Long-term Treatment With Sulindac in Familial Adenomatous Polyposis: A Prospective Cohort Study

GASTROENTEROLOGY 2002;122:641 645 Long-term Treatment With Sulindac in Familial Adenomatous Polyposis: A Prospective Cohort Study MARCIA CRUZ CORREA, LINDA M. HYLIND, KATHARINE E. ROMANS, SUSAN V. BOOKER, and FRANCIS M. GIARDIELLO Department of Medicine, Division of Gastroenterology, The Johns Hopkins University School of Medicine, Baltimore, Maryland Background & Aims: Management of patients with familial adenomatous polyposis (FAP) can consist of colectomy with ileorectal anastomosis (IRA). Sulindac, a nonsteroidal anti-inflammatory drug, causes regression of colorectal adenomas in the retained rectal segment of FAP patients, although long-term use of this therapy has not been studied. We evaluated the long-term effectiveness and toxicity of sulindac in attempting to maintain retained rectal segments free of adenomas. Methods: Twelve FAP patients (5 women), mean age 37.1 years, with IRA received sulindac (mean dosage, 158 mg/day) for a mean period of 63.4 31.3 months (range, 14 98 months). Number, size, and histologic grade of polyps, side effects, and medication compliance were assessed every 4 months. Results: Seven of 12 patients (58%) remained in the study (6 of these polyp-free) for a mean of 76.9 27.5 months. Five of 12 patients (42%) withdrew from the trial after a mean follow-up period of 44 28 months (range, 14 89 months). A significant regression of polyp number was observed in all patients at 12 months (P 0.039) and at a mean of 63.4 31.3 months (P 0.006). Prevention of recurrence of higher-grade adenomas (tubulovillous, villous adenomas) was also observed (P 0.004). At 35 months of follow-up, 1 patient developed stage III cancer in the rectal stump. The most common side effect was rectal mucosal erosions in 6 patients. Conclusions: Long-term use of sulindac seems to be effective in reducing polyp number and preventing recurrence of higher-grade adenomas in the retained rectal segment of most FAP patients. Erosions at the IRA site can preclude adequate dose maintenance. Familial adenomatous polyposis (FAP) is an autosomal-dominant inherited disease caused by a germline mutation of the APC gene on chromosome 5. 1 FAP is characterized by innumerable colorectal adenomas in teenagers and the inevitable development of colorectal cancer. 2 Surgery is the definitive treatment in patients with FAP. Among the surgical alternatives is total colectomy with ileorectal anastomosis (IRA), which is associated with low mortality, low morbidity, and a good functional outcome. 3,4 However, frequent occurrence of cancer in the rectal stump after IRA has been reported. 5 9 Several risk factors have been associated with the development of rectal cancer, including the number of rectal polyps before surgery, time of follow-up, level of IRA, presence of colonic cancer at surgery, surveillance, site of gene mutation, therapy of rectal polyps, and age. 9 11 In 1983, Waddell et al. 12 reported that sulindac, a nonsteroidal anti-inflammatory drug (NSAID), caused regression of rectal adenomatous polyps in patients with FAP. Sulindac reduces prostaglandin synthesis by inhibition of cyclooxygenase-1 and cyclooxygenase-2 enzymes. 13 Subsequent randomized, double-blind, placebocontrolled, short-term studies have confirmed a significant decrease in number and size of polyps in the residual rectum of patients with IRA with the use of this medication. 14,15 However, data on prolonged use of sulindac for regression of colorectal adenomas are limited. 10 Therefore, we evaluated the long-term effectiveness and toxicity of sulindac in maintaining patients with retained rectal segments free of rectal adenomas. Patients and Methods Patients Twelve FAP patients with IRA were enrolled in a long-term study to evaluate the use of sulindac for regression of colorectal polyps. Subjects were recruited from The Johns Hopkins Polyposis Registry. Inclusion criteria were patients with FAP who had undergone colectomy with IRA and had 5 or more rectal adenomas. Exclusion criteria included the following: use of NSAIDs for more than 1 week during the previous 3 months, absence of effective birth control in women of childbearing age, pregnancy, white blood cell count 4000/ ml, platelet count 100,000/mL, blood urea nitrogen 25 mg percent or creatinine 1.5 mg percent, refusal to discontinue use of NSAIDs, history of peptic ulcer disease or gastro- Abbreviations used in this paper: FAP, familial adenomatous polyposis; IRA, ileorectal anastomosis. 2002 by the American Gastroenterological Association 0016-5085/02/$35.00 doi:10.1053/gast.2002.31890

