There is a well-established association between inflammatory

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2006;4: Aminosalicylate Therapy in the Prevention of Dysplasia and Colorectal Cancer in Ulcerative Colitis DAVID T. RUBIN, ANDELKA LOSAVIO, NICOLE YADRON, DEZHENG HUO, and STEPHEN B. HANAUER The Reva and David Logan Center for Research in Gastrointestinal Diseases, University of Chicago, Chicago, Illinois Background & Aims: Aminosalicylates have been suggested as chemopreventive agents for colorectal cancer (CRC) in ulcerative colitis (UC). We studied the effect of aminosalicylate use on dysplasia and CRC risk in chronic UC. Methods: UC patients with dysplasia or CRC were matched with controls by disease duration, extent, and age at diagnosis. The total amount of aminosalicylates over the duration of the disease and the mean daily amount of drug was calculated. Conditional logistic regression was used to examine the relationship of aminosalicylates to the risk of neoplasia; potential confounders were controlled in a multivariable model. Results: Twenty-six cases (8 CRC, 18 dysplasia) were matched with 96 controls. Cases and controls were similar in age (median, 43 vs 42.5 y), age at diagnosis of UC (median, 29.5 vs 30.5 y), duration of UC (median, 11.5 vs 9 y), and extent of disease (58% pancolitis), sex, family history of UC, history of primary sclerosing cholangitis, and smoking history. Cases were more likely to have a family history of CRC than controls (27% of cases, 9% of controls, P.036). Conditional logistic regression adjusted for disease duration, age at diagnosis, and family history of CRC showed that aminosalicylate use of 1.2 g/day or more was associated with a 72% reduction in the odds of dysplasia/crc (odds ratio, 0.28; 95% confidence interval, ). As the total dose of aminosalicylates increased, the odds of dysplasia/crc decreased (P.056). Conclusions: This case-control study shows a significant risk reduction of dysplasia and CRC in UC patients exposed to aminosalicylate therapy. There is a well-established association between inflammatory bowel disease (IBD) and colorectal cancer (CRC), and the risk factors have been identified primarily in ulcerative colitis (UC) patients. 1 3 The risk of CRC in UC is linked closely with the presence of dysplasia, a greater extent and longer duration of disease, the presence of primary sclerosing cholangitis, a family history of CRC, and possibly the degree of inflammation over time. 4 9 The cumulative incidence of CRC is 5% 10% after 20 years of disease and 12% 20% after 30 years Because IBD is a chronic condition and the risks for CRC are identified, prevention strategies are warranted. The recommended prevention approach has been to perform a screening colonoscopy after 8 years of disease, subsequently followed by periodic surveillance examinations, with biopsy examinations to identify dysplasia or colorectal cancer. 13,14 This secondary prevention strategy requires a commitment to colonoscopies and, when dysplasia or CRC is identified, surgical proctocolectomy to remove the at-risk organ. 13 In recent years, there has been interest in the concept of using medical therapy as the primary prevention of CRC in UC by using medical therapy to reduce the risk of dysplasia and cancer. Although a number of agents have been studied, the predominant data have been derived for aminosalicylates, a mainstay of therapy for UC. The rationale for a potential chemopreventive effect for aminosalicylates relates to its antiinflammatory, antioxidant, and pro-apoptotic properties, as well as evidence for inhibition of cell proliferation. 15 Despite this rationale, recent case-control, health-claims database and population-based studies reported different degrees of effect, and the question of aminosalicylates as a chemopreventive agent in UC remains incompletely answered. 