Portal de Periódicos da CAPES

Aspirin and Folic Acid for the Prevention of Recurrent Colorectal Adenomas

2007; Elsevier BV; Volume: 134; Issue: 1 Linguagem: Inglês

10.1053/j.gastro.2007.10.014

1528-0012

Richard F. Logan, Matthew J. Grainge, V Shepherd, Nicholas Armitage, Kenneth Muir,

Gastric Cancer Management and Outcomes

Background & Aims: Although observational studies have found regular aspirin use to be associated with a reduced risk of colorectal neoplasia, results from randomized trials using aspirin have been inconsistent. Dietary folate intake also has been found to be associated with a reduced risk of colorectal neoplasms in observational studies. Methods: A multicenter, randomized, double-blind trial of aspirin (300 mg/day) and folate supplements (0.5 mg/day) to prevent colorectal adenoma recurrence was performed using a 2 × 2 factorial design. All patients had an adenoma (≥0.5 cm) removed in the 6 months before recruitment and were followed-up at 4-month intervals with a second colonoscopy after approximately 3 years. The primary outcome measure was a colorectal adenoma diagnosed after baseline. Results: A total of 945 patients were recruited into the study, of whom 853 (90.3%) underwent a second colonoscopy. In total, 99 (22.8%) of 434 patients receiving aspirin had a recurrent adenoma compared with 121 (28.9%) of 419 patients receiving placebo (relative risk, 0.79; 95% confidence interval [CI], 0.63–0.99). A total of 104 patients developed an advanced colorectal adenoma; 41 (9.4%) of these were in the aspirin group and 63 (15.0%) were in the placebo group (relative risk, 0.63; 95% CI, 0.43–0.91). Folate supplementation was found to have no effect on adenoma recurrence (relative risk, 1.07; 95% CI, 0.85–1.34). Conclusions: Aspirin (300 mg/day) but not folate (0.5 mg/day) use was found to reduce the risk of colorectal adenoma recurrence, with evidence that aspirin could have a significant role in preventing the development of advanced lesions. Background & Aims: Although observational studies have found regular aspirin use to be associated with a reduced risk of colorectal neoplasia, results from randomized trials using aspirin have been inconsistent. Dietary folate intake also has been found to be associated with a reduced risk of colorectal neoplasms in observational studies. Methods: A multicenter, randomized, double-blind trial of aspirin (300 mg/day) and folate supplements (0.5 mg/day) to prevent colorectal adenoma recurrence was performed using a 2 × 2 factorial design. All patients had an adenoma (≥0.5 cm) removed in the 6 months before recruitment and were followed-up at 4-month intervals with a second colonoscopy after approximately 3 years. The primary outcome measure was a colorectal adenoma diagnosed after baseline. Results: A total of 945 patients were recruited into the study, of whom 853 (90.3%) underwent a second colonoscopy. In total, 99 (22.8%) of 434 patients receiving aspirin had a recurrent adenoma compared with 121 (28.9%) of 419 patients receiving placebo (relative risk, 0.79; 95% confidence interval [CI], 0.63–0.99). A total of 104 patients developed an advanced colorectal adenoma; 41 (9.4%) of these were in the aspirin group and 63 (15.0%) were in the placebo group (relative risk, 0.63; 95% CI, 0.43–0.91). Folate supplementation was found to have no effect on adenoma recurrence (relative risk, 1.07; 95% CI, 0.85–1.34). Conclusions: Aspirin (300 mg/day) but not folate (0.5 mg/day) use was found to reduce the risk of colorectal adenoma recurrence, with evidence that aspirin could have a significant role in preventing the development of advanced lesions. See editorial on page 341. See editorial on page 341. In Western Europe and the United States, colorectal cancer remains the second most common fatal cancer after lung cancer,1Ferlay J. Bray F. Pisani P. GLOBOCAN 2002: cancer incidence, mortality and prevalence worldwide. IARC CancerBase no. 5, version 2.0. IARCPress, Lyon2004Google Scholar being responsible for 16,000 deaths in the United Kingdom in 2005.2Cancer Research UKUK bowel cancer mortality.http://info.cancerresearchuk.org/cancerstats/types/bowel/mortality/?a=5441Google Scholar Aspirin is believed to have a chemopreventive role in colorectal cancer based on considerable observational data, which consistently have shown rates of colorectal cancers and adenomas to be 40%–50% lower in aspirin users.3Hawk E.T. Viner J.L. Umar A. et al.Cancer and the cyclo-oxygenase enzyme.Am J Cancer. 