642 CRUZ CORREA ET AL. GASTROENTEROLOGY Vol. 122, No. 3 intestinal hemorrhage, history of malignancy, active bacterial infection, use of dimethylsulfoxide, and history of aspirin allergy. The study was approved by The Johns Hopkins Institutional Review Board, and informed consent was obtained from all patients. At enrollment, none of the patients in the study had been undergoing routine surveillance of the rectum. Study Design Participants were examined by flexible sigmoidoscopy before administration of sulindac (0 months) and with follow-up procedures every 4 months using an Olympus flexible sigmoidoscope (Olympus, Melville, NY). Number and size of polyps were assessed at each visit. Also, at each endoscopic examination, 2 polyps were sampled for histology with care taken not to remove the polyps. When more than 2 polyps were present in the rectum, the 2 largest were sampled in an attempt to obtain the highest dysplasia grade. No effort was made to clear the rectum of polyps either by polypectomy or excisional biopsy. One observer (F.M.G.) performed all assessments. Intestinal preparation was performed by using magnesium citrate or a saline-solution enema to avoid mucosal edema. The endoscopist counted the total number of polyps in the entire circumference of colorectum from the IRA to the anal verge and recorded the examination on videotape. Biopsy specimens for routine histologic examination were obtained at each endoscopy on both the mucosa and polyps. All biopsy specimens were fixed in 10% buffered formalin, embedded in paraffin, and cut at 4- m thickness; all paraffin sections were stained with H&E. Adenomas were classified according to histologic features (tubular, tubulovillous, or villous adenoma). Each patient initially received sulindac 150 mg orally twice a day. Drug dosage was adjusted with an attempt to maintain the rectum free of polyps. Dosage adjustment was as follows: (1) patients were started on sulindac 300 mg/day, but the dosage was reduced by half (150 mg/day) if the rectum was polyp-free on follow-up endoscopy; (2) the dosage was further decreased to 150 mg every other day if the rectum was polyp-free on follow-up endoscopy; (3) if there was recurrence of rectal polyps on subsequent endoscopic examinations, the dosage of sulindac was increased to the prior dosage. Patient compliance was assessed by tablet counts and monthly telephone contact (L.M.H.). Safety was monitored by monthly telephone interviews and at the 4-month visits. Additionally, blood tests, including complete blood count, serum creatinine, serum electrolytes, and bilirubin, were performed at each visit. Adverse events were graded in accordance with the Common Toxicity Criteria of the National Cancer Institute. 16 Dependent and Independent Variables The main dependent variable was the percent change from baseline number of polyps after treatment with sulindac, calculated at 12 months and at last follow-up. Additionally, histologic grade of adenomas encountered at 12 months and at last follow-up was analyzed. Independent clinical variables included age (younger than 35 years vs. 35 years or older), sex, body weight (kg), years post-ira surgery ( 10 years vs. 10 years), dosage of sulindac (mg kg 1 day 1 ), compliance, number of polyps at baseline, and rectal mucosal erosions. Statistical Analysis Wilcoxon signed-rank test was used to estimate the difference in polyp number and histologic grade at 1 year and at last follow-up. Percent change from baseline polyp number was calculated at 1 year and at last endoscopy follow-up. Analysis was performed as intention to treat. The relation between percent change from baseline polyp number (dependent variable) and patient clinical features (independent variables) was evaluated by both univariate and multivariate analysis. Differences among patients who were withdrawn from and those who remained in the study were determined and tested for significance by the Mann Whitney U test. Statistical analyses were performed using STATA 6.0. 