8010 Background: Analysis of minimal residual disease (MRD) in the bone marrow (BM) of patients with multiple myeloma (MM) is accepted by the IMWG to evaluate treatment efficacy and is a well-established prognostic factor. However, there is an unmet need to explore the clinical value of MRD in peripheral blood (PB). Methods: Newly diagnosed MM patients enrolled in the GEM2012MENOS65 trial received six induction (Ind) cycles of bortezomib, lenalidomide, and dexamethasone (VRD) followed by autologous stem cell transplantation (ASCT) and 2 further cycles of consolidation (Cons) with VRD. MRD was analyzed in BM using Next Generation Flow (NGF) and in serum by Mass Spectrometry (MS) using IgG/A/M, κ, λ, free κ and free λ specific beads, both after Ind, at day 100 after ASCT, and after Cons. Sequential samples from the first 184 patients were analyzed. Results: Results of both methods were in agreement (NGF+/MS+ and NGF-/MS-) in 83% of cases post-Ind (152/184), 80% post-ASCT (139/174) and 76% post-Cons (128/169). Stratifying by the log range of MRD by NGF, discordances (NGF+/MS- and NGF-/MS+) seemed to increase at the lower MRD ranges, being 22%, 21% and 19% from ≥10−5 to <10−4 and 21%, 21%, 23% at ≥x10−6(post-Ind, ASCT and Cons, respectively). Analysis of discordances showed that they could be partly explained by the higher percentages of cases found to be positive by MS as compared by NGF at part of the time-points analyzed and at each log range of MRD. From ≥10−5 to <10−4, MRD was detected by NGF in 36%, 28%, 20% of cases post-Ind, ASCT and Cons, respectively vs MS in 37%, 29%, 21% of them; at ≥x10−6, NGF was positive in 11%, 14%, 19% of cases vs MS in 23%, 19% and 16% of them. Considering NGF as a reference, the negative predictive value (NPV) of MS per MRD range (≥10−5 to <10−4 and ≥x10−6, respectively) was: post-Ind: 83% (p<0,0001), 94% (p=0,034); post-ASCT 86% (p<0,0001), 90% (p=0,022); post-Cons 89% (p<0,0001), 85% (p=0,0469). Despite these discordances, the prognostic value of each technique in terms of undetectable MRD and progression-free survival (PFS) was consistent at all time-points (Table) and further, discordant cases (NGF+/MS- and NGF-/MS+) did not display a significantly different PFS as compared to NGF-/MS- cases. Conclusions: The results of MRD assessed by NGF in BM and by MS in PB show a significant concordance and are associated with a similar prognostic value analyzed in terms of PFS. Given its high NPV, MRD in peripheral blood by MS provides a gateway for BM aspiration/biopsy and MRD assessment by NGF.[Table: see text]
Sensitive detection of rare disease-associated cell subsets via representation learning
