Isolated and Dynamic Peripheral Blood Residual Disease Status Characterized By Mass Spectrometry Predicts Outcome in Patients with High Risk Smoldering Multiple Myeloma Treated in the GEM-CESAR Trial

Background: GEM-CESAR trial is a potentially curative strategy for high-risk smoldering multiple myeloma (HRsMM) patients (pts) in which the primary endpoint is the rate of minimal residual disease (MRD) negativity in the bone marrow (BM) assessed by next generation flow (NGF) after autologous stem cell transplantation (ASCT) and sustained 3 and 5 years afterwards. Here, we update the preliminary results presented at ASH2019 about the clinical value of quantitative immunoprecipitation mass spectrometry to assess treatment response in this group of patients.

Methods: Pts received induction with six 4-week cycles of KRd (K: 36mg/m² twice weekly, R: lenalidomide 25mg po od days 1-21 and d: dexamethasone 40mg po weekly), high-dose melphalan followed by ASCT, consolidation with two further cycles of KRd and up to 2 years of maintenance with Rd (R: 10mg po od, d: 20mg po weekly). Peripheral blood Residual Disease (PRD) was analyzed post-ASCT and at treatment completion (after 2 years of maintenance) using Quantitative Immunoprecipitation Mass Spectrometry with anti IgG/A/M, total k and total l beads using the EXENT^® Solution (The Binding Site, part of Thermo Fisher Scientific). From the 90 pts enrolled in the trial, 61 were analyzed post-ASCT and 35 at the end of treatment. Isotype-matched immunoparesia (IMI) was also evaluated post-ASCT using the heavy/light chain pairs. Kaplan-Meier curves were plotted either from the specific time point investigated or from the last one if dynamics were analyzed.

Results: First, PRD was analyzed post-ASCT and at treatment completion and deemed to be positive in 54% (33/61) and 20% (7/35) of patients, respectively. At both time points, PRD assessment discriminated two groups of patients with a very significantly different prognosis: median PFS in PRD+ cases post-ASCT was 4.27 years vs not reached in PRD- cases (p=0.0071) (Fig 1A) and after 2 years of maintenance, the median PFS in PRD- cases was not reached vs 1.43 years in PRD+ cases (p=0.0011) (Fig 1B). Importantly, standard response criteria and, more specifically, the achievement of complete response (CR) or better, was not associated with disease outcome. And further, among samples in ³CR, PRD+ was also able to discriminate a group of patients with poorer outcome, with a median PFS of 4.42 years vs not reached in PRD- cases (p=0.0136)

Interestingly, when we combined the analysis of the M-protein by mass spectrometry with the identification of severe isotype-matched suppression (uHLC < 50%), PRD+ patients post-ASCT could be further stratified in two subgroups with significantly different median PFS (not reached in those with no severe IMI vs 3.85 years in those with severe IMI, p=0.0384) (Fig 1C). A similar tendency was observed among PRD+ cases after 2 years of maintenance (mPFS in those with no severe IMI of 2.5 years vs 9 months in those with severe IMI) although the difference did not reach statistical significance.

Then, we analyzed PRD dynamics after ASCT and at the end of treatment. Sustained PRD negativity during this period was observed in 12 pts (33.3%), and sustained positivity in 6 (16.6%). In 5 pts (13.8%) PRD converted from positive to negative, and in 1 (2.7%) from negative to positive. Whereas sustained PRD+ was associated with a very short median PFS of 1.66 years, patients with sustained PRD- or who converted from PRD+ to PRD- displayed a very favorable outcome with median PFS not reached and significantly different compared to the sustained PRD+ group (p<0,0001) (Fig 1D). Interestingly, the only patient that converted from PRD- to PRD+ had a very poor outcome with a mPFS 3.8 months. By contrast, sustained CR or better according to standard response criteria was not significantly associated with clinical outcome.

Conclusion: In this cohort of high-risk smoldering MM patients treated intensively with a curative intention, whereas standard CR or better was not associated with clinical outcome, PRD status defined by mass-spectrometry segregated two groups of patients with different PFS at the 2 time points analyzed. Furthermore, this outcome was better predicted when PRD dynamics were analyzed and/or with the evaluation of severe IMI on PRD+ patients, highlighting once again the relevance of immune recovery as a determinant factor in the evolution of the disease.