Immune Checkpoint Blockade for the Treatment of Hodgkin Lymphoma

Introduction

Classical Hodgkin lymphoma (cHL) is largely curable with conventional chemotherapy, with an estimated 5-year survival of approximately 88%.¹ Treatment algorithms in cHL depend in part on the clinical stage at the time of diagnosis. In early-stage disease (stage I/II), using the National Comprehensive Cancer Network guidelines, patients are separated into favorable or unfavorable characteristics, depending on several factors including erythrocyte sedimentation rate, bulky disease (greater than one-third of the thoracic width or ≥10 cm in diameter) and number of nodal sites.² Most early-stage patients can be cured by reduced cycles (two to four cycles) of chemotherapy combined with involved field radiation, though the optimal number of cycles and the need for combined modality therapy remain controversial. The chemotherapy regimen used in trials for early-stage disease is mostly ABVD (doxorubicin, bleomycin, vinblastine, and dacarbazine) with possible escalation to BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, and prednisone) if the interim positron emission tomography noted residual or progressive disease. The research focus for the treatment of early-stage disease cHL is to minimize therapy to mitigate the secondary cancer and cardiovascular risks.^3,4

For advanced disease, frontline therapy for cHL has changed over the years. In addition to ABVD, BV-AVD (brentuximab vedotin, doxorubicin, vinblastine, and dacarbazine) has become an option, based on the results of the ECHLON-1 trial.^5,6 Brentuximab vedotin (BV), which is an antibody drug conjugate, replaced bleomycin to reduce the risk of significant lung injury. A more intensive chemotherapy regimen, escalated BEACOPP, is often used in Germany, with similar results to ABVD. Despite this high response rate, a significant number of patients will not respond to frontline therapy or will have relapsed disease. Options for relapsed/refractory cHL patients include high-dose salvage intensive chemotherapy followed by high-dose chemotherapy and autologous hematopoietic stem cell transplantation (HSCT). The optimum second-line salvage chemotherapy regimen is not standardized and is still the subject of debate and investigation. Castagna et al detailed and categorized the second-line chemotherapy options into 1) platinum-based regimens such as ifosfamide, carboplatin, etoposide (ICE),⁷ with an overall response rate (ORR) of 72% and a complete response (CR) rate of 26%; 2) gemcitabine-based regimens such as gemcitabine, vinorelbine, pegylated liposomal doxorubicin (GVD),⁸ with an ORR of 70% and a CR of 19%; and 3) others, namely non-platinum or gemcitabine regimens.⁹ Our institution often uses ICE as salvage chemotherapy as it has the benefit of stem cell collection between cycles.

The unique biology of cHL can drive novel treatments. The malignant cell, the Reed–Sternberg cell (RSC), makes up only a small portion of the involved lymph nodes, while the majority are reactive T cells and other immune cells. Further evaluation of the RSCs found overexpression of PD-L1 and PD-L2 due to genetic amplification of the 9p24.1 locus.¹⁰

The overexpression of the PD-1 ligand leads to a T-cell-exhausted tumor microenvironment.¹¹ The RSCs with copy number alterations of 9p24.1/CD274(PD-L1)/PDCD1LG2(PD-L2) are surrounded by PD-L1-positive tumor-associated macrophages.¹² This is often described as a “castle and moat” model as the RSCs are surrounded by inactivated innate immune cells to protect them from the immune system, specifically, activated T cells. Importantly, alterations of PD-L1 and PD-L2 have correlated with reduced progression-free survival (PFS) when treated with standard chemotherapy regimens.¹³ Also, these RSCs had overall high mutational burden and microsatellite instability (MSI)-associated hypermutation, biological markers that correlate with the response to checkpoint inhibitor (CPI) therapy in other malignancies.¹⁴ Therefore, anti-PD-1/PD-L1 CPI therapy is a rational approach to the treatment of cHL.^15–19 The success of PD-1 blockade suggests that some of these T cells are reversibly exhausted. RSCs with β₂-microglobulin/major histocompatibility complex (MHC) class I loss are noted to have more durable responses.¹¹ In addition to MHC class I, newly diagnosed cHLs had a higher prevalence of NF-kB genetic alterations. Wienand et al also found a mutational signature of spontaneous deamination of cytosine-phosphate-guanines, apolipoprotein B mRNA editing catalytic polypeptide-like, activation-induced cytidine deaminase, and MSI-associated hypermutation.¹⁴

Clinical trial data have shown that anti-PD-1/PD-L1 monotherapy is an effective option for patients who have failed first-line chemotherapy. Thus far, two main CPIs – nivolumab and pembrolizumab, both of which are anti-PD-1 monoclonal antibodies – have been approved by the FDA for relapsed/refractory cHL. In this review, we will discuss the clinical trials for CPI in cHL, organized by the sequence of treatments.

