POSTER PRESENTATIONS Poster presentations will be on display in the exhibit area beginning 10:00 am, Monday, May 3, and during the open exhibit hours.

A n o r e c t a l Disease/AIDS Management of Recalcitrant Perianal Wound in Crohn's Disease by Rectus Abdominis Muscle Flap Booth P1 A. Mitra

.........................

Philadelphia, PA

The incidence of anal p r o b l e m s in Crohn's disease varies greatly among published series anywhere from 8.5 to 94 percent. Although there is evidence that perianal Crohn's disease may heal effectively with treatment of the main disease successfully, persistent perianal w o u n d after protectomy or proctocolectomy has always been a difficult p r o b l e m for surgeons to manage. Poor healing in the perianal area is reported in many series. Despite all aggressive and repeated surgical procedures, wounds often fail to heal. Successful resolution of these wounds has b e e n reported by introduction of space filling vascularized tissue in this area. Various muscles and myocutaneous flaps have been used to cure perianal wounds in perianal recalcitrant wounds. In the present series we describe five such patients; these patients included four males and one female. The average patient underwent at least five d e b r i d e m e n t procedures and basic surgical m a n a g e m e n t for closure of these wounds. The patients' ages ranged from 18 to 34 years. One patient n e e d e d the p r o c e d u r e as a prerequisite for renal transplantation. A mean follow-up of six months to two years was performed. All the above patients were treated by transposition of rectus abdominis muscle to close the perianal wounds after excision and d e b r i d e m e n t of perianal Crohn's. During the present presentation, the operative techniques will be discussed. All patients following the abdominis muscle flap closure remained well healed. In conclusion, we feel that introduction of vascularized tissue to obliterate the dead space might be an answer to recalcitrant perianal wounds in Crohn's disease.

A History of Enema: From Clyster to Irrigation Booth P2 S.A. Walker

.......................

Royal Oak, MI

Recent interest in the use of "high colonic enema" as practiced in "alternative" cancer therapies, by health food faddists, and by klismaphiliacs has p r o m p t e d this review. Instillation of fluid into the rectum is described in some of the earliest known writings. The Indian Susruta Samhita tells of an enema apparatus made from the scrotum of a d e e r and a b a m b o o tube. Egyptian legend, as recorded by Pliny in his Natural History, ascribes the origin of the enema to observations of the Ibis, a bird P23

sacred to their culture. Physicians who administered enemas were known as "shepherds of the anus." In the Western hemisphere, the Mayans used enemas for medicinal and intoxicating purposes. The Catauxi tribe would give hallucinogens via this route to both themselves and their dogs before hunting. Courtly society in 17th century France required the clyster daily. Louis XIV received over 2,000 during his reign. Intestinal irrigation became popular around 1900, both diagnostically and therapeutically. Many reputable doctors utilized the technique, and a sophisticated technology developed. Cleansing of colonic "toxins" was considered to be one of the key benefits of this therapy and supported in the medical literature. Dr. Harvey Kellogg was one of its more colorful proponents.

Convenient and Economic Treatment of Hemorrhoids by Rubber Band Ligation Booth P3 I. Bayer, B. Pickovsky Petah-Tiqva, Israel and New York, NY The pathologic state called hemorrhoids has b e e n known in the medical literature for many years. Clinical experience suggests that very many p e o p l e of both sexes suffer from hemorrhoids. Controversy exists still as to the precise etiology and the exact pathologic findings of the disease. This fact contributes to the disagreement concerning the optimal treatment of hemorrhoids. The aim of this study is to conclude 12 years of experience in the treatment of hemorrhoids by rubber band ligation (RBL) in a proctologic outpatient clinic. During this period, 2,934 patients, 1,728 males and 1,206 females, aged 16 to 86 years, suffering from 2nd and 3rd degree hemorrhoids, were treated by this method. Sixteen hundred fourteen (55 percent) patients n e e d e d two or three visits, and 1,320 (45 percent) n e e d e d four or more visits. All patients were followed up six months and one year after completion of treatment. None of them required narcotics to relieve their pain, 352 (12 percent) n e e d e d some medication to appease the pain, and 205 (7 percent) n e e d e d sick leave of one to three days during the whole p e r i o d of treatment. One h u n d r e d thirty eight patients (4.7 percent) c o m p l a i n e d of recurring b l e e d i n g or prolapse of hemorrhoids, and they were treated by RBL. Forty seven patients (1.6 percent) who still complained after c o m p l e t i o n of RBL were referred for conventional hemorrhoidectomy. One of the greatest advantages of the ambulatory treatment of hemorrhoids by RBL is the quick and painless relief of the patient's agony. This method is safe and convenient and significantly saves hospitalization days, as well as absence from work. It results in cutting down on service cost, vacant beds, and surgery rooms in hospitals. There are decisive advantages of RBL for hemorrhoids treatment.

P24

MEETING ABSTRACTS

How Accurate is the Diagnosis of Anal Ulcers in AIDS Patients?

Booth P4 S.M. Cohen, S.L. Schmitt, F.L. Lucas, S.D. Wexner Ft. Lauderdale, FL Recent reports have suggested that routine microscopic evaluation of anal ulcer tissue from AIDS patients is not the most accurate way to diagnose viral infection. This study was undertaken to determine if either viral culture (VC) or immunohistochemistry (IHC) can improve the diagnostic accuracy as compared with the routine hematoxylin and eosin (H&E) staining. Specifically, we sought to identify inclusion bodies of cytomegalovirus (CMV) or herpesvirus (HSV) to assist in the diagnosis of CMV or HSV. All patients had clinical evidence of an anal ulcer or a nonhealing anal fissure. Duration of symptoms ranged from one week to three months (mean, six weeks). All specimens were submitted for viral culture in addition to routine H&E staining; IHC was also performed. Twenty-five paraffin-embedded anal ulcer biopsies were reviewed over a four-year period (1988-1992) from 23 male patients (age range, 27-73; mean, 37.4 years) with the diagnosis of AIDS or AIDSrelated complex (ARC). Routine H&E staining revealed six (22 percent) specimens with CMV inclusions. Four of these six reacted positively with IHC (67 percent) and one by viral culture (17 percent). In the remaining 19 specimens that did not reveal infection with CMV (78 percent), IHC was positive in two patients (10 percent) and viral culture in one patient (5 percent). Although HSV was not seen in any of the specimens on H&E staining, IHC was positive in one patient (3.5 percent) and viral culture reacted positively in eight (29 percent) specimens. Thus, IHC is a good confirmatory test for CMV inclusions and can be used in equivocal cases. However, neither test can replace the traditional meticulous and careful pathologist in the detection of viral inclusions.

Nonspecific Colon Ulcerations: A Disease of the Immunocompromised Patient Booth P5 EJ. Szilagy, J.B. Ryan . . . . . . . . . . . . . . . . . .

Detroit, MI

Although u n c o m m o n and of unknown etiology, nonspecific ulceration of the colon has been a perplexing problem since the entity was first described in 1832 by Cruveilhier. A review of the literature finds that approximately 150 cases of nonspecific ulcers of the colon have been reported since the turn of the century. Most recently, though, little has been Written about this disease process, especially in immunocompromised and immunosuppressed patients. In addition to a review of the literature to date, this report reviews three such cases: 1) a patient with myeloproliferative syndrome and incomplete Behfet's syndrome, 2) a patient on immunosuppressive agents after cardiac transplantation, and 3) a patient with HIV/AIDS. All patients were diagnosed by colonoscopy. On further workup utilizing multiple modalities, no specific etiology could be identified for any

Dis Colon Rectum, April 1993

of the three cases. Initially, these patients were treated medically in accordance with recent reports that a conservative approach has been successful in the majority of the cases. Two of the three cases reported, however, did not improve and required emergent surgical intervention. Although our experience was limited, this may suggest that immunocompromised patients with nonspecific ulcers of the colon will have a more fulminant clinical course requiring early surgical intervention.

The Use of Endoanal Uhrasonography in the Assessment

of_Anal Fistulas Booth P6 J.G. Williams, K.I. Deen, E. Grant, D. Kumar Birmingham, United Kingdom A reliable method of preoperative delineation of anal fistula tracts would be useful to the surgeon. Fistulography has proved disappointing. We have assessed the accuracy of endoanal ultrasound in identifying primary and secondary tracts in patients undergoing fistula surgery. Sixteen patients (three with Crohn's disease) with anal fistulas were examined preoperatively with a rotating 7mHz ultrasound probe inserted in the anal canal. Accurate differentiation between complex (more than one fistula tract in different planes [n = 12]) and simple fistulas (n = 4) was possible in all patients. In two patients, the clinical impression of a complex fistula was refuted by ultrasound and subsequent surgical exploration. Horseshoe tracts were identified in nine patients, and fluid collections, not evident on clinical examina tion, were identified in eight patients. External sphincter destruction was observed in three patients. Accurate identification of the internal opening of the fistula was only possible in two patients. Surgical findings matched endosonographic appearances in all patients. Uhrasonography was thought to aid surgical treatment in 14 of the 16 patients. Anal ultrasonography is an accurate and minimally invasive method of delineating the relationship of fistula tracts to the anal sphincters. Collections of pus are readily identified, but the internal opening is often difficult to define. The examination can be performed in the operating room at the time of surgery.

Is Pain After Anorectal Surgery Necessary?

Booth P7 O.H. Wiltz, J.E. Garcia, R. Feliciano, J. Martino Santurce, Puerto Rico Postoperative pain and discomfort is a major problem in anorectal surgery. Prior to 1987 we routinely used local, regional, and general anesthesia with significant postoperative pain requiring narcotics for several days. In 1988 we started to use single-dose spinal intrathecal morphine with encouraging results. We now report on 485 patients who underwent anal surgery since 1988. Intrathecal morphine was used in 245 patients, general

Vol. 36, No. 4

MEETING ABSTRACTS

in 100 patients, spinal in 95 patients, and local in 45 patients. The optimal dose of intrathecal morphine was 0.2 mg diluted in five to six mg of tetracaine dextrose solution in a single dose. All patients were followed one month after surgery. Patients treated with conventional anesthesia required postoperative narcotics in 96 percent of the cases and were quite uncomfortable limiting their ability to return to work early. However, 98 percent of patients treated with intrathecal morphine were pain free after surgery and went back to work on acetaminophen in 48 hours. Exposure and dilatation of the anal canal were superb. Complications were minimal; most commonly 20 percent of patients had itching with quick resolution, 0.5 percent had hypoventilation treated with one dose of NARCAN| and 5 percent had urinary retention. These data suggest that single-dose intrathecal morphine is effective and safe. Moreover, this technique eliminates postoperative pain and diminishes the amount of time lost from work in patients undergoing anorectal surgery. Combination Ciprofloxacin and Metronidazole in Severe Perianal Crohn's Disease

Booth P8 M. Solomon, R.S. McLeod, B. O'Connor, H. Steinhart, G. Greenberg, Z. Cohen . . . . . . Toronto, Ontario, Canada Fourteen cases of severe perianal Crohn's disease treated consecutively at the IBD Unit over the past year with a combination of ciprofloxacin (1,000-t,500 mg/ day) and metronidazole (500-1,500 rag/day) were reviewed. There were equal numbers of males and females with a mean age of 34.8 years at presentation. All had quiescent bowel disease. Perianal disease had been present a mean of 28.4 months. Six patients were on metronidazole a t presentation, and seven patients had had previous perianal surgery (mean, three operations/patient). Four patients had previous ileoanal J pouch anastomosis prior to the diagnosis of Crohn's, with a mean of 22.4 months between J pouch and perianal symptoms. All patients were symptomatic. Nine patients had complex fistula, six had anal canal ulceration, one had a rectovaginal fistula, and five had discharging abscesses. Seven patients had multiple perianal lesions. Physician assessment at a mean of 12 weeks after commencing therapy revealed that three patients healed, nine improved, one was unchanged, and one worsened, requiring defunctioning stoma. Thus, 12 of 14 (85 percent) showed benefit from combination therapy. Patients have been followed a mean of 6.4 months since commencing therapy. Five patients had therapy stopped at 12 weeks and have not required further treatment, six patients have required continuous low-dose therapy, three patients stopped therapy at 12 weeks and have subsequently required recommencing therapy for relapse. Thus, 9/14 (64 percent) have required continuous or repeat therapy. Present status is seven quiescent, six mild, one moderate, and zero severe perianal disease. Although these initial results suggest that combination ciprofloxacin and metronidazole is effective in treating severe perianal Crohn's disease, a randomized, controlled trial is necessary to

P25

confirm efficacy and determine the role of continuous therapy.

