breakthrough can and shouId be recognized promptIy, since by so doing serious consequences may be averted. Given a patient with carcinoma of the rectum, or diverticuIitis of the sigmoid, or termina1 iIeitis, who suddenIy deveIops acute uroIogica1 symp-

528

American Journal of Surgery

Lazarus,

Marks-FistuIa

toms, it is encumbent upon the surgeon to seek an expIanation for such symptoms. Symptoms may be extremeIy miId, such as slight increased frequency of urination or miId burning on urination; or the only suggestion of a urinary abnormality may be a varying degree of pyuria which may be constant or intermittent. DIAGNOSIS

Actual Fistulas. In cases of well defined fistulas, the diagnosis can readiIy be estabIished when patients pass urine in the stoo1, or gas and feca1 matter through the urethra. This, unfortunateIy, is not always in which event it becomes the case, necessary to resort to other diagnostic procedures. Cystoscopy usuaIIy reveaIs a marked The site of the comdegree of cystitis. munication appears as an area of edema which surrounds the actua1 point of perforation. The Iatter may or may not be seen. The stoo1 may be seen passing through the actua1 fIstuIa, or Iying free in the bIadder. Occasionally, one may cIinch the diagnosis by fiIIing the bIadder with a soIution of indigo-carmine, and watching it return through a recta1 tube or proctoscope previousIy introduced into the rectum. AIthough theoreticaIIy possibIe, it is extremeIy diffIcuIt to visuaIize the recta1 or sigmoida1 opening of the fistuIa through Instances have been the sigmoidoscope. reported in which the diagnosis was estabIished by means of the barium enema. The diagnosis of a Incipient Fistulas. threatened or incipient fistuIa can onIy be made by proper cystoscopic examinations. The earIiest sign discernibIe by cystoscopy is a smaI1 IocaIized area of redness somewhere on the posterior waI1 of the bIadder, in an otherwise norma appearing viscus. At a sIightIy Iater stage, this area reveaIs itseIf as an isIand of buIIous edema which, when the bIadder is HIed to capacity, may be compIeteIy or partiaIIy ironed out. ShortIy before perforation occurs the Iesion cIoseIy simuIates a bIadder tumor because of the presence of papiIIomatous-Iike ex-

crescences in the region of the buIIous edema. The vesica1 mucosa surrounding this Iesion may, at this time, revea1 a moderate degree of erythema. An important diagnostic feature at this stage is a history of intermittent cIouding of the urine. ShouId the urine during a cystoscopy be found turbid due to pus, the ureters shouId invariabIy be catheterized. Negative urines obtained from both kidneys, microscopicaIIy and cuIturaIIy, serve as confirmative evidence that the source of the pyuria is the bIadder, and Iends additiona1 weight to the suspicion that the bIadder Iesion is due to an incipient communication with a segment of bowe1. LocaIizing the exact site of the f%tuIous opening in the bIadder is of IittIe diagnostic importance because of the extreme mobility of the smaI1 bowe1, sigmoid, transverse coIon and appendix, and aIso because of the fact that in we11 estabIished cases the high grade of accompanying cystitis and reduced vesical capacity make IocaIization diffIcuIt and inaccurate. PROGNOSIS

Prognosis in these cases depends upon a variety of factors. The character of the primary Iesion responsible for the fistuIa is extremeIy important in appraising the uItimate prognosis. If the Iesion is a maIignant one, the outIook is naturaIIy poor. On the other hand, shouId the origina Iesion be inflammatory, such as an appendicea1 abscess, the outIook is good. The presence or absence of genitourinary compIications such as uniIatera1 or biIateraI pyeIonephritis aIso materiaIIy inffuences the prognosis, and this in turn is dependent upon the duration of the disease. In cases of threatened or incipient fistuIas, the prognosis, particuIarIy the immediate prognosis, is far better than in cases of we11 estabIished vesico-intestina1 fistuIas. Another important consideration is the age and physica status of the patient, as younger peopIe stand surgica1 intervention better than oIder ones. The presence of cachexia is not conducive to good healing.