Third or Greater Line of Therapy

In a 2015 study in which 23 patients with previously heavily treated (more than two lines of therapy) relapsed/refractory cHL were administered nivolumab every 2 weeks until they had evidence of CR, tumor progression, or severe toxicity, 80% of patients demonstrated an objective response, including 17% with CR and 70% with partial response (PR), with a rate of PFS of 86% at 24 weeks.²⁰ A multicenter phase II study of nivolumab monotherapy in 17 Japanese patients with relapsed/refractory cHL who had all previously received BV demonstrated an objective response rate of 81.3% among the 16 patients included in the efficacy analysis. Among these 16 patients, four had complete remission and nine had partial remission.²¹ In another phase II study from 34 hospitals across Europe and North America, in which nivolumab monotherapy was administered to 80 patients with recurrent cHL who had progressed after autologous stem cell transplantation (ASCT) and either failed to respond to or relapsed following BV administration, 53 patients (66.3%, 95% CI 54.8–76.4%) demonstrated an objective response based on review from an independent radiological review committee (IRRC).²²

The phase II CheckMate 205 trial also examined the efficacy of nivolumab in relapsed/refractory cHL, but further separated patients into three groups based on treatment history: BV-naïve, autologous hematopoietic cell transplantation (auto-HCT) followed by BV, and BV received before and/or after auto-HCT.²³ In total, 243 patients were treated with nivolumab and the overall objective response rate was 69% (95% CI 63–75%), ranging from 65% to 73% in each cohort. The median response duration among all 243 patients was 16.6 months (95% CI 12.2–20.3 months) and median PFS was 14.7 months (95% CI 11.3–18.5 months). This trial, along with the 2015 trial,²⁰ led to the FDA approval of nivolumab single agent for the treatment of relapsed cHL which had progressed after autologous HSCT and post-transplantation BV. Notably, this was the first FDA application for a PD-1 inhibitor in hematologic malignancies.

Around the same time, in 2017, the FDA also approved pembrolizumab for adults and children with cHL refractory to or which has relapsed after at least three prior therapies based on the initial results from KEYNOTE-087.²⁴ The patients were divided into three cohorts: 1) patients with progression after ASCT and subsequent BV administration (n=69); 2) patients with progression following salvage chemotherapy and BV who were not candidates for ASCT (n=81); and 3) patients with progression after ASCT without exposure to BV (n=60). Cohort 1 had an ORR of 73.9% and a CR rate of 21.7%. Cohort 2 had an ORR of 64.2% and a CR rate of 24.7%. Cohort 3 had an ORR of 70% and a CR rate of 20%. For all the patients included, the ORR was 69% and the CR rate was 22.4%. A 5-year follow-up of KEYNOTE-087 reaffirmed the benefit of pembrolizumab in both BV-naïve patients and those with previous exposure.²⁵ Out of the 210 patients evaluated, 46 patients completed 2 years of pembrolizumab. The ORR was 71%, with a CR rate of 27.6%. The ORR increased to 84.1% for cohort 1 and 67.9% for cohort 2, and there was a slight decrease to 68.3% for cohort 3 at the data cut-off. Furthermore, the median PFS was 56.5 months and the 5-year PFS rate was 44.3%, while the median overall survival (OS) was not reached and the 5-year OS was impressive, at 82.8%.

Before that trial, the phase Ib study KEYNOTE-013 helped open the door for the use of pembrolizumab in this population.^23,26 The study consisted of 31 patients with relapsed/refractory cHL, all of whom had disease progression while on or after treatment with BV, where 55% had more than four previous lines of therapy, and 71% had relapsed following ASCT. Among this group, the ORR was 65% (90% CI 48–79%), with complete remission in 16% (90% CI 7–31%) and partial remission in 48% of patients. PFS was 69% at 24 weeks and 46% at 52 weeks.

Data from clinical trials have shown that anti-PD-1 therapies in combination with other agents are effective in relapsed/refractory cHL. A phase I trial investigated pembrolizumab plus vorinostat in patients with DLBCL, follicular lymphoma, and cHL who had failed one or more prior lines of treatment (with a median of four) and were not candidates for stem cell transplantation.²⁷ In total, 30 patients were enrolled in the trial, 12 of whom had cHL. Among these 12, 11 had previously received BV, seven had been treated previously with anti-PD-1 therapy, and three were refractory to prior anti-PD-1 therapy. Ultimately, among the nine evaluable cHL patients, four achieved CR and five had a PR. Perhaps most notably, these data suggest that patients with refractory disease to anti-PD-1 therapy should not necessarily be excluded from receiving further treatment with anti-PD-1 treatment, particularly when administered in combination with other agents.

