Utility of High-Throughput Sequencing of Immunoglobulin Genes for MRD in Lymphoid Malignancy in the Context of Current Immunotherapeutics

Abstract

Background Measurable residual disease (MRD) assessment is central to prognostication and treatment decisions in acute lymphoblastic leukaemia (ALL), plasma cell myeloma (PCM), mantle cell lymphoma (MCL) and chronic lymphocytic leukaemia (CLL). Historically, multi-parameter flow cytometry (MFC) has been the predominant methodology used, however the use of immunotherapeutic modalities that target antigens that are part of MFC MRD assessment (e.g. CD19, CD22, CD38) can make MFC technically challenging. In addition, biological factors such as KMT2A-rearrangements which are associated with lineage ambiguity may also make MFC more technically difficult. We aimed to assess the performance of high-throughput sequencing of immunoglobulin genes (HTS-Ig) compared with MFC within the contemporary therapeutic landscape of ALL, PCM, MCL and CLL which includes treatment with diverse cellular and immunotherapies that may hamper MFC-based MRD assessment. Methods Samples from patients with B-ALL, PCM, MCL and CLL in morphologic remission with undetectable or MRD <1% in the bone marrow by MFC were selected for retrospective analysis. Baseline and remission samples underwent HTS-Ig using the ClonoSEQ NGS assay (Adaptive Biotechnologies). DNA extraction, library preparation and sequencing (Illumina NextSeq500) were performed at the Peter MacCallum Cancer Centre (Melbourne, Australia). HTS-Ig-MRD results were correlated with standard-of-care MFC-MRD (≥8 colour). Results A clonotype was identified for tracking by HTS-Ig in 97% (115/119) of baseline samples. The 4 samples with no clonotype identified were all from patients with PCM. An HTS-Ig-MRD result was obtained for 131 remission samples (54 B-ALL, 37 PCM, 17 MCL and 23 CLL). Of samples analyzed, 40% (52/131) were post treatment with cellular and/or immunotherapies known to cause challenges in MFC assessment including anti-CD19 CAR T-cell therapy (n=32), blinatumomab (n=14), daratumumab (n=4) and inotuzumab ozogamicin (n=2). HTS-Ig sensitivity correlated strongly with DNA input (r=0.95, 95% CI 0.93-0.96). Sensitivity achieved with HTS-Ig and MFC was 34% (45/131) and 0% (0/131) reaching 10-6, 62% (81/131) and 27% (36/131) reaching 10-5, and 4% (5/131) and 57% (75/131) samples reaching 10-4 respectively. The remaining 15% (20/131) MFC samples achieved a sensitivity of 10-3. A discordant MRD result between MFC and HTS-Ig was observed in 22% (29/131). Of the discordant results 93% (27/29) were HTS-Ig-MRDPOS but MFC-MRDNEG, with 62% (19/27) of these attributable to a greater sensitivity achieved by HTS-Ig-MRD compared to MFC-MRD. Eight samples had discordant (HTS-Ig-MRDPOS and MFC-MRDNEG) results not explained by differences in sensitivity including 1 patient with plasma cell trans-differentiation from previous MCL. There was no significant difference between MRD discordance across the 4 diseases (B-ALL 24% [13/54], PCM 22% [8/37], MCL 24% [4/17], CLL 17% [4/23]). A higher degree of MRD discordance (HTS-Ig-MRDPOS and MFC-MRDNEG) was seen in cases with disease biology and/or recent exposure to treatment (anti-CD19 CAR T-cell therapy, blinatumomab, inotuzumab ozogamicin, daratumumab in preceding 1mo) known to cause challenges in MFC-MRD assessment compared to those without (43% [12/28] vs 16% [15/101], p=0.0029 [Fisher's exact test]). Conclusion In this cohort of patients with diverse lymphoid malignancy, HTS-Ig-MRD demonstrated superior sensitivity and analytical performance to detect MRD when compared to MFC which was particularly pronounced in patients receiving antigen-directed therapies that potentially interfere with MFC assessment. These data support the utility of this approach in the current expanding landscape of cellular and immunotherapies.