2003; 2: 27-55Crossref Scopus (15) Google Scholar Results from randomized controlled trials of aspirin have been inconsistent. In the Aspirin/Folate Polyp Prevention Study (AFPPS), a significant effect of aspirin in preventing colorectal adenoma recurrence after 3 years was observed for a dose of 81 mg/day (19% reduction in risk), but not for a higher dose of 325 mg/day.4Baron J.A. Cole B.F. Sandler R.S. et al.A randomized trial of aspirin to prevent colorectal adenomas.N Engl J Med. 2003; 348: 891-899Crossref PubMed Scopus (1280) Google Scholar In 2 other secondary prevention trials, aspirin 325 mg/day reduced adenoma occurrence in patients undergoing colonoscopic surveillance after colorectal cancer surgery in 1 study,5Sandler R.S. Halabi S. Baron J.A. et al.A randomized trial of aspirin to prevent colorectal adenomas in patients with previous colorectal cancer.N Engl J Med. 2003; 348: 883-890Crossref PubMed Scopus (1033) Google Scholar whereas in the other study, a reduction in adenoma recurrence at 1 year was not sustained at 4 years.6Benamouzig R. Deyra J. Martin A. et al.Daily soluble aspirin and prevention of colorectal adenoma recurrence: one-year results of the APACC trial.Gastroenterology. 2003; 125: 328-336Abstract Full Text Full Text PDF PubMed Scopus (347) Google Scholar, 7Benamouzig R. Deyra J. Martin A. et al.Daily soluble aspirin and prevention of colorectal adenoma recurrence four years results of the APACC trial (abstr).Gastroenterology. 2006; 130: A101Google Scholar In trials where subjects have no neoplasia detected previously, 2 trials using low doses of aspirin found no reduction in the incidence of colorectal cancer up to 10 years after randomization.8Gann P.H. Manson J.E. Glynn R.J. et al.Low-dose aspirin and incidence of colorectal tumors in a randomized trial.J Natl Cancer Inst. 1993; 85: 1220-1224Crossref PubMed Scopus (406) Google Scholar, 9Sturmer T. Glynn R.J. Lee I.M. et al.Aspirin use and colorectal cancer: post-trial follow-up data from the Physicians’ Health Study.Ann Intern Med. 1998; 128: 713-720Crossref PubMed Scopus (283) Google Scholar, 10Cook N.R. Lee I.M. Gaziano J.M. et al.Low-dose aspirin in the primary prevention of cancer: the Women’s Health Study: a randomized controlled trial.JAMA. 2005; 294: 47-55Crossref PubMed Scopus (694) Google Scholar As a consequence the recent US Preventive Services Task Force statement recommended against the use of aspirin to prevent colorectal cancer in average-risk individuals.11Dube C.D. Rostom A. Lewin G. et al.The use of aspirin for primary prevention of colorectal cancer: a systematic review prepared for the U.S. preventive services task force.Ann Intern Med. 2007; 146: 365-375Crossref PubMed Scopus (288) Google Scholar However, in a recent pooled analysis of 2 UK trials using higher doses of aspirin (≥300 mg/day), reduction in colorectal cancer incidence was observed but this was not evident until 10 years after randomization.12Flossmann E. Rothwell P.M. Effect of aspirin on long-term risk of colorectal cancer: consistent evidence from randomised and observational studies.Lancet. 2007; 369: 1603-1613Abstract Full Text Full Text PDF PubMed Scopus (687) Google Scholar These findings support recent data from the Nurses Health Study, which only found a significant reduction in colorectal cancer incidence among participants taking aspirin for more than 10 years, at doses equivalent to 5 or more standard dose (325 mg) tablets a week.13Chan A.T. Giovannucci E.L. Meyerhardt J.A. et al.Long-term use of aspirin and nonsteroidal anti-inflammatory drugs and risk of colorectal cancer.JAMA. 2005; 294: 914-923Crossref PubMed Scopus (403) Google Scholar It has been suggested that folic acid might reduce the risk of colorectal cancer through various mechanisms.14Kim Y.I. Folate and carcinogenesis: evidence, mechanisms and implications.J Nutr Biochem. 1999; 10: 66-88Abstract Full Text Full Text PDF PubMed Scopus (494) Google Scholar, 15Choi S.W. Mason J.B. Folate status: effects on pathways of colorectal carcinogenesis.J Nutr. 2002; 132: 2413S-2418SPubMed Google Scholar The most widely hypothesized of these are that it reduces aberrations in DNA methylation and that it has an effect on DNA damage and repair.14Kim Y.I. Folate and carcinogenesis: evidence, mechanisms and implications.J Nutr Biochem. 