Impact of Minimal Residual Disease in High and Standard Risk Multiple Myeloma

Introduction

Achieving complete remission (CR) improves survival outcomes in multiple myeloma (MM). Applying more sensitive flow-cytometry or PCR based minimal residual disease (MRD) testing further improves prognostication and MRD negativity confers better outcomes among CR patients. In our Total Therapy (TT) trials, gene expression profiling based risk (GEP) identifies 15% patients with high risk (HR) MM whose clinical course is characterized by early relapsing disease and prognosis has remained grim despite advances in therapy. Of interest is whether high-risk (HR) patients fail to achieve MRD negativity leading to relapse. We have determined response rates across the total therapy (TT) trials and in this study we report the impact of molecular negativity by MRD assessment in patients who have achieved at least very good partial response (VGPR) or CR at 4-8 months and 12-24 months, respectively, after enrollment in TT3b-TT5 from 2005-2009. We use next-generation sequencing (NGS)-based MRD assessment (Adaptive Biotechnologies) which is sensitive to 1 MM cell in 10⁶ normal cells. Furthermore, we evaluate MRD status in long term relapse-free survivors (>6 years) to determine the importance of MRD status in prolonged remission.

Materials and methods

Study subjects include 591 patients enrolled in TT3b-TT5 to determine VGPR/nCR/CR by EBMT/IMWG criteria and to analyze PFS and OS of HR and SR patients defined by our GEP model. For MRD testing, we identified 102 patients from the TT3b-TT6 protocols that achieved at least VGPR and had bone marrow (BM) sample available at 4-8 months and 12-24 months after enrollment. Of these 102 patients, 14 patients had HR disease and 88 had SR. Six of 14 HR patients already had clinical relapse between 12-24 months, but were included to determine MRD at 4-8 months. Of all 102 patients, 51 are currently still in CR 6-9 years after protocol enrollment and most recent flow-cytometric MRD status was evaluated.

NGS-based MRD Assessment: Genomic DNA was amplified using locus-specific primer sets for immunoglobulin heavy-chain complete (IGH-VDJ_H) and incomplete (IGH-VD_H) as well as for immunoglobulin κ locus (IGκ). The amplified products were subjected to sequencing and clonal gene rearrangements were analyzed. Final MRD measurement was calculated at a sensitivity level of 1 cancer cell per 1 million cell equivalents.

Flow Cytometry based MRD assessment: Erythrocyte-lysed BM samples were immunophenotyped by use of 8-color (CD138/CD38/CD19/CD45/CD27/CD81/CD56/CD20) staining technique. Myelomatous Plasmacells (PCs) were separated from healthy PCs by different phenotype expression and MRD was deemed negative when no myelomatous PCs were detectable at a sensitivity of <10^-4 to <10^-5.

Results

Response by EBMT/IMWG criteria across our TT3b-TT5 (n=591) protocols showed no significant difference between HR and SR patients by year 1 of enrollment. In the SR group (n=502) 74.9%-77.9% of patients achieve at least nCR/VGPR and 40.3%-50.4% CR. For the HR patients (n=89) the results are similar with 75%-77.9% achieving at least VGPR and 50%-50.4% CR. However, the majority of HR patients start relapsing after one year of enrollment with a relapse rate of 59% for HR at 3 years compared to only 19% for SR at the same time point, indicating that MRD positive disease likely exists at 12-24 months for those patients with relapse. NGS-based MRD assessment will answer this question and data will be available by December 2015.

Fifty one patients of the initial 102 patient group continue to be in complete remission to date >6 years after enrollment. Of these, 47 are MRD negative at a MRD level of <10^-5 (3 HR and 44 SR), two are MRD negative at <10^-4 (SR), leaving only 2 patients (SR) with a detectable MRD burden at >10^-4.

Conclusions

Using conventional response methods, there is no difference between HR and SR MM at 1 year after initiation of treatment. However clinical outcome for HR disease is very poor and characterized by early relapsing disease, indicating that MRD persists even when patients have achieved CR. Sequencing is the most sensitive method for MRD detection and will be used in the present study to evaluate the importance of sequential MRD testing in HR and SR disease. We show that the majority of patients achieving long term remission at 6-8 years are MRD negative indicating the importance of molecular response for long term success.