- Email: [email protected]
Management of severe hemorrhage in ulcerative colitis
Management of Severe Hemorrhage in Ulcerative Colitis John H. Robert, MD, David B. Sachar, MD, Arthur H. Aufses, Jr, MD, Adrian J. Greenstein, MD, New York.New York
Twenty-five patients with ulcerative colitis were treated between 1959 and 1986 at The Mount Sinai Hospital, with severe gastrointestinal hemorrhage as their m a j o r complaint. Twenty-two patients required operation, while three patients were treated medically. Total proctocolectomy with ileostomy was carried out in 5 patients, and subtotal colectomy accompanied by mucous fistula ( 1 4 ) , H a r t m a n n closure ( 2 ) , or ileosigmoidostomy ( 1 ) was performed in 17 patients. Eleven of the patients who underwent operation had emergency colectomies, while the remaining 11 had semielective procedures. Subtotal colectomy was p e r f o r m e d in 10 of the 11 emergency cases. Indications for e m e r g e n c y surgery were massive h e m o r r h a g e alone in seven patients and severe h e m o r r h a g e complicated by toxic megacolon in four patients. One patient died postoperatively of a perforated duodenal ulcer following emergency subtotal colectomy. There were two late deaths f r o m leukemia in one surgically treated patient and one medically treated patient at 9 and 1 8 months, respectively. All 4 of the 25 patients with remaining intact rectums were alive and well at 3- to 12-year follow-up. Subtotal colectomy can be undertaken in patients With massive h e m o r r h a g e from ulcerative colitis for w h o m subsequent ileoanal anastomosis is planned, provided that one recognizes and is prep a r e d for the approximately 12% risk of continued i-ectal hemorrhage.
lcerative colitis is characterized by bloody diarrhea in over 95% of cases, but severe or massive hemorrhage is actually quite rare [1-3], occurring less commonly than other complications such as toxic megacolon [4], colonic perforation [5,6], or neoplastic degeneration [5]. Nonetheless, massive hemorrhage occasionally rep-
U
From the Departments of Surgery (JHR, AHA, AJG) and Medicine (DBS), Divisionof Gastroenterology,Mount Sinai Schoolof Medicine of the City Universityof New York. Requests for reprints should be addressedto Adrian J. Greenstein, MD, Departmentof Surgery, Box 1259, Mount Sinai Medical Center, One GustaveL. LevyPlace, New York.New York 10029. Manuscript submittedFebruary20, 1989,revisedAugust28, 1989. and acceptedSeptember7, 1989. 550
resents the principal indication for emergency surgery. Reports from the mid-1970s suggested that total proctocolectomy was the safest procedure in such circumstances because of the risk of persistent or recurrent bleeding from the rectal segment after subtotal colectomy [7,8]. However, rectal excision adds to the time and risks of the operation and also precludes later secondary ileoanal anastomosis, an option that the less drastic and quicker subtotal colectomy does not foreclose [9]. We have reviewed the therapy of a large series of patients with severe or massive hemorrhage from ulcerative colitis to determine if subtotal colectomy constitutes a rational and safe surgical approach to this problem. MATERIAL AND M E T H O D S Twenty-five patients with severe or massive rectal hemorrhage arising from ulcerative colitis were treated at The Mount Sinai Hospital between 1959 and 1986. Twenty patients were admitted with hemorrhage and 5 patients developed severe hemorrhage within 1 week of the date of admission. Sixteen patients were obtained from a retrospective review of a series of 1356 cases of ulcerative colitis (1.4%) retrieved from the computer files of The Mount Sinai Hospital and admitted between 1959 and 1983. The remaining nine patients were admitted between 1984 and 1986; five patients were derived from private records (AJG) and four were obtained by crossindexing hospital records for ulcerative colitis and gastrointestinal bleeding. Severe hemorrhage was defined as acute bleeding originating in the colon and requiring at least 4 units of blood over a period not exceeding 3 weeks prior to operation, or during the entire course of hospitalization if treatment was nonoperative. An emergency operative procedure was defined as one that was deemed mandatory and could not be deferred because of persistent massive blood loss for which continued transfusion was necessary, with or without associated colonic dilatation. Semielective surgical procedures were those carried out following a severe hemorrhage, in cases where the reduced rate of bleeding no longer made immediate surgery imperative, but in which the severity of the colitis made ultimate colectomy unavoidable. The records of 40 patients were initially collected as potentially eligible for inclusion in this series. Fifteen cases were excluded; in 7, the patients received fewer than 4 units of blood, or the period of transfusions extended beyond 3 weeks. The other eight cases were excluded because the bleeding was not from active colitis: three patients bled from ileostomies long after colectomy had been performed; in four instances, the source of hemorrhage was not the colitis per se (one ease each of hemorrhagic gastritis, duodenal ulcer, esophageal varices, and
THE AMERICAN JOURNAL OF SURGERY VOLUME 159 JUNE 1990
HEMORRHAGE IN ULCERATIVE COLITIS
in situ colonic carcinoma), and one patient developed thrombocytopenic purpura 7 years after a right hemicolectomy. In the remaining 25 patients with severe hemorrhage, the diagnosis and extent of ulcerative colitis were confirmed by clinical [10], radiographic [11], endoscopic [12], and histopathologic [13] criteria at the time of severe hemorrhage. Patients with Crohn's disease were excluded and will be discussed separately. Patients were excluded if any of the classical findings of clinical or pathologic Crohn's colitis (segmental disease, ileal involvement, rectal sparing, fissures, or fistulae) were present. Pathologic examination was carried out on biopsy or surgical specimens in all 25 cases. The extent of colonic involvement at the time of bleeding (i.e., left-sided versus universal colitis) was determined upon histologic examination of the resected specimen for the 22 patients operated upon, while the remaining 3 patients underwent colonoscopy with biopsies once hemorrhage had subsided. First attacks of less than a year's duration were distinguished from relapses using the criteria of Edwards and Truelove [14]. RESULTS There were 16 male and 9 female patients in the series. The mean age at onset of colitis was 25.4 years (range: 5.5 to 62 years, median: 20 years) and at the time of hemorrhage was 28.0 years (range: 9 to 74 years, median: 24 years), with a mean duration from onset of disease to hemorrhage of 2.6 years (range: 7 weeks to 12 years, median: 18 months). Relapses exceeded first attacks by a 3-to-1 ratio. Universal colonic involvement exceeded left-sided ulcerative colitis by a 5-to-1 ratio and was found in all 11 patients who underwent emergency procedures (Table I). There was one death in this series, caused by sepsis from postoperative perforation of a duodenal ulcer. Initial therapy consisted of blood and fluid replacement, together with steroids, in all patients. Twenty-two patients underwent operation, and 3 patients were treated medically with blood transfusion and supportive therapy. All three patients treated conservatively were young. Two female patients aged 10 and 17 years, each with a 1-year history of ulcerative colitis, required 8 and 4 units of blood, respectively; the third patient, aged 23 years, had concomitant acute promyelocytic leukemia that contraindicated operative intervention at the time of hemorrhage despite a sizable blood loss (30 units over a 20-day period). The remaining 22 patients in our series underwent operations that were roughly evenly distributed over intervals from 1 to 19 days after admission, with a mean of 12.5 days (median: 11 days). Emergency colectomies were necessary for 11 patients, 8 male and 3 female, all of whom had universal colitis. The duration of ulcerative colitis ranged from 7 weeks to 6 years, with a mean of 2.4 years. Seven patients were in relapse, but four patients had massive hemorrhage during a first attack between 7 weeks and 6 months (mean: 3 months) after onset of colitis. Indications for