17 Results Clinical Effect Table 1 shows the baseline clinical features of the study population. Seven of 12 patients (58%) were maintained within the study for a mean of 76.9 27.5 months, with 6 of these patients essentially polyp-free on the last visit ( 5 adenomas) (Table 2). Five of 12 patients (42%) were withdrawn from the study after a mean follow-up period of 44 28 months (range, 14 89 months). Reasons for withdrawal included development of rectal cancer (1 patient), progression of histology (1 patient), increase in polyp number (1 patient), refractory rectal mucosal erosions (1 patient), and poor compliance (1 patient). There were no statistically significant differences in age, study follow-up time, compliance, or mean dosage of sulindac between the patients withdrawn and those not withdrawn from the study. Table 1. Clinical Characteristics of Study Population Characteristics Patients (n) 12 Mean age, yr (SD) (range) 37.1 (11.1) (21.5 53.8) Sex (%) Female 5 (42) Male 7 (58) Body weight, kg (range) 74.9 (45 118) Years post-ira surgery, mean SD (range) 12.6 10.3 (1.38 35.5) Dosage of sulindac, mean SD (range) (mg kg 1 day 1 ) 2.5 0.99 (0.07 3.8) Mean compliance (%) SD (range) 86.5 9.7 (70 98) Mean follow-up time SD (mo) (range) 63.4 0.99 (14 98) Patients withdrawn (n) 5

March 2002 LONG TERM TREATMENT WITH SULINDAC IN FAP 643 Table 2. General Characteristics of Patients Submitted to Sulindac Therapy Patient Sex/age (yr) before therapy after 12 months of after long-term Length of (mo) Development of rectal cancer 1 F/32 19 4 4 14 a No 2 F/42 47 0 0 80 No 3 F/52 10 0 4 89 a No 4 F/25 7 0 0 35 a Yes b 5 M/36 80 25 50 52 a No 6 M/21 16 14 22 33 a No 7 M/22 80 28 10 80 No 8 F/24 29 0 3 89 No 9 F/35 16 0 2 86 No 10 F/33 17 7 4 16 No 11 M/46 19 2 0 98 No 12 M/45 7 2 1 89 No a Withdrawn from study. b Rectal carcinoma from an ulcerated lesion after 35 months of therapy. A significant regression of polyp number was observed in all patients at 12 months and at last follow-up (63.4 31.3 months) (Table 3). Mean percentage decrease from baseline polyp number was 76% at 12 months and 72% at last follow-up (P 0.01). There was no significant association between the regression of polyps and any of the clinical variables studied (age, sex, body weight, dosage of sulindac, compliance, and number of polyps at baseline) by univariate or multivariate analysis. A description of adenoma type at the time of enrollment is given in Table 4. At baseline, 58% of the patients had tubular adenomas, whereas 42% had tubulovillous adenomas. At last follow-up, a significant decrease in the recurrence of higher-grade adenomas was noted (P 0.004). At 35 months of follow-up, 1 patient developed stage III cancer in the rectal stump from an ulcer-like lesion. This individual was adenoma-free at the time of diagnosis of rectal cancer. Dosage of Sulindac Patients received a mean sulindac dosage of 2.54 mg kg 1 day 1 (range, 0.07 3.84 mg kg 1 day 1 ). There was no statistically significant correlation between the dosage of sulindac (mg kg 1 day 1 ) and the percentage change from baseline polyp number at 12 months (P 0.23) or at last follow-up (63.4 31.3 months) (P 0.38) (Figure 1). The minimal effective dosage at which there was 90% regression in polyp number from baseline was 2.62 0.54 mg kg 1 day 1. Compliance with drug treatment was excellent; patients took 86.5% 9.7% of the scheduled doses. Adverse Events Few adverse events were reported (Table 5). The most common side effect identified was rectal mucosal erosions, developing in 6 of 12 patients (50%) at a mean interval of 25.3 months (range, 1.44 52.11 months). Mucosal erosions occurred at the IRA site. Perturbations in laboratory studies were not noted. Discussion In this study, short- and long-term use of sulindac was effective in reducing adenomas in the retained rectal segments of patients with FAP. Fifty-eight percent of the patients were maintained on for a mean follow-up period of 77 months (86% of these individuals were polyp-free at last follow-up). Our findings differ from a prior study by Tonelli et al. 10 These investigators reported no significant decrease in rectal adenomas with Table 3. Effect of Sulindac on the Number of Rectal Polyps Number of polyps Mean (SD) Range % Reduction from baseline no. of polyps P value Baseline 28.9 (26.2) 7 80 12 months 6.8 (10.1) 0 28 76 0.002 Last follow-up 8.3 (14.5) 0 50 74 0.004 NOTE. P value determined by Wilcoxon signed-rank test. Polyp regression expressed as percent of baseline polyp number. Table 4. Effect of Sulindac on the Histology of Rectal Polyps Baseline Last follow-up a Tubular adenoma, n (%) 7 (58) 6 (50) Tubulovillous adenoma, n (%) 5 (42) 1 (8) Villous adenoma, n (%) 0 1 (8) Cancer, n (%) 0 1 (8) a P 0.004 determined by Wilcoxon signed-rank test.