NEWSERIESVOL.LIX, No. 3

Lazarus,

Marks-FistuIa

TREATMENT

Vesico-intestina1 fistuIa is an outright surgical disease and treatment may be paIIiative or curative. PaIIiative treatment is indicated in hopeIess cases of carcinoma of the bowel or bIadder, in which event coIostomy or cecostomy is the procedure of choice in diverting the feca1 Aow from the bIadder. PaIIiative colostomy may aIso be tried in certain cases of vesico-intestina1 f%tuIa resuIting from diverticulitis of the coIon. In these cases the coIostomy, once estabIished, must not be cIosed for many months, in order to give the invoIved bowe1 fuI1 opportunity to hea spontaneousIy. CoIostomy is aIso indicated in hopeIess cases of carcinoma of the bIadder associated with vesico-intestina1 f%tuIa. Curative treatment wiI1, of necessity, depend upon the underIying pathoIogica1 condition. In cases of threatened perforation, it is our opinion that preIiminary coIostomy is not necessary, but that one can, after proper preparation of the patient, perform Iaparotomy and direct appropriate treatment to the underIying lesion. By proper preparation of the patient, we mean at Ieast four to six days of treatment with the patient in a hospita1. He is given a high caIoric, Iow residue diet containing high vitamins and mineraIs, and thorough cIeansing of the bowe1 by daiIy use of saIine catharsis. At Ieast 2,000 cc. of 5 per cent gIucose in norma saIine is administered daiIy by sIow intravenous drip, and 4 Gm. of suIfaguanidine given by mouth every eight hours, or suIfadiazine in 15 gr. doses at four hour intervaIs. At Ieast one transfusion of 500 cc. of titrated bIood is given before operation and another transfusion immediateIy after operation. In cases of carcinoma or diverticuIitis of the sigmoid, the affected segment of bowe1 is carefuIIy dissected away from the bladder waI1 to which it is more or Iess firmIy adherent, by bIunt finger dissection, and whenever possibIe a muItipIestage Mikulicz procedure is carried out. When the Iesion is due to carcinoma of the rectum one may,

American Journal of Surgery

529

after carefuIIy dissecting the bowe1 away from the bIadder, carry out a MiIes abdominoperinea1 resection in one or two stages, depending upon the patient’s physica1 condition. The Mummary operation is aIso an exceIIent procedure in poor risks. When the attachment between the bIadder and rectum is due to inffammatory adhesions, the bIadder need not be disturbed. In cases in which the vesica1 waI1 is invaded by carcinoma, a segment of bIadder can be removed aIong with the tumor and the bIadder reconstructed over a Iarge suprapubic drainage tube. In cases of threatened vesico-intestina1 f%tuIas caused by appendicea1 abscesses, the abscesses shouId be drained and the appendicea1 stumps removed whenever possibIe. In cases of termina1 iIeitis, immediate transverse iIeocoIostomy is carried out, after compIeteIy transecting the iIeum at a sufficient distance from the inflamed segment of bowe1 in order to make certain that the anastomosis is performed upon heaIthy intestine. If it is deemed safe to compIete the operation in one stage, the segment of iIeum attached to the bIadder is carefuIIy dissected away from the bIadder and removed aIong with the cecum and ascending coIon. ShouId a two-stage procedure be considered advisabIe, the segment of iIeum attached to the bIadder is Ieft undisturbed, and removed at the time of the second stage of the operation. In the treatment of actua1 or true vesicointestina1 fistuIas, it is our beIief that a preIiminary coIostomy is necessary : (a) when actua1 stoo1 is present in the urine; (6) when pyelonephritis is present; (c) when the basic Iesion is carcinoma, either of the bIadder or of the bowe1; (d) when the bIadder is markedIy inflamed and contracted; (e) in cases of marked debiIity; and fInaIIy (f) in cases in which Iaparotomy discIoses a Iesion of the bowe1, the identity of which is not readiIy discernibIe, or in which the bowe1 is so intimateIy adherent to the bIadder that detachment is considered too hazardous. In performing such a coIostomy for diversion of the feca1