Autonomic Nerves Cross the Plane of Posterior Rectal Dissection

Booth P9 A.P. Meagher, W.J. Adams, D.Z. Lubowski Sydney, Australia The plane of surgical dissection posterior to the mesorectum has been called "embryological," "avascular," "anatomical," "surgical," or more recently "Holy." Yet the developmental anatomy of this region is not well understood, and anterior branches of the median sacral artery are known to cross the plane. We have noted midline anterior branches of the presacral autonomic plexus crossing this plane to supply the rectum. Following an initial cadaveric study, this study was undertaken to evaluate the frequency of these nerves and assess the ability to visualize them operatively. In 10 patients undergoing full mobilization of the rectum, the posterior plane was displayed in a bloodless manner. Structures crossing this plane were traced to their origin and destination, were photographed, and were then biopsied. In all 10 cases, nerves were demonstrated to cross this plane, and they can be found at all levels, from $1 to $5. They originate from the presacral autonomic plexus, which is far broader than described in standard anatomic and surgical texts. Midline, anterior nerves also arise from the pelvic parasympathetic nerves. The posterior "plane" of dissection is crossed by nerves, and the concept that rectal nerve supply only originates from the inferior mesenteric plexus and the lateral pelvic plexus is incorrect. The nerves described here may have functional importance.

Polyps/Endoscopy Adenomatous Colon Polyps (AP) in Patients with a History of Breast Cancer (BC)

Booth P10 C.N. Ellis, H.W. Boggs,* D.J. Coyle, W.S. Blakemore Birmingham, AL and Shreveport, LA The majority of AP are tubular adenomas, and only a very small fraction will become malignant. AP that have a villous component are more likely to progress to cancer. While adenocarcinomas of the breast and colon are associated, there is not an increased incidence of AP in patients with BC. This suggests that the link between colon and breast cancers is during tumor promotion, not tumor initiation. To determine if villous histology is more c o m m o n in the AP of patients with BC, the charts of 1,781 female patients who underwent total colonoscopy were reviewed. One hundred fifty eight of these patients had a history of BC. AP were found in 63 patients with a history of BC (39.8 percent) and 551 (33.9 percent) of the patients without a history of BC (P = NS). The polyps in 13 (20.6 percent) patients with a history of BC had a villous component, compared with 47 (8.5

P26

MEETING ABSTRACTS

percent) of the patients without a history of BC (P < 0.01). These data suggest that, while patients with BC do not have an increased incidence of AP, the polyps are more likely to have a villous component. The etiology of these findings is unclear but may explain the association between colon and BC.

Colonoscopy and Digital Cytometry of DNA in the Clinical Follow-Up o f Patients with Long-Standing Ulcerative

Dis Colon Rectum, April 1993

measurements were performed on all subjects prior to and postsigmoidoscopy. Physiologic recordings ofheart rate and mean arterial pressure (MAP) were recorded before, during, and after the procedure. Patients who listened to self-selected music tapes during the procedure had significantly decreased STAI scores ( P < 0.02), HR ( P < 0.03), and MAP ( P < 0.001) in comparison with the control subjects. The results of the study indicate that music is an effective anxiolytic adjunct to flexible sigmoidoscopy protocol.

Colitis

Booth P l l Colorectal & Anal Cancer

S.C. Nahas, R. E1 Ibrahim, H.W. Pinotti S~o Paulo, Brazil Twenty-one patients having ulcerative colitis (UC) for a period of seven or more years, and considered as being in the risk group for developing large bowel displasia, were studied according to: 1) colonoscopic examination for macroscopic identification of mucosal alterations and obtainment of multiple biopsies; 2) histologic examination of mucosal biopsies according to well-established morphologic criteria; and 3) quantification of mucosalcell DNA using digital cytometry of Feulgen-stained tissue sections. The findings of this study based on the three methods were: 1) colonoscopic examination with multiple biopsies was efficient in providing accurate diagnosis in all cases studied besides allowing the successful endoscopic resection of an aneuploid neoplastic lesion in one patient; 2) histologic examination was helpful in confirming colonoscopy findings; and 3) digital cytometry of DNA was effective in elucidating cases with doubt in histologic examination. Digital cytometry revealed diploid DNA histogram in all phases of UC. Aneuploid histograms occurred only in cases with association of UC and neoplastic development. CONCLUSIONS: This pioneer study in Brazil got to results which showed that these three methods should be utilized in association to increase the accuracy in detecting neoplastic mucosal-cell alterations occurring during the course of long-standing ulcerative colitis.

Effect of Music Therapy on State Anxiety in Patients Undergoing Flexible Sigmoidoscopy Booth P12 K. Palakanis, J. DeNobile, B. Sweeney, C. Blankenship Bethesda, MD Patient anxiety related to flexible sigmoidoscopy can negatively affect acceptability and compliance with screening protocol, complicate and prolong procedure time, and potentially result in prematurely aborted procedures. Music has been recognized through research as a safe, inexpensive, and effective nonpharmaceutical anxiolytic agent. An experimental study was performed on 50 adults scheduled for outpatient sigmoidoscopy. The control group received a standard sigmoidoscopy protocol. Subjects in the experimental group received the standard protocol with the addition of listening to music throughout the procedure. State-Trait Anxiety Inventory (STAI)

Photodynamic Therapy of Adenocarcinoma of the Colon Booth P13 J.H. Sun, G.V. Stiegmann, J.G. Kim, N.W. Pearlman Denver, CO Photodynamic therapy (PDT) using rhodamine-123 and argon laser light causes significant tumor regression in rodent colon adenocarcinoma when compared with treatment with rhodamine alone and with controls. Similar results are seen in adenocarcinoma in vitro cell cultures. Dimethylhydrazine was utilized to induce adenocarcinoma in the rat colon by a well-established multistep biotransformation. Rats were divided into three groups: control, rhodamine-123 alone, and rhodamine + laser treatment. Argon laser was utilized at nonthermal levels, and the tumor size was determined. size in cm Control Rhodamine only Rhodamine + laser P _ 0.005.

1st wk 1.0 1.7 1.7

2nd wk 1.6 0.9 1.0

3rd wk 2.6 0.8* 0.35*

Cell viability counts with trypan blue exclusion tech nique in vitro cultures revealed the following. control Laser only Rhodamine only Rhodamine + laser

100% 98.1% 47.9% 15.6%

CONCLUSION: Photodynamic therapy using rhodamine-123 and argon laser light causes tumor regression in rodent adenocarcinomas of the colon. In vitro studies of the same adenocarcinoma cell line confirm the effect of photodynamic therapy. The mechanisms of cellular injury are consistent with the triplet-mediated Type I (free radical) mechanism of cell injury.

Is DNA Ploidy Pattern of Prognostic Significance in Young Patients with Colorectal Cancer? Booth P14 R.V. Landes, R.C. Hankin, D.C. Barkel, W.L. Beauregard, M.D. Poulik, J.C. Chen, C. Dmuchowski, E. Cho Royal Oak, MI

Vol. 36, No. 4

MEETING ABSTRACTS

Quantitative DNA analysis was performed on 46 patients diagnosed with colorectal cancer at less than 40 years of age. Patients with inflammatory bowel disease and familial polyposis were excluded. Twenty-four patients were classified as diploid, 15 as aneuploid, and three as tetraploid. Stage at presentation was similar in diploid and nondiploid groups (two-tailed P = 0.3432). However, there was a highly significantly greater likelihood of early death for aneuploid tumors (two-tailed P < 0.01) and the collective nondiploid group (twotailed P < 0.01). We conclude that DNA ploidy pattern is an independent predictor of survival in young patients with colorectal cancer.

Colorectal Cancer in Young Patients: Characteristics and Outcome Booth P15 P.Y. Lee, W.S. Fletcher, E.S. Sullivan, J.T. Vetto Portland, OR Controversy still exists regarding the features and prognosis of colorectal cancer in younger patients. We reviewed the records of 62 patients 40 years of age and younger with adenocarcinoma of the colon and rectum, treated and followed at our institution between 1968 and 1991. These patients represented 3.1 percent of our total colorectal patient population during that period. Their mean age was 34.5 years, with the youngest patient being 18 years old. Modified Dukes stages at presentation were 8 percent A, 20 percent B, 23 percent C, and 48 percent D. Underlying inflammatory bowel disease was present in 21 percent of patients and was proportionately distributed between high (C and D) and low (A and B) stages. Half of the Stage D patients had high-grade lesions, compared with only 20 percent of lower-stage patients (P = 0.037). All but two patients had operative exploration; 36 (60 percent) had complete resection of all gross disease. With a mean follow-up of 98.2 months, five-year overall survival for Stage A disease was 100 percent but dropped to 85 percent, 40 percent, and 7 percent for Stages B, C, and D, respectively. Compared with published figures for the general population, younger patients with colon and rectal cancer tend to present at a more advanced stage but have similar stage-related survival.

Direct Contact Radiotherapy: An Option in Select Patients with Rectal Cancer Booth P16 T.L. Hull, I.C. Lavery, J.P. Saxton . . . . . .

P27

anal verge (DAV); mean follow-up (F/U); and recurrence (RECUR). In demographically similar groups, 199 patients received DCR, with 126 being treated with curative intent.

TS

Cure

89

Recur

37

DAV

P/U

(cm)

Well

Mod

TD Poor

(cm)

(months)

2.3 (1-5) 2.5 (1-3)

21

53

3

6

25

1

5.7 (0-12) 5.2 (0 10)

50.2 (0-136) 54.5 (5-132)

NO significant difference with Fisheffs exact test.

With a mean time to recurrence of 16.1 months (156), 27/37 tumors recurred locally and 10/37 distally. Following additional treatment, 17/37 patients were free of cancer at F/U. Direct contact radiotherapy will initially cure 71 percent of select rectal cancers. An additional 13 percent of patients can be cured with secondary treatment. Tumor size, differentiation, and distance from the anal verge do not influence recurrence. Follow-up is needed as some recurrences are detected as long as 56 months post-DCR.

Adjuvant Radiotherapy and Colorectal Anastomosis: Animal Evidence of Cancer Risk Booth P17 J. McCue, J. Sheffield, R. Phillips London, United Kingdom Long-term consequences of radiotherapy on colorectal anastomoses are unknown. We explored the effect of preoperative irradiation on cellular proliferation and carcinogenesis at colorectal anastomoses in rats. METHOD: In 54 adult male F344 rats, the rectum and lower descending colon were exposed to 16 Gy orthovoltage x-rays, v i a anterior and posterior portals. One week later, animals underwent a 5-ram colotomy and sutured closure within the radiation field. Postoperatively, 18 rats were killed at 4, 13, and 26 weeks. Crypt cell production rate (CCPR) at 4 and 13 weeks was assessed by the stathmokinetic method and was compared with values from 36 rats that had undergone operation but not irradiation. RESULTS:Tumors developed at the irradiated anastomosis in 13 animals (27 percent) and in the adjacent irradiated colon or rectum in just two animals (4 percent (P < 0.002, Fisher's exact). Anastomotic proliferative changes were enhanced by irradiation (P < 0.05).

Cleveland, OH

Direct contact radiotherapy (DCR) delivers high-dose irradiation with limited penetration and is an established modality for the curative treatment of select rectal tumors. The purpose of this study is to review the experience from one institution with DCR. All patients with rectal cancer treated with DCR between 1973 and 1992 were studied. Collected data included:.tumor size (TS); tumor differentiation (TD) classified as well, moderate (MOD), and poor; distance from

CCPR (Cells/Crypt/Hr) DXT

No DXT

Weeks 4 13

Anast.

Colon

Anast.

Colon

13.1 11.6

7.0* 8.9*

10.9 8.8

8.0* 7.3**

* P < 0.005, ** P < 0.05

cf.

paired colon.

CONCLUSIONS: Radiation carcinogenesis is greatly enhanced at colorectal anastomoses and may be due to

P28

MEETING ABSTRACTS

exaggerated mucosal proliferative changes. This has disturbing implications in view of the increasing use of adjuvant radiotherapy for rectal carcinoma.

The Value of Endoluminal Ultrasonography in Assessing Visceral Pelvic Invasion in Recurrent Rectal Cancer Booth P18 J. DiPierro, J. W. Milsom, V. W. Fazio, S. A. Strong Cleveland, OH The purpose of this study was to assess the ability of endoluminal ultrasonography (ELUS) to predict extent of pelvic organ involvement in recurrent rectal cancer (RRC). MATERIALS AND METHODS: Seventeen patients (pts) (6 M, 11 F), aged 44 to 76 years, with RRC underwent preoperative ELUS with a 360 ~ MHz transducer, with 15/17 undergoing CT scan. ELUS and CT were c o m p a r e d with operative findings and surgically removed specimens. RESULTS: ELUS provided information regarding pelvic organ invasion not obtained on CT in 10/15 pts and not noted on clincal exam in 12/17. ELUS correctly predicted rectal wall invasion in 8/11, vaginal invasion in 4/5, bladder invasion in 2/3, and prostatic invasion in 1/1, with overall correct prediction of organ invasion in 15/ 20 instances. One of 17 pts was overstaged by ELUS and 0/15 by CT. Two of 17 pts were understaged by ELUS and 6/15 by CT. Two pts had scar formation precluding accurate evaluation by ELUS. All ELUS was done in an office setting. CONCLUSIONS: ELUS is a convenient and accurate method of assessing extent of organ involvement in RRC and appears to be more accurate than pelvic CT. Surgeons should consider ELUS in the preoperative investigation of recurrent rectal carcinoma.

Current Follow-Up Strategies after Resection of Colon Cancer: Results of a Survey of ASCRS Members Booth P19 A.M. Vernava, III, W.E. Longo, T.P. Wade, K.S. Virgo, M.A. Coplin, F.E. Johnson . . . . . . . . . . . . St. Louis, MO BACKGROUND: Follow-up after resection of colon cancer has important clinical and financial implications for patients and society, yet the ideal surveillance strategy is unknown. PURPOSE: To determine the current followup practice pattern of a large, diverse group of experts. METHODS: ASCRS m e m b e r s were asked, v i a a detailed questionnaire, how often they request nine follow-up measures (clinic visit, CBC, liver function tests, CEA, CXR, bone scan, CT scan, colonoscopy, sigmoidoscopy) in their patients treated for cure with TNM Stage I, II, III colon cancer over the first five posttreatment years, RESULTS: Six h u n d r e d eight of 1,663 (37 percent) returned the survey. Sixty-nine (11 percent) were retired and did not c o m p l e t e the questionnaire. Forty-two (7 percent returned patients to their primary physician for postoperative follow-up. Four hundred ninty-seven of 1,663 (30 percent) c o m p l e t e d the survey and provided

Dis Colon Rectum, April 1993

evaluable data. Routine clinic visits and CEA levels were the most frequently performed evaluations for each of the five years. Greater than 80 percent of surgeons see their patients every three to six months for years 1 and two and then every 6 to 12 months for years three, four, and five. CEA level was the most frequently performed test and was obtained by the vast majority of respondents. Eighty percent obtained CEA levels every three to six months for years one and two. Sixty-nine percent requested CEA levels for six to twelve months for year three. Eighty percent requested CEA levels every 6 to 12 months for years four and five. Colonoscopy was performed annually by 47 to 72 percent, d e p e n d i n g on year. CT scan and bone scan were requested by 13 to 46 percent of ASCRS surgeons d e p e n d i n g on year. There was greater variation in the pattern of use of CBC and liver function tests. CONCLUSION: The majority of ASCRS surgeons e m p l o y frequent clinic visits and CEA levcls early in their routine follow-up regimen. Colonscopy and CXR are generally p e r f o r m e d annually. Bone scan and CT scan are not routinely obtained. Liver function tests and CBC are o r d e r e d with widely divergent frequencies. The intensity of follow-up does not vary markedly by TNM stage.

Treatment of Epidermoid Anal Cancer (AC) in the U.S. Veteran: A Nationwide Outcome Study Booth P20 W.E. Longo, A.M. Vernava, III, F.E. Johnson, T.P. Wade, M.A. Coplin, K.S. Virgo . . . . . . . . . . . . . . . St. Louis, MO BACKGROUND: Reports of treatment results in AC patients have generally described small, selected populations. AIM: To evaluate treatment outcome in patients with e p i d e r m o i d AC in Veterans Affairs (VA) Medical Centers (MCs). METHODS: Using national Va computer data sets, we analyzed all patients with the ICD-9 diagnostic code (154.2) for AC in the VA population from 1987 to 1991. Patient demographics, histopathology, tumor size, method of treatment, and survival data were required for patients to be judged evaluable. RESULTS: Four hundred-five AC, were identified by computer search. In 132, patients status of tumor size, lymph nodes, and distant metastases were evaluated. Patients with melanoma (six) and adenocarcinoma (seven) were exCluded. One hundred nineteen patiens with e p i d e r m o i d AC were thus evaluable. The mean age of these 119 patients was 61.3; all were males. Twenty-two (18 percent) were treated by abdominioperinal resection (APR) as primary therapy. Nineteen (16 percent) were initially treated by local excision only. The outcome of 78 patients, who received chemotherapy plus radiation (CR), was analyzed according to tumor size (mo = months).

Size <2 cm 2-5 cm >5 cm

n 46 25 7

Mean Survival 28.4 --. 22 mo. 27.5 + 18 mo. 18.5 -+ 16 mo.

Vol. 36, No. 4

MEETING ABSTRACTS

The mean dose of radiation was 4,329 cGy. Eighty percent of (CR) patients received 5-FU/mitomycin, 14 percent received 5-FU/cis-platinum, and 5 percent received all three drugs. Twenty-four percent sustained toxicity from CR; in !4 percent this required CR termination. Seventeen percent of CR patients developed local recurrent. Mean survival after primary APR was 49 months; mean survival after primary CR was 27 months. CONCLUSIONS: CR is currently used in sixty-six percent of VAMC patients with epidermoid AC. Eighty-six percent of patients complete CR. Tumor size appears to affect prognosis. This is the first national U.S. study of AC and should serve as a benchmark against which other data sets can be measured.

Fecal I n c o n t i n e n c e & C o n s t i p a t i o n

Static-Dynamic Gracilopasty: A Reconstructive Technique of the Anal Sphincter After A.Pe.R. of the Rectum Booth P21 E. Cavina, C. Menconi, G. Ghiselli, M. Seccia Pisa, Italy A surgical technique of sphincteric reconstruction after abdominoperineal resection of the rectum (APeR) for cancer is described. From Jaunary 1985 to October 1992, 76 double electrostimulated graciloplasties (ESG) were performed: 70 simultaneously with an APeR for lower rectal cancer (Group 1) and six as a reconversion of a previous abdominal colostomy in already "free from cancer" patients (Group 2). The admission criteria for Group 1 were lower rectal cancer nonresectable with a sphincter-saving AR, a very poor quality of life, and a verified absence of repetitive lesions for Group 2. The principle of the technique is to obtain a continent perineal colostomy by means of lowering the colonic stump to the perineum and transposing both gracillus muscles from the thigh to encircle the stump. The aim of the double graciloplasty is to stimulate, as closely as possible, the structure of the anal sphincter: one gracillis, drawn behind the lowered colonic stump, "simulates" the puborectal sling, while the other encircles the periheal colostomy in an "alfa" fashion. In 57 patients, both muscles were submitted to ES using electrodes placed on the neurovascular bundle and to protocols of "intermittent" ES, while the last 19 cases underwent "chronic" stimulation with implantable pulse generators (PGs). In the first service, two ESGs had to be reconverted to abdominal colostomies due to a colonic ischemia during the first 12 postoperative hours. In all cases, however, no postoperative mortality (30 days) was observed. The overall complication rate was high, mainly due to mild infections (30 percent); severe complications (ischemia) were observed in 2.6 percent of cases. Over seven years, 19 patients died (14 from cancer recurrence), and one was lost during follow-up; a complete functional evaluation regards 52 of the remaining 56 patients. Four patients had to be withdrawn from our functional study due to abdominal colostomies: one case of late (three years) perineocele, one pelvic recurrence, one colonic

P29

fistula, and one perineal sepsis. The follow-up study was performed at four-month intervals with continence scores (Corman's modified), electromanometry (EMM), and defecography (optional). Continence scores of 73 percent good, 25 percent fair, and 2 percent poor (one case) were shown. It is to be pointed out that, in the last 9 patients (IPG + i.m. electrodes), 100 percent good results were noted. A statistically significant difference was observed in EMM values before and after routine use of i.m. electrodes + quadripolar IPG: 8.7 mm Hg of mean resting pressure and 42.3 m m H g of mean squeeze pressure vs. 40.8 and 122 m m H g pressures.

Endoluminal Ultrasound is Preferable to Needle EMG or 3-D Vector Manometry in Mapping Sphincteric Defects Booth P22 J.J. Tjandra, R. Lowndes, J.W. Milsom, H. McKirdy, T. Schroeder, L.E. Hughes, V.W. Fazio Cardiff, United Kingdom and Cleveland, OH Assessment of complex sphincteric defects by digital rectal examination and intraoperative dissection can be difficult. Adjunctive investigations such as endoluminal ultrasound (ELUS), needle electromyography (EMG), and three-dimensional (3-D) computerized vector manometry may provide useful and objective information. In a series of 22 patients, ELUS of the anal canal accurately detected defects in the internal anal sphincter, (IAS) in 17/17 patients, defects in the external and sphincter (EAS) in 16/16 patients, and integrity of both sphincters in four patients. Presence of sphincteric defects was confirmed by operative findings. Needle EMG accurately detected defects in EAS in all 10 patients examined but was associated with more pain than ELUS (pain score, 10 vs. 4, 10 being most painful). Directional pressures recorded across the anal sphincters using a continuous pullout technique were used to provide vector-volume analysis and a computerized 3-D pressure profile of the anal canal in 9 of 22 patients. This analysis accurately detected the presence of defects in both EAS and IAS in six of nine patients examined. CONCLUSION: ELUS seems preferable to both needle EMG and 3-D computerized vector manometry in mapping anal sphincteric defects and provides a useful anatomic adjunct to physiologic studies of the anal in patients with fecal incontinence.

