Post-treatment circulating tumor DNA in large B-cell lymphoma with an immunoglobulin-based assay: real-world outcomes

Abstract

Measurable residual disease (MRD) analysis using circulating tumor DNA (ctDNA) is a non-invasive method of response assessment in large B-cell lymphoma (LBCL), but data supporting its real-world feasibility and performance are needed. We conducted a multicenter, retrospective study of patients with LBCL assessed in real time with a commercially available, immunoglobulin rearrangement-targeted ctDNA-MRD assay (clonoSEQM). Our primary objective was to assess post-treatment MRD after immunochemotherapy or chimeric antigen receptor T-cell (CAR-T) therapy. Median time from sample collection to MRD result was 9 days. Following frontline treatment (N=102), 12-month progression-free survival (PFS) was 15% versus 85% for detectable versus undetectable MRD (HR 14.5, p<0.001). Detectable MRD was associated with worse PFS for patients with a positive positron emission tomography (PET) scan (HR 4.9, p=0.003). Following CAR-T (N=48), 7 patients with longitudinal testing post-infusion converted from detectable to undetectable MRD by a median 81 days post-infusion without intervention. Detectable MRD by a day 100 landmark after CAR-T was associated with significantly worse 12-month PFS (35% versus 78%, HR 5.0, p<0.001). Detectable MRD was associated with worse PFS for patients with negative PET (HR 4.5, p=0.036). MRD remained associated with PFS after adjustment for PET response, international prognostic index, and histology in immunochemotherapy (p<0.001) and CAR-T settings (p=0.002). Following frontline therapy, MRD demonstrated 64% sensitivity, 96% specificity, 82% positive predictive value, and 90% negative predictive value for progression. Following CAR-T, corresponding values were 63%, 82%, 71%, and 76%, respectively. Real-time ctDNA-MRD assessment is feasible and adds prognostic value after curative-intent LBCL treatment.