530

American JournaI of Surgery

Lazarus,

Marks-FistuIa

stream, it is best to avoid a Ioop coIostomy. A doubIe barre1 coIostomy such as advocated by Lahey or Devine seems best suited for this purpose. When the Iesion is hopeIessIy inoperabIe the colostomy becomes a permanent one. As regards the time interva1 that shouId be permitted to eIapse before attacking the Iesion, it is our opinion that it is best to err on the Iong side rather than to operate too soon. This is particuIarIy true when deaIing with inflammatory Iesions. In cases of carcinoma one can safeIy wait six to eight weeks. A pertinent probIem in this type of case is the method of handIing the opening in the bIadder waI1. We beIieve that when the opening is smaI1 it can safeIy be Ieft undisturbed, and an indweIIing catheter introduced into the bIadder through the urethra and Ieft in situ for a week or ten days. In cases of Iarger openings, we have found that as a ruIe sutures do not hoId because of the friabiIity of the bIadder waI1, and we, therefore, resort to suprapubic cystotomy with adequate drainage of the cuI-de-sac of DougIas to provide for any possible Ieakage. We have found the IocaI use of crystaIIine suIfaniIamide of inestimabIe vaIue in handIing such fistuIas. We do not hesitate using 12 Gm. of the drug in the cuI-de-sac after detaching the bowe1 from the bIadder waI1. FoIIowing operation patients are routineIy given 15 gr. of suIfadiazine at four hour intervaIs for at Ieast four to five days. Since, in most instances, the vesica1 capacity is markedIy reduced in cases of actua1 vesico-intestina1 fistuIas, we consider it advantageous to diIate the bIadder with hydrauIic pressure within five to six weeks after the wound has healed. This procedure must be cautiousIy carried out to avoid throwing too much strain upon the weakened bIadder waI1. CASE REPORTS CASE I. Mr. S. F., aged sixty-one, first seen January IO, 1939, compIained of prostatism and bilatera1 Iumbar pain of two years’ dura-

tion. Fifteen years previousIy he passed several caIcuIi. There were no gastrointestina1 symptoms. FoIIowing a compIete study of the genitourinary tract, the diagnosis made was benign hypertrophy of the prostate. A twostage suprapubic prostatectomy was performed and he was discharged from the hospita1 after twenty-one days. About twenty-one months Iater transurethral revision of the bIadder neck was performed for a smaI1 recurrent adenoma of the right Iobe of the prostate. He Ieft the hospital symptomfree seven days later. On January 31, 1942, approximateIy three months after the revision, he returned compIaining of increased urinary frequency and pyuria. At that time cystoscopy discIosed an edematous circuIar Iesion with a centra1 umbiIication situated on the posterior surface of the bIadder just above the interureteric ridge. Our impression at the time of this examination was that we were deaIing with an extravesica1 Iesion which was threatening to penetrate into the Iumen of the bladder. A carefu1 digital rectal examination was negative. Sigmoidoscopy faiIed to show the presence of a tumor, but reveaIed a sIight reddening and edema of the recta1 mucosa. Barium enema x-ray and ffuoroscopy showed the sigmoid Bexure drawn over toward the cecum, but failed to revea1 a defect. His weight was 177 pounds. When seen ten days Iater he stated that aIthough he noticed no urinary discomfort, the urine appeared very turbid at times, and at other times it was comparativeIy clear. Cystoscopy disclosed the same polypoid Iesion seen at the previous cystoscopy. Although resembIing a bIadder tumor, the Iesion was considered entireIy inflammatory. Overdistending the bIadder faiIed to compIeteIy iron out the lesion. The urine contained much pus. Two weeks Iater hydra&c distention of the bIadder through the cystoscope showed that the Iesion couId be ironed out when the bIadder was fuIIy distended, Ieaving behind a sIight noduIar thickening of the mucosa at the site of the Iesion. Both kidneys were catheterized and the specimens were found negative microscopicaIIy and cuIturaIIy. At this time he was carefuIIy questioned as to whether he had passed any feces or gas in the urine, but he did not recaI1 such an occurrence. One week Iater he returned complaining of vesica1 irritabiIity. A punch biopsy was performed and the tissue

