Trial Information

A Randomized, Placebo-Controlled, Multi-Center Study of Oral Beclomethasone Dipropionate With Ten Days of Prednisone for Treatment of Gastrointestinal Graft Vs. Host Disease

Patient selection. Patients who develop symptoms of GVHD are evaluated with endoscopy and
mucosal biopsy. If biopsy specimens demonstrate histologic findings of GVHD and stool and
mucosal biopsy cultures are negative for pathogens, patients are invited to participate.
Patients are excluded if diarrhea exceeds one liter per day, or if skin or liver GVHD are
present. All patients receive medications for GVHD prophylaxis; patients receiving
corticosteroids within 30 days of study entry are excluded. Patients signed informed
consent documents approved by Institutional Review Boards.

Formulation of BDP. Immediate release tablets and enteric-coated tablets, each contained 1
mg of BDP (orBec, DOR BioPharma, Miami FL). The dosing regimen was one immediate release and
one enteric-coated tablet, taken orally four times daily (total daily dose, 8 mg BDP).

Stratification and randomization. A stratified allocation scheme is used to balance the
treatment groups within study centers. Stratifying variables are HLA-matched sibling and
use of cutaneous corticosteroids at baseline. Patients receive either BDP or identical
placebo tablets.

Treatment plan. Therapy consists of study drugs plus 10 days of prednisone at an initial
prednisone dose of 1 mg/kg/day. In patients with control of GVHD symptoms at Study Day 10,
prednisone is tapered over 7 days, after which patients were maintained on physiologic
replacement doses of prednisone. Patients who do not demonstrate adequate control of GVHD by
Study Day 10 are considered treatment failures. Patients receive study drug for 50 days, or
until they meet the treatment failure endpoint, or are withdrawn from the study. Patients
who are declared treatment failures have study drug discontinued; subsequent treatment for
GVHD is dictated by their physicians.

Definitions of treatment failure and efficacy end-points. Treatment failure is a worsening
or recurrence of GVHD that requires additional immunosuppressive therapy. The primary
efficacy endpoint is the time to treatment failure through Study Day 50. Secondary efficacy
endpoints included time to treatment failure through Study Day 80 (30 days after
discontinuation of study drug); the proportion who experience treatment failure by Study
Days 10, 30, 50, and 80; and the overall survival rate at day-200 days post-transplant.

Evaluation of drug safety and adverse events. Safety assessment is based on cumulative
prednisone exposure, the incidence and degree of hypothalamic-pituitary-adrenal axis
suppression, and rates of adverse events.

Statistical methods. Analysis of time-to-treatment-failure endpoints and overall survival
rate at day-200 post-transplant includes all randomized patients, based on the Kaplan-Meier
method and log-rank test, stratified by source of allograft. For each endpoint, the hazard
ratio for treatment is estimated based on a Cox proportional hazards model that includes a
term for treatment group. Time-to-treatment-failure is right-censored for patients who
discontinue study drug during the first 50 days without meeting the criteria for treatment
failure, provided the reason for discontinuation was not related to uncontrolled GVHD or
study drug-related toxicity. The cumulative probability of treatment failure by Study Days
10, 30, 50, 60, and 80 is estimated by the complement of the Kaplan-Meier estimator. The
Z-test is used to compare the rates of treatment failure between groups at each of these
time points. Sensitivity analyses are performed using the cumulative incidence method to
assess the impact of competing risks. The overall survival rate at day-200 days
post-transplant is analyzed in the same manner as described for the treatment failure rates.
Because of the variable length of time among patients between transplant and randomization,
treatment is defined in the proportional hazards model as a time-dependent covariate. A
counting process approach is used to construct the risk set for this analysis. Multivariate
models are used to determine if the risk reduction attributed to BDP in the univariate model
was still present after accounting for subject-, disease-, and transplant-related factors
thought to modify patients’ risk of mortality during the 200-day period after transplant.
Hypothesis tests of the primary and secondary endpoints are performed using a two-sided
significance level of 0.05. No adjustments are made to the significance level for
inferential tests of the secondary endpoints. All patients who receive at least one dose of
BDP or placebo are included in the assessment of safety. Analyses are done using SAS®
(version 8.2) and R (version 2.0.1) software.