TABLE ! Clinical Material
No. of patients Patients undergoing operation Age at onset of ulcerative colitis (y) Mean Range Age at time of hemorrhage (y) Mean Range First attack/relapse Left-sided/universal colitis
Male
Female
All Patients
16 15
9 7
25 22
30.9 13-62
19.4 5,5-46
25.4 5,5-62
31.7 15-74 4/12 1/15
21.3 9-46 2/7 3/6
28.0 9-74 6/19 4/21
surgery were massive hemorrhage in seven patients, severe hemorrhage followed by toxic megacolon in three patients, and the combination of hemorrhage and toxic megacolon in one patient. Although the amount of bleeding in three patients met our criteria of severity, it was the development of colonic dilatation that dictated emergency surgical intervention in these cases. Two of the four patients with massive hemorrhage during a first attack had complications from the development of toxic megacolon. One patient had a subtotal colectomy, followed by persistent rectal bleeding. He underwent a proctectomy with construction of a pelvic pouch and ileoanal pullthrough 25 days later. He subsequently developed Crohn's disease in his pouch, manifested by fistulae and hundreds of noncaseating granulomas. Careful review of the original colon specimen revealed severe ulcerative colitis with deep ulceration but without a single granuloma. The hospital stay ranged from 16 to 70 days (mean: 39 days) for these patients with massive hemorrhage. Semielective surgical procedures were carried out in the remaining 11 patients in whom intravenous therapy with high-dose steroids had not controlled persistent active bleeding. Patients who underwent emergency colectomy required an average of 17.3 units of blood (range: 5 to 23 units) over a mean 6.4-day period (range: 24 hours to 16 days), compared with 6.8 units (range: 4 to 16 units) over a mean 12.8-day period (range: 3 to 19 days) for those operated upon semielectively. There were no statistically significant differences between blood requirements for first attacks versus relapses (9 units versus 10 units), universal versus left-sided colitis (9.2 units versus 12.5 units), medically versus surgically treated patients (14 units versus 9.2 units), or subtotal colectomy versus total proctocolectomy (8.7 units versus 10.6 units). Both the operated and nonoperated leukemic patients required massive and prolonged transfusion: 23 and 30 units, respectively. It is of interest that the two patients in whom reoperation was carried out for continued or recurrent rectal hemorrhage required more blood in the preoperative period than any of the patients other than those with leukemia; the patient who developed Crohn's disease in his pelvic pouch required 17 units preoperativdy, and the
THE AMERICAN JOURNAL OF SURGERY
VOLUME 159
JUNE 1990
551
ROBERT ET AL
25 Patients
I Operated 22
I Non-Operated I
I
I
I
I
Subtotal Colectomy (Sl'C) 17 (14 MF, 2 H, 1 IS)
I
No further operation 2 i
IS 1
Intact Rectum
I i
Well at 12 yes
Well at 10 yes
I
Total Pmctocolectomy (1"PC)
(4 APR,
Reoperated 14
I
1 Died at 9 mos of leukemia
1
2 Well at 3 and 4 yrs
leukemia
I
I
Died of postop sepsis 1
1 MS + IAA)
1 Died at lS mos of
3 Well at 6.3 yrs (1-15 yes)
Well at 5 yes
Abbreviations:
APR 7
MS + IAA 7
Well at 12.5 yes (3-22 yrs)
6 well at 16.7 mos (1 too-2 yes) 1 anastomotic healing problem at 11 mos
UC = Ulcerative Colitis IS = Ileosigmoidostomy MF = Mucous Fistula H = Hartmann's Closure APR = Abdomino Perineal Resection MS + IAA = Mucosal Stripoing + Ileoanal Anastomosis
Figure 1. Surgical management and outcome of severe hemorrhage in 25 patients with ulcerative colitis.
patient who bled subsequently from a rectosigmoid polyp required 12 units. Five patients underwent total proctocolectomy, four with terminal ileostomy and one with primary mucosal stripping and ileo-anal anastomosis (Figure 1). In one leukemic patient, an emergency total proctocolectomy was performed because of the severity of the rectal bleeding, requiring 23 units of blood over 16 days, with no thought given to preserving the rectum. The remaining four total proctocolectomies were carried out as semielective procedures. Three of the five total proctocolectomies took place in the period from 1980 to 1984 (Figure 1). Seventeen patients had subtotal colectomies, usually with a mucous fistula (14 patients). A Hartmann closure was performed in two patients with severe sigmoid inflammation in order to minimize the risk of recurrent hemorrhage. Finally, one patient with a quiescent rectal mucosa had a one-stage ileosigmoidostomy. Most of these subtotal colectomies (10 of 17) were emergency procedures. Patients operated on for severe bleeding were usually placed in Lloyd-Davies stirrups so that the rectum could be examined during and following the abdominal colectomy. Continued massive rectal hemorrhage requiring immediate proctectomy was not seen in any patient following subtotal colectomy. The only postoperative death occurred on the 41st postoperative day in a male patient with a 4-month history of ulcerative colitis taking high-dose steroids following emergency subtotal colectomy; he had a perforated duo552
T H E A M E R I C A N J O U R N A L OF S U R G E R Y
denal ulcer, duodenocutaneous fistula, and peritoneal sepsis. There were two late deaths caused by acute promyelocytic leukemia 9 and 18 months following the major hemorrhage; the first patient had undergone emergency total proctocolectomy, while the second had been treated nonoperatively. Both died of multiple bleeding complications culminating in cerebral hemorrhage. Twelve patients experienced 15 early postoperative complications, most of them infectious in nature; 6 of these patients, including the one who died, required reoperation. These complications were proportionately distributed between emergency and semielective procedures and between subtotal colectomy and total proctocolectomy (Figure 2). Two of 11 patients who underwent emergency subtotal colectomy had postoperative bleeding from the rectal remnant. The first patient, who had persistent hemorrhage, underwent proctectomy with mucosal stripping, ileal pouch, and ileoanal anastomosis at day 25 while still taking high-dose steroids. Since Crohn's disease developed in the pouch, this patient should be classified as having indeterminate colitis. The second developed a major hemorrhage from a pseudopolyp 4 months after subtotal colectomy, after a relatively quiescent period and underwent emergency abdominoperineal resection in 1979. Two of the three patients treated medically were followed up at 3 and 4 years, respectively; although one is doing well, the other has experienced one flare-up each year requiring transient steroids. The third patient died of
V O L U M E 159
J U N E 1990
HEMORRHAGE IN ULCERATIVE COLITIS
Operated Patients 22
I
I
STC 17
TPC 5
I
I
EMERGENCY 10
I
SEMI-ELECTIVE 7
Persistent/Recurrant Rectal Bleeding*
2
Per/orated DU* --) Septic Death
1
Intraabdominal Abscess
1
Wound Infection Small Bowel Obstruction
1 2
Psychotic Reaction
2
I
EMERGENCY 1
Peritonitis*
1
Sepsis (Fungus)
1
Wound Infections
2
SEMI-ELECTIVE 4
Perineal Wound Bleed*
4 Complications in 3 Patients
Perineal Abscess*
*Complication requiring reintervention
9 Complications in 7 Patients
Figure 2. Complications following surgery in 22 patients with severe hemorrhage from ulcerative colitis.