NEW SERIES VOL. LIX,

No.3

Lazarus,

Marks-Fistula

report was “edematous, chronicaIIy inflamed bIadder waI1.” Recta1 examination was entirely negative, and there were no symptoms referable to the gastrointestina1 tract. It appeared to us then that we were deaIing with a threatened vesico-intestina1 fistula and Iaparotomy was advised and accepted. He entered the hospital March 29, 1942, for preoperative preparation which consisted of thorough cleansing of the bowel, Iow residue, high vitamin diet and sulfaguanidine 4 Gm. every eight hours. The day prior to operation he was given goo cc. of titrated bIood. Operation. (J. A. L.) On April 2, 1942, under spina anesthesia the abdomen was opened through a long left paramedian suprapubic incision extending from the symphysis to a point just above the Ievel of the umbiIicus. There was a large i&Itrating mass about the size of an aduIt’s fist situated deep in the peIvis, intimately adherent to the posterior waI1 of the bIadder. The mass consisted of the Iower sigmoid ffexure direct1y above the rectosigmoidal juncture. The Iiver and para-aortic Iymph-nodes were not involved. With the patisent in high Trendelenburg position the intestines were carefuIIy walled off with hot pads, and by carefuI digita manipulation the sigmoid was easiIy dissected away from the posterior wall of the bladder. AIthough the vesical waI1 was indurated, no opening couId be visuaIized. There was no evidence of free pus during this procedure. Owing to the Iow position of the tumor it was necessary to divide the peritonea1 reflexion on both sides of the mesentery as we11 as the peritonea1 reffexion over the bIadder in order to obtain sufficient mobility of the invoIved segment to perform a MickuIicz operation. The Ioop bearing the tumor was then easiIy deIivered from the abdomen, and after uniting both limbs and cIosing the layers of the BbdominaI waII snugly around the exteriorated loop, the tumor was removed between cIamps. The proxima1 and dista1 cIamps were left in situ, thereby compIete1y obstructing the bowe1. Three Gm. of suIfaniIamide powder were pIaced into the peIvis and around the wound. No attempt was made to reinforce the portion of the bIadder waII freed from the tumor, but an indweIIing urethra1 catheter was introduced and Ieft in situ for seven days. The patient is making an uneventfu1 convaIescence and is awaiting the crushing of the spur.

American

Journalof

Surgery

531

PathoIogicaI diagnosis of the extirpated segment of bowel was “infltrating adenocarcinoma of rectum with extension into adjacent fat tissue.” Comment. This is an exampIe of a case of threatened vesico-intestina1 fistuIa caused by carcinoma of the rectosigmoid, in which the cardinaI symptoms were entireIy vesica1, nameIy, intermittent pyuria with attacks of vesical irritabiIity. CystoscopicaIIy, there was a poIypoid edematous lesion on the posterior vesica1 waI1 simuIating a bIadder tumor, which at first couId be almost compIeteIy ironed out by hydrauIic distention of the bIadder, but Iater couId not be obIiterated. AIthough no distinct communication between the bowe1 and bIadder was found, it was diffIcuIt to expIain the intermittent attacks of pyuria without assuming the presence of some, possibIy minute, sinus between the two viscera, not Iarge enough to permit the passage of gas or stoo1, but capabIe of transmitting bacteria. A tentative diagnosis of a threatened vesico-intestina1 fistuIa was made prior to surgery. This enabIed us to institute proper preoperative preparations. It is our opinion that such preparation was greatIy responsibIe for the patient’s smooth postoperative convaIescence. CASE II. R. C., maIe, aged twenty-four, first seen December 16, 1940, compIained of hematuria and right renaI colic, dysuria and increased urinary frequency of two weeks’ duration. His past history was essentiaIIy unimportant, save for an appendectomy performed in 1937. Cystoscopy reveaIed a Iesion the size of a fifty-cent piece occupying the right posterior wall of the bIadder, which had the appearance of an irreguIar tumefaction with a centra1 umbilication, attached to the bIadder by a broad base. There was no inflammatory reaction of the vesica1 mucosa surrounding the base of the Iesion. The question that arose was whether we were deaIing with an infiItrating bIadder tumor or with a threatened vesicointestina1 fIstuIa. Both kidneys were easiIy catheterized, and the specimens from both sides showed an occasional Ieucocyte. UrinaI-