A 2020 phase Ib dose-escalating study further explored the benefits of combination therapy in 30 patients with relapsed/refractory CHL.²⁸ In this trial, patients received pembrolizumab in combination with AFM13, a bispecific tetravalent antibody that targets CD30 on tumor cells and CD16A on natural killer cells and macrophages to induce tumor cell killing. In this study, patients who had received prior anti-PD-1, anti-PD-L1, or anti-PD-1-L2 treatment were excluded. Overall, the study population had an 83% objective response rate, with 37% showing a complete metabolic response and 47% a partial metabolic response. Among those who received the highest treatment dose, the objective response rate was 88%.

Bolstered by the encouraging data above, two clinical trials investigating anti-PD-1 therapy in combination with radiotherapy for patients with relapsed/refractory cHL are ongoing. NCT04419441 is exploring whether radiotherapy will improve the rate of complete remission in patients receiving anti-PD-L1 therapy. NCT03480334 has the same hypothesis, but its study population is comprised specifically of patients who have recently progressed on anti-PD-1 therapy.

CPI for Maintenance Strategy

A multicohort phase II study by Armand et al evaluated pembrolizumab as a maintenance strategy post-ASCT. These patients consisted of a particularly high-risk group, with 90% of the 30 patients deemed high risk by clinical criteria. The maintenance pembrolizumab therapy lasted for eight cycles every 3 weeks. Among 28 evaluable patients, the rate of PFS was 82% at 18 months, satisfying the investigators’ hypothesis that pembrolizumab would improve PFS from 60% to 80%.²⁹ Similarly, maintenance nivolumab (6 months) post-ASCT was also tested in patients with Hodgkin lymphoma (HL).³⁰ The data are immature, but out of the 37 patients enrolled on this trial, the 6-month or end of therapy PFS is 92.1% and the 12-month OS is 100%. The median PFS and OS have not been reached, with a median follow-up of 9.2 months. Only four discontinued the nivolumab owing to adverse events and two patients progressed while on the nivolumab. Both of these trials suggest that maintenance CPIs are tolerable, and mature data are needed for them to become the standard of care.

Second-Line Therapy

Several studies have also demonstrated the effectiveness of checkpoint inhibition in combination with other agents as second-line therapy in cHL. In one phase I/II study, a total of 91 patients with relapsed/refractory disease were treated with combination BV plus nivolumab. Patients in this study were divided into two cohorts: one group that received BV and nivolumab in a staggered schedule, and a second group that received therapy concurrently. Therapy proved effective in both cohorts; the rate of CR for treated patients was 67% and the objective response rate was 85%.^31,32 Given these encouraging results, BV plus nivolumab is now approved for use as second-line therapy in relapsed/refractory cHL, establishing an alternative to chemotherapy for consolidation prior to ASCT.

Another trial addressed the question of whether immunotherapy in combination with chemotherapy could safely be tolerated prior to ASCT, without adversely affecting peripheral blood progenitor cell mobilization and engraftment. Patients were treated with a regimen of pembrolizumab plus ifosfamide, carboplatin, and etoposide (ICE). Among the 40 patients with relapsed/refractory cHL recruited, 23 were evaluable. All but one patient in this group had successful stem cell mobilization and harvest (this patient had a severe allergic response to filgrastim and instead had a bone marrow harvest), and all patients who underwent stem cell reinfusion successfully engrafted, with a median time of 11 days for absolute neutrophil recovery and 12 days for platelet recovery. Notably, the therapy was well tolerated in this group, with no reports of the more serious adverse immune CPI toxicities of pneumonitis, colitis, hepatitis, or endocrinopathies.³³ Therefore, these data further support the feasibility of immunotherapy as second-line therapy in relapsed/refractory cHL prior to ASCT.

The randomized, open-label phase III trial entitled KEYNOTE-204 demonstrated that pembrolizumab monotherapy was superior to BV in patients with relapsed/refractory cHL.³⁴ This trial was a head-to-head comparison of pembrolizumab and BV, and included both patients with and those without prior BV exposure. All participants were either post-ASCT or ineligible for transplant. In a head-to-head comparison, patients who received pembrolizumab had higher rates of PFS compared to those who had received BV (HR 0.65, 95% CI 0.46–0.85), with PFS rates of 53.9% in patients given pembrolizumab versus 35.6% in the BV group. Pembrolizumab was superior in all of the study’s subgroups, including those who had not previously undergone ASCT, those with primary refractory disease, prior exposure to BV, and no prior exposure to BV. This study led to the extended FDA approval in 2020 of pembrolizumab for adult cHL patients as second-line therapy and pediatric patients as third-line therapy.