TABLE II I n c i d e n c e a n d M o r t a l i t y of S e v e r e H e m o r r h a g e
In U l c e r a t i v e Colitis ( U C )
Patients (n) Author & Reference
Date of Publication
with Hemorrhage
with UC
% with Hemorrhage
n
Deaths %
Sioan et al [ 15] Brown et al [ 16] Truelove & Witts [ 17] Bruce & Cole [18] Edwards & Truelove [ 1] Watts et al [ 19] Jalan et a/[20] Binder eta/ [ 7] Block et al [8] Bobblo eta/ [21] Goligher et al [22] Albrechtsen et al [23] Oakley et al [24] Vecchi et al [25] Present series
1950 1951 1955 1962 1964 1966 1969 1975 1977 1977 1980 1981 1985 1986 1987 1959-1983 1984-1986
22 O 1 5 21 1 6 5 16 2 0 6 11 5
2,000 147 213 118 624 151 399 111 39 46 258 132 311 125
(1.1) (0) (0.5) (4.2) (3.4) (0.7) (1.5) (4.5) (--) (4.3) (O) (4.5) (3.5) (4)
12 NR 1 1 NR NR 3 O -1 -O NR NR
(55) -(100) (20) --
16 9
1,156 --
(1.4) --
1 0
(5.3) (0)
(50) (0) (8.7)* (50) -(0) ---
Type of Operation NR NR None STC ? NR TPC ? NR 4 TPC, 1 STC 5 TPC, 11 STC TPC 4 TPC, 2 STC STC TPC 5 TPC 17 STC
* Patients with ulcerative colitis and Crohn's disease. NR = not reported; STC = subtotal colectomy; TPC = total proctocolectomy.
leukemia 18 months after the episode of severe hemorrhage. As noted previously, 1 of the 17 patients who underwent subtotal colectomy died postoperatively. Seven of the 16 survivors ultimately required abdominoperineal resection (Figure 1), usually for persistent mucous and/ or bloody rectal discharge. Another seven patients underwent subsequent planned rectal mucosal stripping with ileoanal pull-through. The interval between subtotal colectomy and reoperation averaged 24.7 months for abdominoperineal resection (range: 4 months to 5 years, median: 20 months) and 4 months for rectal mucosal
stripping with ileoanal pull-through (range: 25 days to 4 months, median: 2.5 months). The two remaining patients with intact rectal segments are in remission at 8 and 12 years after subtotal colectomy. COMMENTS The frequency of severe hemorrhage reported in the literature ranges from 0 to 4.5% [5] with a mean of 2.2% [15-25], which is close to the 1.4% frequency observed for the 1959 to 1983 period of our series (Table II). However, this relatively rare complication accounts for approximately 10% of all urgent colectomies for ulcer-
THE AMERICAN JOURNAL OF SURGERY
VOLUME 159
JUNE 1990
553
ROBERT ET AL
ative colitis. From 1959 to 1980, 6 emergency or semielective colectomies were carried out at this institution for severe hemorrhage, 48 for toxic megacolon, and 7 for perforation without colonic dilatation, in an overall series of 613 patients [4,6]. This 10% figure is comparable to the 6% reported by Binder et al [7], but lower than the 41% reported by Block et al [8]. Among our patients with severe hemorrhage, there were a number of features similar to those previously reported in the literature (Table II) and in previous publications from this institution [4,26]: Patients were somewhat but not significantly younger than other patients with ulcerative colitis (25 versus 29 years) [18,26]; bleeding was a relatively early complication of ulcerative colitis (mean duration of illness: 2.6 years) [27]; universal coIonic involvement occurred more frequently than leftsided colitis [28] (21 of 25 cases or 84%, versus 60% in our overall series) [26]; and, finally, there was a relatively high occurrence of concomitant toxic megacolon [8,19,22,29,30] (6 of 22 operated cases or 27%, versus 10% in the overall series) [4]. Even though total proctocolectomy is generally recommended in the elective treatment of ulcerative colitis, many investigators advocate a two-stage procedure as being safer in emergency cases [31,32]. Reports in the mid-1970s by Binder et al [7] and Block et al [8,33], however, showed emergency total proctocolectomy to be just as safe as subtotal colectomy, thus prompting many to prefer one-stage total proctocolectomy. In contrast to the less definitive subtotal colectomy, though, total proctocolectomy precludes later restoration of intestinal continuity by ileo-anal anastomosis. Does subtotal colectomy carry an increased risk of recurrent bleeding when it is performed specifically for severe hemorrhage? Various reports suggest that the incidence of postoperative rectal stump bleeding ranges from minimal to 44.1% [23,34]. Two of our 17 patients continued to bleed following subtotal colectomy and required secondary emergency procedures, one abdominoperineal resection at 4 months and one mucosal proctectomy at 25 days after subtotal colectomy. As already noted, the latter patient developed granulomas characteristic of Crohn's disease in the pouch. The other refused early proctectomy and had recurrence of bleeding several months later. Although these two patients represent a 12% bleeding recurrence rate, the unusual pathologic findings in one and the delay in recurrence in the other suggest that the rectum can be preserved, except in the rare case of massive hemorrhage clearly originating in the rectum. A 12.6% incidence of profuse postoperative rectal bleeding has been reported by Korelitz et al [9] in a separate series of 136 patients with residual rectal stump, 126 of whom (93%) underwent secondary abdominoperineal resections. In our series, 14 of 17 patients (82%) underwent subsequent proctectomy. If the severe hemorrhage appears to originate in the rectum or appears to be diffuse in origin, the patient should be placed in LloydDavies stirrups. Continued massive bleeding from the rectal segment observed during surgery, which was not found in any patient in this series, is an indication for 554
immediate total proctocolectomy rather then subtotal colectomy. Although primary mucosal stripping with ileoanal anastomosis was successful in one patient in this series who underwent a semielective procedure, such a procedure should probably not be considered routinely in the surgical handling of acute hemorrhage, especially in emergency surgery for massive hemorrhage. On the other hand, subtotal colectomy still appears to be a feasible alternative in view of the advantage of preserving the option of secondary ileoanal anastomosis. Contraindications might include massive hemorrhage clearly originating from the rectum, certain cases of leukemia with very limited life expectancy, and such advanced age that ileoanal pull-through is an unlikely prospect.
