Harnessing the power of the immune system in non-Hodgkin lymphoma: immunomodulators, checkpoint inhibitors, and beyond

Immune therapies are rapidly becoming a center point for the treatment of many malignancies and for hematological malignancies in particular, including Hodgkin and non-Hodgkin lymphoma. These treatments not only affect the malignant cells but also activate host immune cells to target malignant cells. In lymphomas, these therapies are complicated by the fact that the malignant cell is part of the immune system and nonspecific general immune activation may stimulate the malignant clone. Instead, careful activation of the immune system to specifically stimulate effector cells to target the malignant clone is the goal of immunological therapies. The role of immune pathways in lymphoma is complex and natural mechanisms to regulate these processes are present.¹  These mechanisms include immunological checkpoints that provide a negative inhibitory feedback loop to downregulate the immune response and protect the patient from autoimmunity or significant tissue damage from an overactive immune response.^2-4  These mechanisms are often exploited by malignant cells, which may overexpress immunosuppressive ligands or may secrete immunosuppressive cytokines to dampen an effective antitumor immune response.^5-9  Potential therapies to active the immune system need to overcome these suppressive elements to allow T cells to remain active and effectively lyse malignant cells. Currently, therapeutic mechanisms to either activate immune cells or prevent suppression of immune cells are in clinical testing and very promising results have been seen in a number of clinical trials.

T-cell activation is the first step in an effective immune response. T-cell activation requires presentation of antigens by antigen-presenting cells, allowing T cells to react with antigen-presenting cells through the T-cell receptor engagement with major histocompatibility complex molecules. This interaction allows for antigens foreign to the host to be presented to the immune system for an effective immune response. In lymphoma, this process is complicated by the fact that the tumor-associated antigens are self-proteins and are therefore only weakly immunogenic. An additional second activating signal is then required to allow T cells to become activated and target the presented antigen.¹⁰  This second signal is commonly delivered through CD28, but additional activating signals can be provided through CD27, the inducible T-cell costimulator (CD278), tumor necrosis factor receptor superfamily member 4 (TNFRSF4/OX40), tumor necrosis factor receptor superfamily member 9 (TNFRSF9/CD137/4-1BB), or tumor necrosis factor receptor superfamily member 18 (TNFRSF18/CD357, which is also called the glucocorticoid-induced tumor necrosis factor receptor).¹¹  At the same time, the cell may also receive a negative signal to suppress the immune response. Typically, activated T cells begin to upregulate receptors to receive negative signals to prevent the immune system from overreacting. These negative signals are typically delivered through programmed cell death 1 (PD-1/CD279), cytotoxic T-lymphocyte–associated protein 4 (CTLA4/CD152), lymphocyte-activating 3 (LAG3), CD160, or B- and T-lymphocyte–associated protein (BTLA).¹¹  An effective immune response requires an adequate balance between both positive and negative costimulatory signals.

In lymphoma, the balance between activation and suppression of the immune response is often skewed by the malignant cell toward profound immune suppression.¹²  The malignant cells commonly promote suppressive mechanisms in the tumor microenvironment by recruiting regulatory T (Treg) cells to the tumor or promoting the secretion of multiple immunosuppressive ligands, including transforming growth factor-β.⁶  Transforming growth factor-β, expressed on the surface of the malignant B cell, suppresses effector T-cell function, inducing T cells to differentiate into Treg cells. Treg cells in the tumor microenvironment constitutively express CTLA4 and tumor engagement of the CTLA4 pathway promotes the suppressive function of Treg cells, thereby damping an effective antitumor immune response. Another mechanism of immune suppression is the upregulation of PD-1 ligands, including PD-L1 (CD274/B7-H1) and PD-L2 (CD273/B7-DC), in the tumor.^7,9,13  Upregulation of these ligands has been seen not only on malignant cells but also on other cells such as macrophages within the microenvironment. Furthermore, copy number variation and genetic alterations at chromosome 9p24.1 can result in overexpression of PD-L1 and PD-L2.⁷  This is very commonly seen in Hodgkin lymphoma and results in very high expression of PD-L1 or PD-L2 on the cell surface in these cases, thereby protecting the malignant cell in Hodgkin lymphoma from T-cell–mediated killing. Activated T cells within the tumor microenvironment commonly express PD-1 and PD-1–positive T cells are thereby substantially inhibited and prevented from targeting the malignant cell. These suppressive mechanisms, however, offer an opportunity to modulate the immune response.^2-4,10  Blocking suppressive signals through inhibitory receptors, or providing agonistic signals through stimulatory receptors, is one way in which the suppressed and exhausted immune response can be reactivated.

