In this issue of Blood, Allen and colleagues present encouraging results from one of the first studies evaluating programmed death 1 (PD-1) blockade in the front-line setting for classical Hodgkin lymphoma (cHL).1 In the study by Allen et al, patients with early unfavorable and advanced-stage cHL received 3 cycles of pembrolizumab followed by 4 to 6 cycles of doxorubicin, vinblastine, and dacarbazine (AVD). They enrolled a considerably high number of patients with bulky disease (33% with >10-cm masses), and notable findings included complete response rate after single-agent pembrolizumab of 37%, with another 25% experiencing metabolic tumor volume reduction by >90% (see figure). Furthermore, toxicity was low, and all patients avoided radiation therapy.
Based on the effectiveness and tolerability of PD-1 blockade in relapsed and refractory disease, there has been tremendous interest in evaluating nivolumab and pembrolizumab earlier on in the treatment course for cHL.2,3 Although most patients are cured with doxorubicin, bleomycin, vinblastine, and dacarbazine–based therapy in the front-line setting, there remains considerable room for improvement. First, up to 10% of early-stage patients and up to 25% of advanced-stage patients will require second-line therapy.4,5 Furthermore, treatment-related toxicities, including secondary malignancies and cardiovascular disease due to either chemotherapy or radiation, remain an important concern.6
The desire to evaluate PD-1 blockade in untreated cHL patients seems straightforward but leads to many difficult questions of safety, rationale, and treatment schedule. In particular, will treatment-naive patients experience more toxicity with PD-1 blockade than the heavily treated patients evaluated in the initial studies with pembrolizumab and nivolumab? Even if toxicity rates are similar, we know that immune-mediated adverse events can occasionally be severe. Therefore, is it appropriate to expose all patients with cHL to PD-1 blockade, or should it be reserved for certain patients? Finally, it is unknown whether different chemotherapy agents will mitigate, enhance, or simply add to the efficacy of PD-1 blockade; therefore, should PD-1 blockade and chemotherapy be given concurrently? Sequentially? Or does it depend on the agent?
Allen and colleagues chose to study sequential therapy, which proved to be quite encouraging given that with a median follow-up of 22.5 months (range, 14 to 30.6 months), all patients remain in remission. Furthermore, no patient experienced treatment-related toxicity leading to treatment discontinuation.1 Bröckelmann and colleagues from the German Hodgkin Study Group chose to study both sequential and concomitant therapy in a randomized phase 2 study with nivolumab and AVD.7 In their study, early-stage, unfavorable cHL patients were randomized to receive either 4 cycles of nivolumab plus AVD followed by involved-site radiation (ISRT) or 4 doses of single-agent nivolumab followed by 2 cycles of nivolumab plus AVD, 2 cycles of AVD, and then ISRT. Both treatment strategies were effective, as 94% and 90% of patients achieved complete response following sequential and concomitant therapy, respectively.
Furthermore, 12-month progression-free survival was 98% and 100% for the 2 treatment schedules. In contrast to Allen’s study, Bröckelmann and colleagues’ study reported treatment discontinuation in 5 patients due to significant adverse events, including 1 event each of encephalitis, proctosigmoiditis, pneumonitis, polyneuritis, and infusion-related reaction. All these events resolved with appropriate therapy except for the proctosigmoiditis, which at last follow-up was still present but had improved from grade 4 to grade 1. In Allen’s study, the fact that no patients required treatment discontinuation may have been simply due to the study size (30 patients compared with 109 on Bröckelmann’s study). Interestingly, Bröckelmann noted that most of the serious adverse events on their study were attributed to the combination of nivolumab plus AVD rather than either AVD or nivolumab alone. This suggests that perhaps less toxicity was observed on Allen’s study because AVD was not given concomitantly with PD-1 blockade.
Extensive work is needed to determine whether incorporation of PD-1 blockade into front-line treatment truly improves on our current standard regarding efficacy and safety. The ongoing US intergroup phase 3 study evaluating brentuximab vedotin or nivolumab in combination with AVD plans to enroll ∼1000 patients and therefore will undoubtedly provide valuable information regarding the feasibility and appropriateness of PD-1 blockade in a large, unselected patient population (www.clinicaltrials.gov; #NCT03907488). Planned correlative studies from Allen’s study as well the intergroup study and others will aid in determining which patients are likely to benefit from upfront PD-1 blockade and which are likely to experience serious toxicity. It is hoped we will learn whether PD-1 blockade should be offered to all patients or reserved for select individuals, such as those with a biomarker that predicts PD-1 blockade sensitivity (for example, 9p24.1 amplification) or a factor that predicts poor chemosensitivity (such as early interim PET response, reduction in metabolic tumor volume, or baseline tumor bulk).8 In addition, we will learn how to best assess response to PD-1 blockade. In Allen’s study, one-third of patients had tumor bulk and, although the study is small, the promising early results for this unfavorable patient population suggest that perhaps PD-1 blockade can allow patients to avoid radiation therapy without compromising tumor control. Furthermore, the study shows a dramatic reduction in metabolic tumor volume following single-agent pembrolizumab and demonstrates the potential for this novel method in assessing response to PD-1 blockade. Certainly, larger studies with longer follow-up are needed to study this further; however, Allen’s study represents a very promising early step toward identifying an optimal treatment schedule and patient population for incorporating PD-1 into the front-line setting.
Conflict-of-interest disclosure: The author has received research support from Miragen, Seattle Genetics, Merck, Bristol-Myers Squibb, and Incyte; and honoraria from Imbrium Therapeutics LP, Merck, and Seattle Genetics.