Characterisation of Long-Term Responders to First-Line Myeloma Therapy - Results from the UK Myeloma IX and XI Trials

INTRODUCTION

Features of high risk myeloma (MM) have been studied in detail but patients with longer term responses to first-line therapy are less well characterised. Identification of common features of this group may support optimised management. Here we analysed clinical and genetic characteristics of long-term responders of 4,249 trial patients from the UK MRC Myeloma IX (M-IX) and NCRI Myeloma XI (M-XI) trials.

PATIENTS AND METHODS

In M-IX patients were randomised between alkylating therapy (CVAD or MP) and thalidomide-based induction therapy (CTD). M-XI patients were randomised between thalidomide and lenalidomide based induction (CTD vs CRD) and a response-based bortezomib (CVD) intensification. Fitter patients received HD-Mel+ASCT consolidation. Patients were then randomised to thalidomide (M-IX) or lenalidomide (M-XI) maintenance or observation. Trials included symptomatic, newly diagnosed patients based on CRAB criteria. This analysis included 1,921 My-IX and 2,328 My-XI patients with median follow-up of 73 and 61 months (m), respectively. Genetic profiling was available for 1,866 patients.

Patients with a long-term response post induction (PFS≥48m) were identified and their baseline characteristics, responses and treatment compared to those with PFS<48m. OS difference was compared using the logrank test. Multivariate analysis was performed using logistic regression.

RESULTS

In M-IX, 283 (25.8%) of transplant-eligible (TE) patients had PFS ≥48m whereas 58 (7%) of transplant non-eligible (TNE) patients reached PFS≥48m. In M-XI 410 (34.2%) patients had PFS≥48m for TE and 116 (10.2%) for TNE. Extended progression free survival translated to overall survival (OS) benefit with a median post progression OS of 36.9m for PFS≥48m vs 16.7m for PFS<48m (p<0.0001) for M-IX. For M-XI, OS data had not reached maturity, however the probability of OS at 2 years post progression for those with PFS≥48m was 60% vs 36% for PFS<48m.

Clinical factors including ISS I (P<0.0001) and lower performance status (WHO) (P<0.0001) were positively associated with PFS≥48m. Relative risk by multivariate analyses appeared to be higher for these factors in TNE patients with odds ratio of 1.6 and 1.3 than in the TE group with odds 1.4 and 1.2 across M-IX and XI, respectively.

The proportion of patients with a high risk lesion (Adverse translocation, Gain(1q) or Del(17p)) were lower in the PFS≥48m group than <48m: 34.3% vs. 54.5% and 28.8% vs. 54% for TE and 10% vs. 51.2% and 35.4% vs 52.1% for TNE arms of M-IX and M-XI, respectively. 'Double hit' MM (≥2 high risk lesions) was rare with 5.8% of patients PFS≥48m compared to 16.6% of patients PFS<48m across trials (P<0.0001). Absence of gain(1q) was the only genetic factor retained within a multivariable analysis of baseline parameters associated with PFS≥48m in the TNE group, whereas for the TE group absence of all high risk lesions were associated with PFS≥48m (p<0.0001). Hyperdiploidy was positively associated with PFS≥48m in the TE group (P=0.02) only by univariate analysis.

The majority of patients with PFS ≥48m showed ≥VGPR after induction +/- consolidation: 211 (76.4%) and 340 (84%) of PFS ≥48m patients in the TE arms and 26 (49.1%) and 87 (76.3%) in the TNE arms of M-IX and M-XI, respectively. 86.7% of patients who achieved a ≥VGPR had a PFS ≥48m in the absence of high risk lesions compared to 72.8% with any high risk lesion present (P=0.004). Some patients with PFS≥48m had only reached PR after induction; 56 (20.3%) and 57 (14.1%) of PFS ≥48m patients in the TE arm and 15 (28.3%) and 24 (21.1%) in the TNE arms of M-IX and M-XI, respectively. Baseline factors that were associated with still being able to achieve PFS≥48m from induction after only achieving a PR included the lack of high risk genetic lesions (P<0.0001) and low ISS (P=0.0002). In M-XI, the proportion of patients who only achieved a PR after induction and reached PFS≥48m was 10.6% for patients randomised to observation and 89.4% for patients with lenalidomide maintenance suggesting maintenance may be of particular benefit in this group.

CONCLUSIONS

Response assessment after induction+/-HD-Mel consolidation with baseline factors can define a patient group with superior outcomes in both TE and TNE patients and may influence future treatment strategies of MM patients undergoing first line therapy. Further analyses including modelling of predictors of response duration are ongoing and will be presented at the conference.