Standardized Minimal Residual Disease Detection By Next Generation Sequencing In Multiple Myeloma

Abstract

Background Next generation sequencing (NGS) has been applied to monitor minimal residual disease (MRD) in multiple myeloma (MM). Standardized DNA input and sequencing depth is essential for achieving a uniform sensitivity in NGS-based MRD study. Method We applied a standardized protocol for MRD assessment of four myeloma cases using the LymphoTrack-Miseq platform based on the use of triplicates of 1 μg DNA input for each MRD sample and a sequencing depth of 1 million sequencing reads per replicate. The number of cells contained in 1μg of each sample was validated by the real time PCR standard curve method using plasmids, in which the albumin gene is cloned. Two plasmids containing unique IGH sequences were added to each replicate, one at the concentration of 10-5 (copy number equivalent to 0.001% of the number of total cells in a replicate) for validation of the sensitivity of 10-5, and the other at 5x10-5 or 10- 4 for obtaining an amplification factor. The MRD level in each replicate was calculated from the corresponding reads of the myeloma-specific sequence and the amplification factor. The final MRD level of a sample was defined as the mean MRD levels of the triplicates. In addition, one normal bone marrow sample was used as normal control to evaluate the feasibility of identified clonal sequences as MRD target. Result First of all, five clonal sequences identified in the four myeloma cases were not detected in the normal control except one (sequencing reads of 7), however, of which the complementarity-determining region 3 is in high diversity. The presence of this myeloma-specific sequence in normal control is more likely caused by Abstracts 17th International Myeloma Workshop, September 12-15, 2019 | 274 index misassignment. The spike-in control of 0.001% MRD was consistently detected in all samples, i.e. seven samples of the four myeloma cases, confirming a sensitivity of 10-5. Moreover, the spike-in control at 10-4 appears appropriate for MRD normalization as variation lower then 2.8 folds in frequency among triplicates achieved in 4/5 samples. Furthermore, this standardized NGS approach yielded MRD+ or MRD- results concordant with allele-specific oligonucleotide (ASO) real-time quantitative (RQ)-PCR. NGS showed expected decrease of MRD levels associated with the change in serological response in three follow up samples from one case. NGS showed an improved sensitivity and provided quantification of MRD for cases assigned 'positive but not quantifiable' (PNQ) by ASO RQ-PCR, without the use of patient-specific probes/primers. Conclusion The standardized LymphoTrack-MiSeq-based method is verified to carry a sensitivity of 10-5, hence an effective tool for MRD monitoring in MM. As only a small number of samples are tested here, further study with a larger number of patients is