Various mechanisms to activate suppressed or anergic T cells have been tested in clinical trials. These mechanisms include the use of agonistic antibodies that target costimulatory molecules such as CD27, CD137, or CD40.^14-18  A potential challenge of this approach has been that cells activated by these agonistic antibodies may still subsequently be suppressed by ligands including PD-L1. Therefore, simply activating T cells without preventing suppressive signals may not result in dramatic clinical responses. This is found to be true in a variety of clinical trials. A clinical trial utilizing an agonistic anti-CD27 antibody (varlilumab/CDX1127) was performed in patients with hematological malignancies but included mainly patients with lymphoma.¹⁵  In this study, 24 patients with lymphoma were treated and evidence of immunological activity was seen, including decreased numbers of circulating Treg cells, an increase in soluble CD27 levels, and an induction in proinflammatory cytokines. Antilymphoma activity, however, was modest, with only 1 complete response seen in a patient with Hodgkin lymphoma but the remainder of patients having stable disease as their best response. Similarly, an agnostic anti-CD40 antibody, dacetuzumab, was tested in both phase 1 and phase 2 clinical trials.^16,17  In the initial phase 1 trial, 50 patients with B-cell lymphoma were treated. Overall, the therapy was well tolerated although 2 dose-limiting toxicities were seen, which included liver toxicity with elevated liver function tests and conjunctivitis with transient vision loss. Although tumor decreases were noted in approximately one-third of patients, only 6 objective responses were seen (1 complete response and 5 partial responses).¹⁶  A subsequent phase 2 trial of dacetuzumab as a single agent was performed in patients with diffuse large B-cell lymphoma and very similar results were seen.¹⁷  Approximately one-third of patients had some degree of benefit (the disease control rate was 37%) but the overall response rate was only 9%. On the basis of these results, further development of this antibody has not been pursued. Other therapeutic agents activating stimulatory receptors, including antibodies targeting CD137 (4-1BB), have also yielded modest results.¹⁸  Increases in soluble 4-1BB, memory T cells, and activated natural killer cells were observed. The overall response rate was modest at 21%, although an overall response rate of 29% was seen in patients who were refractory to rituximab; there were 2 complete responses (both in patients with follicular lymphoma). Overall, however, results from trials using agonistic antibodies suggest that although these treatments are promising, it remains challenging to overcome the suppressive elements in the tumor microenvironment. Simply activating intratumoral T cells may be insufficient in view of the multiple inhibitory signals that these cells receive within the lymphoma environment.

Similar to the results seen with agonistic antibodies, blocking inhibitory signals with monoclonal antibodies has provided results that are encouraging but vary by disease and by the antibody used. Simply blocking inhibitory signals may also be insufficient in many diseases. Inhibition of CTLA4 signaling, using anti-CTLA4 antibodies, has a profound effect on Treg cell function and inhibits FOXP3+ Treg cells from suppressing intratumoral effector T cells. Similarly, blocking PD-1 signaling with anti–PD-1 antibodies protected effector T cells at the interface between the tumor and the immune system from inhibition by PD-L1 and PD-L2 and promoted a persistent, effective antitumor response. Promising clinical results have been seen both with anti-CTLA4 therapy as well as with PD-1 blockade in a variety of malignancies. In lymphoma, the anti-CTLA antibody ipilimumab has been tested in patients with relapsed and refractory disease.¹⁹  Eighteen patients were treated and 2 patients responded (1 with a complete response lasting >3 years and 1 with a partial response lasting 19 months). T-cell activation was noted in one-third of patients treated, but not all patients with T-cell immune activation showed clinical benefit.

