Circulating Tumor DNA in the Peripheral Blood As Early Predictor of Clinical Outcome in Relapsed/ Refractory Multiple Myeloma

Introduction: Treatment of multiple myeloma (MM) has improved over the last decade. Although long-term survival is noted in some patients, emergence of resistant disease still prevents cures. The objective of this study was to define "liquid biopsy" parameters that identify patients who do not benefit from a particular treatment before relapse becomes evident by serological markers. Having such parameters at hand will potentially inform changes in treatment.

Methods: Here, we apply low-pass whole genome sequencing to profile a uniform cohort of 45 relapsed and refractory MM (RRMM) patients who have been treated in a multicenter phase II trial evaluating the combination of elotuzumab, pomalidomide, bortezomib and dexamethasone (elo-PVd; NCT02718833). Peripheral blood plasma samples were acquired for circulating tumor DNA (ctDNA) evaluation at four different timepoints (screening, cycle 3 day 1 (C3D1), cycle 5 day 1 and end of treatment). The concentration, relative fraction and copy number profile of myeloma-derived ctDNA were determined across all timepoints.

Results: At the time of this preliminary analysis, 17 patients (35%) continue on treatment whereas 28 patients (58%) have developed progressive disease (PD). Our data suggest that ctDNA levels at screening and C3D1 strongly correlate with progression-free survival (PFS). Patients with available follow-up samples (n=40) were stratified according to ctDNA levels at C3D1 of treatment. Patients with a residual ctDNA level <10% showed a significantly longer PFS (median 17.6 months (95% CI: 1.4-6.5)) as compared to those with ctDNA levels ≥10% (median 5.9 months (95% CI: 0.2-0.7), log-rank Mantel-Cox test P=0.0006). The kinetics of ctDNA were largely concordant with the course of M protein and serum-free light chains (SFLC). To test our hypothesis that ctDNA assessment could be particularly useful for patients with inconclusive serological markers (minimal response (MR)/ stable disease (SD) by IMWG criteria), we performed a subgroup analysis of all patients with MR/SD at first follow-up (C3D1). In this group of 19 patients a residual ctDNA fraction ≥10% translated into a significantly shorter median PFS (1.6 months, 95% CI: 0.1-0.8) as compared to ctDNA levels <10% (5.8 months, 95% CI: 1.2-11.4, P=0.02).

Conclusions: These data indicate that "liquid biopsy" evaluation of ctDNA may refine prognostication and provide added predictive value over serological markers alone. While in the large majority of cases ctDNA has excellent concordance with M protein and SFLC for monitoring of MM disease progression, ctDNA may identify patients where relapse is imminent before it can be detected by serological parameters. This approach may therefore complement our framework for treatment decisions. Notably, this approach is highly scalable, cost-efficient and provides information about the clonal evolution of MM without the need for a bone marrow biopsy.