Early achievement of measurable residual disease negativity in the treatment of multiple myeloma as predictor of outcome

Multiple myeloma (MM) treatment has evolved remarkably with the introduction of new agents such as proteasome inhibitors, immunomodulatory drugs (IMIDS) and therapeutic antibodies. Most patients now achieve very good partial response or better but unfortunately ultimately relapse. Minimal/measurable residual disease (MRD) assessment thus has become essential to assess the depth of response to chemotherapy.1 Several studies have demonstrated that achievement of MRD negativity is associated with improved survival outcomes and could constitute a surrogate therapeutic end-point for progression-free survival (PFS) and overall survival (OS).2-5 Nevertheless, the optimal timing for MRD assessment remains to be defined. To evaluate the impact of MRD negativity achievement on outcomes, MRD was evaluated by multiparameter flow cytometry (MFC) at different time points during treatment. Between January 2011 and October 2017, 155 consecutive patients [eligible to high-dose therapy supported by autologous stem-cell transplantation (ASCT)] with newly diagnosed MM (NDMM) were included in this study. The local ethics committee approved this procedure and all patients provided informed consent. Baseline demographics and disease characteristics are presented in Table 1. Cytogenetic analysis by fluorescence in situ hybridization was performed at diagnosis for 150 patients (97%). The detection of t(4;14) (n = 16), t(14;16) (n = 6) and del(17p) (n = 10) classified patients as high risk (HR, n = 32). The 118 remaining cases were considered standard risk (SR). Treatment included three to six cycles (median = 4) of a triplet induction regimen (bortezomib, IMIDS and dexamethasone) followed by peripheral stem cells collection (after high-dose cyclophosphamide) and ASCT (tandem in 41 cases) conditioned by high-dose melphalan (140–200 mg/m2). One to two months after ASCT, the patients received a median of two consolidation cycles (range 0–4) using the same induction drugs. In addition, 85 patients (55%) underwent maintenance with lenalidomide (10 mg/day) for 12 to 24 months. MFC analysis was performed according to the recommendations of the European Myeloma Network6 on at least 2 × 106 cells, with the ten-colour antibody combination: cyLambda/CD56+CD28/CD20/CD138/CD27/cyKappa/CD19/CD117/CD38/CD45. Positive MRD was defined as at least 20 clustered neoplastic cells, yielding a sensitivity threshold of 10−5. Bone-marrow aspirates were obtained at different time points: prior to ASCT (n = 155) and after ASCT pre and/or post consolidation regardless. Prior to ASCT, median time for sample collection from start of induction was 91 days [95% confidence interval (CI), 85–94]. PFS and OS were evaluated from the first day of induction chemotherapy, using Kaplan–Meier graphical representation and a log-rank test. For significant values in univariate analysis (p < 0.20), the Cox proportional hazards model was used to identify independent predictive factors. Overall, with a median follow-up of 50 months (range 8–101), the median PFS was 45 months (95% CI, 40–54) and the median OS not reached. Before ASCT, negative MRD was obtained for 48 patients (31%). According to the International Myeloma Working Group (IMWG) Uniform Response Criteria for Multiple Myeloma,7 MRD status was negative in 20/23 (87%), 22/63 (35%), 6/58 (10%) and 0/11 in patients with complete response (CR), very good partial response (VGPR), partial response (PR) and stable or progressive disease respectively. Overall (Figure 1A,B), the achievement of negative MRD before ASCT (48/155 patients) was associated with better PFS (median 68 months vs 34 months; p < 0.0001) and more favourable five-year OS (87% vs 71%; p = 0.03) compared to patients with positive MRD (n = 107). For the 86 good responder patients (CR = 23; VGPR = 63), negative MRD (n = 42) was associated with better PFS (median 68 months vs 32 months; p = 0.002) and superior five-year OS (90% vs 73%; p = 0.04). After ASCT, 28 patients had missing data for MRD. Among the 127 remaining patients, 101 (80%) had reached negative MRD (48 prior to ASCT and 53 after ASCT). The value of early achievement of MRD negativity was confirmed since the 48 patients with negative MRD prior to ASCT had better PFS (68 months vs 45 months, p = 0.0007) and superior five-year OS (87% vs 65%, p = 0.04) compared to those who only obtained negative MRD after ASCT and/or consolidation (n = 53) (Figure 1C,D). The 26 patients with persistent positive MRD had worse PFS compared to the negative MRD post- and pre-ASCT cases (24 months vs 45 months vs 68 months, p < 0.0001). No significant difference was found in OS between negative post-ASCT and persistent positive MRD patients. The 31/118 SR cytogenetic patients (26%) with a negative MRD after induction had a better PFS (88 months vs 39 months; p = 0.0001) and five-year OS (92% vs 76%; p = 0.03) (Figure 1E,F) than the 87 with positive MRD. Identically, the 14 patients with HR cytogenetic and negative pre-ASCT MRD had better PFS than the 18 remaining (54 months vs 28 months; p = 0.0017). HR cytogenetic patients with positive MRD (n = 18) also had both worse PFS (28 months vs 39 months, p = 0.017) and five-year OS (55% vs 76%, p = 0.006) than MRD-positive SR patients (n = 87). Interestingly, in univariate analysis, PFS and OS were not statistically different for patients with single or tandem ASCT (PFS: 45 months vs 43 months; p = 0.12; five-year OS: 78% vs 71%; p = 0.39) or for patients with or without maintenance therapy (PFS: p = 0.22; OS: p = 0.67). Multivariate analysis included cytogenetics risk, International Staging System, age at diagnosis, age ≥65 years, single or tandem ASCT and MRD status prior to ASCT. MRD positivity before ASCT [hazard ratio (HaR) 3.90; 95% CI, 2.19–6.94; p < 0.0001] and adverse cytogenetic profiles (HaR, 2.33; 95% CI, 1.42–3.84; p = 0.0009) were independent variables with a negative impact on PFS. The same variables (MRD positivity: HaR, 3.67; 95% CI, 1.44–9.36; p = 0.006 and HR cytogenetics: HaR, 3.39; 95% CI, 1.64–7; p = 0.001) were also independent negative predictors of OS. MFC, which has become readily available in a short time, is shown here to be a powerful tool to assess MRD and predict survivals. This supports the fact that MFC-based MRD is now being considered as an end-point to assess treatment effectiveness in clinical trials.8 This paper is a retrospective study highlighting the significantly improved outcomes in NDMM with early achievement of MRD negativity prior to ASCT, regardless of treatment after ASCT. Furthermore, early MRD response stands out as a prognostic factor independently of the cytogenetic risk. MRD status after induction, combined with cytogenetic diagnostic characterization, could thus play a pivotal role in risk assessment in first-line treatment of MM patients and lead to the development of risk-stratified therapeutic approaches. Julien Guy, Denis Caillot, Marion Baumelou and Alexandre Payssot performed the research and analysed the data with considerable help from Celine Row, Ingrid Lafon, Jean-Noél Bastie, Steeve Chevreux, Chrétien Marie-Lorraine, Marc Maynadié and Denis Caillot who provided patient data and contributed to the concept and design of the study. Jessica Racine contributed technical support. All authors participated in drafting the paper, revising it critically and approved the final version.