We conducted a two-stage phase II study to assess safety and activity of sorafenib in patients with relapsed aggressive DLBCL. Response assessment was based upon the criteria from the International Workshop to Standardize Criteria for non-Hodgkin Lymphoma . The study was conducted through the Eastern Cooperative Group and was approved by the respective Institutional Review Boards. Patients with de novo or transformed DLBCL were eligible if they had previously received therapy with curative intent and had relapsed greater than 2 months after their last treatment. Patients were required to have progressed after or be ineligible for autologous stem cell transplant. Eligibility criteria included age greater than 18 years old, ECOG performance status (PS) of 0–1, measurable disease by computed tomography, absolute neutrophil count count ≥ 1,000/mm3, platelet count ≥ 75,000/mm3, normal serum creatinine, total bilirubin ≤ 2.0 times institutional upper limit of normal, AST ≤ 2.5 × institutional upper limit of normal, ALT ≤ 2.5 times institutional upper limit of normal, and normal PT/INR.
Patients received sorafenib at a dose of 400 mg PO BID continuously in 28-day cycles. Patients who showed no disease progression at the end of cycle 2 were to receive an additional 4 cycles (for a total of 6 cycles) of sorafenib. Patients who were responding or stable at the end of cycle 6 were to continue to receive 28-day cycles of sorafenib until progressive disease or excessive toxicity.
Patients were instructed to take the tablets every 12 hours with an 8 oz. glass of water, with or without food. If sorafenib was taken with meals, patients were instructed to take sorafenib with a moderate to low-fat meal. To track compliance, patients were required to complete a pill calendar. Adverse events reporting requirements and appropriate dose modifications in case of toxicities were described in the protocol. Patients were restaged for response after 2 and 6 cycles using the International Workshop Criteria. Patients who progressed or had unacceptable toxicity at any time discontinued therapy. Patients with stable disease after 6 cycles continued treatment at physician’s discretion. Responding patients were to continue on medication.
Statistical design and method
The study used a two-stage Simon design  to assess the clinical efficacy of sorafenib in patients with relapsed DLBCL. A total of 37 eligible patients were required to test the null hypothesis that the true response rate for this regimen is at most 5% versus the alternative hypothesis that the true overall response rate is 20% or greater. In first stage, 13 patients (12 eligible) were to be entered. If at least 1 response was observed among the 12 eligible patients, an additional 28 patients (25 eligible) were to be entered. Treatment would be considered promising with at least 4 responders out of the 37 eligible patients.
Descriptive statistics were used to characterize patients at study entry. Toxicities were assessed using the NCI Common Terminology Criteria for Adverse Events (CTCAE) Version 3.0. Exact binomial confidence intervals were used to describe response rate. Progression-free survival (PFS) was defined as the time from study entry to progression or death. Overall survival (OS) was defined as the time from study entry until death from any cause. PFS and OS were estimated using the Kaplan-Meier method.