The authors present a small number of patients to address the important clinical problem of severe hemorrhage in ulcerative colitis. This complication is relatively uncommon, but is occasionally the principal indication for emergency surgery. The issue of whether to remove the rectum or preserve it for subsequent rectal mucosal stripping is the major point of discussion. REFERENCES 1. Edwards FC, Truelove SC. The course and prognosisof ulcerative colitis. Gut 1964; 5: 1-26. 2. Farmer RG. Clinical features and natural history of inflammatory bowel disease. Med Clin North Am 1980; 64: 1103-15. 3. Smith JN, Winship DH. Complications and extraintestinal problems in inflammatory bowel disease. Med Clin North Am 1980; 64: 1161-71. 4. Greenstein AJ, Sachar DB, Gibbs A, et al. Outcome of toxic dilatation in ulcerative colitis. J Clin Gastroenterology 1985; 7: 137-44. 5. Goligher JC, de Dombal FT, Watts J McK, Watkinson G. Ulcerativecolitis. Baltimore:Williams & Wilkins, 1965:123, 15074, 227-8. 6. GreensteinAJ, AufsesAH Jr. Differencesin pathogenesis,incidence and outcome of perforation in inflammatory bowel disease. Surg Gynecol Obstet 1985; 160: 63-9. 7. Binder SC, Miller HH, Deterling RA. Emergencyand urgent operations for ulcerative colitis. Arch Surg 1975; 110: 284-9. 8. BlockGE, Moossa AR, SimonowitzD, Hassan SZ. Emergency colectomyfor inflammatoryboweldisease. Surgery 1977; 82:5316. 9. KorelitzBI, Dyck WP, Klion FM. Fate of the rectum and distal colon after subtotal colectomyfor ulcerative colitis. Gut 1969; 10: 198-201. 10. Cello JP. Ulcerative colitis. In: Sleisenger MH and Fordtran JS, eds. Gastrointestinal disease. Philadelphia: WB Saunders, 1983: 1122-67. 11. Marshak RH, Lindner AE, Maklansky D. Radiology of the colon. Philadelphia: WB Saunders, 1980: 64-119. 12. Waye JD. The role of colonscopyin the differentialdiagnosisof inflammatoryboweldisease. Gastrointest Endosc 1977;23:150-4. 13. Morson BC. Pathology of ulcerative colitis. In: Kirsner JB, Shorter RG, eds. Inflammatory bowel disease. Philadelphia: Lea and Feibiger, 1980: 281-95. 14. Edwards FC, TrueloveSC. The course and prognosisof ulcerative colitis. Gut 1963; 4: 299-315. 15. Sloan WP, Bargen JA, BaggenstossAH. Local complications of chronic ulcerativecolitis based on the study of 2000 cases. Proc Mayo Clin 1950; 25: 240-4. 16. Brown ML, Kasich AM, Weingarten B. Complications of
THE AMERICAN JOURNAL OFSURGERY VOLUME 159 JUNE 1990
HEMORRHAGE IN ULCERATIVECOLITIS
chronic ulcerative colitis. Am J Dig Dis 1951; 18: 52-4. 17. Truelove SC, Witts LJ. Cortisone in ulcerative colitis. Final report on a therapeutic trial. Br Med J 1955; 2: 1041-7. 18. Bruce D, Cole WH. Complications of ulcerative colitis. Ann Surg 1962; 155: 768-81. 19. Watts JMK, de Dombal FT, Goligher JC. Early results of surgery for ulcerative colitis. Br J Surg 1966; 53: 1005-14. 20. Jalan KN, Sircus W, Card WI, et al. A n experience of ulcerative colitis. I. Toxic dilation in 55 cases. Gastroenterology 1969; 57: 68-82. 21. Bobbio P, Zola C, Salcuni PF. La colectomia totale d'urgenza nella colite ulcerosa. Acta Bio-Med Ateneo Parmense 1977;48: 93100. 2 2 . Goligher JC, Duthie HL, Nixon HH. Surgery of the anus, rectum and colon. London: Bailliere Tindall, 1980: 689-826. 23. Albrechtsen D, Bergan A, Nygaard K, Gjone E, Flatmark A. Urgent surgery for ulcerative colitis: early colectomy in 132 patients. World J Surg 1981; 5: 607-15. 24. Oakley SR, Lavery IC, Fazio VW, Jagelman DG, Weakley FL, Easley K. The fate of the rectal stump after subtotal colectomy for ulcerative colitis. Dis Colon Rectum 1985; 28: 394-6. 25. Vecchi R, Arena N, Cunsolo A, Gozzetti G. Emorragie massive in corso di colite ulcerosa e di morbo di Crohn. Minerva Chir 1986; 41: 653-6. 26. Greenstein AJ, Sachar DB, Smith H, et al. Cancer in universal
and left-sided universal colitis: factors determining risk. Gastroenterology 1979; 77: 290-4. 27. Jalan KN, Prestcott RJ, Sircus W, et al. Ulcerative colitis. A clinical study of 399 patients. J R Coil Surg Edinb 1971; 16: 133851. 28. Both H, Torp-Pedersen K, Kreiner S, Hendriksen C, Binder V. Clinical appearance at diagnosis of ulcerative colitis and Crohn's disease in a regional patient group. Scand J Gastroenterol 1983; 18: 987-91. 29. Philippe I, Philippe M, Girard M. Les colectasies aigues au tours des recto-colites hemorragiques. Donnees actuelles. Lyon Medical 1972; 227: 491-503. 30. Huizenga KA. Medical treatment and prognosis of some local complications of chronic ulcerative colitis and Crohn's disease. In: Kirsner RS, Shorter RG, eds. Inflammatory bowel disease. Philadelphia: Lea & Febiger, 1980: 466-7. 31. Lulu DJ, Dragstedt LR. Primary total proctocolectomy for ulcerative colitis. Ann Surg 1974; 40: 417-20. 32. Flatmark A, Fretheim B, Gjone E. Early colectomy in severe ulcerative colitis. Scand J Gastroenterol 1969; 4: 505-10. 33. Simonowitz D, Block GE, Moossa AR, Hassan SZ. Emergency colectomy for inflammatory bowel disease. Review of Surgery 1977; 34: 359-62. 34. Moss GS, Keddie N. Fate of rectal stump in ulcerative colitis. Arch Surg 1965; 91: 967-70.
