P-038: Postinduction minimal residual disease (MRD) within stem cell graft (gMRD) correlates with marrow MRD (mMRD) and progression free survival (PFS) following autologous stem cell transplantation

Abstract

Listen

Translate article

<h3>Background</h3> Earlier studies have reported the impact of clonal plasma cell contamination within stem cell grafts as a negative prognostic factor. In this study we aimed to quantify circulating plasma cells (cPCs) by flow in apheresis products (gMRD) and compare with marrow (mMRD) and outcome. <h3>Methods</h3> All consecutive (September 2006 - July 2020) newly diagnosed transplant eligible multiple myeloma (MM) patients were included prospectively. In the sample drawn for HPSC quantification of the graft and bone marrow (only if response is ≥VGPR), the number of cPCs were quantified by Flow. As described earlier (Montero et al.) undetectable MRD is defined as absence of cPCs at a sensitivity of 10-4 (n=54) between 2006-2016 and 10-5 (n=37) after 2017. MRD assessment is similar in the graft and marrow. Statistical analysis was performed on merged data regardless of sensitivity of MRD assessment. Results were reported in the intention-to-treat (ITT) population for mMRD. <h3>Results</h3> Patients were given either Bortezomib based triplet without immunomodulatory drug (IMID) ((85.7%)) or with an IMID (17.6%) as induction. The median age was 62 years (range:37-75 years). 86 (94.5%) patients had cytogenetics by FISH at diagnosis, of which 14% had high risk. Extramedullary disease was detectable among 14.3%. Post-induction complete response (CR) was achieved in 28.6% (n=26). Following mobilization, gMRD was detectable in 57.1% of patients. CR rate among gMRD (+) vs. (-) patients were: 14/26 (26.9%) vs 12/39 (30.8%). Kappa coefficient (SE): -0.036, p:0.8) pointing to lack of correlation between gMRD and biochemical response. Undetectable mMRD was reached among 26.4%. Among the patients with CR mMRD detectable vs. undetectable: 15/67 (22.4%) vs 11/24 (45.8%) (SE: -0.151, p:0.03) points to a correlation between mMRD levels and biochemical response. Among 39/91 gMRD(-) patients, undetectable mMRD(-) was also observed in 46.2%. A statistically significant correlation between gMRD and mMRD (SE: 0.364, p<0.001) was calculated. Patients with undetectable gMRD displayed a better PFS compared to patients with detectable gMRD (NR vs 37.5 months). There were insignificant differences in median PFS according to the undetectable mMRD (n=24) at level 10-5 vs 10 -4 (NR vs 37.5 months; p=0.9). In addition, achievement of post-induction undetectable mMRD plus gMRD was also associated with improved PFS (HR: 0.04 p=0.01) with no relapse. Having detectable MRD in either graft or marrow estimates a 2 years-PFS of 61.1 % (p=0.07). <h3>Conclusions</h3> Our real-world triplet drug induction-based experience shows for the first time postinduction marrow and graft MRD to be correlated with each other and with PFS. A non-invasive cellular residual disease measurement within graft at either 10-4 or 10-5 level may be an additional end point for therapeutic efficacy.