Patterns of Male Fecal Incontinence Booth P23 S.M. Sentovich, L.J. Rivela, A.G. Thorson, G.J. Blatchford, M.A. Christensen . . . . . . . . . . . . . . . . . . . . . . Omaha, NE Fecal incontinence has been investigated thoroughly to only female patients. Recognition of clinical and manometric patterns of male fecal incontinence may be valuable for planning treatment and evaluating clinical outcome in male patients. A retrospective review of 26 men complaining of fecal incontinence but without any obvious physical or proctoscopic findings consistent with

P30

MEETING ABSTRACTS

incontinence was undertaken to characterize these patterns. Twelve men (46 percent) had true incontinence characterized by a loss of control over gas, liquid, and/ or stool. Fourteen men (54 percent) were "leakers" and complained about smears of stool staining their underclothes a few hours after a bowel movement. In both the incontinent and leaker groups, abnormal anorectal sensation, prolonged pudendal .nerve terminal motor latencies, and a history of prior anorectal operations occurred with an equal frequency. Manometry revealed lower mean maximum resting (MMR) pressures and mean maximum squeezing (MMS) pressures in both the incontinent (P < 0.001) and leaker (P < 0.05) groups as compared with normal controls. The leakers, though, had significantly higher MMR pressures (P < 0.05) and MMS pressures (P < 0.005) than the incontinent men. The resting sphincter length in the leaker group was significantly longer than the resting sphincter length in both the incontinent men (P < 0.005) and normal controis (P < 0.05). Despite significantly higher MMR and MMS pressures when compared with incontinent men, the leaker sphincter is a long, low-pressure sphincter at rest that may result in incomplete evacuation of the anal canal and stool leakage after a bowel movement. Leader and incontinent men have unique clinical and manometric profiles that should be considered in their treatment and evaluation of clinical outcome. Prudendal Nerve (PN) Compression: A Possible Etiology of Fecal Incontinence (FI) Booth P24 C.N. Ellis, D.J. Coyle, W.S. Blakemore Birmingham, AL FI is more c o m m o n in elderly multiparous w o m e n and, while not life-threatening, can diminish the quality of life. Causes of FI include direct muscle damage from surgical or obstetric trauma or neurologic or muscular conditions. Anal sensation and voluntary sphincter contraction, which are dependent upon normal PN function, are decreased in FI. The PN arises from sacral nerve routs 2, 3, and 4 and traverses the lesser sciatic foramen (LSF). Compression of nerve does not cause histologic abnormalities but does lead to numbness, muscle weakness, and eventual atrophy. To determine if compression of the PN could occur in the LSF, the course of the PN was dissected in 79 cadavers (61 female and 18 male). Bilateral compression of the PN between the sacrospinous and sacrotuberous ligaments was identified in 20 of the females (33 percent) but none of the males. Pregnancy results in marked hypertrophy of the pelvic ligaments and could possibly cause compression of the PN contributing to FI. If this mechanism is confirmed by further studies, surgical decompression may restore nerve function. Are Pudendal Nerve Latencies a Predictive Factor in the Success of Biofeedback for Fecal Incontinence? Booth P25 L.L. Jensen, A.C. Lowry . . . . . . . . . . . .

Minneapolis, MN

Pudendal nerve latencies (PNL) are a predictive factor in the outcome of plication sphincteroplasty for trau-

Dis Colon Rectum, April 1993

matic fecal incontinence. To evaluate whether pudendal nerve latencies would predict the success with biofeedback for fecal incontinence, we reviewed the charts of 43 patients who had PNL assessed prior to treatment. Thirty-eight were females with an average age of 55 years (range, 28-92 years). Eighteen patients had idiopathic incontinence, 11 had previous known obstetric injuries, five had previous anorectal surgery, five were S/P colon resection, one (male) was S/P trauma, and three had other miscellaneous etiologies. Success was defined by patient assessment, a minimum of 90 percent decreased frequency of incontinence, and improvement in their incontinence score. Pudendal Nerve Latency O u t c o m e of Biofeedback

Normal

Unilateral Abnormal

Bilateral Abnormal

Success Failure

17 t

14 1

9 1

In our review, pudendal nerve latency did not appear to be a factor in predicting success of biofeedback for fecal incontinence.

Long Term Results of Surgery for Fecal Incontinence Booth P26 R. Miller, A. Mills, P. Durdey Bristol, United Kindgom Fifty-four patients operated on between 1985 and 1988 were contacted. Forty-seven answered a questionnaire (33 anterior sphincter repair (ASR), 14 postanal repair (PAR), median age 55 [25-85], M:F 2:45). Thirty-seven returned for anorectal physiology studies (26 ASR, 11 PAR, 19 idiopathic, 18 traumatic injuries). All patients had been assessed preoperatively and at one year postoperatively. Continence was scored on a scale of 1 to 18, which reflected the severity and frequency of incontinence episodes. A score of 3 represented incontinence to flatu s only, a score of 8 represented incontinence to flatus and occasional soiling but continence to solid stool. The median duration of follow-up was five years (range, 4-7 years). Results, Total G r o u p ( n = 47) 1 year

5 years

Score 3 8

Number

Percent

Number

Percent

24 31

51 65

10 18

21 38

Anterior Sphincter Repair ( n = 33) 1year

5years

Score 3 8

Number

Percent

Number

Percent

17 23

52 70

10 14

30 42

In the total group there was a significant improvement in maximum resting pressure and upper anal canal sensation at one year postoperatively (MRP: preoperatively 45 vs. 60 cm HaO at one year, P < 0.05; sensation: 23 ma

Vol. 36, No. 4

MEETING ABSTRACTS

preoperatively v s . 10.7 ma at one year, P < 0.005). Both parameters deteriorated significantly at five years followup compared with one year postoperatively (MRP: 60 v s . 47 cm Ha0, P < 0.05, upper anal sensation 10.7 ma v s . 13 ma, P < 0.005). Squeeze pressure and sphincter length were unaltered. In conclusion, continence deteriorates considerably with time. Maximum resting pressure and upper anal canal sensation appear to be the most important factors in maintaining continence after surgery.

Colpocystodefecography (CCD): An Original and Complete Morphodynamic Study of the Female Pelvic Organs and Floor Under Physiologic Stress Conditions Booth P27 D. Hock-Salve, R. Lombard, C. Jehaes, S. Markiewicz, L. Penders, F. Fontaine, P. Cusumano, G. Nelissen Li&ge, Belgium CCD combines micturating cystography and defecography in a patient whose vagina has been opacified; it allows one to study all the pelvic structures including the Douglas pouch. CCD performs better than clinical examination to diagnose prolapses and particularly elytroceles. The reason for this is easily understandable. Indeed, the sitting position and the defecating efforts constitute a much stronger strain upon the pelvic floor than that which is achieved during clinical examination performed on a prone patient. CCD documents the morphology, function, and dynamic interactions of the pelvic organs. Altogether, this information is most helpful to choose the best surgical therapy. In particular, CCD may provide grounds to widen the surgical procedure to neighboring structures in order to avoid possible late postoperative uterus, bladder, or Douglas pouch prolapse, complications which a classical defecography alone would never have predicted.

Postchildbirth/Hysterectomy Constipation--The Hindgut Neuropathy Identified by Dynamic Radioisotope Scanning and Treated by Rectosigmoidectomy Booth P28 A. MacDonald, J.N. Baxter, R.G. Bessent, H.W. Gray, I.G. Finlay . . . . . . . . . . . . . . . . . . . . . . . . . Glasgow, Scotland Patients with severe constipation who attribute the onset of symptoms to childbirth or hysterectomy (PC/ PH) may have a hindgut neuropathy. We used dynamic radioisotope scanning (DRS) to identify this abnormality and select patients with constipation for segmental rectosigmoidectomy. Ten patients with PC/PH constipation and 10 normal controls had a 2-m PVC water perfusion catheter passed by a colonscope to the right colon. Five milliliters of Tcm-99 10 mMq-labeled DTPA was injected into the right colon, followed by 5 mg of bisacodyl. Serial gammacamera pictures were taken for 45 minutes, Response time and transit of isotope to the splenic flexure and

P31

sigmoid colon were comparable in patients and controls (see table). PC/PH Patients

Response time (rain) Transit to: Splenic flexure (rain) Proximal sigmoid (rain)

Controls

Mean

(SD)

Mean

(SD)

12.5

(6)

13.1

(4)

P< 0.9

19.1

(7)

20.3

(4)

P< 1

28.3

(6)

29.9

(5)

P< 0.38

In all 10 patients, isotope stopped in the sigmoid, failure to progress to the rectum during the next two hours. In contrast, 9 of the 10 controls had isotope in the rectum within 35 minutes. Six patients identified by DRS to have rectosigmoid dysmotility underwent rectosigmoidectomy with coloanal anastomosis. Symptomatic relief was obtained in six patients. These data suggest that patients with PC/PH constipation have hindgut dysmotility alone. This simple technique may allow the selection of patients with PC/PH constipation who may benefit from segmental hindgut resection.

Rectal Motor Complexes are Reduced in Slow Transit Constipation Booth P29 G.S. Duthie, R. Farouk, D.C.C. Bartolo Edinburgh, Scotland Patients with slow-transit constipation have a colonic motility disorder. We have used ambulatory rectal pressure recordings to determine if rectal motor complex activity is also disordered in these patients. Recordings from 11 normal subjects (N) aged 36 years (range, 2574), 15 with slow-transit constipation (STC) aged 35 (18-64), and 10 with obstructed defecation (OD) aged 38 (24-72) were analyzed. Patients with STC had significantly (P<0.001) fewer rectal motor complexes (RMC) (total 18) than did normal subjects (total 90) and those with OD (total 82). Although RMCs were fewer in STC, they exhibit similar waveforms (N = 3.1/min [range, 2.34.1]; STC = 3.35 [2.5-3.7]; OD = 3.1 [2.1-7.3)], associated with similar rectal pressure peaks (N = 25 cmH20 [10451; STC = 25 [15-451; OD = 25 [15-70]). Anal canal pressure and internal and external sphincter activity were unchanged during rectal motor complexes. These results confirm that slow-transit constipation affects the whole of the large bowel. Patients with obstructed defecation have normal rectal function.

Physiologic Measurement of External Sphincter Pressure: To Squeeze or Cough? Booth P30 A.P. Meagher, D.Z. Lubowski . . . . . . . . Sydney, Australia Rises in intra-abdominal pressure such as during coughing or movement are c o m m o n and, to prevent

MEETING ABSTRACTS

P32

incontinence, are accompanied by reflex contraction of the external sphincter. Voluntary contraction of the sphincter is an u n c o m m o n event yet is routinely used to measure sphincter strength. Seventy-five patients referred for anorectal physiology underwent a standardized manometry protocol using a four-channel perfused catheter (three anal and one rectal side-hole) to compare cough pressure (CP) and squeeze pressure (SP). Maximum anal CP was higher than SP (mean, 158 vs. 133 cmH20; P = 0.001). Each pressure was measured three times in all cases, and intraindividual variance was less using CP (mean, 20 percent vs. 29 percent, P = 0.05). There was significant correlation between CP and SP (P <0.001), although there were some wide differences, suggesting that use of both CP and SP in manometry will better assess sphincter strength. Measurement of the rectoanal pressure gradient during coughing correlated significantly better with degree of incontinence than CP~ SP, or a combination of these two. The presence of a positive gradient was 100 percent specific for incontinencel The sensitivity was only 43 percent, demonstrating that factors other than simple mechanical sphincter weakness are involved in incontinence. In conclusion, measurement of CP has practical and research relevance and should be added to standard manometric protocols.

Rectal Prolapse

Perineal Excision of Rectal Procidentia in Elderly High Risk Patients: A Ten Year Experience Booth P31

P.S. Ramanujam, K.S. Venkatesh

........

Phoenix, AZ

This is a 10-year study assessing the morbidity, mortality, recurrence rates, and anal incontinence in elderly high-risk patients who underwent perineal excision for a complete rectal procidentia. Seventy-two elderly highrisk patients were seen with a complete rectal prolapse. Sixty-five patients presented with a chronic recurrent rectal prolapse and underwent a perianal excision with a posterior levator ani repair. Seven patients presented with an acute incarcerated prolapse. These patients had perineal excision only. Sixty-nine patients were women. The prolapsed portion of the rectum was removed through the perineum. A posterior levator ani repair was performed to recreate the anorectal angle (levator ani repair was not done on seven patients with incarcerated prolapse). The follow-up period ranged from six months to eight years. There were no mortalities, and the morbidity was low. Two patients in the incarcerated group developed an anastomotic leak requiring a diverting colostomy. There were five recurrences. Eighty-three percent of the patients had a marked improvement in anal continence. In conclusion, perineal exicision of rectal prolapse was well tolerated by elderly high-risk patients. Posterior levator ani repair which recreates the anorectal angle seems to improve anal continence. The incarcerated rectal prolapse group is at a higher risk for postoperative anastomotic leaks.

