Dynamics of Circulating Tumor DNA in Relapsed/Refractory DLBCL Patients

Introduction

The outcome of relapsed /refractory (R/R) diffuse large B-cell lymphoma (DLBCL) is poor. The majority of patients (74%) do not respond to salvage chemotherapy with a median overall survival of 4.6 months. Therefore, there is a need for novel biomarkers to identify these poor-responding patients and to allow changes in treatment strategy. Circulating tumor DNA (ctDNA) has potential as prognostic biomarker during first-line treatment in DLBCL patients. However, data on the prognostic significance of ctDNA in R/R DLBCL patients receiving salvage chemotherapy is limited. The aim of this study was to evaluate the dynamics of ctDNA in R/R DLBCL patients by low-coverage whole genome sequencing, an efficient, rapid, and cost-effective sequencing methodology.

Methods

Twenty five patients with R/R DLBCL diagnosed between 2019 and 2022 in the University Medical Center Groningen were included. Plasma samples were collected at the time of R/R disease diagnosis (baseline, n = 25) and during second-line chemotherapy (n = 86). Clinical data, including total metabolic tumor volume (TMTV) were retrieved from patients' files. Low-coverage whole genome sequencing of cell-free DNA was performed on all plasma samples. Both IchorCNA and QDNAclinic were used to identify copy number variants (CNVs) and IchorCNA was used to determine the estimated tumor fraction (ETF).

Results

The median age of R/R DLBCL patients was 61 years (range: 28-72); 17 patients were male (n = 17, 68%), and 16 (64%) presented with Ann Arbor stage III-IV disease. Second-line chemotherapy consisted of rituximab combined with dexamethasone, cytarabine, and cisplatin (R-DHAP) for 23 patients (92%), while two patients (8%) with concurrent central nervous system involvement received high-dose methotrexate based chemotherapy. The outcomes of the second-line treatment were: complete remission (CR) in 9 patients (36%), partial remission (PR) in 2 patients (8%), progressive disease (PD) in 10 patients (40%), and relapse after CR in 4 patients (16%).

At the time of R/R disease, ctDNA was detected in 14 of the 25 cases (56%). A modest correlation was observed between ETF and TMTV (r² = 0.35). Detectable ctDNA (ETF > 0) was associated with poor outcomes in second-line therapy. Specifically, 8 of 10 patients (80%) with both ETF > 0 and TMTV > 200ml experienced PD/PR. In contrast, 7 of 10 patients (70%) with no ETF and low TMTV achieved long-term remission. Furthermore, patients with a higher number of CNVs at baseline had worse treatment outcomes compared to those with fewer CNVs, with a median number of 8.5 CNVs (range: 0-16) in CR, and 15 (range: 0-30) in PD/PR group.

For patients with detectable ctDNA at baseline, ctDNA dynamics were monitored at baseline, interim, and the end of treatment. Inferior prognosis was associated with the presence of detectable ctDNA during or at the end of treatment, whereas patients who achieved ctDNA clearance demonstrated more favorable outcomes. Notably, 5 out of 6 (83%) patients with detectable ctDNA at the interim showed PD, and all 6 (100%) patients with detectable ctDNA at the end of treatment experienced PD. We observed 1 patient with detectable ctDNA at interim who reverted to CR at the end of treatment. This patient remained in CR.

Conclusion

A low number of CNVs or undetectable ctDNA at baseline are associated with a favorable outcome in patients with R/R DLBCL. Moreover, absence or presence of detectable ctDNA during or at the end of treatment has the potential to serve as a biomarker for identifying R/R DLBCL patients likely to benefit from second-line chemotherapy, as well as those who may not.

Diepstra:Takeda: Research Funding.