Liquid biopsy for Hodgkin: a game changer?

In this issue of Blood, Spina et al evaluate the role of circulating tumor DNA (ctDNA) in newly diagnosed and relapsed patients undergoing treatment of classical Hodgkin lymphoma (cHL).¹ 

Response to therapy and detection of relapsed disease are typically assessed by computed tomography and fluorodeoxyglucose positron emission tomography (FDG-PET) imaging in aggressive B-cell lymphoma.^2,3  Although helpful, conventional imaging has several limitations, and alternative modalities for disease assessment are needed. Recently, molecular monitoring of disease looking at ctDNA dynamics has been tested, particularly in diffuse large B-cell lymphoma (DLBCL), in which results demonstrate that the kinetics of ctDNA during therapy can predict clinical outcome.^4,5  Additionally, ctDNA monitoring in remission can detect relapse before the onset of radiologically detectable disease. Studies have also shown that genotyping ctDNA through the application of a technology called cancer personalized profiling by deep sequencing has the ability to identify tumor biological factors that underpin genetic heterogeneity of tumors.⁶  This is a particularly helpful concept in DLBCL, considering the disease comprises many molecular subtypes and offers the opportunity to noninvasively identify early emergence of resistant mutations to various therapies.⁷  In cHL, the role of ctDNA monitoring is not well studied but hypothetically could be very helpful in response assessment particularly. Currently, interim FDG-PET scanning is used alone to identify chemorefractory disease and guide therapy intensification decisions. Although negative interim FDG-PET accurately predicts excellent outcomes, the positive predictive value of this technique is low in this setting and tools such as ctDNA may aid interim response interpretation.⁸  Elucidating the molecular heterogeneity and mutational spectrum of cHL has been much more challenging than in DLBCL because of the paucity of neoplastic cells in the infiltrate. Again, genotyping ctDNA represents a potentially interesting approach.

Spina et al report on a retrospective study in which they used a highly sensitive and deep next-generation sequencing ctDNA assay to analyze specimens from both newly diagnosed (80) and relapsed/refractory (32) patients with cHL. First, having demonstrated that ctDNA mirrors the genetics of microdissected Reed-Sternberg cells using cancer personalized profiling by deep sequencing, they assayed ctDNA in 80 newly diagnosed cases to characterize the mutational landscape of cHL. Among their findings was that ∼40% of cases had mutations of STAT6, with TNFAIP3 and ITPKB being the other most common mutations. Using a probabilistic classifier that was derived from differential gene expression, they compared the ctDNA signatures of cHL, primary mediastinal B-cell lymphoma, and DLBCL; as has already been demonstrated with gene expression profiling of tumors, cHL and primary mediastinal B-cell lymphoma were genetically closely related and cHL and DLBCL were very distinct. In relapsed/refractory cases in which longitudinal ctDNA monitoring was performed, treatment-related clonal evolution was demonstrated with interesting patterns in a small number of patients on immune-checkpoint inhibitors. Finally, in a subset of newly diagnosed patients who received Adriamycin, bleomycin, vinblastine, and dacarbazine chemotherapy, a 100-fold fall or 2-log drop in ctDNA following 2 cycles of therapy predicted a significantly better outcome; in addition, ctDNA quantification complemented FDG-PET imaging assessment.

This study is helpful because it establishes that, in patients with cHL, ctDNA can be used as a source of tumor DNA to profile mutations and characterize disease biology. Technically, in cHL, tumor biopsies represent a huge challenge because, first, mediastinal disease is common, is a challenging site to access, and diagnostic biopsies are frequently small core samples. Second, the low representation of tumor cells compared with background cells and high degree of fibrosis in cHL can make diagnostic interpretation exceeding difficult. Hence alternative “liquid” biopsies may be particularly helpful in this disease, with potentially many roles that include guiding therapy escalations and de-escalations and monitoring clonal evolution patterns, particularly in the setting of novel agents.