644 CRUZ CORREA ET AL. GASTROENTEROLOGY Vol. 122, No. 3 Figure 1. Relationship between polyp regression and dosage of sulindac (mg kg 1 day 1 ) at last follow-up. *Polyp regression on sulindac expressed as percent of baseline polyp number after 12 months of treatment with sulindac (numbers correspond to the identification numbers of each patient from Table 2). the use of sulindac for FAP patients with IRA after a mean period of 48.6 28.7 months. The longer follow-up period in the present study may contribute to the differences observed. Analysis between clinical characteristics and change in polyp number showed that baseline (initial) number of polyps was not a determinant of polyp regression. Furthermore, none of the clinical variables studied was independently associated with reduction of rectal polyps. Also, no significant correlation was found between the dosage of sulindac (mg kg 1 day 1 ) and the mean percentage change from baseline polyp number. The minimal effective dosage at which a 90% reduction in adenomas occurred was 2.62 mg kg 1 day 1, an oral dosage lower than in previous studies. 15,18 However, use of a rectal formulation of sulindac has been reported to be effective at a mean dosage of 67 mg/day, 19 which is a mean dosage less than used orally in our investigation (151.4 56.8 mg/day [range, 69 238 mg/day]). Issues regarding compliance with rectal and the unavailability of a standard rectal formulation prevented our use of this route of administration. The mechanism of action of NSAIDs (including sulindac) in chemoprevention is unknown but may be linked to effect on mucosal prostaglandin levels. Cyclooxygenase-2 overexpression has been shown to be associated with increased prostaglandin E 2 biosynthesis and angiogenesis. 18 Sulindac decreases the biosynthesis of prostaglandin E 2 by cyclooxygenase-2 inhibition. 20 However, individual patient response to sulindac is heterogeneous. In a previous study, we showed interpatient heterogeneity of response to sulindac with a wide range of reduction of prostaglandin E 2 levels. 21 The reason for this heterogeneity is still unclear. Treatment with sulindac seemed to significantly decrease the recurrence of higher-grade adenomas. Similar findings were reported by Winde et al., 19 who performed a nonrandomized, controlled phase II clinical trial in 25 FAP patients with IRA. Fifteen patients received a rectal formulation of sulindac for 132 weeks. Dysplasia reversion (significant change to low-grade dysplasia) induced by rectal sulindac maintenance therapy was noted in all but 2 patients who persisted with medium or severe dysplasia. To our knowledge, the correlation between reversion of polyp dysplasia grade and oral sulindac therapy has not been described previously. Of concern, 1 patient developed a rectal carcinoma. The occurrence of cancer in the rectal remnant of patients with FAP during has been described in 3 other patients. 10,22,23 In 2 of these patients, the tumor appeared endoscopically as a flat, ulcerated lesion, as noted in our patient. Although several studies have estimated the risk of cancer development in FAP patients with IRA, 24 27 comparison with the rate of rectal cancer in patients treated with sulindac chemoprevention will require additional follow-up. In this patient cohort, use of sulindac over a prolonged period produced virtually no adverse events or perturbations in laboratory values. The most common side effect was asymptomatic mucosal erosions at the IRA site, which required adjustment of sulindac dosage. Two of these patients were removed from the protocol when lower dosages of medication failed to prevent recurrence Table 5. Incidence and Toxicity Grade of Adverse Reactions in the Study Group Adverse reaction n (%) Toxicity grade Gastrointestinal erosions (IR) 6 (50) a G2 Gastrointestinal other 1 (8) b G2 (abdominal bloating) Hepatic hyperbilirubinemia 2 (17) c G2 Pulmonary other (bronchitis) 2 (17) G1 Syndromes other (flu-like) 2 (17) d G1 2 Auditory/hearing other (tinnitus) 1 (8) e G1 Metabolic/laboratory (hypokalemia) 1 (8) f G1 Neurologic (dizziness) 1 (8) g G1 NOTE. Grading score is specific for each system/symptom: G0, no adverse reaction; G1, asymptomatic, not requiring treatment; G2, symptomatic, may require medical therapy; G3 and G4, more advanced adverse reactions varying according to the symptoms. a Ileorectal anastomosis site erosions. b Abdominal bloating after starting medication (lasted 10 days). c Gilbert s disease. d Flu-like syndrome consists of fever and myalgia with or without nausea and/or vomiting and/or diarrhea and/or headache and/or abdominal cramps. e Associated with bronchitis/upper respiratory infection. f Patient dehydrated. g Related to pressure in ears/earache. IR, ileorectal.

