Gareth J Morgan, MD, PhD1, Faith E. Davies, MD2, Walter M. Gregory, PhD3*, Sue E. Bell, PhD3*, Alex J. Szubert, MSc3*, Nuria Navarro-Coy, MPhil3*, Mark T. Drayson, MD, PhD4*, Roger G. Owen, MD5, Gordon Cook, MD5*, Fiona M. Ross, PhD6, Graham H. Jackson, MD7* and J. Anthony Child, MD3*
1Institute of Cancer Research, The Royal Marsden Hospital, London, United Kingdom
2Institute of Cancer Research, Royal Marsden Hospital, London, United Kingdom
3Clinical Trials Research Unit, University of Leeds, Leeds, United Kingdom
4Myeloma Clinical Trials Unit, University of Birmingham, Birmingham, United Kingdom
5St. James' University Hospital, Leeds, United Kingdom
6Wessex Regional Genetics Laboratory, University of Southampton, Salisbury, United Kingdom
7Department of Haematology, University of Newcastle, Newcastle-upon-Tyne, United Kingdom
Background: Although a meta-analysis has suggested that a consistent PFS benefit is seen with maintenance thalidomide therapy in multiple myeloma (MM) patients, the impact on OS remains unclear (Hicks LK, et al. Cancer Treat Rev 2008;34:442-452). This could be due to a lack of effect, differing biological subgroups in relatively small studies with short follow-up, or lack of effective relapse treatment.
Aims: The study was set up to evaluate the effect of thalidomide maintenance therapy in newly diagnosed MM patients aged ≥ 18 years with OS and PFS as end points, and to examine differential effects in fluorescence in situ hybridization (FISH) identified molecular subgroups.
Methods: Following induction treatment in an intensive pathway for younger fitter patients and a non-intensive pathway for the remaining patients, eligible patients were randomized to open-label thalidomide maintenance until progression (50 mg/day escalating to 100 mg/day after 4 weeks assuming good tolerance) or no maintenance. Patients of all ages were included in the randomization. FISH was performed using standard approaches. Adverse FISH groups were defined as any of t(4;14), t(14;16), t(14;20), 1q+, 17p–, or 1p32– (in the intensive pathway only); the remainder were defined as favorable.
Results: Overall, 820 patients were eligible of which 818 were evaluable. Median patient age was 65 years (range, 31-89). Median follow-up from maintenance was 38 months (range 12-66 months). Median time on maintenance was 7 months (range 0-50 months). Maintenance thalidomide significantly improved PFS, with a difference between the curves of 13% (95% confidence interval [CI] 6%-20%) established by 24 months and maintained till the current maximum follow-up at 66 months (hazard ration [HR] 1.36; 95% CI 1.15-1.61; logrank P< 0.001). However, in the initial analysis there was no apparent impact on OS (P = 0.40). These findings were consistent regardless of prior intensive or non-intensive induction treatment. At 66 months follow-up, a PFS benefit was seen in the favorable FISH group (P = 0.004) with no effect on OS (P= 0.6). In the adverse FISH group there was no effect on PFS (P = 0.48) and a negative effect on OS (P = 0.009). Subsequently, we evaluated the effect of relapse treatment on outcomes and used this data in a mathematical model to determine if OS benefit would have accrued from the prolonged PFS if all patients had received effective treatment at relapse. A total of 523 patients have progressed to date and of these, 47% received thalidomide as initial relapse treatment (either as a single agent or in combination), 30% received novel agents (either bortezomib or lenalidomide), and 27% received alkylating agents or steroids. Median survival after progression was significantly worse in patients who received thalidomide maintenance than those who did not (P < 0.005); this could, at least partly, be attributed to patients who received thalidomide at progression. In those patients, the median OS after progression was significantly greater in the no maintenance group versus the thalidomide maintenance group (P = 0.004). Among patients treated with novel agents at progression, prior thalidomide maintenance therapy had no impact on OS and these patients were effectively salvaged. The mathematical model employed to examine the effect of effective salvage therapy at progression suggested that a significant survival benefit of 5.5% at 3-years in favor of thalidomide maintenance would have accrued if all patients had received effective therapy at progression (HR 0.77; 95% CI 0.60-0.99; P = 0.04). Next, we examined whether continuous thalidomide therapy was associated with a consolidation or maintenance effect. Upgrading of response with thalidomide maintenance was not common. Importantly, there was no difference in response over time between maintenance therapy and no maintenance. In contrast to what has been reported previously, these findings suggest a maintenance effect.
Conclusions: Maintenance thalidomide significantly improves PFS and if effective relapse treatment is available this translates into an OS advantage. The median duration of maintenance thalidomide therapy of 7 months in the present study was short. The clinical impact of maintenance would be improved if patients could remain on therapy for longer, suggesting that the use of other agents such as lenalidomide, with better tolerability profiles, may produce better results.