532 ysis:

American Journalof Surgery

Lazarus,

Marks-FistuIa

specific gravity 1.025; aIbumin I PIUS; sugar negative; micro-mass of white bIood ceIIs. The bIadder specimen showed Gramnegative bacihi on smear and Bacillus coIi aerogenes on cuIture. The renaI specimens were negative on smear and steriIe on cuIture. Al1 specimens were negative for the mycobacterium tuberculosis. urea nitrogen 11.5 mg. BIood Chemistry: per cent; creatinin 1.07 mg. per cent; urea cIearance 127.5 per cent. The genitourinary tract faiIed to show any abnormaIity on the Bat and pyelographic x-rays. The fohowing day biopsy was performed upon the vesica1 Iesion. This was reported inflammatory with no evidence of mahgnancy. PhysicaI examination was essentiaIIy negative, except for a large, boggy prostate which was feIt per rectum. Preoperative Diagnosis : Threatened vesicointestina1 IistuIa. The patient was sent to the hospita1 and thoroughIy prepared for 6 days prior to operation. Operation. (J. A. L.) On January g, 1941, under spinal anesthesia the abdomen was opened through a six-inch midsuprapubic incision. Starting at approximateIy six inches from the iIeoceca1 vaIve, the termina1 iIeum was intimateIy adherent to the posterior waI1 of the bIadder. At this point the ileum was anguIated and the proximal Iimb of this Ioop, which was adherent to the bIadder, was dilated. The waI1 of this portion of iIeum was very friabIe. The proxima1 Iimb of the Ioop was mobiIized at a distance of about tweIve inches from its attachment to the bladder and divided between crushing clamps, along with its mesentery. Both ends of the severed intestine were inverted with three layers of sutures. The proxima1 segment was then anastamosed side-to-side with the midportion of the transverse colon. The wound was cIosed in Iayers. On January 27, 1941, eighteen days after he was given 500 cc. of the iIeocoIostomy, titrated blood and the foIlowing day, under spina anesthesia the abdomen was opened through a Iong right rectus incision extending from the IeveI of the Costa1 arch to the symphysis. There were dense adhesions binding the terminal iIeum to the posterior bIadder wall. After carefuIIy separating this Ioop of intestine from the bIadder by bIunt dissection, a smaI1 perforation was seen in the center of an area of indurated vesica1 waI1. The affected termina1

segment of iIeum, ascending and proxima1 part of the transverse colon were resected en masse, and the end of the colon inverted with three layers of sutures. Raw surfaces were obIiterated by suturing the cut edges of the posterior parietal peritoneum. A smaI1 Penrose drain was introduced into the retroperitonea1 space and permitted to emerge from the upper angIe of the wound. An attempt to cIose the fistuIous opening into the bIadder was unsuccessful due to induration and friabiIity of the tissue. A rubber dam was introduced into the pelvis behind the bladder and permitted to emerge from the Iower angIe of the wound. The wound was cIosed in layers around drains. An indweIIing urethral catheter was introduced and Ieft in situ for seven days. FolIowing a compIeteIy uneventfu1 recovery, he was discharged from the hospitaI February 12, 1941. When Iast seen September 18, rgq.1, he was compIeteIy symptom-free and had gained fifteen pounds in weight.

Comment. This is a case of threatened vesico-intestina1 fistula due to terminaI iIeitis, in which the diagnosis was made prior to operation by cystoscopic examination. The symptoms were entireIy uroIogica1. CompIete recovery folIowed a two-stage resection and transverse iIeocoIostomy. The opening in the bIadder couId not be cIosed by suture, but healed spontaneousIy with the aid of an indweIIing urethra1 catheter. CASE III. November

G. W., maIe, aged forty, first seen

1932, compIained of severe Iower abdomina1 pain, with compIete obstipation of two days’ duration. Urinary symptoms consisted of sIight dysuria and increased urinary frequency. Preoperative diagnosis : IntestinaI obstruction. Laparotomy performed on November I, 1932, reveaIed a Iarge carcinoma involving the sigmoid, which was intimateIy adherent to the posterior vesica1 waI1. The termina1 four inches of iIeum were aIso gIued to the bIadder waI1. A transverse coIostomy was performed. Death followed one week Iater, apparentIy from severe hemorrhage in the abdomina1 wound. Autopsy showed no abnormaIity of the stomach and smaI1 intestine. A colostomy opening was present in the transverse coIon. I,

NEW SERIES VOL.. LIX. No.

