Immune checkpoint inhibitors in Hodgkin and non-Hodgkin lymphoma: how they work and when to use them

KEYWORDS:

• Lymphoma
• checkpoint inhibition
• biomarker
• clinical trial
• microenvironment
• Hodgkin lymphoma

1. Targeting immune checkpoints

The introduction of immune checkpoint inhibitors to the armamentarium of anticancer drugs has revolutionized cancer therapeutics in solid and hematological malignancies, and select inhibitory antibodies of the programmed death ligand axis have already found their way into clinical practice following US FDA approval. However, despite convincing objective clinical responses in some cancer types, major obstacles to clinical implementation and questions around optimal use remain. In lymphoid cancers, the most dramatic clinical responses have been observed in classical Hodgkin lymphoma (CHL). In contrast, to date results in non-Hodgkin lymphoma (NHL) have been less striking and response biomarkers remain elusive.

Primarily based on preclinical evidence, the clinical activity of immune checkpoint inhibitors is thought to be mediated by blockade of the interaction of specific ligands (e.g. programmed death ligand 1, PDL1 /CD274; programmed death ligand 2, PDL2 /PDCD1LG2; B7-1 /CD80; B7-2 /CD86) expressed on tumor cells, with receptors (e.g. CTLA4 and PD1) expressed on subsets of nonmalignant T cells in the tumor microenvironment [1]. PD1 and CTLA4 are co-inhibitory receptors that provide accessory signaling to the T cell receptor, and physiologically act to regulate immune responses, limit autoimmunity and induce T cell anergy and exhaustion. The most prominent therapeutic approaches currently include anti-PD1, anti-PDL1, and anti-CTLA4 antibodies that were shown to reengage exhausted T cells in preclinical models and have demonstrated clinical efficacy across a range of malignancies [2]. Interestingly, and supporting the hypothesis of reengagement of PD1 + T cells, the most striking clinical results of immune checkpoint inhibition have been observed in melanoma, which is characterized by a high mutational load (related to a higher likelihood of generating immunogenic neo-epitopes) [3,4] and CHL, which features a prominent immunomodulatory tumor microenvironment [5,6]. However, further studies including serial biopsies of patients treated with checkpoint inhibition are needed in these and other tumor types to elucidate the precise mechanisms of action, the drug target cell population, and spatial relationship of cells expressing the PD1 receptor and ligands.

2. Clinical trials

Pidilizumab was the first anti-PD1 antibody evaluated in a phase I study of relapsed/refractory (R/R) lymphomas and provided ‘proof of principle’ that checkpoint blockade should be further explored [7]. With abundant expression of PDL1 as a cornerstone of immune evasion in CHL, there is a strong rationale for PD1 blockade as a therapeutic approach. On this basis, in the pivotal phase I study of the PD1 antibody nivolumab (3 mg/kg week 1,4 and then Q 2 weeks) in relapsed hematologic malignancies, an expansion cohort of patients with R/R CHL was planned due to strong preclinical rationale, and striking responses were observed in 23 treated patients with an overall response rate (ORR) of 87% and complete remissions (CRs) in 17% of patients [6]. Similarly, in the multi-cohort phase Ib study, KEYNOTE 013, evaluating the anti-PD1 antibody pembrolizumab (10 mg/kg Q 3 weeks) in a range of lymphoma subtypes, an ORR of 65% was demonstrated in the cohort of patients with R/R CHL (n = 31) and a CR rate of 16%. With a median follow-up of 17 months, the 12 month progression-free survival (PFS) rate was 46% [8]. In the latter study, correlative studies were performed and flow cytometry studies demonstrated a significant increase in the total T cells, CD4 and CD8 T cell subsets, and natural killer (NK) cells in the peripheral blood comparing baseline and post cycle 7, as well as an increase in an IFNγ gene signature using the NanoString platform. However, neither of these biomarkers correlated with objective response. Building on these preliminary results, multi-cohort phase II studies in R/R CHL were initiated with both nivolumab (CheckMate 205) and pembrolizumab (KEYNOTE 087). Cohort B of CheckMate 205 evaluated nivolumab in CHL patients with disease recurrence following autologous stem cell transplantation (ASCT) and subsequent brentuximab vedotin (BV). The independent radiological review demonstrated an ORR of 66.3% (CR 8.8%) and the investigator ORR was 72.5% with a CR of 27.5% [9]. On the basis of these results and the phase I study, on 17 May 2016, the FDA approved the use of nivolumab in R/R CHL with failure following ASCT and BV. Similar results were reported in the phase II study using fixed-dose pembrolizumab with an investigator reported ORR [10] of 70% (CR 20%) in patients with treatment failure after ASCT and BV (cohort 1) and an ORR of 80% (CR 27%) in patients considered transplant ineligible due to failure of standard salvage therapy followed by BV (cohort 2) [11]. As such, pembrolizumab has received FDA breakthrough designation in this setting. To date, safety data was consistent with what has been reported in solid tumors with no grade 4 immune-related events in these studies.