1999; 10: 66-88Abstract Full Text Full Text PDF PubMed Scopus (494) Google Scholar, 15Choi S.W. Mason J.B. Folate status: effects on pathways of colorectal carcinogenesis.J Nutr. 2002; 132: 2413S-2418SPubMed Google Scholar High levels of dietary folate and total folate have been associated with reductions in the risk of colorectal cancer in cohort and case-control studies.16Sanjoaquin M.A. Allen N. Couto E. et al.Folate intake and colorectal cancer risk: a meta-analytical approach.Int J Cancer. 2005; 113: 825-828Crossref PubMed Scopus (301) Google Scholar However, in a recent report describing the folate results from the Aspirin/Folate Polyp Prevention Study, significant harmful effects of folate were reported for 2 of the secondary outcome measures (advanced lesion and ≥3 outcome adenomas) during an extended follow-up period of 3–5 years from the time of the second colonoscopy.17Cole B.F. Baron J.A. Sandler R.S. et al.Folic acid for the prevention of colorectal adenomas—a randomized clinical trial.JAMA. 2007; 297: 2351-2359Crossref PubMed Scopus (767) Google Scholar This observation is consistent with data from animal studies that suggest that although folate initially has a suppressive effect it may accelerate progression of existing preneoplastic lesions.18Ulrich C.M. Potter J.D. Folate supplementation: too much of a good thing?.Cancer Epidemiol Biomarkers Prev. 2006; 15: 189-193Crossref PubMed Scopus (207) Google Scholar The adenoma-carcinoma sequence is known to account for most colorectal cancers19Leslie A. Carey F.A. Pratt N.R. et al.The colorectal adenoma-carcinoma sequence.Br J Surg. 2002; 89: 845-860Crossref PubMed Scopus (504) Google Scholar and colorectal adenomas are regarded as a logical surrogate end point for trials in this area, particularly in patients with a previous history of adenomas where rates of recurrence are known to be high. The United Kingdom Colorectal Adenoma Prevention (ukCAP) trial was set up to determine whether treatment with 300 mg/day aspirin or 0.5 mg/day folic acid could reduce the risk of recurrence over 3 years among patients who had colorectal adenomas removed during endoscopic surveillance. The ukCAP study was a double-blind, randomized, placebo-controlled trial performed in 10 study centers (9 in the UK and 1 in Denmark, see list in Appendix). A 2 × 2 factorial design was used to allow separate assessments of both aspirin and folic acid in preventing the recurrence of colorectal adenomas. Recruitment for this study occurred between December 1997 and November 2001; the last patient completed the study in July 2005. Ethical approval for the study was gained independently for each study center. The trial received International Standard Randomized Controlled Trial Number (ISRCTN) registration during July 2001 (http://www.controlled-trials.com/ISRCTN22669537). Eligible patients were those younger than age 75 who had a colorectal adenoma 0.5 cm or greater (after fixation) removed either in the 6 months before recruitment or before this time if then followed by removal of one or more adenomas of any size in the 6 months before randomization. All patients were therefore required to have a clean colon as determined by a complete colonoscopy or barium enema if colonoscopy was incomplete, at the time of commencing trial medication. Exclusion criteria included serious medical conditions that might preclude successful completion of the trial (eg, hepatic cirrhosis, renal failure, unstable heart conditions), the need for regular treatment with nonsteroidal anti-inflammatory drugs (including aspirin), known intolerance or sensitivity to aspirin, or active ulcer disease, bleeding disorders, or anticoagulant treatment. Patients who had undergone a resection of the large bowel, or in whom adenoma removal was thought to have been incomplete, also were excluded from the study. For all patients recruited a study entry form was forwarded to the study coordinating center, where entry criteria including the reported histology was checked by a senior trial nurse (VCS) and patients found not to meet