9,16 19 The purpose of this study was to use a case-control design to determine if the use of aminosalicylate medications is associated with a reduced risk of dysplasia or CRC in patients with chronic UC. Materials and Methods Patients This was designed as a matched case-control study comparing cases with chronic UC who developed dysplasia or CRC with controls with chronic UC without neoplasia. Patients in this study were obtained from the University of Chicago IBD Registry, a tertiary clinical database of all IBD patients evaluated at the University of Chicago from 1985 through the present. We identified UC patients with colitis-related dysplasia or CRC from 1985 to 2000, as well as control patients with UC who did not develop neoplasia. At our institution, colitis-associated dysplasia or cancer is diagnosed by an experienced gastrointestinal pathologist and confirmed by a second experienced gastrointestinal pathologist. Controls were patients with UC who had an intact colon and no CRC at the time of cases diagnosed with CRC (index date). Controls were chosen randomly within the IBD registry to match to the identified cases based on the following criteria: (1) age at diagnosis of UC within 5 years, (2) same endoscopic and microscopic extent of disease at the time of diagnosis of UC, and (3) duration of disease within 5 years. The duration of disease was calculated as the elapsed time from diagnosis of UC to the index date. The extent of UC was confirmed endoscopically and histologically. Patients with a primary diagnosis of CRC with UC as an incidental finding were excluded from this study. Subjects Abbreviations used in this paper: CRC, colorectal cancer; IBD, inflammatory bowel disease; UC, ulcerative colitis by the AGA Institute /06/$32.00 doi: /j.cgh

2 November 2006 AMINOSALICYLATE PREVENTS NEOPLASIA IN UC 1347 Table 1. Aminosalicylate Formulations and Their Mesalamine Equivalents Medication Mesalamine (Asacol, Pentasa, Rowasa enema, Canasa suppositories) Olsalazine (Dipentum) Sulfasalazine (Azulfidine) Mesalamine equivalents 1 g mesalamine 1 g mesalamine (reference) 1 g olsalazine 1 g mesalamine 1 g sulfasalazine g mesalamine whose medical charts were incomplete also were excluded. The study was approved by the Institutional Review Board at the University of Chicago. Data Collection Two investigators (N.Y. and A.L.) independently reviewed medical charts and extracted the following data from each patient: sex, confirmation of UC, age at diagnosis of UC, duration of UC, presence of dysplasia/crc, aminosalicylate dose and duration, folic acid use, number of barium enemas and colonoscopies during follow-up evaluation of UC, extent of UC, smoking history, presence of primary sclerosing cholangitis, and family history of IBD or CRC. In regards to family history of CRC, patients were classified as either positive or negative for CRC in a first-degree relative. Disagreements in collected data were resolved by a third investigator (D.T.R.). When information was unclear or incomplete, telephone calls to patients were made to clarify data and ensure accuracy. Analysis of Mesalamine Use Dosages for individual aminosalicylates were converted to equivalent dosages of mesalamine. In this study, aminosalicylates included: ph-release mesalamine (Asacol; Procter and Gamble, Mason, OH), delayed-release mesalamine (Pentasa; Shire Pharmaceuticals, Wayne, PA), olsalazine sodium (Dipentum; Celltech Pharmaceutical, Rochester, NY), sulfasalazine (Azulfidine EN-tabs; Pharmacia & Upjohn, Kalamazoo, MI), mesalamine enemas (Rowasa; Solvay Pharmaceuticals, Inc, Marietta, GA), and mesalamine suppositories (Canasa; Axcan Pharma Inc, Mont Saint-Hilaire, Quebec, Canada). Calculations were performed to determine mesalamine equivalent dosage and the following 2 parameters were reported: (1) the total amount of aminosalicylate over the duration of disease, and (2) the mean daily amount of drug (g/day). Mesalamine equivalents were calculated for each aminosalicylate using the conversions shown in Table 1. The total combined intensity of all aminosalicylate drugs over the course of illness was calculated by dividing the total grams of aminosalicylate over the entire duration of illness, measured in terms of days. When exact dosage changes were unknown it was assumed that the change occurred halfway between the 2 clinic visits. Calculations were performed and reviewed by 2 different individuals to ensure accuracy. Based on Table 2. Characteristics of Cases With Colorectal Cancer or Dysplasia and Controls Characteristics Cases (n 26) Controls (n 96) Median IQR Median IQR P value a Age, y Age at UC diagnosis, y Years of UC Colonoscopies N % N % Sex Female Male Location Proctitis Left-sided Pancolitis Family history of UC Yes No Family history of colorectal cancer Yes No Smoking Yes No Folic acid supplementation (0.4 mg/d or 1.0 mg/d) Primary sclerosing cholangitis Yes No IQR, interquartile range. a P value from conditional logistic regression.