Results with PD-1 blockade, although particularly promising in Hodgkin lymphoma, have varied in other diseases and are more modest in other histologies. In view of the overexpression of PD-1 ligands within the tumor microenvironment, blockade of PD-1 signaling appears to inhibit a signaling pathway that is critical to the tumor-induced suppression of the immune response. Clinical results, particularly in Hodgkin lymphoma, have been quite dramatic. In the initial phase 1 clinical trials of the anti–PD-1 antibodies pembrolizumab and nivolumab, very high response rates were seen in patients who had experienced multiple relapses.^20,21  The phase 1 clinical trial of pembrolizumab showed an overall response rate of 65% in patients with Hodgkin lymphoma and most of the responses (70%) lasted >24 weeks.²⁰  Similarly, the phase 1 trial of nivolumab showed an overall response rate of 87% in patients with Hodgkin lymphoma.²¹  However, the response rates were quite discrepant between patients with Hodgkin lymphoma and patients with other lymphomas. The response rates in other lymphoma subtypes were far more modest and ranged between 15% and 40%. In patients with multiple myeloma, the best response to nivolumab was stable disease.²² 

Additional confirmatory phase 2 studies have been performed in Hodgkin lymphoma as well as other lymphoma subtypes.^23,24  The phase 2 trial of pembrolizumab in Hodgkin lymphoma (KEYNOTE-087) confirmed an overall response rate of 69% in patients with relapsed and refractory disease who had also previously received brentuximab vedotin. Of note, responses seen with pembrolizumab were similar in patients with primary refractory disease to those of patients with relapsed disease after multiple lines of therapy.²³  In the phase 2 trial of nivolumab, similar promising results were seen in patients with Hodgkin lymphoma post-transplantation who had either not responded to brentuximab vedotin or had not yet received the drug; responses were seen in two-thirds of these patients (66%).²⁴  Most encouragingly, the responses have been durable, with long-term follow-up confirming that patients could continue to tolerate treatment and continued to benefit from therapy.

In an attempt to use both CTLA4 blockade and PD-1 blockade in combination, a phase 1 clinical trial in lymphoma and multiple myeloma using ipilimumab and nivolumab has been completed.²⁵  In the Hodgkin lymphoma cohort, the overall response rate was 74%. Of note, however, the complete response rate was not different from that of patients treated with nivolumab alone. The response rate was only 20% in patients with B-cell lymphoma and only 9% in patients with T-cell lymphoma. The best response seen in patients with myeloma was stable disease. These results do not suggest synergy for dual immune checkpoint blockade in lymphoid malignancies.

The efficacy of immune modulation and immune checkpoint inhibition is promising and future studies will need to test the efficacy of these antibodies in combination with other active agents. Multiple clinical trials utilizing a variety of different approaches are currently in progress, and many of these trials are combining various immunological strategies. These strategies include combining >1 immune checkpoint blocking antibody, using immune checkpoint blockade in combination with antibodies that induce immune activation, or using immune checkpoint blockade in combination with other immunotherapies such as monoclonal antibodies, vaccines, or viral therapies. An additional approach being tested includes using immune checkpoint blockade in combination with chimeric antigen receptor T cells to prevent suppression of the activated chimeric antigen receptor T cells. Immune checkpoint blockade is also being tested in combination with small molecule inhibitors, including Bruton’s tyrosine kinase inhibitors, phosphoinositide 3-kinase inhibitors, or immune-modulatory drugs (e.g., lenalidomide) that result in a more favorable immune environment. Furthermore, many of these agents are being used with standard chemotherapy or in combination with antibody drug conjugates such as brentuximab vedotin. As one can tell, the future of immunotherapy, although promising, is complicated by the fact that many combinations could be tested. It will require a greater understanding of the underlying biology, thoughtful clinical trial design, and the use of relevant biomarkers to identify the most effective immunological combinations for patients with lymphoma.

Stephen M. Ansell, Mayo Clinic, 200 First St SW, Rochester, MN 55902; e-mail: ansell.stephen@mayo.edu.