March 2002 LONG TERM TREATMENT WITH SULINDAC IN FAP 645 of adenomas. In this setting, the use of celecoxib, a selective cyclooxygenase-2 inhibitor with reportedly fewer gastrointestinal side effects, may be of interest. Several clinical implications can be drawn from this study, albeit with caution given the sample size. First, sulindac is an effective short- and long-term therapy for the treatment of adenomas in most FAP patients with IRA. Second, treatment with sulindac of FAP patients requires frequent endoscopic surveillance to assess polyp response and adjust medication dosage. Third, although sulindac is well tolerated with minimal patient side effects, colorectal mucosal erosions may preclude adequate dosing. In summary, most experts would recommend that young patients with classic FAP undergo total proctocolectomy to eliminate the risk of colorectal cancer and need for frequent endoscopy. However, in patients with IRA with retained rectum, use of may reduce the complexity and difficulty of follow-up examinations by decreasing the number of polyps required to be sampled and/or ablated. Nevertheless, because of the occurrence of rectal cancer in this setting, endoscopic surveillance should be accomplished with a high degree of vigilance. References 1. Kinzler KW, Nilbert MC, Su LK, Vogelstein B, Bryan TM, Levy DB, Smith KJ, Preisinger AC, Hedge P, McKechnie D, Finniear R, Markham A, Groffen J, Boguski MS, Altschul SF, Horii A, Ando H, Miyoshi Y, Miki Y, Nishisho I, Nakamura Y. Identification of FAP locus genes from chromosome 5q21. Science 1991;253:661 665. 2. Bussey HJR. Familial polyposis coli. Family studies, histopathology, differential diagnosis and results of treatment. Baltimore: The Johns Hopkins University Press, 1975. 3. Jagelman DG. Report form the 4th biennial meeting of the Leeds Castle Polyposis Group. Int J Colorectal Dis 1991;6:224 236. 4. Dozois RR. Restorative proctocolectomy and ileal reservoir. Mayo Clin Proc 1986;61:283 286. 5. Bess MA, Adson MA, Elveback LR, Moertel CG. Rectal cancer following colectomy for polyposis. Arch Surg 1980;115:460 467. 6. De Cosse JJ, Bulow S, Jarvinen H, Alm T, Hultcrantz R, Moesgaard F, Costello C. Rectal cancer risk in patients treated for familial adenomatous polyposis. The Leeds Castle Polyposis Group. Br J Surg 1992;79:1372 1375. 7. Iwama T, Mishima Y, Utsumomiya J. The impact of familial adenomatous polyposis on tumorigenesis and mortality at the several organs. Ann Surg 1993;217:101 108. 8. Heiskanen I, Jarvinen HJ. Fate of the rectal stump after colectomy and ileorectal anastomosis for familial adenomatous polyposis. Int J Colorectal Dis 1997;12:9 13. 9. Bullow C, Vasen H, Jarvinen H, Bjork J, Bisgaard ML, Bulow S. Ileorectal anastomosis is appropriate for a subset of patients with familial adenomatous polyposis. Gastroenterology 2000; 119:1454 1460. 10. Tonelli F, Valanzano R, Messerini L, Ficari F. Long-term treatment with sulindac in familial adenomatous polyposis: is there an actual efficacy in prevention of rectal cancer? J Surg Oncol 2000; 74:15 20. 11. Friedl W, Caspari R, Sengteller M, Uhlhaas S, Lamberti C, Jungck M, Kadmon M, Wolf M, Fahnenstich J, Gebert J, Moslein G, Mangold E, Propping P. Can APC mutation analysis contribute to therapeutic decisions in familial adenomatous polyposis? Experience from 680 FAP families. Gut 2001;48:515 521. 