THE AMERICAN JOURNAL OF SURGERY VOLUME 159 JUNE 1990 555
Twenty-five patients with ulcerative colitis were treated between 1959 and 1986 at The Mount Sinai Hospital, with severe gastrointestinal hemorrhage as their m a j o r complaint. Twenty-two patients required operation, while three patients were treated medically. Total proctocolectomy with ileostomy was carried out in 5 patients, and subtotal colectomy accompanied by mucous fistula ( 1 4 ) , H a r t m a n n closure ( 2 ) , or ileosigmoidostomy ( 1 ) was performed in 17 patients. Eleven of the patients who underwent operation had emergency colectomies, while the remaining 11 had semielective procedures. Subtotal colectomy was p e r f o r m e d in 10 of the 11 emergency cases. Indications for e m e r g e n c y surgery were massive h e m o r r h a g e alone in seven patients and severe h e m o r r h a g e complicated by toxic megacolon in four patients. One patient died postoperatively of a perforated duodenal ulcer following emergency subtotal colectomy. There were two late deaths f r o m leukemia in one surgically treated patient and one medically treated patient at 9 and 1 8 months, respectively. All 4 of the 25 patients with remaining intact rectums were alive and well at 3- to 12-year follow-up. Subtotal colectomy can be undertaken in patients With massive h e m o r r h a g e from ulcerative colitis for w h o m subsequent ileoanal anastomosis is planned, provided that one recognizes and is prep a r e d for the approximately 12% risk of continued i-ectal hemorrhage.
lcerative colitis is characterized by bloody diarrhea in over 95% of cases, but severe or massive hemorrhage is actually quite rare [1-3], occurring less commonly than other complications such as toxic megacolon [4], colonic perforation [5,6], or neoplastic degeneration [5]. Nonetheless, massive hemorrhage occasionally rep-
U
From the Departments of Surgery (JHR, AHA, AJG) and Medicine (DBS), Divisionof Gastroenterology,Mount Sinai Schoolof Medicine of the City Universityof New York. Requests for reprints should be addressedto Adrian J. Greenstein, MD, Departmentof Surgery, Box 1259, Mount Sinai Medical Center, One GustaveL. LevyPlace, New York.New York 10029. Manuscript submittedFebruary20, 1989,revisedAugust28, 1989. and acceptedSeptember7, 1989. 550
resents the principal indication for emergency surgery. Reports from the mid-1970s suggested that total proctocolectomy was the safest procedure in such circumstances because of the risk of persistent or recurrent bleeding from the rectal segment after subtotal colectomy [7,8]. However, rectal excision adds to the time and risks of the operation and also precludes later secondary ileoanal anastomosis, an option that the less drastic and quicker subtotal colectomy does not foreclose [9]. We have reviewed the therapy of a large series of patients with severe or massive hemorrhage from ulcerative colitis to determine if subtotal colectomy constitutes a rational and safe surgical approach to this problem. MATERIAL AND M E T H O D S Twenty-five patients with severe or massive rectal hemorrhage arising from ulcerative colitis were treated at The Mount Sinai Hospital between 1959 and 1986. Twenty patients were admitted with hemorrhage and 5 patients developed severe hemorrhage within 1 week of the date of admission. Sixteen patients were obtained from a retrospective review of a series of 1356 cases of ulcerative colitis (1.4%) retrieved from the computer files of The Mount Sinai Hospital and admitted between 1959 and 1983. The remaining nine patients were admitted between 1984 and 1986; five patients were derived from private records (AJG) and four were obtained by crossindexing hospital records for ulcerative colitis and gastrointestinal bleeding. Severe hemorrhage was defined as acute bleeding originating in the colon and requiring at least 4 units of blood over a period not exceeding 3 weeks prior to operation, or during the entire course of hospitalization if treatment was nonoperative. An emergency operative procedure was defined as one that was deemed mandatory and could not be deferred because of persistent massive blood loss for which continued transfusion was necessary, with or without associated colonic dilatation. Semielective surgical procedures were those carried out following a severe hemorrhage, in cases where the reduced rate of bleeding no longer made immediate surgery imperative, but in which the severity of the colitis made ultimate colectomy unavoidable. The records of 40 patients were initially collected as potentially eligible for inclusion in this series. Fifteen cases were excluded; in 7, the patients received fewer than 4 units of blood, or the period of transfusions extended beyond 3 weeks. The other eight cases were excluded because the bleeding was not from active colitis: three patients bled from ileostomies long after colectomy had been performed; in four instances, the source of hemorrhage was not the colitis per se (one ease each of hemorrhagic gastritis, duodenal ulcer, esophageal varices, and
THE AMERICAN JOURNAL OF SURGERY VOLUME 159 JUNE 1990
HEMORRHAGE IN ULCERATIVE COLITIS
in situ colonic carcinoma), and one patient developed thrombocytopenic purpura 7 years after a right hemicolectomy. In the remaining 25 patients with severe hemorrhage, the diagnosis and extent of ulcerative colitis were confirmed by clinical [10], radiographic [11], endoscopic [12], and histopathologic [13] criteria at the time of severe hemorrhage. Patients with Crohn's disease were excluded and will be discussed separately. Patients were excluded if any of the classical findings of clinical or pathologic Crohn's colitis (segmental disease, ileal involvement, rectal sparing, fissures, or fistulae) were present. Pathologic examination was carried out on biopsy or surgical specimens in all 25 cases. The extent of colonic involvement at the time of bleeding (i.e., left-sided versus universal colitis) was determined upon histologic examination of the resected specimen for the 22 patients operated upon, while the remaining 3 patients underwent colonoscopy with biopsies once hemorrhage had subsided. First attacks of less than a year's duration were distinguished from relapses using the criteria of Edwards and Truelove [14]. RESULTS There were 16 male and 9 female patients in the series. The mean age at onset of colitis was 25.4 years (range: 5.5 to 62 years, median: 20 years) and at the time of hemorrhage was 28.0 years (range: 9 to 74 years, median: 24 years), with a mean duration from onset of disease to hemorrhage of 2.6 years (range: 7 weeks to 12 years, median: 18 months). Relapses exceeded first attacks by a 3-to-1 ratio. Universal colonic involvement exceeded left-sided ulcerative colitis by a 5-to-1 ratio and was found in all 11 patients who underwent emergency procedures (Table I). There was one death in this series, caused by sepsis from postoperative perforation of a duodenal ulcer. Initial therapy consisted of blood and fluid replacement, together with steroids, in all patients. Twenty-two patients underwent operation, and 3 patients were treated medically with blood transfusion and supportive therapy. All three patients treated conservatively were young. Two female patients aged 10 and 17 years, each with a 1-year history of ulcerative colitis, required 8 and 4 units of blood, respectively; the third patient, aged 23 years, had concomitant acute promyelocytic leukemia that contraindicated operative intervention at the time of hemorrhage despite a sizable blood loss (30 units over a 20-day period). The remaining 22 patients in our series underwent operations that were roughly evenly distributed over intervals from 1 to 19 days after admission, with a mean of 12.5 days (median: 11 days). Emergency colectomies were necessary for 11 patients, 8 male and 3 female, all of whom had universal colitis. The duration of ulcerative colitis ranged from 7 weeks to 6 years, with a mean of 2.4 years. Seven patients were in relapse, but four patients had massive hemorrhage during a first attack between 7 weeks and 6 months (mean: 3 months) after onset of colitis. Indications for