3

Lazarus,

Marks-FistuIa

AmericanJournalof Surgery

533

“ bIadder epitheIium showing extensive acute and chronic inff ammation.” X-ray of tbe Genitourinary Tract. The right renaI siIhouette was much enlarged whiIe the left appeared normal. Intravenous pyeIography disclosed deIayed fiIIing of a large right hydronephrotic kidney. The left renaI peIvis showed no abnormaIity. Cystogram showed a fiIIing defect invoIving the vauIt and right haIf of the bIadder. Barium enema showed a smaI1 amount of barium fiIIing the ampuIIa and rectosigmoid. The barium appeared to pass over to the right side of the peIvis, where it seemed to fray out and blend into the adjacent tissue. Sigmoidoscopy discIosed a markedIy thickened and edematous recta1 mucosa but no Comment. This patient presented himevidence of tumor. self with symptoms of acute intestina1 obPhysicaI examination was essentiaIIy struction due to carcinoma of the sigmoid. negative. The urologica symptoms were miId, conLaboratory Findings. UrinaIysis: specific sisting of sIight dysuria and increased gravity 1.012; albumin 2 pIus; sugar negative; urinary frequency. There was no evidence microscopy-pus and bacteria. BIood chemisof stoo1 or of the passage of gas through try: urea n. 12.5 mg. per cent; creatinin I.0 the urethra. Yet at operation it was noted mg. per cent; urea clearance 136 per cent. that the tumor of the sigmoid was intiBacterioIogy : bIadder urine showed StaphymateIy adherent to the bIadder. At autopsy Iococcus aIbus on cuIture. BIood count: an actua1 fistuIa between the sigmoid and hemogIobin I I. I Gm. per 100 cc. of bIood; red bIood ceIIs 3,22~,000; white bIood ceIIs 8,250 per bIadder was found. cu. mm.; coIor index 1.2. Differential: poIyCASE IV. J. L., maIe, aged forty-three, a morphonuclear ceIIs 66 per cent; Iymphocytes I I per cent; rods 19 per cent; mononucIears resident of Honduras, CentraI America, first 4 per cent. Wassermann was 2 plus. 2 hour seen JuIy IO, 1940, compIained of passing phenoIsuIphonphthaIein 30 per cent. stoo1 and gas through the urethra. IIIness began Preoperative Diagnosis. ActuaI vesico-intessix years previousIy with diffIcuIty in voiding tina fistuIa, possibIy due to carcinoma of and the passage of stoo1 in the urine. Since sigmoid, with right infected hydronephrosis. then he had had intermittent bouts of chiIIs Operation I. (J. A. L.) On JuIy 20, 1940, and fever. He voided frequentIy day and night, under spina anesthesia the abdomen was had dysuria, and passed ribbon-like stooIs. His opened through a six-inch Ieft reversed Kamappetite was poor and boweIs costive. There merer incision. Situated in the peIvis was a Iarge was a Ioss of fifteen pounds in weight during mass consisting of rectosigmoid intimateIy the past six months. The patient admitted adherent to the bIadder. AIthough a specimen that he had contracted syphilis twenty-five for biopsy couId not be obtained, it was beyears previousIy and gonorrhea eighteen years Iieved that the Iesion was inflammatory rather ago. then neopIastic. A tranverse coIostomy was Cystoscopy. A No. 21 F. cystoscope was performed. passed with diffIcuIty due to an obstruction at Operation 11. (J. A. L.) On August 6, 1940, the vesica1 neck. The bIadder capacity was owing to the patient’s marked intoIerance to very much reduced and there was a marked an indweIIing urethra1 catheter and to the degree of cystitis. There was a large tumefaction severe grade of cystitis present, a suprapubic invoIving the posterior waI1 of the bIadder, extending into the vesical lumen. A we11 cystotomy was performed. Upon opening the bladder a smaI1 poIypoid Iesion was noted on defined median bar was aIso present. Biopsy the posterior waI1 of the bladder. In the center of the posterior vesica1 waI1 was reported