3 1348 RUBIN ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 11 Table 3. Use of Aminosalicylates and the Risk of Colorectal and Dysplasia Mesalamine use Cases (n 26) Controls (n 96) N % N % Odds ratio 95% confidence interval P value Yes No Reference Total dose, g a Reference Daily dose, g 0to Reference to Daily dose, g Reference Mean SD Mean SD Total dose, g b Daily dose, g c a Categories were grouped according to quartile dose of controls. b Odds ratio was calculated as per 1000-g increment. c Odds ratio was calculated as per 1-g increment. previously published results, the predefined primary categorization of mesalamines for the analysis was 1.2 g/day. 18 Statistical Analysis Conditional logistic regression was used to examine the relationship of aminosalicylate to risk of dysplasia or CRC and to balance the variables between cases and controls. The effect was expressed as odds ratios along with the 95% confidence interval. The use of aminosalicylate in total dose and dose per day was analyzed as both continuous and categoric variables. The total dose of aminosalicylate was categorized according to quartile doses of controls. The daily dose was categorized into less than 1.2 g/day, g/day, and 2.4 g/day or more according to clinical aminosalicylate prescriptions. Potential confounding was controlled by using multivariable conditional logistic regression models. A P value of less than.05 was considered statistically significant. Results We identified 26 cases, of whom 8 had CRC and 18 were diagnosed with dysplasia. Ninety-six controls were matched to the 26 cases, with 3.7 controls per case, on average; the number of controls per case ranged from 1 to 6. The median age of cases at the time of diagnosis of CRC or dysplasia was 43 years (range, y). The average age of controls when cases were diagnosed was 42.5 years (range, y). Cases and controls were similar in age, age at diagnosis of UC, and duration of UC, and identical in the extent of UC. The median within matching set difference between controls and cases was 0 years (interquartile range, 3 to 2) in duration of UC and 0 years (interquartile range, 3 to 2.5) in age when cases were diagnosed with cancer or dysplasia. Cases and controls also were similar in the distribution of sex, family history of UC, smoking habit, and use of folic acid. However, cases were more likely to have a family history of CRC as compared with controls (P.036) (Table 2). Because some patients used multiple aminosalicylate formulations during their course, it was not possible to separate the analysis by formulation. However, the total mesalamine use (in grams) for this population was assessed and pooled. Sulfasalazine accounted for 53% of the total aminosalicylate use, phrelease mesalamine 29%, time-release mesalamine 12%, and olsalazine sodium accounted for 6% of the total aminosalicylates in this study. Aminosalicylates were analyzed in both total dose and dose per day, and as both continuous and categoric variables (Table 3). In the univariate analysis, controls were significantly more likely to have used at least 1.2 g/day of mesalamines than cases (P.011). There was no strong evidence of a dose-response relationship because the odds ratios for different dose levels (total or daily) were not in a monotonic decrease trend, although the tests for trend were statistically significant (P.025 for total dose, P.036 for daily dose) (Figure 1). We also used multiple conditional logistic regression to control for confounding (Table 4). Because of the limited sample size, logistic regression cannot include all of the factors simultaneously, hence age and duration of UC were chosen to remove any residual confounding caused by imperfect matching. Family history of CRC also was chosen because it was not distributed evenly between cases and controls. After controlling for these factors, there was a statistically significant association between use of at least 1.2 g/day of mesalamines and risk of CRC or dysplasia in patients with UC (P.024). When the total dose of mesalamines was treated as a continuous variable, the risk of CRC or dysplasia decreased by 16% per 1000-gram increment (P.056). When the daily dose of mesalamines was treated as a continuous