Dis Colon Rectum, April 1993

The DeLorme Procedure for Complete Rectal Prolapse in Severely Debilitated Patients: An Analysis of 42 Cases Booth P32

G.C. Oliver, D. Vachon, T.E. Eisenstat, R.J. Rubin, E.P. Salvati . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plainfield, NJ Over a 10-year period, 42 patients were selected to undergo the DeLorme procedure for complete rectal prolapse. Initially, we selected this procedure when advanced age and poor overall health mitigated against an abdominal repair. However, our early results stimulated us to offer this procedure in younger individuals. The mean age of this patient population is 78 years (range, 28-95). Females outnumber males 6:1. All procedures were performed in the prone jackknife position utilizing intravenous sedation and a perianal block. Patient followup averaged 7 months to 10 years (mean, 40 months). Recurrences have occurred in eight patients (19 percent). The mean time to recurrence was four months (range, 1-8 months). Subsequent repairs of failed DeLorme operations have included five Silastic anal encirclement procedures and two perineal proctosigmoidectomies (Altemeier procedure). The remaining recurrence is pending re-repair. One death occurred in an 81year-old patient on postoperative day one of cardiopulmonary arrest. One anastomotic dehiscence occurred, while minor complications were noted to occur in 20 percent of cases (UTI, retention, and atelectasis primarily). Pitfalls in performing this procedure relate primarily to associated perineal and colonic conditions. Most notable among these problems are weak or absent anal sphincter tone, perineal descent, and extensive diverticular disease prohibiting effective and complete proximal mucosectomy. An incomplete mucosectomy of the prolapsing bowel sets the stage for early recurrence of prolapse.

Surgical Treatment of Rectal Prolapse Associated with Systemic Sclerosis Booth P33

K.R. Dorsey-Tyler, G. Harmon, S.J. Medwell Seattle, WA Systemic sclerosis is a connective tissue disease involving multiple organs. Anorectal dysfunction (fecal incontinence, constipation, rectal prolapse) is frequently described. Current literature fails, however, to address surgical treatment of rectal prolapse in these patients. Five patients with severe and long-standing (12-22 years) systemic sclerosis were treated for debilitating full-thickness rectal prolapse. All were female, with ages ranging from 55 to 73 years. All suffered from incontinence of liquid feces. Treatment in four of the five cases consisted of sigmoidectomy and rectopexy. The fifth patient had a rectopexy of Hartmann's pouch with subtotal colectomy after a colonic volvulus with associated ischemia. One of the patients died 19 months after surgery without evidence of recurrent prolapse, but with residual incontinence of liquid stool. Of the four survivors, one

MEETING ABSTRACTS

Vol. 36, No. 4

suffered a small recurrent prolapse successfully treated with a Thiersch repair; the others remain free of recurrence. Some incontinence of liquid stool remains in three patients, but all report a significant improvement in quality of life following repair. Sigmoidectomy and rectopexy appear to offer a significant benefit to these patients. Although some incontinence, particularly of liquid stool, may remain, the quality of life can be improved appreciably by this procedure.

Laparoscopy

Must Early Postoperative Oral Intake Be Limited to Laparoscopy? Booth P34 S.R. Binderow, S,M. Cohen, S.D. Wexner, S.L. Schmitt, J.J. Nogueras, D.G. Jagelman . . . . . . Ft. Lauderdale, FL This prospective, randomized study was designed to evaluate whether or not early postoperative feeding can be solely claimed as a unique benefit of laparoscopic surgery. A prospective, randomized trial was performed between July I and September 30, 1992. All patients (51) underwent laparotomy with either a colonic or an ileal resection; in all cases the nasogastric tube was removed immediately after the operation. Group 1 consisted of 27 patients (age range, 15-81; mean, 52 years) who received a regular diet on the first postoperative morning. Group 2 consisted of 24 patients (age range, 15-87; mean, 49 years) who were fed in a traditional manner; regular food was permitted after resolution of ileus as defined by resumption of bowel movements in the absence of abdominal distention, nausea, or vomiting. The rate of nasogastric tube reinsertion for distention with persistent vomiting, was 22 percent (six patients) for Group 1 and 8.3 percent (two patients) for Group 2. Although vomiting was experienced more frequently by patients in Group 1 (48 percent vs. 17 percent), there was no difference between the two groups with regard to the duration of postoperative ileus (3.6 vs. 3.8 days, respectively). In the 21 patients (78 percent) from Group 1 who did not require nasogastric tube reinsertion, there was a trend toward shorter hospitalization (6.7 vs. 8.0 days, respectively). Thus, the laparoscopic surgeons' claims of shorter hospitalization and earlier tolerated oral intake may not be unique to laparoscopy.

Laparoscopic Oncologic Abdominoperineal Resection Booth P35 J.W. Milsom,* C. Decanini,'~ B. Bohm,* V.W. Stolfi* *Cleveland, OH and 5-Mexico City, Mexico The purpose of this study is to define a standardized technique for laparoscopic oncologic abdominoperineal resection (LOAPR) in a cadaver model. In 10 fresh cadavers (three females and six males), a laparoscopic APR with ligation of the origin of inferior mesenteric artery, removal of complete mesorectum and adjacent lymph nodes (LN), and wide lateral clearance

P33

was carried out. After surgery, autopsy was performed, and the number of remaining inferior mesenteric LN, amount of remaining mesorectum, and length of remaining IMA were evaluated. No major intraoperative complications were recorded.

Operation time (rain) LNin specimen Remaining LN

Median 135 13 0

(Range) (105-180) (12-41) (0)

A brief video will be presented which shows our standard approach of performing a laparoscopic oncological abdominoperineal resection. Oncologic abdominoperineal resection with high vascular ligation of IMA and complete clearance of lymph nodes and mesorectum can be performed using our standardized technique.

Laparoscopic v s . Open Colorectal Resections for Malignancy: Assessment of Lymphatic Resection by Mesenteric Lymphangiography Booth P36 P.A. Cataldo, C. Hadick, P. Resnikov, J,D. Mellinger, B. Cunningham . . . . . . . . . . . . . . . . . . . . . . . . Dayton, OH Recent advances in surgical techniques and instrumentation have made laparoscopic colorectal resections possible. Application of this technology in the setting of colorectal neoplasia has been criticized because of concerns over the extent of lymphatic resection achieved via laparoscopy. Inadequate mesenteric resections have been associated with increased local recurrence rates and decreased survival in previous reports. The present study compares laparoscopic and standard open low anterior resections in a porcine model, with particular emphasis on the extent of intestinal and mesenteric resection. Endoscopic mesenteric lymphangiography was performed preoperatively to label rectosigmoid lymph nodes. Twenty animals were then randomized to either laparoscopic or open low anterior resection. All animals were sacrificed and autopsied 14 days postoperatively. Length of intestinal resection, total number of lymph nodes per specimen, number of labeled nodes per specimen (evaluated both radiographically and microscopically), and number of unresected, labeled lymph nodes (evaluated by postoperative KUB and autopsy data) were then compared. No differences in any of these parameters were noted between groups. We conclude that equivalent resections are possible in this model v i a laparoscopic and traditional open methods. Differences in comparative anatomy preclude direct application of these data to the human. Mesenteric lymphangiography may be useful in future comparisons of these techniques in the clinical and laboratory setting. Further studies will be necessary to assess the appropriateness and long-term implications of laparoscopic colorectal resections for malignant disease.

MEETING ABSTRACTS

P34

Cost Effectiveness of Laparoscopic Assisted Colectomy Booth P37

A.J. Vayer, Jr., S.W. Larach, P.R. Williamson, A. Ferrara, M. Salomon . . . . . . . . . . . . . . . . . . . . . . . . . Orlando, FL While laparoscopic-assisted colectomy is technically feasible, the cost-effectiveness of this new procedure has not been evaluated. This study was undertaken to examine the cost-effectiveness of laparoscopic-assisted colectomy vs. open resection, studying our initial experience with 31 laparoscopic colectomies as c o m p a r e d with a matched group of open resections by the same surgeons. The results are as follows:

Total Cost O.R. Time O.R, Equip. Anesthesia POD B.S. POD Flatus POD B.M. POD D/C

Open

Lap. Asst.

P Value

$24,219 $3,061 $2,324 $640 2.96 5.54 6.92 9.06

$27,387 $4,268 $5,550 $943 1.74 3.16 4.30 6.83

0.49 0.003 <0.001 0.002 <0.001 <0.001 <0.001 0.01

Despite the higher cost of O.R. time and equipment, the total cost of laparoscopic-assisted c o l e c t o m y is not statistically different from open resection. The cost savings are realized in a shortened hospital stay.As proficiency with this n e w procedure is obtained and when reusable instruments are acquired, actual cost savings may be found with the laparoscopic technique.

Does Laparoscopy Confer an Advantage over Standard Colectomy? Booth P38

S.L.Schmitt, S.D. Wexner, J.J. Nogueras, D.G. Jagelman Fort Lauderdale, FL Laparoscopic colorectal surgery has b e e n said to expedite both return of bowel function and, consequemly, hospital discharge, as c o m p a r e d with standard colectomy. However, few comparative data exist to support these claims. Therefore, this study was undertaken to compare the time to oral intake and to subsequent discharge from the hospital after laparoscopic-assisted (LAC) and after open laparotomy (OC) colonic resections. From August 1991 to September 1992, 36 patients underwent LAC and were prospectively evaluated. Thirty five age, sex, diagnosis, and procedure-matched OC controis were reviewed for comparison. Procedures in these 71 patients included: right colectomy for carcinoma (n = 10), total c o l e c t o m y w i t h end ileostomy for toxic colitis (n = 2), sigmoid colectomy for diverticulitis (n = 2) or carcinoma (n = 4), a b d o m i n o p e r i n e a l resection for carcinoma (n = 2), total colectomy with ileorectal anastomosis for Crohn's disease (n = 3) or colonic inertia (n = 5), ileocectomy for Crohn's (n = 4), and restorative proctocolectomy with ileoanal reservoir for ulcerative colitis (n = 22) or familial polyposis (n = 9). Oral intake was begun on postoperative days zero to nine (mean, 3.7) after LAC and postoperative days two to seven

Dis Colon Rectum, April 1993

(mean, 4.1) for OC. The patients were discharged from the hospital on postoperative days 5 to 16 (mean, 7.7) for LAC and postoperative days 5 to 18 (mean, 8.2) for OC. At this time, LAC does not appear to confer any advantage over OC in time to oral intake or length of hospitalization.

Ileoanal Pouch/Kock Pouch

Can Minimal Anal Manipulation with Mucosectomy Preserve Anal Function in Ileal Pouch Surgery? Booth P39

J. Sullivan, J.W. Milsom, V.W. Fazio

. . . Cleveland, OH

Previous ileal pouch-anal anastomosis (IPAA) techniques utilized a traumatizing Gelpi or other retractor which forcefully dilates the anal canal to accomplish hand-sewn (HS) IPAA. The purpose of this study is to evaluate whether a minimally dilating technique preserves anal sphincteric function c o m p a r e d with an IPAA circular stapled (CS) technique. METHODS: We have utilized a technique for HS IPAA (HS-new) which places only a small (25-ram) anoscope within the anal canal, yet effectively allows for accurate suturing. We also c o m p a r e d a group of historic control patients of similar age who underwent HS IPAA (HSold) prior to d e v e l o p m e n t of the new technique. Thirtytwo patients were randomized prospectively into HS vs. CS IPAA and c o m p a r e d with the HS-old group; median follow-up was 9.5 months (range, 5-14 months). RESULTS: Decrease in maximum resting pressure (,~MRP) (pre- vs. postoperatively) quality of life (QL) (scale of 1-10, 10 highest), minimal morning leakage (MinAML), major morning leakage (MajAML), minimal evening leakage (MinPML), and major evening leakage (MajPML) were compared:

~,MRP QL MinAML MajAML MinPML MaiPML

HS-new 52~ 8.7 33% 0% 50% 16%

CS 33% 8.7 20% 0% 40%

HS-old 38% N/A 16% 2% 43%

0%

12%

CONCLUSIONS: CS patients have less traumatic injury to their sphincters, less decrease in MRP, and improved functional results even when attempts are made to minimize anal manipulation during m u c o s e c t o m y and handsewing.