Of note, there is limited information on the efficacy of PD1 inhibitors in NHL. In the initial phase I study with nivolumab, excluding patients with CHL, 81 patients with R/R hematological malignancies were treated including 10 patients with DLBCL and 11 patients with follicular lymphoma with only modest objective response rates of 40% and 36%, respectively [12]. Since primary mediastinal large B cell lymphoma (PMBCL) frequently harbors genetic alterations of the PDL locus on chromosome 9p24.1 with resulting high expression of PDL1 and PDL2, a separate cohort was evaluated in the phase Ib pembrolizumab trial and 4 of 9 patients have had an objective response [13]. With a number of new trials underway evaluating checkpoint inhibitors in lymphoma, there remain a number of unanswered questions. The longest follow-up is still under 2 years in all studies, and thus, response durability remains unknown. Further, the optimal treatment duration remains uncertain which may impact toxicities particularly as these agents are incorporated earlier in the disease course, but also has significant cost implications.

The next study phase of checkpoint therapy will be to evaluate combinations with other systemic therapies particularly as there is emerging data that they may enhance immunogenicity. For example, the antibody drug conjugate BV is FDA approved in R/R CHL following ASCT and as consolidation therapy in high-risk patients post ASCT [14,15]. The active agent in BV is monomethyl auristatin E (MMAE), a microtubule disrupting agent that ultimately causes apoptotic cell death. Although the role of BV in directly inducing cell death has been well described, a recent study of BV in HL cell lines demonstrated induction of the hallmark immunogenic cell death markers calreticulin and HSP90 providing a rationale for combining BV with checkpoint inhibitors [16]. Further, BV can also enhance the expression of CD86 and MHC class II antigens further inducing immunogenic cell death [17]. A phase I industry sponsored study is currently ongoing evaluating BV in combination with Nivolumab in the pre-ASCT setting in R/R CHL. Combination therapy is particularly attractive in NHL where response rates with checkpoint inhibitors are more modest [12]. As such, BV and nivolumab are also under evaluation in a phase 1/2 study of CD30+ relapsed/refractory DLBCL, peripheral T cell lymphomas and mycosis fungoides/ Sézary syndrome. A similar rationale could be extended to other antibody drug conjugates, including those that target B and T cell markers.

Another example for a promising combination therapy is the ongoing multi-cohort phase I ECOG-ACRIN study (E4412) evaluating BV in combination with the anti-CTLA4 inhibitor ipilimumab in R/R CHL, demonstrating an ORR of 72% and CR rate of 50% with manageable toxicity [18]. Ipilimumab activates suppressed T cells at an earlier phase than PD1 inhibition and may have an independent effect. Studies in metastatic melanoma demonstrate striking responses combining ipilimumab and nivolumab [19] and the ongoing E4412 study will include cohorts treated with this combination in addition to BV and nivolumab as well as all three agents.

3. Biomarker considerations

Another important, but largely open question remains if clinical decisions regarding checkpoint inhibition might be guided by biomarker considerations. The hope for effective clinically relevant biomarkers is fueled by the finding that PDL1 and PDL2 expression is heterogeneous in specific lymphoma entities [20–24], and abundance and spatial patterns of PD1 expressing (exhausted) T cells have been reported as variable and correlated with treatment outcome [25,26]. Moreover, in a subset of B cell lymphomas, including CHL, PMBCL, and diffuse large B cell lymphoma (specifically primary central nervous system lymphoma and primary testicular lymphoma), structural genomic alterations (copy number amplification, translocations, and intrachromosomal rearrangements) are correlated with elevated gene expression of PDL1 and PDL2[22,27,28], suggesting that these somatically acquired changes underlie an immune privilege phenotype that might be preferentially targetable with immune checkpoint inhibition in the subset of patients with PDL1 and PDL2 alterations. However, despite this intriguing biology, there is not yet convincing evidence that PDL mutation status, elevated expression, or PD1 + T cell abundance is linked to treatment response of checkpoint inhibitors in lymphoma. As an example, Hodgkin Reed Sternberg cells, the malignant cells in CHL, harbor amplifications of the PDL locus on chromosome 9p and are positive for PDL1 surface expression in the majority of cases [24]. However, in the pivotal phase I clinical trial [6], treatment response was not associated with any investigated biomakers, which may reflect limited statistical power or incomplete biological characterization. Interestingly, in melanoma, mutational load, neoantigen load, and expression of cytolytic markers in the immune microenvironment were significantly associated with clinical benefit to CTLA-4 blockade [29,30] which might provide a template for biomarker development in lymphoma, although likely connected with great technical challenges for clinical implementation.

4. Future directions

Despite the encouraging results of the above described clinical trials, and FDA approval of nivolumab in R/R CHL, a number of challenges remain: (1) More precise definition of indications related to clinical and molecular heterogeneity; (2) Rational selection of combination therapies in future clinical trials; (3) Effective biomarker approaches and reevaluation of existing biomarkers to guide treatment decisions; and (4) Study of drug resistance and treatment failure. Given the rapid emergence of other therapeutic approaches in CHL and NHL, that will be assessed in parallel and create competition for trial design and patient accrual, prioritization strategies and biomarker-driven study designs will be paramount to maximize clinical benefit. Importantly, biomarker studies have to be firmly integrated into prospective trial design, and anticipating the emergence of predictive biomarkers in the rapidly expanding field of checkpoint inhibitor therapy, future indications might be best tied to simultaneously developed companion diagnostics.

Declaration of interest

KJ Savage has been a consulting advisor and received honoraria from Bristol-Myers Squibb, Merck and Seattle Genetics. The remaining author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Additional information

Funding

This paper has been funded by the Michael Smith Foundation for Health Research and the Canadian Institutes of Health Research.