Next Generation Flow (NGF) for High Sensitive Detection of Minimal Residual Disease (MRD) in Multiple Myeloma (MM)

Introduction: The clinical and prognostic utility of MRD monitoring in MM bone marrow (BM) by first generation (4-6-color) flow cytometry (flow-MRD), has been now demostrated for more than a decade. Thereby, flow-MRD is considered to be a well-suited technique for MRD monitoring in MM, due to its high applicability and specificity, and its broad availability in diagnostic laboratories. However, recent results have confirmed that 1^st generation flow-MRD has a lower sensitivity than molecular techniques such as allele-specific oligonucleotyde (ASO)-PCR and next generation sequencing (NGS); in addition, the lack of standardization of conventional flow-MRD approaches, also has a negative impact on its reproducibility. Here we report on the validation of the next generation flow (NGF)-MRD approach developed by the EuroFlow Consortium and the International Myeloma Foundation (IMF) for ultrasensitive and standardized detection of MRD in MM.

Methods: A total of 275 BM samples were included in the study: 1) a group of 31 normal/reactive and 63 diagnostic MM BM samples were evaluated to identify the most efficient candidate markers for the NGF panel, using a multivariate analysis-based approach; 2) next, a total of 181 BM samples from 15 healthy donors (HD), 36 MGUS, 15 MM and 3 solitary plasmacytoma (SP) patients studied at diagnosis, plus 112 MM follow-up samples, most of which (n=71) corresponded to BM samples from MM patients in very good partial response, complete remission (CR) or stringent CR were analyzed. These samples were evaluated by the EuroFlow-IMF NGF-MRD method. The method was based on a (standardized) lyse-wash-and-stain sample preparation protocol, the measurement of high numbers of BM cells (≥5x10⁶ cells/tube) and an optimized 8-color, 2-tubes, antibody panel, for accurate identification of BM plasma cells (PCs) and discrimination between phenotypically aberrant (aPC) and normal PC (nPC): tube 1: CD138_BV421 CD27_BV510 CD38_FITC CD56_PE CD45_{PerCP Cy5.5} CD19_{PE Cy7} CD117_APC CD81_{APC C750}, and; tube 2: identical to tube 1 except for cytoplasmic (Cy) Immunoglobulin (Ig) k_APC/CyIgl_{APC C750}). Results obtained with the NGF-MRD MM method in the 71 VGPR, CR and sCR samples, were then compared with a 2^nd generation (routine) 8-color flow-MRD approach which involved a single 8-color combination staining for the same markers described above for tube 1, but in the absence of full optimization of the positions for the antibody-fluorochrome conjugates and no selection for treatment-independent antibody CD38 clones. In a subset of 16 of these 71 MRD samples in which enough material was available, comparison with NGS was also performed in parallel.

Results: In all MGUS, MM and SP cases analyzed at diagnosis, aPC showed aberrant phenotypes vs. nPC from HD BM, based on multivariate analysis of individual cells from each of the patients against a reference data base of normal/reactive PCs (100% applicability). For the MM MRD BM samples, a median of 9.8x10⁶ (range: 2.4-15.3) events were acquired (tube 1 plus 2) vs. 1x10⁶  (range: 0.03-5) events for the  2^nd generation flow-MRD approach with an (estimated) 10 times lower limit of detection and 10 times lower limit of quantitation of 3x10^-6 and 5x10^-6 vs. 3x10^-5 and 5x10^-5 for the NGF-MRD vs. the 2^nd generation flow-MRD approaches, respectively. This led to a higher rate of MRD+ samples with the NGF-MRD method: 14/48 (29%) cases that were flow-MRD-negative with the 2^nd generation 8-color flow-MRD method became MRD+ (median percentage of residual aPC of 0.0007%; range: 0.0002 to 0.02%) (Figure 1A). In contrast, 4/38 (11%) samples were negative by NGF, while positive by 2^nd generation flow-MRD; one of these four proved to be MRD-negative by cytoplasmic immunoglobulin light chain restriction analysis. Further comparison of NGF vs NGS showed 9 of the 16 samples evaluable were MRD-negative by NGS; from them, one third (3/9) were positive by NGF with a median number of residual aPCs of 0.005% (range: 0.0002-0.006%) (Figure 1B).   

Conclusions:The EuroFlow-IMF NGF-MRD approach provides a fast, highly applicable, ultrasensitive, standardized and accurate approach for the assessment of MRD in BM samples from MM patients and overcomes the current limitations of both 1^st and 2^nd generation flow-MRD approaches; preliminary results showed higher sensitivity than NGS.