The Iower bowe1 presented many adhesions and the sigmoid was densely adherent to the bIadder. There was a ragged, necrotic, annular, infihrating tumor in the sigmoid, which had invaded the pelvic fascia and had perforated into the vesical lumen. Microscopic examination of the sigmoidal Iesion showed a papiilary tumor forming irreguIar acini. There was great variation in the size and shape of the cells. The stroma was moderate in amount. Microscopic examination discIosed the bIadder wal1 thickened and diffusely infiltrated with tumor ceIIs of sigmoida1 origin. The vesica1 mucosa in the region of the fistula was markedIy hyperpIastic.

534

American Journal of Surgery

Lazarus,

Marks-FistuIa

of this Iesion there was a smaII opening through which a ribbon of pus couId be seen escaping into the bIadder cavity. Operation III. (J. A. L.) On August 23,Ig4o, under spinal anesthesia a right nephrectomy and subtotal ureterectomy was performed for infected hydronephrosis. He was discharged from the hospital on September 23, 1940, and returned to Honduras with a we11 reguIated functioning coIostomy. He returned to New York City on ApriI 25, 1941, and was admitted to the hospita1 in preparation for a segmental resection of the sigmoid. Operation IV. (J. A. L.) On May 6, 1941, with the patient in high TrendeIenburg position, the abdomen was opened through a ten-inch right reverse Kammerer incision extending from the symphysis to a point two inches above the umbiIicus. The sigmoid Aexure was found indurated and adherent to the posterior waII of the bIadder. By gentle finger dissection the sigmoid was carefuIIy peeIed away from the bladder. A segmenta resection of the invoIved portion of the sigmoid was then performed, and an end-to-end anastomosis accomplished. No effort was made to cIose the defect in the bIadder waI1 because of the indurated state of the tissues around the fistuIa. An indweIIing urethra1 catheter was introduced. SuIfaniIamide crystaIs were pIaced into the peIvis and appropriate drainage instituted. FoIIowing the operation he was given 500 cc. of titrated blood. FoIIowing an uneventfu1 convalescence he was prepared for cIosure of the coIostomy which was performed on June 13, 1941. He was discharged from the hospita1 symptom-free and all wounds heaIed on JuIy 5, 1941.

Comment. We have here discussed a case of actuaI vesico-intestina1 fistuIa resuiting from diverticuhtis of the sigmoid, complicated by a severe infection of one kidney. The stages empIoyed in handIing such a probIem have been carefuIIy reviewed. No attempt was made to close the vesicaI opening, which heaIed spontaneousIy with the aid of an indweIIing urethra1 catheter. It is our impression that the use of suIfonamides both oraIIy and IocaIIy pIayed an extremeIy important r6Ie in the smoothness of the postoperative

convaIescence foIIowing the segmenta resection of the invoIved sigmoida1 segment of coIon. CASE v. s. R., maIe, aged sixty-four, was admitted to the hospital on the night of ApriI of urinary 2% 1941% with an acute attack retention of twenty-four hours’ duration. During the past few months he had been compIaining of increased diurna1 and nocturna1 urinary frequency. On carefu1 questioning he admitted a Ioss of twenty-five pounds in weight during the past year. Two weeks previousIy he began having nausea and vomited on a few occasions, and also compIained of alternating bouts of diarrhea and constipation. PhysicaI examination disclosed an emaciated, cachectic looking individua1 who appeared in great distress. Recta1 examination faiIed to discIose a Iesion within the Iower bowel. The prostate feIt neither hard nor enIarged, but a boggy fuIIness couId be made out in the cuI-desac. On abdomina1 paIpation an indefinite mass was feIt in the region of the Ieft iIiac fossa. The rest of the examination was essentiaIIy unimportant. Cystoscopy reveaIed a we11 defined median bar and moderate contracture of the vesical neck. There was no evidence of IateraI lobe enIargement. Situated on the posterior waI1 of the bIadder near the vauIt was an area of redness and edema suggesting the possibifity that this portion of the bIadder was in contact with some affected intra-abdominal viscus. Sigmoidoscopy. Situated four and a haIf inches from the anus and involving the anterior waI1 of the rectum there was an uIcerating carcinoma which, on biopsy, proved to be adenocarcinoma. Operation. (J. A. L.) On May 13, 1941, under spina anesthesia the abdomen was opened through a Iong left rectus incision, extending from the symphysis to the IeveI of the umbilicus. Situated in the peIvis there was a Iarge indurated mass about the size of an orange, consisting of rectosigmoid hopeIessIy adherent to the posterior waI1 of the bladder, and a loop of ileum. The, entire mass was aIso adherent to the anterior parieta1 peritoneum. A permanent colostomy was performed, using the sigmoida1 Ioop. FoIIowing the operation a11 urinary symptoms promptIy disappeared, and the patient Iived in relative comfort for seven months before he succumbed.

