Standardized minimal residual disease detection by next-generation sequencing in multiple myeloma

Abstract

Bone marrow (BM) based minimal residual disease (MRD) is one of the most powerful prognostic factors in multiple myeloma (MM). Therefore, standardization and easy operation of MRD testing is crucial. Regarding MRD detection by next- generation sequencing (NGS), the majority of data came from LymphoSIGHT platform which is now called clonoSEQ, an expensive customer service provided by a single company in the USA. However, the reported sensitivity of 10-6 of LymphoSIGHT platform by serial dilution experiment in selected patients was not experimentally documented and guaranteed in each & every MRD sample. On the other hand, LymphoTrack, provided by Invivoscribe provides customer with kits comprising multiplex primers compatible with Illumina MiSeq. Therefore, the first aim of this study was to devise a standardized NGS-based MRD detection approach with a uniform sensitivity. In a pilot experiment of 7 samples, the sensitivity of 10-5 of LymphoTrack-MiSeq platform was achieved by using a standardized protocol of triplicates of one μg DNA input and one million sequencing reads per replicate. Two plasmids cloned from immunoglobulin rearrangements were spiked into MRD sample DNA in parallel with MRD measurement, one for marking sensitivity of 10-5 and the other for MRD normalization. This platform was further simplified and validated in 19 samples. The plasmid spike-in controls were replaced by genomic DNA from myeloma cells to make the procedure more convenient and accurate. Sensitivity of 10-5 was consistently reached. Reproducibility of MRD detection was confirmed by the small inter-run variation tested in 3 samples. Compared with allele-specific oligonucleotide real-time quantitative-PCR, this approach does not require any patient-specific primer or probe, thus, much less labor- and time-intensive. Utility of this standardized NGS-based MRD approach in prognostication was evaluated in fifty MM patients who achieved very good partial response or better post autologous stem cell (ASCT) transplant. MRD was detected in 40/50 (80%) of these patients. MRD at the threshold 10-4 was prognostic of both progression-free survival and overall survival. A larger study is warranted to evaluate the prognostic impact of MRD using the cut-off of 10-5. Furthermore, in four patients, change of MRD from undetectable to positivity was observed in one patient after ASCT despite persistent clinical complete response. More cases are needed to thoroughly evaluate the dynamic change in MRD post-ASCT and the optimal time point for assessing MRD post-ASCT. Finally, efficacy of a 3-weekly daratumumab-based regimen was evaluated by MRD assessment in 10 relapsed and 3 newly diagnosed MM patients. MRD negativity was achieved in 7/10 (70%) relapsed patients and 2/3 (67%) newly diagnosed patients. In conclusion, we standardized an NGS-based MRD approach with a sensitivity of 10-5 which is sufficiently sensitive and feasible for MRD measurements. MRD post-ASCT was a prognostic factor of survivals. 3-weekly daratumumab in combination with proteasome inhibitor and/or immunomodulatory agent are effective and rendered MRD-negativity in the majority of relapsed MM.