TABLE ! Clinical Material
No. of patients Patients undergoing operation Age at onset of ulcerative colitis (y) Mean Range Age at time of hemorrhage (y) Mean Range First attack/relapse Left-sided/universal colitis
Male
Female
All Patients
16 15
9 7
25 22
30.9 13-62
19.4 5,5-46
25.4 5,5-62
31.7 15-74 4/12 1/15
21.3 9-46 2/7 3/6
28.0 9-74 6/19 4/21
surgery were massive hemorrhage in seven patients, severe hemorrhage followed by toxic megacolon in three patients, and the combination of hemorrhage and toxic megacolon in one patient. Although the amount of bleeding in three patients met our criteria of severity, it was the development of colonic dilatation that dictated emergency surgical intervention in these cases. Two of the four patients with massive hemorrhage during a first attack had complications from the development of toxic megacolon. One patient had a subtotal colectomy, followed by persistent rectal bleeding. He underwent a proctectomy with construction of a pelvic pouch and ileoanal pullthrough 25 days later. He subsequently developed Crohn's disease in his pouch, manifested by fistulae and hundreds of noncaseating granulomas. Careful review of the original colon specimen revealed severe ulcerative colitis with deep ulceration but without a single granuloma. The hospital stay ranged from 16 to 70 days (mean: 39 days) for these patients with massive hemorrhage. Semielective surgical procedures were carried out in the remaining 11 patients in whom intravenous therapy with high-dose steroids had not controlled persistent active bleeding. Patients who underwent emergency colectomy required an average of 17.3 units of blood (range: 5 to 23 units) over a mean 6.4-day period (range: 24 hours to 16 days), compared with 6.8 units (range: 4 to 16 units) over a mean 12.8-day period (range: 3 to 19 days) for those operated upon semielectively. There were no statistically significant differences between blood requirements for first attacks versus relapses (9 units versus 10 units), universal versus left-sided colitis (9.2 units versus 12.5 units), medically versus surgically treated patients (14 units versus 9.2 units), or subtotal colectomy versus total proctocolectomy (8.7 units versus 10.6 units). Both the operated and nonoperated leukemic patients required massive and prolonged transfusion: 23 and 30 units, respectively. It is of interest that the two patients in whom reoperation was carried out for continued or recurrent rectal hemorrhage required more blood in the preoperative period than any of the patients other than those with leukemia; the patient who developed Crohn's disease in his pelvic pouch required 17 units preoperativdy, and the
THE AMERICAN JOURNAL OF SURGERY
VOLUME 159
JUNE 1990
551
ROBERT ET AL
25 Patients
I Operated 22
I Non-Operated I
I
I
I
I
Subtotal Colectomy (Sl'C) 17 (14 MF, 2 H, 1 IS)
I
No further operation 2 i
IS 1
Intact Rectum
I i
Well at 12 yes
Well at 10 yes
I
Total Pmctocolectomy (1"PC)
(4 APR,
Reoperated 14
I
1 Died at 9 mos of leukemia
1
2 Well at 3 and 4 yrs
leukemia
I
I
Died of postop sepsis 1
1 MS + IAA)
1 Died at lS mos of
3 Well at 6.3 yrs (1-15 yes)
Well at 5 yes
Abbreviations:
APR 7
MS + IAA 7
Well at 12.5 yes (3-22 yrs)
6 well at 16.7 mos (1 too-2 yes) 1 anastomotic healing problem at 11 mos
UC = Ulcerative Colitis IS = Ileosigmoidostomy MF = Mucous Fistula H = Hartmann's Closure APR = Abdomino Perineal Resection MS + IAA = Mucosal Stripoing + Ileoanal Anastomosis
Figure 1. Surgical management and outcome of severe hemorrhage in 25 patients with ulcerative colitis.
patient who bled subsequently from a rectosigmoid polyp required 12 units. Five patients underwent total proctocolectomy, four with terminal ileostomy and one with primary mucosal stripping and ileo-anal anastomosis (Figure 1). In one leukemic patient, an emergency total proctocolectomy was performed because of the severity of the rectal bleeding, requiring 23 units of blood over 16 days, with no thought given to preserving the rectum. The remaining four total proctocolectomies were carried out as semielective procedures. Three of the five total proctocolectomies took place in the period from 1980 to 1984 (Figure 1). Seventeen patients had subtotal colectomies, usually with a mucous fistula (14 patients). A Hartmann closure was performed in two patients with severe sigmoid inflammation in order to minimize the risk of recurrent hemorrhage. Finally, one patient with a quiescent rectal mucosa had a one-stage ileosigmoidostomy. Most of these subtotal colectomies (10 of 17) were emergency procedures. Patients operated on for severe bleeding were usually placed in Lloyd-Davies stirrups so that the rectum could be examined during and following the abdominal colectomy. Continued massive rectal hemorrhage requiring immediate proctectomy was not seen in any patient following subtotal colectomy. The only postoperative death occurred on the 41st postoperative day in a male patient with a 4-month history of ulcerative colitis taking high-dose steroids following emergency subtotal colectomy; he had a perforated duo552
T H E A M E R I C A N J O U R N A L OF S U R G E R Y
denal ulcer, duodenocutaneous fistula, and peritoneal sepsis. There were two late deaths caused by acute promyelocytic leukemia 9 and 18 months following the major hemorrhage; the first patient had undergone emergency total proctocolectomy, while the second had been treated nonoperatively. Both died of multiple bleeding complications culminating in cerebral hemorrhage. Twelve patients experienced 15 early postoperative complications, most of them infectious in nature; 6 of these patients, including the one who died, required reoperation. These complications were proportionately distributed between emergency and semielective procedures and between subtotal colectomy and total proctocolectomy (Figure 2). Two of 11 patients who underwent emergency subtotal colectomy had postoperative bleeding from the rectal remnant. The first patient, who had persistent hemorrhage, underwent proctectomy with mucosal stripping, ileal pouch, and ileoanal anastomosis at day 25 while still taking high-dose steroids. Since Crohn's disease developed in the pouch, this patient should be classified as having indeterminate colitis. The second developed a major hemorrhage from a pseudopolyp 4 months after subtotal colectomy, after a relatively quiescent period and underwent emergency abdominoperineal resection in 1979. Two of the three patients treated medically were followed up at 3 and 4 years, respectively; although one is doing well, the other has experienced one flare-up each year requiring transient steroids. The third patient died of
V O L U M E 159
J U N E 1990
HEMORRHAGE IN ULCERATIVE COLITIS
Operated Patients 22
I
I
STC 17
TPC 5
I
I
EMERGENCY 10
I
SEMI-ELECTIVE 7
Persistent/Recurrant Rectal Bleeding*
2
Per/orated DU* --) Septic Death
1
Intraabdominal Abscess
1
Wound Infection Small Bowel Obstruction
1 2
Psychotic Reaction
2
I
EMERGENCY 1
Peritonitis*
1
Sepsis (Fungus)
1
Wound Infections
2
SEMI-ELECTIVE 4
Perineal Wound Bleed*
4 Complications in 3 Patients
Perineal Abscess*
*Complication requiring reintervention
9 Complications in 7 Patients
Figure 2. Complications following surgery in 22 patients with severe hemorrhage from ulcerative colitis.