the inclusion criteria then were withdrawn from the trial. Initially, endoscopy lists, pathology records, and case notes were reviewed to identify patients from the participating study centers. Potentially eligible patients were contacted by letter informing them of the study along with an invitation to attend a consent clinic. This was followed-up by telephone if no response was received within 10 days. General practitioners of patients were contacted once interest had been expressed, to confirm they had no objections to the patient entering the trial. At the consent clinic a brief medical questionnaire was administered where information was taken on past medical history and current medication use to confirm suitability, after which written consent was obtained. Information on dietary folate and calcium intake at baseline was obtained shortly after commencement of the study via an interviewer-administered food frequency questionnaire, which was completed at the patient’s home. In addition, approximately two thirds (n = 600) of randomized patients subsequently agreed to participate in a case-control study from which more detailed information on a number of demographic and lifestyle factors was obtained. From this study we were able to derive details on alcohol intake and family history of colorectal cancer before trial entry. Randomization to treatment group was performed centrally (Queen’s Medical Centre, Nottingham, UK). Patients were randomized to 1 of 4 treatment arms based on the 2 × 2 factorial design using a computer-generated randomization list with a block size of 8, with randomization stratified by center. The randomization schedule for each center was made available to regional pharmacists. The researchers and all clinical staff involved with patient recruitment were blind to this treatment-allocation schedule. Active treatments were 300 mg of enteric-coated aspirin (Bayer Healthcare, Leverkusen, Germany) and 0.5 mg of folic acid supplementation (Surepharm, Burton on Trent, UK), with both medications to be taken daily. Aspirin and folate placebo tablets were identical in appearance to the active treatments. All trial medication was bottled, labeled, and supplied from the central trial pharmacy and dispensed to patients through regional pharmacies at each participating center. Study participants, clinical staff involved with patient surveillance, and investigators at the coordinating center in Nottingham were all blinded to treatment. Only trial pharmacists had knowledge of the treatment administered. Patients were contacted by the research nurses by telephone initially after 1 month and then subsequently at intervals of 4 months to encourage compliance with study medication and to check that they were having no problems tolerating study medication. This was supplemented by home visits at intervals of 12 months. Compliance with study medication was assessed during these contacts and calculated on the basis of self-report, with validation by a tablet count of unused medication either returned to the pharmacy or collected during the 12-monthly (yearly) home visits. Compliance was defined as the percentage of aspirin and folate tablets taken while on trial medication. If the patient permanently withdrew from taking aspirin and/or folate medication before the date of their follow-up colonoscopy, time spent beyond treatment cessation was not included in the denominator for calculating treatment compliance. Patients were advised to contact their research nurse if they experienced any adverse events that could be caused by study medication. In addition, patients routinely were asked during the 4 monthly contacts the question, “Since the last visit have you had any medical problems?” and details of any relevant events were recorded. Information obtained from these contacts was used to supplement events spontaneously reported by the patient. Patients also specifically were asked about dyspepsia, which may have been of insufficient severity to prompt contact with the study team. An adverse event form was completed for all notifications, which was sent to the study coordinating center. All patients were scheduled for a second colonoscopy consistent with existing surveillance practice. At the beginning of the trial this usually was after 3 years for patients diagnosed with colorectal adenomas, although on occasion clinicians would request a follow-up examination after 1 year. All polyps retrieved during this follow-up examination were sent for routine review by local histopathologists. All polyps that were described as adenomatous were classified as either tubular, tubulovillous, or villous, and according to size and presence of dysplasia. The primary outcome measure was the percentage of patients who developed one or more recurrent colorectal adenomas or cancers during the follow-up period (combined outcome). Prespecified secondary outcomes were the number of adenomas detected during follow-up evaluation and the percentage of patients who developed advanced colorectal neoplasia. Advanced neoplasia was defined as adenomas that were either 1 cm or larger in diameter, villous or tubulovillous, or showed severe dysplasia. Any colorectal cancers that developed among trial participants also were included in this category. Only adenomas detected at least 6 months after confirmation of a clean colon at baseline were considered to be recurrent, with three quarters detected at the scheduled end-of-study examination 3 years after randomization. Adenomas or cancers detected during earlier scheduled examinations or from examinations arranged after presentation of symptoms were termed midtrial adenomas and cessation of study medication was agreed at this point, with the patient having reached the trial end point. If no polyps were discovered during midtrial examinations or only metaplastic polyps were found, then the patient continued on trial medication until the scheduled follow-up examination after 3 years. A factorial design allowed examination of 2 interventions. Four treatment groups resulted from this design (aspirin + folate [AF]; aspirin only [AO]; folate only [FO]; and double placebo [P]). We were interested primarily in the main effects of aspirin (AF + AO vs FO + P) and folate (AF + FO vs AO + P) with respect to the rate of adenoma recurrence, hence the data collected were analyzed using an at-the-margins approach (which we have termed a main effects approach).20McAlister F.A. Straus S.E. Sackett D.L. et al.Analysis and reporting of factorial trials: a systematic review.JAMA. 2003; 289: 2545-2553Crossref PubMed Scopus (217) Google Scholar At the outset of the trial it was not anticipated that folate and aspirin would act in a nonadditive manner to prevent adenoma development in our study group, based on current understanding of the mechanisms through which these agents are believed to reduce colorectal cancer risk. The intention at the outset of this trial was to recruit 1000 patients in total over the 10 study centers. For the main effects analysis described earlier, this proposed sample size would provide more than 90% power to detect a one-third reduction (33%) in the risk of adenoma recurrence for both aspirin and folate treatments (independent main effects); assuming equal numbers being assigned to each treatment, a 30% adenoma recurrence rate in untreated patients, and a 10% drop-out rate. The estimated one-third reduction in risk was at the lower end of the range of possible risk reductions that were believed to result from treatment with both aspirin and folate supplements based on epidemiologic evidence available around the time of study inception.21Ferraroni M. La Vechia C. D’Avanzo B. Selected micronutrient intake and the risk of colorectal cancer.Br J Cancer. 1994; 70: 1150-1155Crossref PubMed Scopus (193) Google Scholar, 22Giovannucci E. Rimm E.B. Hunter D.J. et al.Aspirin and the risk of colorectal cancer in women.N Engl J Med. 1995; 333: 609-614Crossref PubMed Scopus (979) Google Scholar Crude relative risks (risk ratios) and 95% confidence intervals (CIs) were calculated to compare adenoma recurrence rates on the basis of aspirin and folate treatment. The number of adenomas per patient with respect to aspirin and folate treatment was compared using the Mann–Whitney U test. All efficacy analyses were performed on the basis of intention to treat (ie, all patients who had a follow-up colonoscopy were included in the analysis regardless of compliance with treatment). Adverse events were recorded and summarized for all study participants, not just those who