RESUMO
PURPOSE: Current clinical challenges in Hodgkin lymphoma (HL) include difficult-to-treat relapsed/refractory disease and considerable long-term toxicities of treatment. Since clinical risk factors lack discriminatory power, intensity of therapy is mainly based on tumor burden. Exploring HL genetics and tumor microenvironment (TME) might provide valuable insights for improved risk stratification. MATERIALS AND METHODS: In this study, we applied circulating tumor DNA sequencing to 243 patients obtained from pivotal German Hodgkin Study Group trials to identify subtypes of HL. Independent validation of the subtypes was performed in 96 patients treated in the EuroNet-PHL-C2 study. Outcome differences of subtypes were assessed in an event-enriched clinical validation cohort comprising 72 patients from the HD21 trial, using a refined, validated, and clinically feasible assay. RESULTS: We propose a biologic classification of HL consisting of three distinct subtypes: inflammatory immune escape HL is characterized by frequent copy-number variations including immune escape variants such as high-level amplifications of the PD-L1 locus and an inflammatory TME. Virally-driven HL is associated with Epstein-Barr virus and/or human herpesvirus 6 and an inflammatory TME with neutrophils and macrophages, while the tumor mutational burden (TMB) is low. Oncogene-driven HL is defined by a high TMB, recurrent mutations in oncogenic drivers such as TNFAIP3, ITPKB, and SOCS1, and a cold TME. A refined and validated assay version aiming at clinically feasible risk stratification showed significant progression-free survival differences between subtypes. In addition, assessment of minimal residual disease (MRD) allowed for the detection of patients at very high risk of relapse within the subtypes. CONCLUSION: We propose a clinically feasible, noninvasive method for individualized risk stratification and MRD monitoring in patients with HL on the basis of circulating tumor DNA sequencing.
RESUMO
BACKGROUND: Individualization of treatment in Hodgkin's lymphoma is necessary to improve cure rates and reduce treatment side effects. Currently, it is hindered by a lack of genomic characterization and sensitive molecular response assessment. Sequencing of cell-free DNA is a powerful strategy to understand the cancer genome and can be used for extremely sensitive disease monitoring. In Hodgkin's lymphoma, a high proportion of cell-free DNA is tumor-derived, whereas traditional tumor biopsies only contain a little tumor-derived DNA. METHODS: We comprehensively genotype and assess minimal residual disease in 121 patients with baseline plasma as well as 77 follow-up samples from a subset of patients with our targeted cell-free DNA sequencing platform. FINDINGS: We present an integrated landscape of mutations and copy number variations in Hodgkin's lymphoma. In addition, we perform a deep analysis of mutational processes driving Hodgkin's lymphoma, investigate the clonal structure of Hodgkin's lymphoma, and link several genotypes to Hodgkin's lymphoma phenotypes and outcome. Finally, we show that minimal residual disease assessment by repeat cell-free DNA sequencing, as early as a week after treatment initiation, predicts treatment response and progression-free survival, allowing highly improved treatment guidance and relapse prediction. CONCLUSIONS: Our targeted cell-free DNA sequencing platform reveals the genomic landscape of Hodgkin's lymphoma and facilitates ultrasensitive detection of minimal residual disease. FUNDING: Mildred Scheel School of Oncology Aachen-Bonn-Cologne-Düsseldorf MD Research Stipend, Next Generation Sequencing Competence Network grant 423957469, Deutsche Krebshilfe grant 70112502, Deutsche Forschungsgemeinschaft (DFG) grant EN 179/13-1, the HL MRD consortium, and the Frau-Weiskam und Christel Ruranski-Stiftung.