Lenalidomide, Adriamycin and Dexamethasone (RAD) Versus Bortezomib, Lenalidomide and Dexamethasone (VRD) in Newly Diagnosed Multiple Myeloma (MM) - Post-Induction Response and MRD Results By Flow Cytometry and NGS from a Phase 3 Randomized Controlled Clinical Trial (RCT)

IntroductionHigh-dose chemotherapy and stem cell transplant (SCT) remains a standard of care in medically fit patients (pts) with newly diagnosed (ND) MM. Induction triplets with at least one of the newer compounds are recommended. Bortezomib (V), lenalidomide (R) and dexamethasone (D; VRD) ranks among the most effective regimens and VRD/SCT was superior to VRD alone in an RCT. In a phase 2 study, we demonstrated RAD induction (lenalidomide 25 mg d1-21; Adriamycin 9 mg/m² iv d1-4; dexamethasone 40 mg d 1-4 and 17-20 every 4 weeks) followed by SCT to be safe and effective (Knop et al., Leukemia 2017). Therefore, we decided to compare RAD versus (vs) VRD (lenalidomide 25 mg, d1-14; subcutaneous bortezomib 1.3 mg/m² d 1, 4, 8, 11; dexamethasone 20 mg d 1+2, 4+5, 8+9, 11+12 every 3 weeks) induction (3 cycles each) in an RCT. MethodsThe current study was set up according to a double 2x2-factorial design to enrol transplant-eligible pts up to 65 years. The post-induction (PI) complete response (CR) rate as per IMWG criteria was the efficacy co-primary endpoint. We hypothesized the CR rate following RAD to be non-inferior to VRD which was estimated to be 20%. The study was powered to confirm non-inferiority of RAD at a margin of 10% with a one-sided alpha level of .05. Cytogenetic characterization was performed by fluorescence in situ hybridization (FISH) from CD138-enriched plasma cells. Minimal residual disease (MRD) was assessed by second-generation eight-color flow cytometry (FC; EuroFlow protocol). Bone marrow (BM) samples from baseline and defined restaging time points were analyzed for an acquisition of ⩾10⁷cells/sample. In a subgroup of 103 pts, we evaluated the applicability of comprehensive immunoglobulin (Ig) amplicon next generation sequencing (NGS) to detect molecular MRD markers and to compare the results with FC. NGS-based marker screening was performed in baseline BM. Sequencing libraries were prepared using 2-step PCR employing multiplex primer sets for IGH V-D-J (FR1, FR2 and FR3), IGH D-J and IGK loci (V-J and KDe). For MRD detection, we used 1-step library preparation with the same primer sets. Results476 pts with a median age of 55 (range, 32-65) years were randomized between 05/2012 and 06/2016 and 469 received at least one dose of study drug. High-risk (HR) FISH abnormalities comprised del17p (11.3% of pts); t(4;14) (11.7%); and t(14;16) (4.5%). 232 pts were randomized to receive RAD, and 237 to VRD, respectively. 90.5% of RAD vs 93.7% of VRD pts completed all 3 cycles. PI CR rate was 13.5% (95% CI, 9.4%-18.7%) with RAD vs 13.4% (95% CI, 9.3-18.5) with VRD, (P=.971). Rates of ≥VGPR were 40.6% (50% CI, 34.2%-47.3%) with RAD vs 48.9% (95% CI, 42.3-55.6%) with VRD (P=.076). In pts with HR cytogenetics, rates of ≥VGPR were 43.3% (RAD) vs. 59.3% (VRD; P=.096). From 317 pts with paired samples, 33/151 (21.9%) of RAD vs 45/166 (27.1%) of VRD pts were FC MRD negative (P=.169) following induction at a median sensitivity level of 6.73x10^-6. 197/239 positive pts (82.4%%) had MRD levels above 0.01%, and 42 (17.6%), between 0.0001 and 0.01%. Flow MRD negativity as per IMWG MRD criteria (Kumar et al, Lancet Oncol 2016) was seen in 8/151 (5.2%) pts with RAD vs 6/166 (3.6%) with VRD (P=.27). The remainder of pts did not (yet) fulfil IMWG CR for various reasons. NGS marker screening identified at least 1 Ig marker in 98/103 evaluable patients. To date, 47/98 pts were analyzed for NGS MRD following induction. Four out of 47 (8.5%) subjects were sequencing negative (3/4 post-VRD) with all of them also being IMWG FC MRD negative. One VRD patient died during induction for a mortality rate of 0.2 %. 62.1% of RAD vs 55.3% of VRD pts experienced at least one serious adverse event (SAE; p=.16). SAEs with relationship to study drugs of at least °3 severity occurred in 26.3% (RAD) vs 23.6% (VRD) pts (p=.523). ConclusionsTo the best of our knowledge, this is the first RCT to compare two R-based triplets in SCT-eligible pts. The co-primary efficacy endpoint was met with identical PI CR rates of around 13% for RAD and VRD, respectively. However, a trend emerges to favor VRD over RAD in terms of at least VGPR including HR FISH subjects. Analysis of MRD by multicolor FC showed 5% of pts to be already IMWG flow MRD-negative. Results for all 98 pts evaluable for NGS MRD will be presented. As of yet, too few progression events have occurred to estimate the second co-primary endpoint, 3-year progression-free survival. Longitudinal response and MRD analysis are ongoing.