Bortezomib and Low-Dose Dexamethasone with or without Continuous Low-Dose Oral Cyclophosphamide for Primary Refractory or Relapsed Multiple Myeloma: Final Results of a National Multicenter Randomized Controlled Phase III Study

Background

Multiple myeloma (MM) typically follows a relapsing course and many patients require multiple lines of therapy. The combination of bortezomib and dexamethasone (VD), alone or with cyclophosphamide (VCD) has shown good efficacy and tolerability in patients with relapsed or refractory (rel/ref) MM in non-randomized trials (Mikhael Br J Haematol 2009, Kropff Br J Haematol 2007). This phase III, randomized, controlled study (NCT00813150) aimed to evaluate VCD vs VD in patients with rel/ref MM.

Methods

Adult patients with rel/ref MM (1st–3rd relapse) with Karnofsky performance status ≥60% were eligible for inclusion; patients with grade ≥2 peripheral neuropathy (PN) were excluded. Patients were randomized 1:1 to receive VD (up to 8 x 3-week cycles of intravenous bortezomib 1.3 mg/m² on days 1, 4, 8, and 11 plus oral dexamethasone 20 mg on the day of bortezomib administration and the following day) or VCD (as VD with the addition of oral cyclophosphamide 50 mg daily). The study was designed with 80% power to detect a difference of 2.8 months in the primary endpoint of time to disease progression (TTP) at the 5% significance level (two-sided); a sample size of 270 patients was required. Secondary endpoints included overall survival (OS), response rate (RR), and safety. Disease progression was assessed per IMWG guidelines and adverse events (AEs) were assessed using NCI-CTCAE v3.0. TTP and OS were calculated using Kaplan-Meier methodology and compared using two-sided log-rank test.

Results

In total, 96 patients were randomized; 93 began study medication and were included in the safety population, and 90 received at least one dose of study drugs and one efficacy assessment and were included in the intention to treat (ITT) population. The planned sample size was not reached due to lack of recruitment and the study was stopped prematurely as a result of a joint decision of sponsor and coordinating investigator. The VD and VCD arms were well balanced in terms of baseline characteristics. Median age was 71 years (range 30–85) with most patients (97%) suffering from relapse of myeloma disease (3% refractory disease). One, 2, and 3 prior lines of therapy were reported by 57%, 30%, and 9% of patients, respectively; 14% had received prior bortezomib. In total, 19%, 24%, and 29% of patients were ISS disease stage I, II, and III, respectively. Patients in both VD and VCD arms received a median of 5 cycles of therapy. Reasons for premature termination of treatment included AEs (35% and 32% of patients in the VD and VCD arms, respectively) and progressive disease (2% in the VD arm only). After a median follow-up of 24 months, median TTP was 12.6 months in the VD arm vs 9.9 months in the VCD arm (p=0.192); the corresponding hazard ratio (HR) was in favor of the VD arm (HR 0.71, 95% CI 0.43–1.19), although this was not statistically significant (p=0.196). Median OS could not be determined in the VD arm vs 42 months in the VCD arm. The resulting Kaplan-Meier plots were superimposable with a HR of 0.85 (95% CI, 0.41–1.73; p=0.645). In total, 32 (74%) and 33 (70%) patients in the VD and VCD arms achieved ≥PR; 23 (54%) and 22 (47%) patients achieved ≥VGPR. The most common treatment-emergent non-hematologic AEs included PN (28% vs 47% in the VD vs VCD arms), fatigue (28% vs 40%), diarrhea (24% vs 36%), and constipation (22% vs 28%). Grade ≥3 PN occurred in 7 (15%) patients in the VCD arm vs 2 (4%) in the VD arm. Higher rates of grade ≥3 leukopenia (13% vs 0%) and grade ≥3 thrombocytopenia (34% vs 24%) were observed in the VCD arm; however, rates of clinically relevant infections were comparable (grade ≥3 infections 9 [19%] in the VCD arm vs 8 [17%] in the VD arm) and no clinically relevant bleeding events occurred.

Conclusions

These data indicate that both treatment regimens were similar in efficacy and safety. However, due to limited enrollment, the study is underpowered to draw firm conclusions. Further studies will be required to address the potentially enhanced neurotoxicity of VCD compared with VD. The dosing of cyclophosphamide also warrants further investigation, as other studies suggest that a dose intensified regimen may result in higher CR rates.