New SERIES VOL. LIX.

No. 3

Lazarus,

Marks-FistuIa

Comment. This iIIustrates another case of incipient vesico-intestinal fistuIa due to carcinoma of the rectosigmoid, giving rise to symptoms of prostatism and compIete urinary retention. Owing to the marked discrepancy between the urinary symptoms and the cystoscopic findings, and aIso because of the presence of a characteristic area of edema and congestion on the posterior vesical waI1, a tentative diagnosis of an intra-abdominal Iesion secondariIy affecting the bIadder was made. This was confirmed at operation. CompIete diversion of the feca1 flow from the affected segment of bowe1 by permanent coIostomy Ied to prompt abatement of urinary symptoms, indicating thereby the fact that irritation of a portion of the detrusor muscle by an extravesica1 Iesion can and does give rise to symptoms of prostatism indistinguishabIe from a simiIar syndrome caused by a Iesion of the prostate gIand. The differentia1 diagnosis can onIy be made by carefu1 cystoscopic interpretation. SUMMARY

AND

CONCLUSIONS

I. Vesico-intestina1 fistuIas may be divided into actua1 fistuIas and incipient or threatened fistulas. 2. CarefuI anaIysis of the Iiterature indicates that inflammatory Iesions of the bowe1 constitute the underIying cause of fistuIas in the majority of cases (51 per cent). Tumors of the bowe1 were present in 21.3 per cent of the coIIected cases. A review of the inff ammatory intestina1 group shows that diverticuIitis was the primary cause in 65.8 per cent of the cases. The

AmericanJournalofSurgery535

most frequent neopIastic Iesion of the bowel was carcinoma of the rectum (50.8 per cent). As regards the Iocation of the fistuIa, it was noted that the most frequent site was between the bIadder and rectum (43.5 per cent). 3. The cardina1 symptoms of actua1 fistuIa are the passage of gas or stoo1 through the urethra. Symptoms of an incipient or threatened fistula are more suggestive than pathognomonic. In cases in which the primary Iesion is in the bIadder, the appearance of intestina1 symptoms is suggestive of a contact infIammation between the bIadder and intestine. When the primary lesion is in the bowe1, the appearance of vesical symptoms shouId Iikewise suggest this same condition. 4. Diagnosis, especiaIIy in cases of incipient fistulas, depends primariIy upon cystoscopic examination. 5. Proper preoperative preparation of the bowe1 is probabIy one of the most important requirements for successfu1 surgery in this type of case. In cases of actua1 fistuIas, graded stage operations are necessary. The use of an indweIIing urethra1 catheter obviates the necessity of cIosing the f%tuIous opening in the bladder. The systemic and IocaI empIoyment of suIfonamides has been found of great vaIue in this group of cases. 6. Prognosis depends entireIy upon the underlying pathologica condition. REFERENCES

BALCH, J. F. J. Ural., 86: 706707, 1931. HIGGINS, C. C. J. Ural., 86: 694-705, 1931. KELLOGG, W. A. Am. J. .Swg., qr: 135-186, 1938. PASCAL,A. ThPse de Paris, Igoo.