In a similar vein, a subsequent phase II study investigated whether pembrolizumab plus gemcitabine, vinorelbine, and liposomal doxorubicin (GVD) would be effective as second-line therapy prior to ASCT. Thirty-seven patients were recruited for the study, and 34 were eligible for analysis. Among this group, 31 achieved a CR after two cycles and three achieved a PR. One of these three partial responders ultimately achieved CR after four cycles of therapy. At the time of publication, 32 patients have undergone ASCT, one is awaiting transplantation, and one declined transplantation and instead opted for maintenance pembrolizumab therapy. All patients were still in remission at the time of publication.³⁵ No dose-limiting toxicities were noted in the safety phase of the trial. Grade 3 toxicities include elevated liver enzymes (four patients), neutropenia (four patients), oral mucositis (two patients), rash (one patient), and hyperthyroidism (one patient).

One study pursued a unique angle by having a cohort of patients who received a combination of immunotherapy agents: nivolumab and ipilimumab. Ipilimumab is a monoclonal antibody targeted against CTLA-4, an immune checkpoint receptor on T cells.³⁶ Using multiplex immunofluorescence microscopy with digital image analysis, Patel et al found that the tumor microenvironment in cHL is enriched for non-T-regulatory, cytotoxic T-lymphocyte-associated protein-4 (CTLA-4)-positive T cells that outnumber PD-1-positive T cells.³⁷ Unlike the tumor-associated macrophages that express PD-L1, T cells that surround the RSCs are more often positive for CTLA-4 than PD-1. The patients included in this trial had wide variance in prior treatments; some had as few as one prior line of therapy while others had received nine lines of therapy (median of 2). Of the 64 patients enrolled in this phase I trial, 61 were evaluable. These patients were separated into three cohorts: one group received BV and ipilimumab; a second received BV and nivolumab; and the final group received BV, ipilimumab, and nivolumab. Each of these cohorts demonstrated robust objective response rates and CR rates: 76% and 57%, respectively, in the BV/ipilimumab group; 88% and 61%, respectively, in the BV/nivolumab group; and 82% and 73%, respectively, in the BV/nivolumab/ipilimumab group.

The most significant toxicities occurred in the BV/nivolumab and BV/nivolumab/ipilimumab groups, each of which had one grade 5 pneumonitis death and five dose-limiting toxicities between the two of them. Common grade 1–2 toxicities included fatigue, elevation of liver transaminases, peripheral neuropathy, and diarrhea. For grade 3 or 4 adverse events, 10 (43%) of the ipilimumab group, three (16%) of the nivolumab group, and 11 (50%) of the triple-therapy patients experienced high-grade toxicities.

Frontline Therapy

More recently, Allen et al pursued CPI therapy as a component of frontline therapy, given its effectiveness in relapsed/refractory disease.³⁸ In this multicenter phase II study, 30 patients with untreated, early, unfavorable, or advanced-stage disease were treated with pembrolizumab and doxorubicin, vinblastine, and dacarbazine (AVD) chemotherapy. These patients first received three cycles of pembrolizumab monotherapy, followed by AVD for four to six cycles depending on the stage and bulk of disease. After pembrolizumab monotherapy, 11 patients had a complete metabolic response, and an additional seven patients had >90% reduction in metabolic tumor volume based on PET/CT. All patients in the study attained a complete metabolic response after two cycles of AVD, and at median follow-up of 22.5 months there were no cases of disease progression, death, or further therapy. Furthermore, the therapy was well tolerated, with the most notable adverse effects being six patients with grade 1 rash, four patients with grade 2 transfusion reactions, one patient with reversible grade 4 transaminitis, and one patient with reversible Bell’s palsy.³⁹

This trial by Allen et al used a lead-in or sequential strategy of CPIs prior to chemotherapy, with high response rates. However, the question of concurrent versus sequential use of CPIs with chemotherapy remains. The German Hodgkin Study Group led by Bröckelmann et al performed a phase II study looking at early-stage unfavorable disease patients, with a 1:1 assignment to either concurrent treatment with four cycles of nivolumab and AVD or sequential treatment with four doses of nivolumab, two cycles of N-AVD, and two cycles of AVD, followed by involved-site radiotherapy.⁴⁰ There were 109 patients included in this study, with a 90% CR rate in the concurrent arm and 94% CR in the sequential arm. The one-year PFS rates were 100% and 98% for the concurrent arm and sequential arm, respectively. These results suggest that both options are reasonable, but the single-agent nivolumab (four cycles) lead-in had a 96% response rate, which, along with the pembrolizumab data from Allen et al,³⁹ will require further investigation to minimize toxicities.