TABLE II I n c i d e n c e a n d M o r t a l i t y of S e v e r e H e m o r r h a g e
In U l c e r a t i v e Colitis ( U C )
Patients (n) Author & Reference
Date of Publication
with Hemorrhage
with UC
% with Hemorrhage
n
Deaths %
Sioan et al [ 15] Brown et al [ 16] Truelove & Witts [ 17] Bruce & Cole [18] Edwards & Truelove [ 1] Watts et al [ 19] Jalan et a/[20] Binder eta/ [ 7] Block et al [8] Bobblo eta/ [21] Goligher et al [22] Albrechtsen et al [23] Oakley et al [24] Vecchi et al [25] Present series
1950 1951 1955 1962 1964 1966 1969 1975 1977 1977 1980 1981 1985 1986 1987 1959-1983 1984-1986
22 O 1 5 21 1 6 5 16 2 0 6 11 5
2,000 147 213 118 624 151 399 111 39 46 258 132 311 125
(1.1) (0) (0.5) (4.2) (3.4) (0.7) (1.5) (4.5) (--) (4.3) (O) (4.5) (3.5) (4)
12 NR 1 1 NR NR 3 O -1 -O NR NR
(55) -(100) (20) --
16 9
1,156 --
(1.4) --
1 0
(5.3) (0)
(50) (0) (8.7)* (50) -(0) ---
Type of Operation NR NR None STC ? NR TPC ? NR 4 TPC, 1 STC 5 TPC, 11 STC TPC 4 TPC, 2 STC STC TPC 5 TPC 17 STC
* Patients with ulcerative colitis and Crohn's disease. NR = not reported; STC = subtotal colectomy; TPC = total proctocolectomy.
leukemia 18 months after the episode of severe hemorrhage. As noted previously, 1 of the 17 patients who underwent subtotal colectomy died postoperatively. Seven of the 16 survivors ultimately required abdominoperineal resection (Figure 1), usually for persistent mucous and/ or bloody rectal discharge. Another seven patients underwent subsequent planned rectal mucosal stripping with ileoanal pull-through. The interval between subtotal colectomy and reoperation averaged 24.7 months for abdominoperineal resection (range: 4 months to 5 years, median: 20 months) and 4 months for rectal mucosal
stripping with ileoanal pull-through (range: 25 days to 4 months, median: 2.5 months). The two remaining patients with intact rectal segments are in remission at 8 and 12 years after subtotal colectomy. COMMENTS The frequency of severe hemorrhage reported in the literature ranges from 0 to 4.5% [5] with a mean of 2.2% [15-25], which is close to the 1.4% frequency observed for the 1959 to 1983 period of our series (Table II). However, this relatively rare complication accounts for approximately 10% of all urgent colectomies for ulcer-
THE AMERICAN JOURNAL OF SURGERY
VOLUME 159
JUNE 1990
553
ROBERT ET AL
ative colitis. From 1959 to 1980, 6 emergency or semielective colectomies were carried out at this institution for severe hemorrhage, 48 for toxic megacolon, and 7 for perforation without colonic dilatation, in an overall series of 613 patients [4,6]. This 10% figure is comparable to the 6% reported by Binder et al [7], but lower than the 41% reported by Block et al [8]. Among our patients with severe hemorrhage, there were a number of features similar to those previously reported in the literature (Table II) and in previous publications from this institution [4,26]: Patients were somewhat but not significantly younger than other patients with ulcerative colitis (25 versus 29 years) [18,26]; bleeding was a relatively early complication of ulcerative colitis (mean duration of illness: 2.6 years) [27]; universal coIonic involvement occurred more frequently than leftsided colitis [28] (21 of 25 cases or 84%, versus 60% in our overall series) [26]; and, finally, there was a relatively high occurrence of concomitant toxic megacolon [8,19,22,29,30] (6 of 22 operated cases or 27%, versus 10% in the overall series) [4]. Even though total proctocolectomy is generally recommended in the elective treatment of ulcerative colitis, many investigators advocate a two-stage procedure as being safer in emergency cases [31,32]. Reports in the mid-1970s by Binder et al [7] and Block et al [8,33], however, showed emergency total proctocolectomy to be just as safe as subtotal colectomy, thus prompting many to prefer one-stage total proctocolectomy. In contrast to the less definitive subtotal colectomy, though, total proctocolectomy precludes later restoration of intestinal continuity by ileo-anal anastomosis. Does subtotal colectomy carry an increased risk of recurrent bleeding when it is performed specifically for severe hemorrhage? Various reports suggest that the incidence of postoperative rectal stump bleeding ranges from minimal to 44.1% [23,34]. Two of our 17 patients continued to bleed following subtotal colectomy and required secondary emergency procedures, one abdominoperineal resection at 4 months and one mucosal proctectomy at 25 days after subtotal colectomy. As already noted, the latter patient developed granulomas characteristic of Crohn's disease in the pouch. The other refused early proctectomy and had recurrence of bleeding several months later. Although these two patients represent a 12% bleeding recurrence rate, the unusual pathologic findings in one and the delay in recurrence in the other suggest that the rectum can be preserved, except in the rare case of massive hemorrhage clearly originating in the rectum. A 12.6% incidence of profuse postoperative rectal bleeding has been reported by Korelitz et al [9] in a separate series of 136 patients with residual rectal stump, 126 of whom (93%) underwent secondary abdominoperineal resections. In our series, 14 of 17 patients (82%) underwent subsequent proctectomy. If the severe hemorrhage appears to originate in the rectum or appears to be diffuse in origin, the patient should be placed in LloydDavies stirrups. Continued massive bleeding from the rectal segment observed during surgery, which was not found in any patient in this series, is an indication for 554
immediate total proctocolectomy rather then subtotal colectomy. Although primary mucosal stripping with ileoanal anastomosis was successful in one patient in this series who underwent a semielective procedure, such a procedure should probably not be considered routinely in the surgical handling of acute hemorrhage, especially in emergency surgery for massive hemorrhage. On the other hand, subtotal colectomy still appears to be a feasible alternative in view of the advantage of preserving the option of secondary ileoanal anastomosis. Contraindications might include massive hemorrhage clearly originating from the rectum, certain cases of leukemia with very limited life expectancy, and such advanced age that ileoanal pull-through is an unlikely prospect.