12. Waddell WR, Loughry RW. Sulindac for polyposis of the colon. J Surg Oncol 1983;24:83 87. 13. Nonsteroidal anti-inflammatory agents. In: McEvoy G, ed. AHFS drug information. Bethesda, MD: The American Society of Hospital Pharmacists, 1993:1201 1205. 14. Giardiello FM, Hamilton SR, Krush AJ, Piantadosi S, Hylind LM, Celano P, Booker SV, Robinson CR, Offerhaus GJ. Treatment of colonic and rectal adenomas with Sulindac in familial adenomatous polyposis. N Engl J Med 1993;328:1313 1316. 15. Labayle D, Fisher D, Vielh P, Drouhin F, Pariente A, Bories C, Duhamel O, Trousset M, Attali P. Sulindac causes regression of rectal polyps in familial adenomatous polyposis. Gastroenterology 1991;101:635 639. 16. Cancer Therapy Evaluation Program. Common toxicity criteria. Bethesda, MD: National Cancer Institute, March 1998. 17. StataCorp. Stata Statistical Software: release 6.0. College Station, TX: Stata Corp., 1999. 18. Nugent KP, Farmer KC, Spigelman AD, Williams CB, Phillips RK. Randomized controlled trial of the effect of sulindac on duodenal and rectal polyposis and cell proliferation in patients with familial adenomatous polyposis. Br J Surg 1993;80:1618 1619. 19. Winde G, Schmid KW, Schlegel W, Fischer R, Osswald H, Bunte H. Complete reversion and prevention of rectal adenomas in colectomized patients with familial adenomatous polyposis by rectal low-dose sulindac maintenance treatment. Dis Colon Rectum 1995;38:813 830. 20. Uefuji K, Ichikura T, Mochizuki H. Cyclooxygenase-2 expression is related to prostaglandin biosynthesis and angiogenesis in human gastric cancer. Clin Cancer Res 2000;6:135 138. 21. Giardiello FM, Spannhake EW, DuBois RN, Hylind LM, Robinson CR, Hubbard WC, Hamilton SR, Yang VW. Prostaglandin levels in human colorectal mucosa: effects of sulindac in patients with familial adenomatous polyposis. Dig Dis Sci 1998;43:311 316. 22. Niv Y, Fraser GM. Adenocarcinoma in the rectal segment in familial polyposis coli is not prevented by. Gastroenterology 1994;107:854 857. 23. Lynch HT, Thorson AG, Smyrk T. Rectal cancer after prolonged sulindac chemoprevention. Cancer 1995;75:936 938. 24. Bulow C, Vasen H, Jarvinen H, Bjork J, Bisgaard ML, Bulow S. Ileorectal anastomosis is appropriate for a subset of patients with familial adenomatous polyposis. Gastroenterology 2000; 119:1454 1460. 25. Bussey HJR, Eyers AA, Ritchie SM, Thomson JPS. The rectum in adenomatous polyposis: the St. Mark s policy. Br J Surg 1985; 72(suppl):S29 S35. 26. Bess MA, Adson MA, Elveback LR, Moertel CG. Rectal cancer following colectomy for polyposis. Arch Surg 1980;115:460 467. 27. Sarre RG, Jagelman DG, Beck GJ, McGannon E, Fazio VW, Weakley FL, Lavery IC. Colectomy with ileorectal anastomosis for familial adenomatous polyposis: the risk of rectal cancer. Surgery 1987;101:20 26. Received June 18, 2001. Accepted November 2, 2001. Address requests for reprints to: Marcia Cruz-Correa, M.D., Department of Medicine, Division of Gastroenterology, The Johns Hopkins University School of Medicine, 1830 East Monument Street, Room 420, Baltimore, Maryland 21205. e-mail: macruzco@jhsph.edu; fax: (410) 614-8337.