4 November 2006 AMINOSALICYLATE PREVENTS NEOPLASIA IN UC 1349 OR (95% CI) <1.2g g 2.4+ g Daily Dose of Aminosalicylate Figure 1. Risk of CRC or dysplasia in UC and aminosalicylate exposure. Average daily dose of aminosalicylate (5-ASA) equivalents were calculated and grouped into conventional dose ranges. Patients receiving between 1.2 and 2.4 g/day of aminosalicylates had a significantly decreased risk of dysplasia or CRC than those receiving 0 to less than 1.2 g/day. variable, the risk of CRC/dysplasia decreased by 56% per 1-g/day increment (P.03). Discussion This case-control study supports the hypothesis that aminosalicylate therapy is associated with a statistically significant reduced risk of dysplasia or CRC in UC patients. In this study, we show that patients taking a (cumulative) average dose of aminosalicylates of 1.2 g/day or more had a 72% risk reduction of neoplasia. We also show a statistically significant doseresponse relationship for the cumulative dose of mesalamines decreasing the risk of CRC. Our data also confirm that a family history of CRC is associated with an increased risk of neoplasia. Our study results add to a growing list of publications examining the potential chemopreventive effects of aminosalicylate therapy and CRC in UC. 9,16 18 However, publications continue to report varying degrees of benefit or not. Eaden et al 18 performed a case-control study of similar design to this study and reported an 81% risk reduction for CRC in patients taking at least 1.2 g/day of mesalamine. Their study differed from ours in several ways. First, cases were obtained from a variety of centers throughout the United Kingdom, and controls were all from the same referral center. Second, they included additional variables of steroids and physician visits, which also were shown to protect against CRC. van Staa et al 19 reported the association between mesalamine exposure and CRC in patients with UC using the General Research Practice Database (also in the United Kingdom), and also confirmed a 33% risk reduction effect of mesalamine therapy. In their study, sulfasalazine use was not associated with a statistically significantly decreased risk of CRC. Bernstein et al 17 reported that mesalamine therapy was not protective using the Manitoba Canada population database. In their study, variables such as extent of disease and family history of CRC could not be controlled. In addition, the pharmacy data was only of 2 years duration. Furthermore, the number of patients using mesalamines was surprisingly low (119 of a sample size of 373), suggesting either that patients in this population were not using mesalamine therapy in most cases or that the database was not accurately capturing drug use. Rutter et al 9 reported an increased risk of neoplasia based on degree of inflammation, and as a secondary analysis looked at a variety of factors including mesalamine exposure. Their study also did not show a statistically significant benefit to mesalamine therapy. Recently, Velayos et al 20 reported the results of a metaanalysis combining the results of 9 case-control or cohort studies of mesalamine chemoprevention in UC, including the previously unpublished results of this study. They concluded that mesalamine therapy provided a 54% risk reduction for CRC in UC, and when combining end points of dysplasia and CRC together, provided a 51% risk reduction. When using dysplasia alone as an end point, there was not a statistically significant benefit, although the investigators postulated that this was the result of a small number of total cases that were evaluable using this end point. The strengths of our study were that all cases and controls were from a single institution, controls were similar to cases in key variables, and the detailed chart review and data acquisition. In addition, our study collected sufficient information on total mesalamine exposure, as compared with other studies that tried to approximate this by using limited pharmacy data by number of prescriptions. Limitations to this study were that it was a retrospective chart review at a referral center and was exposed to the inaccuracies of data inherent in such studies. As with any such review, it remains possible that there was inaccurate data capture related to the duration of follow-up evaluation at our institution or possibly owing to aminosalicylate dosing changes over time, although we believe that these potential limits were minimized by our methods and exclusion criteria. In addition, we did not attempt to correct for exposure to steroids because we found the record of steroid exposure to be highly variable and unreliable. We believe, however, that the similar characteristics of cases and controls at baseline make it likely that exposure to steroids would be similar, although we cannot be certain. In addition, although we recorded the amount of aminosalicylate in the medical record or based on patient recall, we did not adjust for potential nonadherence to therapy. If patients were Table 4. Multiple Conditional Logistic Regressions for Aminosalicylates and the Risk of Colorectal or Dysplasia Aminosalicylates 95% confidence (5-ASA) AOR a interval P value Ever used aminosalicylates Total dose per 1000-g increase Daily dose per 1-g increase Daily dose 1.2 g AOR, adjusted odds ratio. a After adjusting for age, duration of UC, and family history of CRC.