The authors present a small number of patients to address the important clinical problem of severe hemorrhage in ulcerative colitis. This complication is relatively uncommon, but is occasionally the principal indication for emergency surgery. The issue of whether to remove the rectum or preserve it for subsequent rectal mucosal stripping is the major point of discussion. REFERENCES 1. Edwards FC, Truelove SC. The course and prognosisof ulcerative colitis. Gut 1964; 5: 1-26. 2. Farmer RG. Clinical features and natural history of inflammatory bowel disease. Med Clin North Am 1980; 64: 1103-15. 3. Smith JN, Winship DH. Complications and extraintestinal problems in inflammatory bowel disease. Med Clin North Am 1980; 64: 1161-71. 4. Greenstein AJ, Sachar DB, Gibbs A, et al. Outcome of toxic dilatation in ulcerative colitis. J Clin Gastroenterology 1985; 7: 137-44. 5. Goligher JC, de Dombal FT, Watts J McK, Watkinson G. Ulcerativecolitis. Baltimore:Williams & Wilkins, 1965:123, 15074, 227-8. 6. GreensteinAJ, AufsesAH Jr. Differencesin pathogenesis,incidence and outcome of perforation in inflammatory bowel disease. Surg Gynecol Obstet 1985; 160: 63-9. 7. Binder SC, Miller HH, Deterling RA. Emergencyand urgent operations for ulcerative colitis. Arch Surg 1975; 110: 284-9. 8. BlockGE, Moossa AR, SimonowitzD, Hassan SZ. Emergency colectomyfor inflammatoryboweldisease. Surgery 1977; 82:5316. 9. KorelitzBI, Dyck WP, Klion FM. Fate of the rectum and distal colon after subtotal colectomyfor ulcerative colitis. Gut 1969; 10: 198-201. 10. Cello JP. Ulcerative colitis. In: Sleisenger MH and Fordtran JS, eds. Gastrointestinal disease. Philadelphia: WB Saunders, 1983: 1122-67. 11. Marshak RH, Lindner AE, Maklansky D. Radiology of the colon. Philadelphia: WB Saunders, 1980: 64-119. 12. Waye JD. The role of colonscopyin the differentialdiagnosisof inflammatoryboweldisease. Gastrointest Endosc 1977;23:150-4. 13. Morson BC. Pathology of ulcerative colitis. In: Kirsner JB, Shorter RG, eds. Inflammatory bowel disease. Philadelphia: Lea and Feibiger, 1980: 281-95. 14. Edwards FC, TrueloveSC. The course and prognosisof ulcerative colitis. Gut 1963; 4: 299-315. 15. Sloan WP, Bargen JA, BaggenstossAH. Local complications of chronic ulcerativecolitis based on the study of 2000 cases. Proc Mayo Clin 1950; 25: 240-4. 16. Brown ML, Kasich AM, Weingarten B. Complications of
THE AMERICAN JOURNAL OFSURGERY VOLUME 159 JUNE 1990
HEMORRHAGE IN ULCERATIVECOLITIS
chronic ulcerative colitis. Am J Dig Dis 1951; 18: 52-4. 17. Truelove SC, Witts LJ. Cortisone in ulcerative colitis. Final report on a therapeutic trial. Br Med J 1955; 2: 1041-7. 18. Bruce D, Cole WH. Complications of ulcerative colitis. Ann Surg 1962; 155: 768-81. 19. Watts JMK, de Dombal FT, Goligher JC. Early results of surgery for ulcerative colitis. Br J Surg 1966; 53: 1005-14. 20. Jalan KN, Sircus W, Card WI, et al. A n experience of ulcerative colitis. I. Toxic dilation in 55 cases. Gastroenterology 1969; 57: 68-82. 21. Bobbio P, Zola C, Salcuni PF. La colectomia totale d'urgenza nella colite ulcerosa. Acta Bio-Med Ateneo Parmense 1977;48: 93100. 2 2 . Goligher JC, Duthie HL, Nixon HH. Surgery of the anus, rectum and colon. London: Bailliere Tindall, 1980: 689-826. 23. Albrechtsen D, Bergan A, Nygaard K, Gjone E, Flatmark A. Urgent surgery for ulcerative colitis: early colectomy in 132 patients. World J Surg 1981; 5: 607-15. 24. Oakley SR, Lavery IC, Fazio VW, Jagelman DG, Weakley FL, Easley K. The fate of the rectal stump after subtotal colectomy for ulcerative colitis. Dis Colon Rectum 1985; 28: 394-6. 25. Vecchi R, Arena N, Cunsolo A, Gozzetti G. Emorragie massive in corso di colite ulcerosa e di morbo di Crohn. Minerva Chir 1986; 41: 653-6. 26. Greenstein AJ, Sachar DB, Smith H, et al. Cancer in universal
and left-sided universal colitis: factors determining risk. Gastroenterology 1979; 77: 290-4. 27. Jalan KN, Prestcott RJ, Sircus W, et al. Ulcerative colitis. A clinical study of 399 patients. J R Coil Surg Edinb 1971; 16: 133851. 28. Both H, Torp-Pedersen K, Kreiner S, Hendriksen C, Binder V. Clinical appearance at diagnosis of ulcerative colitis and Crohn's disease in a regional patient group. Scand J Gastroenterol 1983; 18: 987-91. 29. Philippe I, Philippe M, Girard M. Les colectasies aigues au tours des recto-colites hemorragiques. Donnees actuelles. Lyon Medical 1972; 227: 491-503. 30. Huizenga KA. Medical treatment and prognosis of some local complications of chronic ulcerative colitis and Crohn's disease. In: Kirsner RS, Shorter RG, eds. Inflammatory bowel disease. Philadelphia: Lea & Febiger, 1980: 466-7. 31. Lulu DJ, Dragstedt LR. Primary total proctocolectomy for ulcerative colitis. Ann Surg 1974; 40: 417-20. 32. Flatmark A, Fretheim B, Gjone E. Early colectomy in severe ulcerative colitis. Scand J Gastroenterol 1969; 4: 505-10. 33. Simonowitz D, Block GE, Moossa AR, Hassan SZ. Emergency colectomy for inflammatory bowel disease. Review of Surgery 1977; 34: 359-62. 34. Moss GS, Keddie N. Fate of rectal stump in ulcerative colitis. Arch Surg 1965; 91: 967-70.
THE AMERICAN JOURNAL OF SURGERY VOLUME 159 JUNE 1990 555