State of the Defunctionalized Sphincter in Patients Undergoing Ileal Pouch-Anal Anastomosis Booth P40

R.J. Staniunas, J.O. Keck, T. Counihan, P. Marcello, R.C. Barrett, M. Oster, P.L. Roberts, D.J. Schoetz, J.J. Murray, M.C. Veidenheimer, J.A. Coller . . . . . . . Burlington, MA A number of patients who have defunctionalized anal sphincters as a result of long-term proximal fecal diver-

Vol. 36, No. 4

MEETING ABSTRACTS

sion presented for restorative surgery. The physiologic profiles of 13 patients defunctionalized (DEF) for one year or more (range, 1-17.5), who subsequently underwent IPAA, were compared with 26 patients with no previous colectomy (NPC) or proximal fecal diversion prior to formation of the pouch. There was no difference in age or sex. Significantly different preoperative manometric data included mean resting pressure (mrp), mean squeeze pressures (rasp), and squeeze pressure volume

P35

result in better postoperative physiologic profiles. Since the pressure changes were symmetrical, it suggests a denervation injury mostly affecting the internal sphincter. The distal level ofrectal transection could affect the intrinsic neurologic pathways located in the proximal internal sphincter. * Paired t-test. ** Unpaired t-test.

(spv). The Critical Level of the Ileoanal Anastomosis mrp rasp spy

DEF

NPC

71.6 + 30.6 mmHg 109.34- 52.4 mmHg 487,925mmHga

91.5 --- 25.6 mmHg ( P = 0.05)* 171.74- 64.5 mm Hg (P= 0.005)'~ 1,282,684mmHg3 ( P = 0.007)*

* U0paired t-test. t Mann-Whimey test. Fisher's exact test.

Despite significant differences in the preoperative manometric profiles, DEF and NPC pouch patients had similar functional results. Both groups had a mean of 6.1 BM/24 hours and could defer defecation for a mean of two hours. Leakage occurred in 20.3 percent of DEF and 19.8 percent of NPC (P= 0.69).~ These data suggest that the defunctionalized anal sphincter can adequately support a restorative procedure without regard to the timing of pouch creation.

The Three Dimensional Profile of the Anal Sphincter Before and After Rectal Mucosectomy Booth P41 R.J. Staniunas, J.O. Keck, R.C. Barrett, M. Oster, J.A. Coller . . . . . . . . . . . . . . . . . . . . . . . . . . . . Burlington, MA The etiology of postoperative changes in the anal sphincter after rectal mucosectomy is poorly understood. We analyzed the three-dimensional manometric pressure profiles of 38 patients who underwent mucosectomy and IPAA, to determine structural and functional relationships. There was an average postoperative reduction of mean resting pressure and mean squeeze pressure of 49.3 percent (P < 0.0001)* and 21.6 percent (P = 0.0003)*, respectively. The resting sphincter length also decreased significantly by an average of 0.7 -+ 1.03 cm (P = 0.0003)*. Linear regression analysis of manometric variables shows that the percent change in resting sphincter length significantly correlates with resting pressure (r = 0.44, P -- 0.008). Analysis of patients with a greater than 50 percent decrease in postoperative resting pressure shows that the average decrease in physiologic sphincter length was 1.11 cm, which was significantly different from the average decrease in length of 0.27 cm in patients with resting pressure less than 50 percent (P = 0.01).** Although there were significant changes in the three-dimensional profiles, specifically pressure volume and asymmetry, these changes did not significantly affect sphincter resting and squeeze pressure measurements. These data show that preservation of anal sphincter length in the range of only 1 cm may

Booth P42 P.M. Sagar, W. Lewis, M. Williamson, P.J. Holdsworth, D. Johnston . . . . . . . . . . . . . . . . . . Leeds, United Kingdom Restorative proctocolectomy (RP) with a stapled ileoanal anastomosis has received criticism because of the risk of retained rectal mucosa. To counter this, we have used eversion of the anorectal stump to facilitate accurate transection at the top of the anal transitional zone. Our hypothesis was that anorectal eversion would not damage the anal sphincter. One hundred fifty-four patients were studied after RP. Transanal mucosectomy was carried out in 36 patients, standard stapled ileoanal anastomosis in 76 patients, and stapled ileoanal anastomosis with eversion of the anorectum in 42 patients. Laboratory tests of anal motor and sensory function and clinical interview were carried out at least eight months after operation.

Mucosectomy n Resting anal pressure (cmH20) Rectoanat inhibitory reflex Seepage or soilage

Stapled

Everted and Stapled

36 54* (27-103) Absent

76 80 (36-145) Present

42 73 (34-156) Present

17/36" (47%)

6/76 (8%)

4/42 (10%)

Medians (ranges). * P < 0.01.

Whereas mucosectomy resulted in decreased RAP and impaired clinical results, anorectal eversion in the course of stapled RP did not significantly alter manometric or functional outcome.

Optimalization of Sphincter Function After Restorative Proctocolectomy--A Prospective Randomized Trial Booth P43 J.M.N. Jorge, S.D. Wexner, P.J. Morgado, Jr., K. James, J.J. Nogueras, D.G. Jagelman . . . . . Fort Lauderdale, FL Impairment of sphincter function in patients (pts) who undergo restorative proctocolectomy (RP) is usually most severe immediately after ileostomy closure. Therefore, a prospective, randomized trial was undertaken to assess the value of preileostomy closure sphincterstrengthening exercises (SSE) in improving early functional outcome. Twenty-six pts who underwent RP between July 1991 and June 1992 were studied. They included 16 males and 10 females of a mean age of 38

P36

MEETING ABSTRACTS

(17-69) years. Pts were randomized either to undergo a five-week pelvic floor exercise program or to a control group. An incontinence score (IS) of 0 to 20 was used to clinically assess the functional results. Anorectal manometric assessment included: high-pressure zone length (HPZ), mean resting pressure (MR), highest resting pressure (HR), mean squeeze pressure (MS), and highest squeeze pressure (HS). The paired t-test was used to compare the preoperative manometric and functional results with those at the time of ileostomy closure (A). This latter time c o r r e s p o n d e d with the conclusioli of the exercise program or the equivalent time p e r i o d for the control group.

SSE

IS

HPZ

MR

HR

MS

HS

(cm)

(mmHg)

(mmHg)

(mmHg)

(mmHg)

No ( n = 13)

2

-0.08

-20.8

-22.2

-18.02

--15.8

Yes ( n = 13)

0

-0.3

--27.1

--41.9

--10.9

--21.9

Pvalues

0.2

0.7

0.5

0.7

0.8

0.05

In conclusion, SSE before ileostomy closure did not minimize the transient impairment of functional results observed ion patients who underwent RP.

Ileal Pouch-Anal Anastomosis (IPAA): Is Preoperative Anal Manometry Predictive of PostoPerative Functional Outcome? Booth P44 P. Morgado, Jr., S.D. Wexner, K. James, D.G. Jagelman, J.J. Nogueras . . . . . . . . . . . . . . . . . . Fort Lauderdale, FL The aim of this study was to determine the value of preoperative anal manometry in predicting postoperative continence. Anal manometry was p e r f o r m e d in 73 patients before IPAA surgery ( m l ) , before loop ileostomy closure (m2), and at follow-ups of one (m3) and two (m4) years. Mean and maximum resting and squeeze pressure were d o c u m e n t e d at each occasion. Two years after surgery, pressures were correlated (r) with an incontinence score (IS) (0 = perfect control, 20 -- complete incontinence). (mmHg) mRP mSP r (mSP) r (mRP)

ml 66 114 0.02 -0.06

m2 42.8 102.9 0.04 -0.14

m3 53.8 103.4 0.06 -0.13

m4 54.7 95.8 0.062 -0.15

A significant ( P < 0.05) decrease in mean resting pressures (mRP) was observed after IPAA, followed by a significant ( P < 0.05) improvement of mRP after loop ileostomy closure. These pressures continued to significantly ( P < 0.05) improve for at least two years. Mean squeeze pressures (mSP) did not Change ( P > 0.05) at any time during the study. There was no correlation between preoperative mRP and IS (mean -- 4.83). In conclusion, although anal manometry showed a characteristic trend in internal anal sphincter recovery after IPAA, it failed to prove helpful in the prediction of

Dis Colon Rectum, April 1993

clinical outcome. Thus, although this study supports the continued use of manometry in a research setting, it challenges the value of routine manometry in a clinical context,

Recovery of Anal Sphincter Function After' Restorative Proctocolectomy (RP) in Patients over the Age of Fifty Booth P45 J.M.N. Jorge, S.D. Wexner, K. James, J.J. Nogueras, D.G. Jagelman . . . . . . . . . . . . . . . . . . . . . Fort Lauderdale, FL Despite initial reluctance to perform RP in o l d e r patients (pts), newer techniques may result in an acceptable clinical outcome. This study was undertaken to assess the progression of anal sphincter function and clinical outcome in pts ->50 years old (Group I) as c o m p a r e d with <50 years old (Group II). Group I consisted of 22 pts (mean = 56 years) and Group II, 50 (mean = 32 years). An incontinence score (IS; range 0-perfect continence to 20-total incontinence) assessed after ileostomy closure included type and frequency of incontinence, use of pad, and lifestyle alteration. Manometry was performed preoperatively (MN1) and before (MN2) and after (MN3) ileostomy closure. This assessment included high-pressure zone length (HPZ), mean and highest resting pressure (MR, HR), and mean and highest squeeze pressure (MS, HS). Wilcoxon and paired t-test were used to compare these results.

Group

HPZ (cm)

MR (mmHg)

HR (mmHg)

MS (mmHg)

HS (mmHg)

IS (0 20)

l (n = 22)

MN1 MN2 MN3

3.4 • 1.0 2.4• 3.3 • 1.1

72 • 24 37+22 * 53 • 17

100 • 30 53--+24 ~ 72 • 27

126 • 69 112• 93 + 32

163 • 76 148• 115 --+33

5.9

II (n = 50)

MN1 MN2 MN3

3.4• 2.4 + 1.2 2.9 • 1.2

64• 45 • 17" 55 • 20

91-+24 65 • 24* 81 • 33

109+54 99 + 48 108 -+ 48

152 :t: 65 138 • 58 153 4- 69

3.9

9 P < 0.05; paired t-test.

No significant difference was found in either MN1 or IS between the two groups ( P > 0.05). However, in the MN2, both MR and HR were significantly lower in Group I. In conclusion, the transient impairment of IAS function after RP is more severe in older pts ( P = 0.01), but it does recover after ileostomy closure.

The Mechanism of Continence and Defecation Following Ileal Pouch-Anal Anastomosis Booth P46 R. Farouk, G.S. Duthie, D.C.C. Bartolo Edinburgh, Scotland Panproctocolectomy and ileal pouch-anal anastomosis does not usually disrupt continence. To assess why, we have studied 72 patients (median age, 39 years; 51 male) by ambulatory electromyography and manometry 4 to 12 months after surgery. Resting anal pressure was 48 cmH20 in the alert state and 12 cmH20 during sleep. Daytime pouch pressure was 18 cmH20 and 6 cmH20 during sleep. The mean internal sphincter EMG frequency was 0.38 Hz.

Vol. 36, No. 4

MEETING ABSTRACTS

Alert-state pouch contractions (1-5 per hour) were associated with anal sphincter recruitment (maximum pressure: pouch = 58 cmH20, anal = 71 cmH20). Defe cation was associated with p r o l o n g e d pouch contractions and a reversal of the anal-pouch pressure gradient. Nocturnal pouch contractions were infrequent except in eight patients who reported leakage. Pouch pressures in these patients e x c e e d e d anal pressures during contractions (pouch pressure = 48 cmH20, anal pressure = 18 cmH20). Nocturnal pouch contractions are associated with incontinence. Isolated daytime pouch contractions result in anal sphincter recruitment. Frequent pouch contractions result in defecation.