5 1350 RUBIN ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 4, No. 11 nonadherent with their aminosalicylate therapy, the results of this analysis may have, in fact, been an underestimate of the true effect. In conclusion, this case-control study supports our hypothesis that aminosalicylate therapy is chemopreventive for dysplasia or CRC development in UC. A prospective study of this therapy in high-risk individuals is unlikely to occur because of logistic and ethical limitations, so consideration of the clinical impact of these data should be based on the best available data and safety concerns for our patients. 21 Because the aminosalicylates are a safe and accepted maintenance therapy for UC, the possibility that they provide a chemopreventative benefit provides additional information for clinicians to consider in their recommendation of maintenance therapies and in discussions with patients about adherence to relapse-preventing maintenance therapy. References 1. Gyde SN, Prior P, Allan RN, et al. Colorectal cancer in ulcerative colitis: a cohort study of primary referrals from three centers. Gut 1988;29: Lennard-Jones JE. Cancer risk in ulcerative colitis: surveillance or surgery. Br J Surg 1985;72(Suppl):S84 S Collins RH, Feldman M, Fordtran JS. Colon cancer, dysplasia, and surveillance in patients with ulcerative colitis. A critical review. N Engl J Med 1987;316: Greenstein AJ, Sachar DB, Smith H, et al. Cancer in universal and left-sided ulcerative colitis: factors determining risk. Gastroenterology 1979;77: Sugita A, Sachar DB, Bodian C, et al. Colorectal cancer in ulcerative colitis. Influence of anatomical extent and age at onset on colitis-cancer interval. Gut 1991;32: Levin B. Inflammatory bowel disease and colon cancer. Cancer 1992;70: Brentnall TA, Haggitt RC, Rabinovitch PS, et al. Risk and natural history of colonic neoplasia in patients with primary sclerosing cholangitis and ulcerative colitis. Gastroenterology 1996;110: Nuako KW, Ahlquist DA, Mahoney DW, et al. Familial predisposition for colorectal cancer in chronic ulcerative colitis: a casecontrol study. Gastroenterology 1998;115: Rutter M, Saunders B, Wilkinson K, et al. Severity of inflammation is a risk factor for colorectal neoplasia in ulcerative colitis. Gastroenterology 2004;126: Ekbom A, Helmick C, Zack M, et al. Ulcerative colitis and colorectal cancer: a population-based study. N Engl J Med 1990;323: Katzka I, Brody RS, Morris E, et al. Assessment of colorectal cancer risk in patients with ulcerative colitis: experience from a private practice. Gastroenterology 1983;85: Mir-Madjlessi SH, Farrer RG, Easley KA, et al. Colorectal and extracolonic malignancy in ulcerative colitis. Cancer 1986;58: Kornbluth A, Sachar DB. Ulcerative colitis practice guidelines in adults (update): American College of Gastroenterology, Practice Parameters Committee. Am J Gastroenterol 2004;99: Itzkowitz SH, Present DH. Consensus conference: colorectal cancer screening and surveillance in inflammatory bowel disease. Inflamm Bowel Dis 2005;11: Giovannucci E, Egan KM, Hunter DJ, et al. Aspirin and the risk of colorectal cancer in women. N Engl J Med 1995;333: Pinczowski D, Ekbom A, Baron J, et al. Risk factors for colorectal cancer in patients with ulcerative colitis: a case-control study. Gastroenterology 1994;107: Bernstein CN, Blanchard JF, Metge C, et al. Does the use of 5-aminosalicylates in inflammatory bowel disease prevent the development of colorectal cancer? Am J Gastroenterol 2003;98: Eaden J, Abrams K, Ekbom A, et al. Colorectal cancer prevention in ulcerative colitis: a case-control study. Aliment Pharmacol Ther 2000;14: van Staa TP, Card TR, Leufkens HG, et al. 5-aminosalicylate use and colorectal cancer risk in inflammatory bowel disease: a large epidemiological study. Gut 2005;54: Velayos FS, Terdiman JP, Walsh JM. Effect of 5-aminosalicylate use on colorectal cancer and dysplasia risk: a systematic review and metaanalysis of observational studies. Am J Gastroenterol 2005;100: Rubin DT, Lashner BA. Will a 5-ASA a day keep the cancer (and dysplasia) away? Am J Gastroenterol 2005;100: Address requests for reprints to: David T. Rubin, MD, 5841 South Maryland Avenue, MC 4080, Chicago, Illinois drubin@ medicine.bsd.uchicago.edu; fax: (773) Supported by Procter and Gamble Pharmaceuticals (D.T.R.). D.T.R. is a consultant and receives financial support from Salix Pharmaceuticals and Shire Pharmaceuticals; S.H. is a consultant and receives financial support from Procter and Gamble Pharmaceuticals, Shire Pharmaceuticals, and Salix Pharmaceuticals. Presented at Digestive Disease Week, Orlando, Florida, May 19, 2003.