Circulating Tumor DNA in Untreated Classical Hodgkin Lymphoma Patients Treated with Pembrolizumab and Chemotherapy: Dynamic Response Assessment and Correlation with Baseline Metabolic Tumor Volume

Introduction

We have previously presented early results of untreated CHL patients treated with concurrent pembrolizumab + AVD (Lynch et al. ASH 2021) which was highly effective but had increased rates of false positive PET/CT results leading to additional scans and/or biopsies. Circulating tumor DNA (ctDNA) may represent a more sensitive and specific method of response assessment. Herein we report updated clinical results with correlation of dynamic ctDNA monitoring as well as data on total metabolic tumor volume (TMTV).

Methods

This was a single arm pilot study combining pembrolizumab with AVD in untreated CHL of any stage. Pembrolizumab 200 mg IV was administered every 21 days concurrently with AVD chemotherapy. AVD was administered at the standard dosing and schedule. Plasma samples were analyzed for ctDNA at baseline, post cycle 1 (if available), post cycle 2, and end of treatment. ctDNA levels were quantified as haploid genome equivalents/mL plasma using CAPP-Seq and PhasED-Seq (Kurtz et al. Nat Biotech 2021). Baseline, post cycle 2, and end of treatment PET/CT were analyzed for TMTV using a threshold based semiautomated technique that included any tumor with standardized uptake value (SUV) above the liver background SUVmax using MIM Encore (version 7.1.3 Cleveland, OH).

Results

Among the 30 patients enrolled, 29 are evaluable for response and/or exploratory analysis (one patient withdrew from study treatment prior to any response assessment). With median follow-up of 21 months, 2-year PFS and OS were 97% and 100%, respectively. While 5 patients had residual FDG uptake at EOT, only 1 (20%) has developed recurrent Hodgkin lymphoma. ctDNA data was available for 18 patients. Pretreatment ctDNA levels were significantly correlated with TMTV (R_S=0.68, p=0.003). 17/18 patients had detectable ctDNA at baseline and were therefore evaluable for MRD assessment. Among 8 PET2+ patients, only 1 patient had detectable ctDNA, and ctDNA became undetectable by the end of treatment. Importantly, of these 8 patients, none have relapsed. Only 1/9 (11%) patients who were PET2- had detectable ctDNA after 2 cycles, and the cancer ultimately recurred. At EOT for those who received > 2 cycles, 2/14 had detectable ctDNA (one CHL relapse, one with secondary DLBCL 11 months later). ctDNA analyses for the remaining patients are in process and will be presented at the meeting.

Conclusion

Concurrent pembrolizumab + AVD represents highly effective frontline therapy for CHL, but results in spurious PET findings in a significant proportion of patients despite most patients having undetectable ctDNA and no relapse. ctDNA may represent a more sensitive and specific response assessment tool to be studied in larger datasets.