underwent a second colonoscopy. Small numbers for most events of interest meant that exact tests were used to determine whether any differences in numbers of events reported for aspirin (folate) and nonaspirin- (nonfolate-) treated patients were more extreme than what would be expected by chance alone. All statistical analyses were performed using SAS versions 8.2 and later (SAS Institute, Cary, NC), with a P value of less than .05 used to denote statistical significance. A summary of recruitment is provided in Figure 1. A total of 945 patients were recruited into the trial (307 recruited from East Midlands, 82 from Birmingham, 103 from Bristol, 77 from Liverpool, 43 from Manchester, 33 from Blackpool, 79 from Belfast, 59 from South Yorkshire, 112 from South Wales, and 50 from Denmark). Six patients were classified as having been recruited in error. This included 1 patient for whom consent was withdrawn after 2 months of medication. In the remaining cases, patients were found to be ineligible on the basis of size and histology of baseline polyps during a uniform review of histology records. These patients were withdrawn from the trial very shortly after commencing medication. Baseline characteristics were very similar between the randomization groups overall (Table 1). Overall, the mean age of patients at baseline was 57.8 years (range, 27.6–74.6 y); 67.3% of patients qualified for the study with an adenoma 1 cm or greater in diameter, and 54.9% had baseline adenomas that were villous or tubulovillous. Although a higher percentage of males were randomized to groups receiving both treatments and no active treatment (double placebo), these discrepancies balanced out when comparisons were made with respect to aspirin and folate treatment.Table 1Characteristics of ukCAP Participants: Baseline Measures and Trial DurationCharacteristicFO (n = 234)AO (n = 236)AF (n = 236)P (n = 233)Mean age, y (SD)58.28 (9.36)57.92 (9.68)56.85 (9.04)58.00 (9.37)Male sex, N (%)120 (51.3)128 (54.2)144 (61.0)142 (60.9)Mean no. of qualifying baseline adenomas (SD)1.39 (0.82)1.41 (0.85)1.34 (0.77)1.38 (0.74)Villous or tubulovillous baseline adenomas, N (%)125 (53.7)124 (52.8)124 (53.9)139 (60.2)Baseline adenomas ≥1 cm, N (%)166 (70.9)162 (68.6)157 (66.5)149 (63.9)Mean dietary folate intake, μg/day (SD)aInformation on baseline calcium and folate intake only available for 765 trial participants.309.2 (126.1)310.8 (107.8)311.7 (107.0)297.6 (105.4)Mean dietary calcium intake, mg/day (SD)aInformation on baseline calcium and folate intake only available for 765 trial participants.1095.6 (601.3)1145.0 (608.1)1181.9 (647.8)1131.2 (624.8)Mean weekly units of alcohol consumed, (SD)bInformation on alcohol consumption and family history only available for 600 trial participants., cWeekly units of alcohol consumed averaged over lifetime.10.47 (15.99)10.29 (16.05)10.56 (12.85)10.81 (14.48)Colorectal cancer in first-degree relative, (%)bInformation on alcohol consumption and family history only available for 600 trial participants.30 (19.2)28 (19.6)36 (24.0)38 (25.2)Mean trial duration, y (SD)dTime between baseline and end-of-study colonoscopies., eAnalysis restricted to 853 patients who had a second colonoscopy.3.38 (1.03)3.43 (0.94)3.26 (0.99)3.42 (1.07)Mean duration of trial medication, y (SD)eAnalysis restricted to 853 patients who had a second colonoscopy.2.61 (1.16)2.60 (1.13)2.55 (1.14)2.72 (1.18)Abbreviations: FO, folate only; AO, aspirin only; AF, aspirin and folate; P, double placebo.a Information on baseline calcium and folate intake only available for 765 trial participants.b Information on alcohol consumption and family history only available for 600 trial participants.c Weekly units of alcohol consumed averaged over lifetime.d Time between baseline and end-of-study colonoscopies.e Analysis restricted to 853 patients who had a second colonoscopy. Open table in a new tab Abbreviations: FO, folate only; AO, aspirin only; AF, aspirin and folate; P, double placebo. A total of 853 patients underwent a follow-up endoscopic examination (90.8% of those randomized). Reasons for a follow-up examination not being performed included death, poor health, and failure to contact patients who previously had withdrawn from taking study medication (Figure 1). Of the patients who had