The CheckMate 205 trial referenced above (in the section “Third or Greater Line of Therapy”) also included a cohort of patients who received immunotherapy with chemotherapy as frontline therapy for newly diagnosed cHL.^41,42 In this study, 51 patients with stage IIB with unfavorable risk, III, or IV cHL underwent treatment. The investigators planned for four doses of nivolumab monotherapy followed by nivolumab plus doxorubicin, vinblastine, and dacarbazine combination therapy for six cycles. After two combination cycles, the rate of complete remission was 51% (and 71% complete metabolic response) via an independent review committee and 71% per investigator. On a more recent trial result update, the ORR was 84% (95% CI 71–93%), including 67% of the patients reaching a CR.⁴³ At 21-month follow-up, the rate of PFS was 83% per investigator. Furthermore, Roemer et al found that responses in this trial correlated with a higher level of 9p24.1 copy gain and increased PD-L1 expression on RSCs.⁴⁴ Surprisingly, major histocompatibility complex (MHC) class II on the RSCs was predictive of the responses, whereas β₂-microglobulin/MHC class I was not.

Other Checkpoint Inhibitors

Though other anti-PD-1 immunotherapy agents have not been approved by the FDA for use in relapsed/refractory cHL, studies indicate that sintilimab, tislelizumab, and camrelizumab are effective in these patients. Sintilimab, a fully humanized IgG₄ monoclonal antibody toward PD-1, was administered to patients taking part in ORIENT-1, a phase II trial consisting of 96 patients from 18 hospitals in China who had relapsed/refractory disease following two or more lines of therapy.⁴⁵ In total, 74 patients among the 92 analyzed (80.4%, 95% CI 70.9–88.0%) demonstrated an objective response via review from an independent committee. In another phase II study among 70 Chinese patients with relapsed/refractory HL who received tislelizumab, a humanized anti-PD-1 antibody that is designed to bind less to Fc gamma receptors, 61 (87.1%) of the patients assessed by an independent review committee had an objective response, with a 9-month PFS rate of 74.5%.⁴⁶

Camrelizumab, another anti-PD-1 antibody, was administered to 75 Chinese patients with relapsed/refractory cHL after ASCT or who had received two or more lines of chemotherapy, among whom 57 (76.0%, 95% CI 64.7–85.1%) achieved an objective response based on review by an independent review committee.⁴⁷ A two-arm, open-label phase II study consisting of 86 relapsed/refractory cHL patients who had already undergone at least two lines of therapy reaffirmed the efficacy of camrelizumab, though it also suggested that co-administration of decitabine improves the clinical response. Among the patients who were naïve to anti-PD-1 therapy, six of 19 (32%) achieved complete remission with camrelizumab monotherapy, compared with 30 of 42 patients (71%) who received decitabine plus camrelizumab. In patients who had previously been treated with anti-PD-1 therapy, 28% achieved complete remission, while PR was seen in 24% of these patients, suggesting that co-administration of decitabine could bypass PD-1 resistance in patients who had previously been treated with anti-PD-1 therapy.⁴⁸

JAVELIN Hodgkins, a phase Ib trial for patients with relapsed/refractory cHL, showed that avelumab, an anti-PD-L1 monoclonal antibody, is also effective in this patient population.⁴⁹ In this study of 31 heavily pretreated patients, nine had received three prior treatments and 20 had received four or more treatments. The objective response rate among all patients was 41.9%, with a CR rate of 19.4%, and the objective response rate among patients who had previously received an allogeneic hematopoietic stem cell transplant was even more significant, at 55.6%.

Conclusion

CPIs are destined to be active agents in HL owing to the biological nature of 9p24.1 amplification. Both pembrolizumab and nivolumab have been successful in treating cHL patients as single agents or as combination therapy with brentuximab vedotin or chemotherapy in the third line. Growing evidence suggests that CPIs in the first or second line may improve outcomes. Additional clinical trials are needed and are ongoing to evaluate the best time to use CPIs in these patient populations (Table 1). CPIs in combination with newer targeting agents will also be a point of further trials, especially since the PD-1/PD-2 pathway is not the only method of immune evasion by RSCs.⁵⁰ Furthermore, a better understanding of the systemic and tumor microenvironment will help us predict which patients will respond to CPI therapy.⁵¹

Table 1 Active or Recently Completed Clinical Trials Using Checkpoint Inhibitors (CPIs) in Hodgkin Lymphoma