Targeted Next Generation Sequencing of Circulating Tumor DNA in Multiple Myeloma

Introduction: Immunophenotypic and chromosomal analysis among patients with multiple myeloma (MM) has traditionally been performed on plasma cells enriched from bone marrow aspirates. However, due to the patchy nature of MM, a bone marrow aspirate may not accurately represent the total body burden of disease nor does it reflect the true heterogeneity of the cancer cells. Additionally, since it is an invasive procedure, frequent resampling to monitor clonal evolution is infeasible. Next generation sequencing of circulating tumor DNA is a promising new approach that may address these shortcomings. Circulating tumor DNA, either from cell-free DNA or circulating tumor cells, originates from multiple different sites of disease and may be obtained from a single blood sample. This non-invasive approach may enable clinicians to detect early signs of resistance, monitor minimal residual disease, identify therapeutic targets, and provide prognostic information to help guide treatment decisions.

Methods: Bone marrow mononuclear cells (BMMC), peripheral blood mononuclear cells (PBMC), and plasma samples were obtained from 45 patients treated at our center at the time of MM diagnosis, remission, or disease relapse. All samples were obtained through an Institutional Review Board approved protocol. DNA was extracted and library preparation for next generation sequencing was performed using unique molecular identifiers (UMI). The incorporation of UMIs allows for the identification and removal of errant base calls introduced during PCR amplification and Illumina sequencing thereby increasing the sensitivity of the assay to detect rare mutations. Target enrichment was performed using a panel of primers specific to exons of genes known to be significantly, recurrently mutated in MM as well as genes with predictive or prognostic value. Illumina sequencing was performed to a depth of 10,000X. Comparison of single nucleotide variants (SNV) and indels between matched BMMC, PBMC and plasma samples was determined. Correlation of clinical outcomes with mutation status and variant allele frequency was performed.

Results: While good concordance was found between BMMC, PBMC, and plasma samples among the SNVs and indels detected, distinct variants unique to each tissue compartment were also identified. Indicating the assay may have utility for disease monitoring, variants detected in the blood pre-treatment were no longer detectable in post-treatment remission samples. Additionally, new variants in blood samples not present prior to initiation of therapy were detected at later time points providing evidence for clonal evolution. Finally, driver mutations were found in genes for which known targeted therapy is available or in genes that may theoretically confer drug resistance such as CUL4A that binds to Cereblon, a target of the immunomodulatory drug lenalidomide.

Conclusions: While bone marrow biopsy represents the standard of care for diagnosis and disease monitoring in patients with multiple myeloma, the procedure is subject to sampling bias and frequent serial biopsies are impractical. A noninvasive approach using targeted next generation sequencing of circulating tumor cells is an appealing alternative. We demonstrate the feasibility of this technique in MM PBMC and plasma samples and find evidence that the assay may support clinical decision making.