Are Patients Enrolled in Randomized Controlled Trials (RCTs) Representative of the General Inflammatory Bowel Disease (IBD) Patient Population?

Abstract

Background: Over the past decade, a number of pivotal RCTs have been conducted to determine efficacy and allow regulatory approval of new biologic agents for the treatment of IBD. However, the extent to which inclusion and exclusion criteria from these RCTs may relate to patients in “real-life” practice has not been defined. The aim of this study was to determine the external validity of these RCT results by determining the proportion of “reallife” patients who would have qualified for enrollment into one of the pivotal IBD RCTs. Methods: We performed an observational study of consecutive adult patients aged with an established diagnosis of CD or UC presenting to the outpatient clinics at a tertiary IBD referral center from 1/2008 to 6/2009. Eligible patients had moderate to severe symptoms of CD (Harvey-Bradshaw Index, HBI, score 8-16) and UC (Mayo score 6-12) requiring an escalation in their current medical management. Inclusion and exclusion criteria were extracted from 8 published RCTs of the biologic agents approved for use in CD and UC and applied to the study population to determine trial eligibility. Results: Only 64 (31.1%) of the 206 patients would have qualified for enrollment into any of RCTs. There were no differences in age, gender, disease duration, severity and distribution between trial eligible and ineligible patients. Among the 125 CD patients (HBI score 10.4 ± 3.4), 65.6% would not have qualified for enrollment into any of the biologic RCTs. The majority (74.1%) of the 81 UC patients (Mayo score 8.3 ± 1.4) would not have been eligible for enrollment into the ACT trials. The primary reasons for trial ineligibility are outlined in Table 1.Of the antiTNF naive patients (n=74), only 44.6% would have been eligible for the ACCENT, 35.1% for the CLASSIC, and 16.2% for the SONIC trials. For patients who had not received biologics during the prior 12 weeks (n=98), 34.7% would have qualified for the CHARM and PRECISE trials, 28.6% for the ENCORE and 33.7% for the ENACT trials. Of the 82 CD patients deemed “trial ineligible”, 49.4% ultimately required surgical resection. However, 49.4% received biologics with a 60% response rate to anti-TNF therapy. Among the 60 “trial ineligible” UC patients, 23.3% had colectomies while 41.7% received infliximab or cyclosporine (CSA). There was a 63.2% response rate to anti-TNF therapy and 33.3% response to CSA. Conclusions: Only 1/3 of patients with moderate to severe CD or UC seen in a tertiary care IBD outpatient practice would qualify for enrollment into one of the above pivotal RCTs. However, the anti-TNF response rates among these “trial ineligible” patients were similar to those reported in the RCT setting. Prospective studies investigating external validity of RCT trial data including both community and referral center patients are needed to truly ascertain generalizabilty of trial results. Table 1. Primary reasons for clinical trial eligibility.