a second colonoscopy, 144 (16.9%) had withdrawn from taking both study medications by the time of their follow-up examination (Figure 1). An additional 81 patients withdrew from the aspirin arm of the study only, the most common reason being that the patient had a clinical indication for either aspirin or a medication that potentially could contraindicate aspirin treatment (n = 53). The number of patients withdrawing from the aspirin arm of the trial was slightly higher among those who were assigned to aspirin than assigned to placebo aspirin treatment (49 vs 32 patients). Only 1 patient withdrew from the folate arm of the trial only. During the time spent on trial medication (ie, before formal withdrawal from aspirin and/or folate treatment), compliance reported with study medication was good. Among all patients who underwent a second colonoscopy, 76% reported taking more than 95% of their prescribed aspirin medication, and 81% reported taking more than 95% of their folate medication (Table 2). Less than 80% compliance was reported for only 7% of patients with respect to aspirin treatment and 4% of patients for folate treatment (3% and 1%, respectively, when restricting analysis to patients who completed trial medication). Levels of compliance reported with aspirin and folate medication were very similar according to actual treatment assignment (Table 2).Table 2Summary of Compliance With Study Medications by Aspirin and Folate TreatmentComplianceAspirin complianceFolate complianceAll patients, N (%)Aspirin (AF + AO), N (%)Nonaspirin (FO + P), N (%)All patients, N (%)Folate (AF + FO), N (%)Nonfolate (AO + P), N (%)All patients with second colonoscopy <80%55 (6.7)33 (8.0)22 (5.5)33 (4.0)16 (3.8)17 (4.2) 80% to <90%58 (7.1)27 (6.5)31 (7.7)53 (6.4)27 (6.5)26 (6.4) 90% to <95%84 (10.3)42 (10.1)42 (10.4)75 (9.1)41 (9.8)34 (8.4) ≥95%620 (75.9)312 (75.4)308 (76.4)665 (80.5)335 (79.9)330 (81.2) TotalaExcluded from analysis: aspirin: compliance unknown (n = 15), spent fewer than 28 days on aspirin medication (n = 21); folate: compliance unknown (n = 15), spent fewer than 28 days on aspirin medication (n = 12).817414403826419407Patients who completed aspirin/folate medication <80%17 (2.8)5 (1.7)12 (3.8)8 (1.1)3 (0.9)5 (1.4) 80% to <90%34 (5.5)17 (5.6)17 (5.4)38 (5.5)17 (4.9)21 (6.0) 90% to <95%67 (10.8)34 (11.3)33 (10.4)64 (9.2)33 (9.5)31 (8.8) ≥95%500 (80.9)246 (81.5)254 (80.4)587 (84.2)293 (84.7)294 (83.8) TotalbExclusions from analysis as a result of unknown compliance: aspirin, n = 10; folate, n = 11. Compliance was calculated as follows: (no. of days on medication – no. of missed tablets)/no. of days on medication × 100.618302316697346351a Excluded from analysis: aspirin: compliance unknown (n = 15), spent fewer than 28 days on aspirin medication (n = 21); folate: compliance unknown (n = 15), spent fewer than 28 days on aspirin medication (n = 12).b Exclusions from analysis as a result of unknown compliance: aspirin, n = 10; folate, n = 11. Compliance was calculated as follows: (no. of days on medication – no. of missed tablets)/no. of days on medication × 100. Open table in a new tab A summary of trial outcomes is provided in Table 3. A total of 220 patients (25.8%) had 1 or more colorectal adenomas detected during follow-up evaluation, of whom 104 patients (12.2%) developed 1 or more advanced adenomas. The mean number of adenomas detected per patient was 0.39 (SD, 0.82), ranging from 0.29 (SD, 0.59) in patients taking both active medications to 0.48 (SD, 0.88) among those taking folate only. Of the 220 patients who developed a recurrent adenoma, in 65 patients (29.5%) the adenoma was detected before the scheduled end of trial colonoscopy (median time until detection of midtrial adenomas from baseline examination, 1.2 y; interquartile range, 1.0–2.0 y).Table 3Trial Outcomes by Treatment GroupTreatment groupNo. of patients completing trialPatients with any adenoma(s), N (%)Patients with advanced adenoma(s),aCriteria: adenomas ≥1 cm in diameter, adenomas with villous or tubulovillous histology, adenomas with signs of severe dysplasia, or diagnosis of colorectal cancer. N (%)No. of outcome adenomas per patient, mean (range)FO21565 (30.2)33 (15.4)0.48 (0–5)AO21749 (22.6)22 (10.1)0.33 (0–6)AF21750 (23.0)19 (8.8)0.29 (0–3)P20456 (27.5)30 (14.7)0