Minimally invasive assessment of peripheral residual disease (PRD) during maintenance or observation in patients (Pts) with multiple myeloma (MM): An update from the GEM2017FIT clinical trial

Abstract

Abstract Background. The expectation of a treatment-free interval in pts achieving deep and durable MRD responses, as well as the possibility of using MRD progression to resume treatment or initiate a new line of therapy, implies the need of dynamic MRD monitoring. Bone marrow (BM) aspirates may be considered every few years for MRD assessment, but in between there is an unmet need of evaluating PRD to have a more frequent readout of tumor kinetics while preserving pts' quality of life. Aim. Investigate the independent prognostic value of minimally-invasive PRD assessment in MM pts undergoing maintenance or observation. Methods. This study included 247 transplant-ineligible pts who were enrolled in the GEM2017FIT trial. Pts were randomized to 18 induction cycles with VMP-Rd (n=77), KRd (n=80) or Dara-KRd (n=90), after which there was a second randomization to maintenance with Dara-Len (n=141) vs observation (n=106). PRD was evaluated after the second randomization in large peripheral blood (PB) volumes using BloodFlow, a method that combines CD138+ enrichment with flow cytometry for ultra-sensitive monitoring of circulating tumor cells (CTCs) at 10-7. In 150/247 pts, PRD was also evaluated in serum using EXENT mass spectrometry (MS). Periodic assessments of PRD in PB/serum and of MRD in BM using next-generation flow (NGF) were respectively performed every 6 and 12 months. Progression-free survival (PFS) was landmarked at the latest PRD assessment. Results. Among the 247 pts in whom PRD was assessed by BloodFlow, 32 (13%) had detectable CTCs. The median number of CTCs/µL was 0.007 (range, 0.0002 – 4.7). Detection of PRD with BloodFlow resulted in a 15-fold increased risk of progression and/or death (HR: 15.4, 95% CI 7.3 - 32.5, p<.001). Median PFS from the latest PRD assessment was not reached vs 2 months in pts with absence vs presence of CTCs. The association between positive PRD and dismal prognosis was observed in pts undergoing maintenance (HR: 18.4, 95% CI 7 – 42.3, p<.001) and observation (HR: 13.7, 95% CI 2.9 – 63.4, p=.001). A multivariate analysis of PFS including induction therapy, the Revised International staging system, MRD and PRD assessments, showed that only positive MRD by NGF (HR: 11.8, 95% CI 2.2 – 61.9, p=.004) and positive PRD by BloodFlow (HR: 7.1, 95% CI 2.2 – 23.3, p=.001) had independent prognostic value. BloodFlow was performed in 1,175 longitudinal samples, of which 1066 (91%) were PRD negative and 109 (9%) PRD positive. Of the 247 pts, 214 (86.6%) were systematically PRD negative, 21 (8.5%) converted from negative to positive, 1 (0.4%) converted from positive to negative and 11 (4.5%) were systematically PRD positive. Median PFS was not reached in pts starting or becoming PRD negative and was of 3 months in those starting or becoming PRD positive (p<.001). Of note, only 4% of pts with 24-month sustained PRD negativity progressed thus far. Of the 1,175 PB samples with PRD assessment by BloodFlow, 719 had simultaneous MRD testing in BM. Considering pts' MRD status in BM as the reference, PRD assessment using BloodFlow showed positive and negative predictive values of 83% and 93%, respectively. The concordance between PRD and MRD was 84%, with 536/719 (75%) double-negative and 67/719 (9%) double-positive results. Among discordant assessments, 111/719 (15%) were PRD-/MRD+ and 5/719 (1%) were PRD+/MRD-. Median PFS from the latest assessment was not reached in PRD-/MRD- pts. By contrast, those who were PRD+/MRD+, PRD-/MRD+ and PRD+/MRD- showed inferior PFS (median of 10, 6 and 2 months, respectively, p<.001). Of the 725 PB and serum samples with paired PRD and MS assessments, 499 (69%) were double-negative, 145 (20%) were BloodFlow-/MS+, 18 (2%) were BloodFlow+/MS-, and 63 (9%) were double-positive. Median PFS from the latest assessment was not reached in double-negative and BloodFlow-/MS+ pts, and was of 3 months in BloodFlow+/MS- and double-positive pts (p<.001). Of note, among pts with double-negative PRD by BloodFlow and MS, only 5% were MRD positive in BM. Conclusions. This study confirms the independent prognostic value of dynamic PRD monitoring with BloodFlow and its complementarity with MRD assessment in BM to identify MM pts that, either during maintenance or observation, are at imminent risk of progression. The combined assessment with BloodFlow and MS enables the identification of patients with double-negative PRD in whom the risk of being MRD positive in BM and of disease progression is ≤5%.