Sequencing-Based Detection of Minimal Residual Disease Is Associated with Outcomes after Allogeneic Hematopoietic Stem Cell Transplantation in Patients with Lymphoid Malignancies

Introduction: Monitoring for minimal residual disease (MRD) in hematologic malignancies can identify patients at high risk of relapse after standard therapy and hematopoietic stem cell transplantation (HSCT). However, techniques are limited for MRD monitoring of most lymphoma subtypes. The LymphoSIGHT^TM method (Sequenta, Inc.) is a high-throughput sequencing-based MRD assay that detects small amounts of circulating tumor DNA (CTD) in patients with lymphoid malignancies. This assayhas shown promise in the monitoring of patients with chronic lymphocytic leukemia (CLL), mantle cell lymphoma (MCL), and diffuse large B-cell lymphoma (DLBCL). We evaluated whether CTD measured by LymphoSIGHT^TM pre- and post-allogeneic HSCT predicts relapse and outcome in patients with lymphoid malignancies.

Methods: We retrospectively studied patients enrolled between 2009-2012 on a multicenter phase 3 trial comparing sirolimus-containing to non-sirolimus containing GVHD prophylaxis regimens in patients with lymphoma undergoing allogeneic HSCT. Only patients with available archival tumor tissue were included. Patients underwent conventional restaging at 3, 6, 12, 18 and 24 months per protocol. Genomic tumor DNA was extracted from formalin-fixed paraffin-embedded tissue or bone marrow aspirate mononuclear cells. PCR amplification of IGH and IGK regions was performed followed by high-throughput sequencing to determine the tumor clonotype(s). As part of the trial, peripheral blood mononuclear cell (PBMC) and plasma samples were collected pre-HSCT and at days 30, 60, 100, 180, 270, 365, 545 and 730 after HSCT. DNA from plasma (DLBCL) and PBMCs (other histologies) was amplified and sequenced to determine the level of CTD per million diploid genomes.

Results: We studied 68 patients with B-cell non-Hodgkin lymphoma, classical Hodgkin lymphoma (HL), or CLL. Clonotype calibration was successful in 51 patients (75%; 82% when excluding diagnostic samples with insufficient DNA). MRD evaluation was performed at >=1 time point in 40 patients (13 CLL, 4 DLBCL, 4 transformed indolent lymphoma (TIL), 8 follicular lymphoma (FL), 7 MCL, 4 HL). Overall, 9/40 patients in the cohort relapsed, with a 2y cumulative incidence of relapse/progression (CIR) of 18%. Thirteen of 26 patients (50%) tested were MRD+ pre-HSCT, including 27% of patients in complete remission (CR) and 71% of patients in partial remission (PR) by conventional criteria. The 2y CIR in MRD+ vs MRD- patients at baseline was 38% vs 7% (p=0.2), 2y non-relapse mortality (NRM) was 15% vs 8% (p=0.6), and 2y progression-free survival (PFS) was 46% vs 85% (p=0.12). At day 100, 12 of 32 patients (38%) tested were MRD+. The 2y CIR in day 100 MRD+ patients was 25% vs 0% for MRD- patients (p=0.25), 2y NRM was 25% vs 5% (p=0.05), and 2y PFS was 50% vs 95% (p=0.01). In multivariable models with histology and disease status at HSCT as covariates, the hazard ratio (HR) for progression or death associated with baseline MRD+ was 2.8 (p=0.3); for day 30 MRD+, HR=6.0 (p=0.1); and for day 100 MRD+, HR=8.3 (p=0.07). When entered into the models as a time-dependent covariate, MRD-positivity was associated with a significantly increased risk of progression or death (HR 5.2, 95% CI 1.8-15.4; p=0.03) and of disease relapse/progression (HR 12.7, 95% CI 2.5-64.3, p=0.002). Histology and disease status were not significantly associated with either outcome in the models.

Nine patients relapsed or progressed at a median of 4 (range 1-28) months post-HSCT and 7 (78%) had an MRD+ PB sample at a median of 6 months prior to relapse (Figure). Five patients were persistently MRD+ in all PB samples tested prior to relapse and 2 were initially MRD- then became MRD+. All 5 patients who relapsed after day 100 were MRD+ prior to relapse. Among 31 patients who did not relapse: 16 were MRD- in all PB samples tested; 9 were initially MRD+ then became MRD- by day 365 (4 by d100, 3 by d180, 2 by d365); 2 had a mixed MRD pattern, initially MRD-, then MRD+, then MRD- thereafter; the remaining 4 patients, all of whom had CLL, were persistently MRD+ but had marrow or PB evidence of persistent disease without progression at MRD+ time points.

Conclusions: MRD status based on CTD detection in patients with lymphoid malignancies is predictive of outcome after allogeneic HSCT. This technique may provide important prognostic information throughout the HSCT course and identify a target population for early intervention to prevent disease relapse.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal