To the Editor,

The assessment of bone marrow measurable residual disease (MRD) has consistently shown a significant prognostic value in patients with multiple myeloma (MM), with a benefit in survival outcomes associated with MRD negativity surpassing the value of complete response [1, 2]. Next-generation sequencing and Euroflow on bone marrow reach higher sensitivity than standard flow, increasing the predictive potential [3, 4]. Thus, MRD was included in the consensus criteria for response [5] and its role as a surrogate marker for survival outcomes is under consideration [1, 6]. Preliminary studies suggest that MRD dynamics could demonstrate greater prognostic value than just the MRD status at a single time point [7, 8].

MRD assessments are performed in MM to assess the quality of response and to make prognostic statements, but one can imagine using such results to make clinical decisions, much as one does with M-spikes. Unlike other hematological malignancies, therapeutic decisions (treatment escalation, de-escalation or discontinuation) based on MRD is a pending topic in MM.

Korde et al. [9] published a trial where MRD testing impacted the number of cycles of therapy as it is also planned in the MASTER trial [10]. Following IFM2009 trial, some have postulated the usefulness of post-induction MRD status to decide between early/delayed autologous stem cell transplantation [11]. Also, ongoing trials as REMNANT or PREDATOR-MRD are evaluating the role of MRD conversion (from negative-to-positive) as a trigger for pre-emptive therapy [12]. However, these results are still preliminary and conclusive data are scarce.

We analyzed how MRD results could guide clinical decision-making through the retrospective analysis of outcomes in 400 MM patients with extensive MRD monitoring during frontline therapy (patients at least in VGPR and ≥ 1 MRD assessments during follow-up according to our clinical practice). NGS of Ig genes or second-generation flow or next generation flow at level of 10−5 was employed for MRD assessment and 92% of patients were in CR. In 67 patients, a clinical decision was made based on MRD results, mostly during maintenance (83%). Thirty-three out of these 67 were MRD-negative cases (treatment was reduced in 3 and stopped in 30), while 34 were MRD-positive when a therapy change was made (intensification in 27 and new treatment in 7 cases). None of them met criteria for progressive disease according to IMWG consensus. Twelve out of 34 MRD-positive patients subsequently achieved MRD negativity after intensification or a change in therapy.

Globally, 186 patients achieved MRD negativity showing a prolonged progression-free survival (PFS) versus those who did not achieve MRD negativity (mPFS 104 vs. 45 months, p < 0.0001). No differences were observed when MRD was assessed by NGS or MFC (p = 0.2).

Patients in whom a clinical decision was made based on MRD (n = 67) had a prolonged PFS versus those in whom a clinical decision was not made (n = 333) (mPFS from the first MRD datapoint was 104 months [73–165] vs. 62 months [46–80], p = 0.005); statistical significance persisted in a landmark analysis at 12 months (p = 0.04) or from the start of induction (p = 0.05) (Fig. 1a). No differences in major clinical features were found between both subgroups (Table 1a–d). In the MRD-negative group, those in whom treatment was stopped did just as well as those whose therapy was continued (mPFS, 120 vs. 82 months, p = 0.1). Patients with an MRD-positive marrow, in whom therapy was changed or intensified, exhibited prolonged PFS versus those who continued therapy without change (mPFS, Not reached vs. 39 months, p = 0.02) (Fig. 1b, c). Only making clinical decisions based on MRD (HR 0.5; 95%CI 1.41–6.87) and age (HR 1.2; 95%CI 1.1–1.5) were significant in a multivariate analysis (including age, sex, myeloma isotype, cytogenetic risk, hemoglobin, response, creatinine and clinical decision-making).

Fig. 1
figure 1

Kaplan–Meier curves showing the impact of making clinical decisions based on MRD. a PFS from the first MRD datapoint, comparing patients who underwent a change in therapy based on MRD with those in whom no change in therapy was made. b MRD-negative patients: treatment discontinuation (maintenance or transplant) vs. no change in therapy. c MRD-positive patients: beginning a new therapy or intensifying therapy vs. no change in therapy

Full size image
Table 1 Main characteristics at diagnosis and therapy at frontline, all patients (n = 400)
Full size table

Depth of MRD is commonly considered the best prognostic factor in MM [2, 4, 7], and a good surrogate marker for survival in clinical trials [1]. However, some myeloma experts question the employment of MRD to guide MM treatment due to the lack of evidence. Our results suggest that the use of MRD to make clinical decisions has a positive impact on survival outcomes.

The achievement of MRD negativity had a relevant impact on PFS. Interestingly, PFS improved when treatment was modified in patients who were MRD-positive; while PFS was not different according to the discontinuation or persistence of therapy when MRD negativity was achieved. The main limitations of this study are its retrospective nature, the heterogeneity of the time of MRD assessment, and the lack of specific pre-defined rules regarding when and how to make these decisions and the small sample size of the MRD making decisions population, for these reasons, results should be interpreted carefully. Nevertheless, this study has several strengths including the large sample size and the multinational and multi-institutional approach with superimposable results between methodologies and institutions.

In conclusion, the use of MRD to guide treatment in MM is potentially as useful as Serum Protein Electrophoresis and light chain measurement, especially in patients who are in stringent Complete Response. Prospective randomized clinical trials currently ongoing may provide new evidence in this setting.