Methods
Patients
Patients aged 18 years and older with active UP or UPS extending at least 5 cm, but no further than 40 cm from the anal verge, were eligible for enrollment. Patients had mild to moderate disease, with a baseline Modified Mayo Disease Activity Index score (hereafter referred to as "Mayo score") between 5 and 10, inclusive, with subscale ratings of ≥2 for endoscopic appearance and rectal bleeding. The Mayo score is the sum of 4 subscale scores: stool frequency, rectal bleeding, endoscopic findings, and a physician's global assessment. Since publication of the original Mayo Disease Activity Index, the endoscopy subscale was modified such that patients with any degree of friability are classified as having a subscale score of 2.
Exclusion criteria included evidence of Crohn's disease or indeterminate colitis, significant comorbid condition, a positive stool test for bacterial pathogens (Clostridium difficile toxin, or ovum and parasites), and adrenal insufficiency, defined as a measurement of <18 μg/dL serum cortisol after adrenocorticotropic hormone (ACTH) challenge. Medication restrictions included use of systemic, oral, topical, or rectal corticosteroids; laxatives; enemas; treatments for irritable bowel syndrome (eg, alosetron, lubiprostone); anticoagulants; rectal mesalamine therapies; oral mesalamine therapies at dosages of >4.8 g/d; narcotics; antibiotics; and antidiarrheal medications (eg, loperamide, bismuth subsalicylate).
The protocol was approved by institutional review boards and ethics committees. All patients provided written informed consent. All authors had full access to the study data and reviewed and approved the final manuscript.
Study Design
Two identically designed, phase 3, randomized, double-blind, placebo-controlled, multicenter studies (Study 1 [ClinicalTrials.gov ID: NCT01008410] and Study 2 [ClinicalTrials.gov ID: NCT01008423]) were conducted in the United States and Russia during November 2009 to March 2013 (Study 2) or to April 2013 (Study 1). Patients were assigned to a treatment group via a randomization schedule, stratified by study center, generated by an interactive voice response system/interactive web response system. Patients were randomized in a 1:1 allocation to receive budesonide rectal foam 2 mg/25 mL or placebo twice daily for 2 weeks, then once daily for 4 weeks. Concomitant use of oral mesalamine drugs at a stable dosage of up to 4.8 g/d was permitted. Each study consisted of a screening phase (completed within 7 days of randomization), a single-blind run-in/stabilization phase of 4 to 7 days, a 6-week double-blind treatment phase, and a 2-week follow-up phase (Supplementary Figure 1). Via administration of a placebo, the single-blind run-in/stabilization phase allowed patients to practice and familiarize themselves with appropriate use of the foam delivery device before the treatment phase of the study. Patients were required to meet inclusion criteria after the run-in/stabilization phase to continue in the study. A colonoscopy was required for patients newly diagnosed or without a confirmed diagnosis of UC within 12 months of the screening visit. Colonoscopy, if needed, was performed no more than 10 days, and no less than 4 days, before randomization. If a colonoscopy was not required, patients were scheduled for sigmoidoscopy 4 to 7 days before randomization. Histology results from the colonoscopy were required from patients with newly diagnosed UC, before randomization, to confirm active UP or UPS.
(Enlarge Image)
Supplementary Figure 1.
Study design. A colonoscopy was required for patients newly diagnosed or without a confirmed diagnosis of UC within 12 months of the screening visit. Colonoscopy, if needed, was performed no more than 10 days, and no less than 4 days, before randomization. If a colonoscopy was not required, patients were scheduled for sigmoidoscopy 4 to 7 days before randomization. BID, twice daily; EOS, end of study; EOT, end of treatment; QD, once daily.
Assessments
The primary efficacy end point was the percentage of patients achieving remission at week 6 (defined as an endoscopy subscore ≤1, rectal bleeding subscore of 0, and improvement or no change from baseline in the stool frequency subscore of the Mayo score). Scores ranged from 0 to 3 for each subscore of the Mayo score (endoscopy subscore: 0 = normal or inactive disease, 1 = mild disease, 2 = moderate disease, 3 = severe disease; rectal bleeding subscore: 0 = no blood seen, 1 = streaks of blood with stool less than half the time, 2 = obvious blood with stool most of the time, 3 = blood alone passed; stool frequency subscore: 0 = normal number of stools per day for each individual patient, 1 = 1 to 2 stools more than normal, 2 = 3 to 4 stools more than normal, 3 = ≥5 stools more than normal; physician's global assessment subscore: 0 = normal, 1 = mild disease, 2 = moderate disease, 3 = severe disease). Endoscopic disease extent and activity were determined by local investigators.
Key secondary efficacy end points included the percentage of patients achieving a Mayo rectal bleeding subscore of 0 at week 6, the number of scheduled assessments (weeks 1, 2, 4, and 6) in which patients had a rectal bleeding subscore of 0, and the percentage of patients achieving a Mayo endoscopy subscore of 0 or 1 at week 6. Safety assessments included monitoring of AEs, clinical laboratory tests (including morning cortisol concentrations and ACTH challenge tests), and vital signs. For purposes of reporting laboratory-derived AEs, adrenal insufficiency was defined as having a serum cortisol of ≤18 μg/dL at 30 minutes post-ACTH challenge.
Pharmacokinetic Analysis
Blood samples for budesonide pharmacokinetic assessments were collected on multiple visits from patients in the United States. The time of administration of the most recent dose of study drug and the time of blood collection were recorded. Plasma budesonide concentrations were determined using a validated high-performance liquid chromatography/dual mass spectrometry method. Plasma budesonide concentrations were summarized by descriptive statistics and analyzed by population-based pharmacokinetic methods using NONMEM version 7.2 (ICON plc, Hanover, MD). Statistical comparisons of pharmacokinetic data were performed using the extended rank-sum test.
Statistical Analyses
All patients randomized to treatment were included in the intention-to-treat population. The safety population included patients in the intention-to-treat population who received ≥1 dose of the study drug. Baseline characteristics, clinical laboratory values, and AEs were summarized descriptively. Differences in treatment arms for the primary efficacy end point and secondary efficacy end points with percentages as outcomes were analyzed using a logistic regression model after adjusting for analysis center effect. Subgroup analyses of the primary end point were conducted based on the data combined from the 2 studies, by the same method that was used to analyze the primary end point. Secondary efficacy outcomes with categorized changes (eg, change from baseline in Mayo score) were analyzed using an ordinal logistic regression test, adjusting for analysis center effect. Analysis of change from baseline was performed by fitting fixed effects linear models to the data. Analyses of ordinal data (eg, the number of scheduled assessments with rectal bleeding responder classifications) were performed using the proportional odds model for ordinal outcome (ie, PROC LOGISTIC in SAS/STAT 9.3 software; SAS Institute Inc, Cary, NC), adjusting for country effect. Multiplicity of the key secondary efficacy outcomes was addressed by statistical testing of the end points in a hierarchical manner: the percentage of patients achieving a Mayo rectal bleeding subscore of 0 at week 6, the number of scheduled assessments (weeks 1, 2, 4, and 6) in which patients had a rectal bleeding subscore of 0, and 3), and the percentage of patients achieving a Mayo endoscopy subscore of 0 or 1 at week 6. Mean compliance percentages were calculated using the following equation:
Sample size estimates assumed remission rates of 40% and 23% for budesonide foam and placebo, respectively, for both studies. Based on the assumed remission rates and a significance level of α = .05, it was determined that 133 patients were needed in each treatment arm for each study to test the primary efficacy end point with a power of 85%.