Secondary Proctectomy and Ileal Pouch-Anal Anastomosis After Ileorectal Anastomosis for Familial Adenomatous Polyposis Booth P47 C. Penna, A. Kartheuser, E. Tiret, R. Parc Paris, France Colectomy with ileorectal anastomosis (IRA) is the p r o c e d u r e of choice for prophylaxis of colorectal cancer in familial adenomatous polyposis (FAP) for many centers. However, a secondary proctectomy with ileal pouch anal anastomosis (IPAA) has to be p e r f o r m e d in a certain n u m b e r of cases. Between 1983 and 1991, a conversion of IRA into IPAA has been indicated in 29 FAP patients for uncontrollable rectal polyposis in 16 cases, patient's wishes in eight cases, rectal cancer in two cases, and stenosis of the IRA in two cases. Previously unsuspected d e s m o i d tumors p r e c l u d e d the proctectomy in three cases (10 percent). After IPAA following IRA, there was no postoperative mortality; four patients (14 percent) p r e s e n t e d a postoperative compli cation and four a late complication. The pathologic examination of proctectomy s p e c i m e n s revealed unsus pected neoplastic lesions in four patients (14 percent). One year after IPAA, the mean stool frequency was 4.4 ___ 0.4, 44 percent of the patients had nocturnal stooling, and daytime and nighttime continence were normal in 91 percent and 87 percent of patients, respec tively. These results did not differ from those of previous IRA and from those obtained in 120 FAP patients who received IPAA as the first surgical procedure. We conclude that conversion of IRA into IPAA is a safe and effective procedure that should be offered to FAP patients with IRA and presenting a high risk of rectal cancer.

Indeterminate Colitis: Long-Term Outcome in Patients After Ileal Pouch-Anal Anastomosis Booth P48 P.B. Mclntyre, J.H. Pemberton, B.G. Wolff, R.R. Dozois, R.W. Beart, K.A. Kelley . . . . . . . . . . . . . . Rochester, MN Controversy persists about the long-term results of ileal pouch-anal anastomosis (IPAA) p e r f o r m e d for indeterminate colitis (IC). PURPOSE: To compare functional outcomes of IPAA for typical chronic ulcerative

P37

colitis (CUC) and IC. METHOD: Of 1,400 IPAA patients (pts), 71 (40 women, 31 men; mean age, 31 years) were identified with a diagnosis of IC (preoperative clinical diagnosis of CUC but inconclusive histologic features in the resected specimen). Mean follow-up was 56 months (1-120 months). Outcomes were c o m p a r e d with 1,232 CUC pts (578 women, 654 men; mean age, 31 years) after IPAA. Mean follow-up was 60 months (0-132 months). RESULTS: There was no difference in the frequency of daily bowel movements (IC = 7 -+ 3 vs. CUC -- 7 + 2). Daytime and nighttime incontinence rates were similar. The prevalence of pouchitis was identical at 33 9 percent. However, the failure rate was higher in the IC group (IC = 19 percent, vs. CUC = 8 percent; P = 0.03). Of the 13 IC failures, four were related to perioperative complications, while nine d e v e l o p e d perianal disease. Mean time to failure in IC pts was 35 months (2--80 months). Five other pts have b e e n diagnosed with Crohn's, but p o u c h function remains good. CONCLUSIONS: At a mean of nearly five years after operation, failure appears to occur more frequently in pts with IC than in pts with CUC. However, the great majority of IC pts have long-term functional results identical to those of pts with CUC. In pts with strictly defined IC, IPAA remains the p r o c e d u r e of choice.

Fertility, Sexual and Gynecologic Function After IPAA: Results of a Questionnaire Booth P49 T. Counihan, P.L. Roberts, D.J. Schoetz, J.A. Coller, JO. Murray, M.C. Veidenheimer . . . . . . . . . . Burlington, MA Pelvic surgery is associated with infertility and dyspareunia in women. To d e t e r m i n e the incidence of infertility, gynecologic problems, and sexual dysfunction after ileal pouch-anal anastomosis (I PAA), a questionnaire was sent to 209 w o m e n who underwent p o u c h surgery at a single institution from 1980 to 1991. At present, the response rate was 33 percent (70/209). The mean age at pouch construction was 31 years. Forty-four w o m e n had 97 children prior to pouch surgery, and there were seven pregnancies after IPAA (two vaginal delivery, five cesarean section). Five w o m e n c o m p l a i n e d of infertility after IPAA. Seventeen w o m e n had dyspareunia after IPAA, which resolved in two. Six patients have u n d e r g o n e surgery for ovarian cysts; two have had a tubo-ovarian abscess. Two patients have had peritoneal inclusion cysts. CONCLUSION: These results suggest that the incidence of infertility, gynecologic problems, and sexual dysfunction may be higher after IPAA than previously reported.

Stapled Nipple Valve to Pouch Improves Results in Continent Ileostomy Patients Booth P50 J.B. Harrison, V.W. Fazio, J.W. Milsom, S.E. Hockenberry Cleveland, OH Incontinence of the Kock ileal reselwoir has b e e n due primarily to slippage of the intussuscepted nipple valve

P38

MEETING ABSTRACTS

(NV). Several surgical techniques have been devised to decrease NV slippage. The purpose of this study was to review our experience over a three-year period with a modification of stapling the NV to the anterior wall of the pouch. METHODS: All patients undergoing NV-pouch stapling (NV-PS) were compared with an earlier group undergoing a routine Kock pouch procedure between 1977 and 1985. Postoperative morbidity, hospital stay, incontinence, fistula, intubation difficulties, stenosis, and major reoperation rates were compared. Results

Median f/u (too) Hosp. days Incom.2~valveslippage Intub. difficult. Stomal stenosis Fistula Pouch-to-valve Cutaneous Majorreoperation

NV-PS (n = 35)

K pouch (n = 252)

15 8 1 (2.8%) 0 3 (8.5%i

36 12 24 (9.5%) 34 (13%) 17 (6.7%)

1 (2.8%) 0 2 (5.7%)

8 (3.2%) 12 (4.8%) 82 (33.7%)

CONCLUSIONS: The NV-PS modification leads to significantly less NV slippage when compared with nonNV-PS patients in our institution. The postoperative morbidity and fistula rate were comparable. This technique should be the procedure of choice in patients who are candidates for continent ileostomy.

Prospective Randomized Trial of Pouch Design in Restorative Proctocolectomy (RP): Early Results of J vs. W, Big v s . Little Booth P51

M.E.R. Williamson, W.G. Lewis, P. Sagar, P.J. Holdsworth, D. Johnston . . . . . . . . . Leeds, United Kingdom Controversy persists over the ideal pouch design in RP. Larger pouches may result in reduced bowel frequency. Several designs have been used, but only two, the J pouch and the W pouch, are now in common use. The early results of a prospective trial are presented, which aims to compare the effects both of pouch design and of length of ileum used, on functional outcome. Patients were randomly assigned to either a J or a w pouch, constructed from either 30 or 40 cm of ileum, creating four groups; J30, J40, W30 and W40 cm (seven, seven, eight, and eight patients, respectively). Four patients have yet to have the ileostomy closed. Median follow-up is six months (3-18). Operative technique was standard, with the eversion-stapling technique as reported previously. In each patient, clinical and physiologic assessments were carried out before and 1.5, 3, 6, and 12 months after operation. ComparingJ with W pouches, no significant difference was found with regard to either maximum tolerated volume: 230 ml (110-330) vs, 175 ml (20-680); bowel frequency in 24 hours: seven (3-10) vs. six (2-16); resting anal pressure: 68 cmH20 (21-87) and 67 cmH20 (18-153), respectively. Only one patient in each group

Dis Colon Rectum, April 1993

reported urgency. Fecal leakage was reported by 3 (21 percent) patients in the J group and 6 (38 percent) in the W group (NS). No difference was found between big and little pouches. The J pouch is simpler to construct and, so far at least, yields results as good as the W pouch.

Colorectal Disease

Erythromycin Management of Colonic Pseudo-Obstruction Booth P52 D.N. Armstrong, J.H. Pemberton . . . . . .

Rochester, MN

Current treatment of colonic pseudo-obstruction entails colonoscopic decompression of the cecum. Erythromycin is a prokinetic agent in the gut, its actions being mediated by activation of motilin receptors, or smoothmuscle Ca channels. PURPOSE: A pilot study, to determine if erythromycin is effective in treating colonic pseudo-obstruction. METHODS: Nine patients with colonic pseudo-obstruction were treated with a trial of erythromycin (400 mg p.o.~500 i.v.q. 6 hours). Pseudo-obstruction was secondary to spinal trauma in two patients and metabolic abnormalities in four and was postoperative in three. Cecal diameter was measured by plain abdominal radiograph every 24 hours. Supportive measures were initiated in all patients. RESULTS: Within 24 hours, cecal diameter decreased dramatically in eight of the nine patients (mean 13.7 + 0.99 cm before vs. 7.3 --- 0.38 cm after erythromycin; P < 0.0001, paired t-test). Moreover, each patient experienced symptomatic relief. No patient required decompressive colono'scopy. One patient failed to respond; colonoscopy revealed pancolonic pseudomembranous colitis. CONCLUSION: Erythromycin is an effective noninvasire therapeutic option for patients with colonic pseudoobstruction. A prospective, randomized trial is in progress.

Reversal of Hartmann's Procedure: The Effect of Timing on Ease and Safety Booth P53

J.O. Keck, B.C. Collopy, P.J. Ryan, R. Fink, J.R. Mackay, R.J. Woods . . . . . . . . . . . . . . . . . . . Melbourne, Australia The optimal time for reversal of Hartmann's procedure remains controversial. Significant operative difficulty and morbidity have been reported for Hartmann's reversal. This study examines 11 years of experience with Hartmann's reversal with attention to reversibility, operative difficulty, and timing of reversal. Hartmann's procedure was performed on 111 patients, mostly for advanced cancer and complicated diverticular disease. Of 96 patients who survived, 50 underwent reversal (52 percent). Of those with diverticular disease, 40 of 48 underwent reversal (83 percent). Mortality for Hartmann's reversal was 2 percent; anastomotic leak rate

Vol. 36, No. 4

MEETING ABSTRACTS

4 percent; and overall complication rate 26 percent. Patients undergoing reversal were divided into two groups according to timing of surgery: early (before 15 weeks, 13 patients) and late (after 15 weeks, 37 patients). There was no difference between these groups in mortality, morbidity or anastomotic leakage. However, bed stay was longer in the early group and graded operative difficultywas greater. In particular, cases where adhesion density was most severe, and where accidental enterotomy occurred, were more c o m m o n in the early group. (P = 0.02, Miettinen's modification of Fisher's exact test). Hartmann's reversal can be safely performed in most survivors of complicated diverticular disease. Operative difficulty is less after a delay of 15 weeks.

Does Abnormal Histology at the Pouch Closure Site Predict Postoperative Morbidity After Hartmann's Procedure? Booth P54 H.K. Nazarian, L.B. Kong, P.R. Fleshner Los Angeles, CA

Integrity of rectal stump closure must be assured for safe execution of Hartmann's procedure. We reviewed the records of 166 patients who underwent Hartmann's procedure during the eight-year period ending October 1992 to determine whether inflammatory changes at the distal resection margin influence postoperative morbidity. Thirty-three of these patients had histologic changes of colitis at the resection margin. Three patients (9 percent) with abnormal histology had pouch leaks compared with one (1 percent) of 133 patients with normal histology (P = 0.02). Abnormal histology was also associated with a higher incidence of postoperative peritonitis, 9 percent vs. 1 percent (P = 0.05). Three of nine patients with an abnormal resection margin and preoperative hypoalbuminemia developed an intra-abdominal abscess compared with none of the 24 patients with an abnormal resection margin and a normal preoperative albumin (P = 0.02). All other preoperative factors in~ cluding steroids, adequate bowel prep, hypotension, and anemia were not associated with additional postoperative morbidity in the subgroup of patients with abnormal histology. Intraoperative factors such as perforation and method of pouch closure also did not significantly correlate with higher morbidity in this subgroup. Patients undergoing Hartmann's procedure with abnormal histology at the resection margin are at higher risk of developing postoperative pouch leak and peritonitis than patients with histologically normal resection margins. National Audit of Complicated Diverticular Disease: Analysis of Index Cases and Natural History over Five Years After Hospital Admission Booth P55 M.R.B. Keighley, N. Farmakis, R. Tudor Birmingham, United Kingdom

Three hundred six cases of complicated diverticular disease were entered into a national audit from 22 hos-

P39

pitals between 1985 and 1988. Admission complications included acute phlegmon (n = 74), pericolic abscess alone (n -- 51), purulent peritonitis (n = 46), large bowel obstruction (n --- 40), fecal peritonitis (n = 35), pericolic abscess complicated by fistula (n = 18), and hypovolemia from lower GI bleeding (n = 42). The overall mortality was 12 percent (4 percent: acute phlegmon; 8 percent: peritonitis; 2 percent: pericolic abscess; 18 percent: fecal peritonitis; 5 percent: large bowel obstruction; 2 percent: bleeding). The outcome in a cohort of 120 patients followed for five years or more was available. Although 67 (56 percent) of these patients had no further symptoms from diverticular disease, 7 still had a persistent colostomy (6 percent), 14 patients (12 percent) remained sympto: matic, and 39 (32 percent) had further complications after hospital admission, 10 of which were fatal. These data imply that nearly half of all patients with complicated diverticular disease have persistent symptoms and a third developed serious complications within five years of their index admission. There is, therefore, a high risk of mortality and serious complications in complicated diverticular disease if it is treated conservatively without a subsequent elective resection.

