A sobering “check” on immune checkpoint inhibitors

In this issue of Blood, Rauch et al analyze primary cells from human T-cell leukemia virus-1 (HTLV-1)–associated adult T-cell leukemia/lymphoma (ATLL) patients treated on a phase 2 clinical trial with nivolumab to elucidate mechanisms of hyperprogression that halted the trial after just 3 patients received a single treatment.¹  Small numbers limit the report as only 2 of 3 patients provided consent for posttreatment sample collection, but the hypotheses generated contain important information for both a rare disease and an increasingly used cancer therapy. Notably, in ATLL, PD-1 blockade may function as a tumor suppressor, disrupt tolerogenic suppression, upregulate additional checkpoints, and/or cause ATLL cells to express growth factors.

Immune evasion, now established as a fundamental hallmark of malignancy, also plays a role in chronic viral infection. In ATLL, the precise mechanism of oncogenesis is not known; however, the expanded clone of malignant ATLL arises from an oligoclonal expansion of HTLV-1–infected cells with high mutational frequency, including recurrent mutations related to T-cell development, activation and migration, transcriptional regulation, and immunosurveillance.²  In fact, a major determinant of the risk of ATLL is the host’s cytotoxic T lymphocyte response to both the HTLV virus and the ATLL clones.³  One possible escape mechanism is the PD-1 pathway, and nivolumab, a fully human monoclonal antibody against PD-1, is an immune checkpoint inhibitor with a strong preclinical rationale in this incurable virally mediated malignancy.

Unfortunately, preclinical data did not align with outcomes for the first few patients treated. ATLL has a broad clinical spectrum with presentations that range from the aggressive (acute and lymphoma subtypes) to the more indolent (chronic and smoldering subtypes), and the study enrolled a single patient each with chronic, smoldering, and acute ATLL, with 2 of the 3 patients having low tumor burden by imaging.⁴  All 3 patients rapidly progressed, and available correlative studies may provide an explanation. First, PD-1 may have functioned as a tumor suppressor, which is supported by evidence that all ATLL clones expanded, not just a specific or most abundant clone. Furthermore, 2 of the 3 patients had indolent subtype disease that became aggressive after therapy. PD-1 may uniquely govern tolerance in patients with indolent ATLL, and ATLL progression may be determined by factors in the immune microenvironment rather than factors intrinsic to the ATLL cell itself. Another hypothesis relates to the role of regulatory T cells (Tregs). Tregs are a key contributor to the maintenance of immune tolerance, and gene expression patterns among tissue-resident, peripheral-resident, and tumor-resident Tregs each differ.⁵  ATLL cells, regardless of compartment, share the Treg phenotype (typically CD4⁺ and CD25⁺), and in 2 patients, the gene expression profiles between tumor-resident Tregs and ATLL cells were similar after treatment with nivolumab. The authors hypothesize compensatory upregulation of additional checkpoints or promotion of ATLL growth factors occurs. Further gene expression profiles of a larger number of patients may shed light on the delicate balance between the ATLL cell and its microenvironment. Limitations of these analyses are the very small numbers and lack of a control population.

Immune checkpoint inhibitors have produced dramatic responses in several advanced cancers; however, accelerated tumor growth is a growing concern. In solid tumors, ∼4% of patients treated with immune checkpoint inhibitors demonstrate “pseudoprogression” (an initial flare-up followed by tumor shrinkage), and 4% to 29% of patients exhibit hyperprogression (acceleration of tumor growth during treatment).⁶  Although definitions and measurements of progression are not uniform, the concern has prompted new standardized approaches for measuring solid tumors in patients treated with immune checkpoint inhibitors (iRECIST).⁷  In lymphoid malignancies, the clinical outcomes of immune checkpoint inhibitors in relapsed and refractory Hodgkin lymphoma (HL) are particularly impressive, and nivolumab is US Food and Drug Administration approved for classical HL refractory to autologous stem cell transplantation and brentuximab. There may be an increased risk of graft-versus-host disease in patients who go on to allogeneic stem cell transplantation, and caution is advised.⁸  Clinical evaluation of immune checkpoint inhibitors is ongoing for HL in the upfront setting, as maintenance after autologous stem cell transplantation, and in multiple NHL subtypes. A recent series demonstrates high response rates in NK/T cell lymphoma.⁹ 

Rauch et al are the first to investigate immune checkpoint therapy in patients with ATLL. A key question is whether the rapid progression represents the natural history ATLL, or is immune checkpoint inhibitor related? Reports suggest ATLL diagnosed in North America among primarily patients of Caribbean descent has a distinct mutational signature and outcome compared with ATLL in Japan.¹⁰  Nevertheless, evaluation of biomarkers predictive of response will guide optimal patient selection for prospective trials in ATLL and other malignancies. In the interim, a sobering pause is required to reassess those that who will benefit from immune checkpoint inhibitors.