The immunotherapy real estate of Hodgkin lymphoma

In this issue of Blood, Gottschlich et al¹ present an in silico analysis of genomics data to highlight CD86 as a novel immunotherapeutic target in classic Hodgkin lymphoma (cHL). The authors explore the CD86-CTLA4 crosstalk axis between malignant Hodgkin Reed-Sternberg (HRS) cells and T cells in the tumor microenvironment (TME), and describe a CD86-28z chimeric antigen receptor (CAR) T-cell approach with promising preclinical in vivo efficacy.

CHL was known for, and continues to be, paradigmatic for treatment success of polychemotherapy and radiation in hematological malignancies. Now, cHL is in focus once again for impressive responses to immunotherapy modalities that currently include antibody-drug conjugates (brentuximab vedotin) and anti–programmed death protein-1 (PD-1) checkpoint inhibitor therapy as part of routine clinical management. Other modalities, such as CAR T-cell therapy, are the subject of clinical trials.² Progression-free survival of pivotal phase 3 clinical trials testing both brentuximab vedotin³ and nivolumab⁴ containing regimens in first-line treatments, respectively, suggest cure of the lymphoma in the vast majority of patients. These regimens are seen as part of a solution to reduce significant toxicities of polychemotherapy and radiation, which include increased risk for secondary malignancies, infertility, and long-term impaired organ function, typically in adolescents and young adults.⁵ With immunotherapies coming with their own set of toxicities, a productive debate about reduction of treatment intensity, avoidance of well-described toxicities of chemotherapy drugs, and radiation has been initiated. A key to this discussion is the careful selection of patients who can ideally be matched to treatment regimens based on clinical risk profiles, interim response assessment, and biomarkers that capture the biological heterogeneity of cHL at initial diagnosis.

Gottschlich et al add to the potential choices of available immunotherapeutic options at disease relapse by providing a preclinical rationale for CD86 targeting using CAR T cells. The study builds on a foundation of previously published genomics data, including microarray gene expression analysis of laser microdissected HRS cells, as well as bulk and single-cell transcriptomes of the TME in cHL. For context, genomic interrogations of cHL have been notoriously difficult due to the rarity of malignant HRS cells and the predominance of various types of immune cells in tissue biopsies.⁶ Therefore, the authors chose a multimodal in silico target screening approach to reveal CD86 as a highly expressed target on both the malignant HRS cells and tumor-associated macrophages. They next describe the development of CD86-28z CAR T cells, tested in vitro (HL-derived cell lines) and in vivo (cell-line xenotransplantation into NSG mice), to establish strong antigen-dependent CAR T-cell proliferation and effective T-cell–mediated killing of both malignant cells and CD86⁺ macrophages. Importantly, the in vivo study also encompassed assessment of infection risk and CAR T-cell–mediated off-tumor toxicity in syngeneic mouse models, wherein acceptable safety profiles were established.

The current pathogenesis models postulate derivation of the HRS cells from germinal center or postgerminal center B cells with clonal, but unproductive, B-cell receptor editing, constitutively active NF-κB and JAK-STAT signaling driven by somatic gene mutations, and various hallmarks of immune escape.⁷ Although pathogenic insights are growing, longstanding open questions include the causes of the specific HRS-cell morphology (large, often multinucleated cells), the precise HRS cell of origin, and the lack of knowledge about early drivers of tumor-supporting TME ecosystems. Although their study leaves many questions unanswered regarding pathogenesis unresolved, placing the CD86-CTLA4 ligand-receptor pair more prominently on the map opens the door for follow-up studies and additional (pre)clinical development of immunotherapies targeting this interaction. A particular strength of the study is the dissection of TME function using CD86-directed interventions in coculture and humanized mouse xenotransplantation models, where key results include the reprogramming of cHL-associated T cells with reduction of both CTLA4 and PD-1 expression. Of note, treatment modalities in scope are not necessarily limited to CAR T cells but may also include bispecific antibodies or more traditional direct targeting of, for example, CTLA4, currently being explored in clinical trials.⁸ The rational focus on CTLA4 or CTLA4 interaction partners is also supported by multiple single-cell studies that identified CTLA4 (in combination with LAG3 in a subset of cells) as the most cHL-typic T-cell subtype.⁹ 

Dual-malignant-cell and TME targeting with CAR T cells is not unprecedented and has been tested preclinically with CD123,¹⁰ another surface expression marker found on HRS cells and tumor-associated macrophages. However, the only CAR T-cell target currently being tested in human clinical trials remains CD30, and thus growing the armamentarium of available immunotherapy targets remains attractive in cHL. Given the very high treatment response rates of first and subsequent lines of treatment, the clinical utility landscape for new immunotherapeutic approaches, such as presented in this study, remains tight. However, targeted modalities should remain high on the agenda so we may provide options for the still significant number of patients with cHL who have treatment-refractory disease and as we work toward the goal of toxicity reduction.

Conflict-of-interest disclosure: C.S. receives research support from Trillium and Epizyme and has performed consulting work for Bayer and Eisai. He is also coinventor of a patent owned by British Columbia Cancer for a subtyping assay for aggressive lymphomas (Nanostring).