Anastomotic Dehiscence: Is Surgical Exploration Still Necessary? Booth P56

O. Wiltz, W.D. Wong, S.M. Goldberg, D.A. Rothenberger Minneapolis, MN Anastomotic dehiscence is a major cause of mortality and morbidity. We performed 2,990 colonic and 845 colorectal anastomoses in a five-year period (19871992). Records of 31 (0.8 percent) patients with a clinical leak following colonic (0.5 percent) and colorectal (1.8 percent) anastomoses were reviewed. Surgical indications were cancer (n = 23), IBD (n = 6), diverticular disease (n = 1), and prolapse (n = 1). Stapled anastomoses were performed in 68 percent (n = 21) and singlelayer hand-sewn anastomoses in 32 percent (n = 10). Abdominal pain and tenderness, distention, ieukocytosis, fever, shock, diarrhea, and obstipation were the most common clinical findings. One patient died (3 percent). Immediate exploration was undertaken in three patients. Nineteen patients had Gastrograf in enemas and five (26 percent) did not show a leak, whereas CT scans were done in 13 patients and two (16 percent) failed to demonstrate an abscess or leak. However, in all cases there was either an abnormal CT scan or Gastrograf in enema. All patients with colonic and 10 with colorectal leaks required surgical exploration, drainage, and proximal diversion. Six of 16 patients (38 percent) with colorectal leaks were treated conservatively; two had CTguided drainage, and four had drainage per rectum. These results suggest that either CT scans or Gastrograf in enemas alone may be insufficient to rule out a leak, whereas both studies will maximize diagnostic capabilities. Clinical judgment will identify a subset of patients

MEETING ABSTRACTS

P40

with colorectal leaks that can be managed nonoperatively.

The CT Scan

vs.

the Surgeon--Who Do You Trust? Booth P57

M.W. Arnold, S. Schneebaum, E.W. Martin, Jr., D.C. Young . . . . . . . . . . . . . . . . . . . . . . . . . . . . Columbus, OH Over 156,000 new cases of colorectal cancer are diagnosed yearly. In most hospitals, a preoperative CT scan of the abdomen and pelvis is obtained, at a cost of $600 per patient ($90 million nationwide), to rule out metastatic disease. This study of 27 colorectal cancer patients was performed to determine if this additional expense is warranted. All had preoperative CT scans of the abdomen and pelvis. The accuracy of the preoperative CT scan was compared with that of the initial surgical exploration, based on positive histologic (H&E) confirmation of all suspect tissues found at surgery. Specific sites evaluated were liver, primary tumor, pelvis, and abdomen exclusive of liver. The statistician was blinded to the sites and techniques being evaluated to eliminate bias. True positives and negatives were considered accurate, and false positives and negatives were considered inaccurate. Resuits are shown below:

Liver PrimaryTmr. Pelvis Abdomen (excl. lvr.)

CT Scan Accuracy 70% 56% 81% 74%

P 0.004 0.0003 NS 0.01

Surgical Explor. Accuracy 89% 93% 66% 89%

P NS NS 0.0018 NS

Except for the pelvis, where CT scan was more accurate, surgical exploration was significantly more accurate than preoperative CT scan. This study suggests that routine CT scan could be abandoned except in cases where pelvic disease is likely.

CT-Scan Drainage of Intra-abdominal Abscesses: Pre: and Postoperative Modalities in Colon and Rectal Surgery Booth P58 S. Schechter, T.E. Eisenstat, G.C. Oliver, R.J. Rubin, E.P. Salvati . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plainfield, NJ CT-scan-guided percutaneous drainage of intra-abdominal abscesses has changed the colon and rectal surgeon's approach to pre- and postoperative intra-abdominal infections. In an effort to prove its efficacy, a retrospective study was performed on 70 patients who underwent CT-scan drainage of intra-abdominal abscesses over a six-year period. Fifty-one patients had underlying colon and rectal disease. Eighteen of these patients (35 percent) had spontaneous abscesses and underwent drainage as a preoperative modality, while 33 patients (65 percent) were drained postoperatively. The average severity of illness scores using the APACHE II system were 6 and 9, respectively. The spontaneous group comprised perforated sigmoid colon (eight),

Dis Colon Rectum, April 1993

proximal colon (three), Crohn's disease (four), appendix (two), and small bowel (one). In 56 percent of patients, surgery was avoided by adequate drainage. The postoperative group of patients comprised partial colectomy (27) and ileal pouch procedure (six). Eighty three percent of the postoperatively drained group safely avoided surgery; 17 percent required a repeat drainage procedure. Half of the patients discharged had drainage catheters in place. Mortality among the spontaneous and postoperative groups was 0 and 8 percent, respectively. Mortality was associated with an elevated APACHE II score. CT-scan abscess drainage is an important adjunct to colon and rectal surgery in that 56 percent of spontaneous and 83 percent of postoperative patients avoided emergent surgery.

The Colon and Rectal Disorders of Ehlers-Danlos Syndrome Booth P59 M.L. Snyder, B.A. Orkin, L.E. Smith .. Washington, D.C. The Ehlers-Danlos syndrome comprises nine inherited connective tissue disorders characterized by defects in the synthesis of collagen. Poor wound healing, abnormal scars, recurrent hernias, and incisional hernias discourage surgeons from recommending surgery. Unfortunately, few data are available regarding associated colon and rectal problems. Two hundred eighteen members of the Ehlers-Danlos Foundation, 171 (78 percent) female and 47 (22 percent) male, responded to a survey assessing demographics, past history of colon and rectal disorders, and present bowel habits. The mean age when surveyed was 35 years (2-75 years), and the mean age at diagnosis was 23 years (birth to 68 years). One hundred eighty eight (87 percent) subjects categorized their level of disability as mild-86 (40 percent), moderate-75 (35 percent), or severe-27 (13 percent). A variety of complex colon and rectal problems were reported. Nine (4 percent) subjects reported spontaneous bowel perforations, two (1 percent) had an intussusception , 34 (16 percent) had experienced diverticulitis. Four patients had colostomies created for treatment of perforation. Fifty one (23 percent) subjects reported coccygeal fractures or chronic pain. Altered bowel function included tissue prolapse per anum with evacuation in 65 (30 percent), complete rectal prolapse in 18 (8 percent), blood in the stool in 36 (17 percent), and constipation at some time in 60 (28 percent). Colorectal complaints and complications are c o m m o n in patients with EhlersDanlos syndrome. Because of poor tissue healing, surgery should be reserved for emergencies or chronic conditions that significantly interfere with quality of life. Vascular Expression of Cell Adhesion Molecules in Inflammatory Bowel Disease Booth P60 P.A. Dean, P.S. Ramsey, H. Nelson

. . . . Rochester, MN

Cell adhesion molecules (CAMs) are cell surface glycoproteins that play a central role in immune and inflam-

Vol. 36, No. 4

MEETING ABSTRACTS

matory cell migration from the vascular space. Intercellular adhesion molecule-1 (ICAM-1) and endothelial leukocyte adhesion molecule-1 (ELAM-1) are inducible CAMs involved in the regulation of lymphocyte and neutrophil migration to sites of inflammation, though few studies have investigated their role in colonic disease. The influx of inflammatory cells characteristic of inflammatory bowel disease (IBD) suggests an important role for these CAMs in the pathophysiology of Crohn's and ulcerative colitis. To evaluate this, we used immunohistochemistry to detect the vascular expression of ICAM-1 and ELAM-1 in frozen sections of normal colonic tissue (n -- 23) and in Crohn's and ulcerative colitis specimens (n = 12). Microvessels were detected using the endothelial marker UEA, and CAMs were detected using anti-ICAM-1 and anti-ELAM-1 monoclonal antibodies. Positively stained microvessels (>5 urn) were quantitated at 200x optical field by two independent observers and averaged for a total of 10 fields per specimen. There was no significant difference in the number of microvessels between normal and IBD-involved colon, but a dramatic increase was seen in the number of vessels expressing ICAM-1 in IBD. Specimen

UEA V e s s e l s

ICAM-1 V e s s e l s

Normal IBD

89.8 + 32 81.5 --- 31

2.6 --+-0.4 39.4 +-- 9.1"

* P < 0.001 vs. n o r m a l .

Qualitative changes were noted in ELAM-1 expression, which were undetectable in normal colon but readily detectable in the mucosa, submucosa, and muscularis of colitis samples. These studies indicate that the expression of inducible CAMs is markedly enhanced in active IBD and suggest that these molecules participate in the inflammatory cell influx of IBD. CAMs offer an attractive new therapeutic target for control of active IBD.

Familial Adenomatous Polyposis Is Not Associated with Increased Cell Turnover or Major Changes in Bile Acid Secretion Booth P61

G. Barker, J.P. Neoptolomos, M.R.B. Keighley Birmingham, United Kingdom It has been suggested that bile acids and rapid colonic cell turnover are associated with large bowel cancer. Changes in bile acid secretion and intestinal cell turnover have not been widely explored in precancer. We have collected duodenal bile after proctocolectomy and three-day fecal stool samples in patients with

P41

an intact colon to compare fecal bile acids in familial adenomatous polypsis (n = 11) and in age- and sexmatched controls (n = 10). There was no difference in the total fecal bile acid output between FAP (3.690 #mol/ g dry weight) and controls (3.31 #mol/g dry weight). The only significant difference in component bile acids was in 5-alpha-cholonic acid: 0.049/~mol/g wet weight in FAP compared with an absence of 5-alpha-cholonic acid in controls. Similarly, there was no difference in the duodenal bile acid profile in FAP compared with controls after proctocolectomy. Crypt cell production rates (CCPR) were measured from duodenal and rectal biopsy. There was no significant increase in duodenal or rectal CCPR in FAP compared with controls. These data imply that changes in duodenal bile acids and cell proliferation are not implicated in duodenal malignancy in FAP. Furthermore, it appears unlikely that bile acids are important in the pathogenesis of colorectal cancer in FAP.

The Prevalence of Anti-Endothelial Cell Antibody in Inflammatory Bowel Disease: Immunologic Similarities with Systemic Vasculitis Booth P62 R.T. Patel, A. Pall, D. Adu, M.R.B. Keighley Birmingham, United Kingdom Antiendothelial cell antibody (AECA) and antineutrophil cytoplasmic antibody (ANCA) are known to occur in systemic vasculitis. ANCA has been described in inflammatory bowel disease (IBD), but little is known of the role of AECA in IBD. Steroids remain the cornerstone in the treatment of both groups of diseases. Our aims were to assess both the prevalence and effect of total colectomy on AECA in IBD. Sera from 109 patients were tested for AECA by an ELISA using human umbilical vein endothelial cells: 40 with ulcerative colitis (UC), 30 with UC who had undergone restorative proctocolectomy (RPC), 30 with Crohn's, and 9 controls. Overall, 25/100 patients with IBD were AECA positive: 8/40 with UC, 11/30 who had undergone RPC, and 6/30 with Crohn's. All five patients who had pouchitis were AECA positive. Twenty three of 25 patients who had AECA were also found to have ANCA. These results show that AECA occurs in IBD and that immunologic similarities of IBD and systemic vasculitis are not just coincidence. Despite total ablation of the disease by RPC, AECA persists. Chronic pouchitis may be a possible reason. AECA and ANCA may represent two major antibody systems in IBD which need further investigation.