Blockade of programmed death protein 1 (PD-1), known as “checkpoint blockade,” with monoclonal antibodies such as nivolumab and pembrolizumab has been shown to be particularly effective in patients with relapsed/refractory (R/R) classical Hodgkin lymphoma (cHL), including those who have relapsed following autologous hematopoietic stem cell transplantation (HSCT).1,2 However, disease in most responding patients will progress within one to two years of starting anti-PD-1 antibody therapy, revealing that checkpoint blockade alone is unlikely to be curative in the majority of patients.3,4 Historically, allogeneic HSCT (allo-HSCT) using a reduced-intensity conditioning regimen is recommended as a potential curative treatment in this setting, with long-term disease-free survival in 25 to 40 percent of patients.5-7 Because of the established role of PD-1 inhibitor therapy in R/R cHL, many patients who are candidates for allo-HSCT will have received anti–PD-1 monoclonal antibody therapy during their pretransplant treatment course, yet the potential of immunomodulatory effects of checkpoint blockade to increase rates and severity of subsequent graft-versus-host disease (GVHD) has remained a central safety concern, especially given the young age of the majority of patients. The optimal transplant strategy is not defined for these patients owing to the scarcity of prospective data to guide these important treatment decisions. Thus, a better understanding of methods to render allo-HSCT safe for patients who received checkpoint inhibition therapy before allo-HSCT, as well as continued exploration of the use of PD-1 blockade following transplantation, are of ongoing clinical interest.
Multiple retrospective analyses have reported the outcomes of patients with R/R cHL who received pretransplant PD-1 blockade therapy.8-11 The initial data suggest that pretransplant PD-1 blockade therapy may be associated with increased early transplant-related complications, including several cases of fatal acute GVHD, noninfectious febrile syndromes, and severe hepatic sinusoidal obstructive syndrome, despite reduced-intensity conditioning regimens. However, a lower relapse rate compared to historical controls was noted, supporting the efficacy of this treatment strategy. Furthermore, a single-center report of 14 patients who received PD-1 blockade as a final salvage treatment before allo-HSCT with post-transplant cyclophosphamide (PTCy) as GVHD prophylaxis, did not demonstrate severe GVHD,12 suggesting that PTCy may be an effective prophylaxis regimen for patients who received pretransplant PD-1 blockade. Based on these published reports, the U.S. Food and Drug Administration issued a “warning and precaution” regarding the immune-mediated adverse responses associated with allo-HSCT after nivolumab and pembrolizumab, and consensus recommendations were developed to reduce the risk of immune complications. These recommendations included discontinuation of checkpoint inhibitor therapy six or more weeks before transplantation and the use of PTCy-based GVHD prophylaxis.13,14
In their report, Dr. Chiara DePhilippis and colleagues retrospectively reviewed toxicity and outcome data on 59 adult patients (>19 years; median age, 30 years) with relapsed cHL who underwent haploidentical SCT followed by PTCy, cyclosporine A, and mycophenolate mofetil as GVHD prophylaxis at three medical centers in Europe between 2014 and 2018. They analyzed the effect of pretransplant PD-1 blockade on outcomes by comparing data from patients who received PD-1 blockade (n=29) prior to allo-HSCT versus patients who did not receive PD-1 blockade prior to allo-HSCT (n=30). The 100-day cumulative incidence of acute GVHD was 33 percent for the entire cohort, and no statistically significant difference was demonstrated between patients who received PD-1 blockade prior to allogeneic bone marrow transplantation (allo-BMT) and those who did not (41% vs. 33%, respectively; p=0.45). Regarding toxicity, the one-year cumulative incidence of moderate to severe chronic GVHD was similar between patients who received PD-1 blockade before allo-BMT and those who did not (7% vs. 8%; p=0.673). No patients who received PD-1 blockade before allo-BMT developed veno-occlusive disease. Three patients who received PD-1 blockade before allo-BMT developed a noninfectious febrile syndrome after engraftment that required steroid therapy, but no patients who had not received PD-1 blockade prior to allo-HSCT developed a similar febrile syndrome.
Regarding treatment outcome, at a median follow-up of 26 months, the two-year overall survival (OS) and progression-free survival (PFS) rates for all patients were 74 percent and 65 percent, respectively. No difference was demonstrated in OS between patients who received PD-1 blockade prior to allo-HSCT and those who did not (77% vs. 71%, respectively). A non–statistically significant higher PFS was demonstrated in the cohort of patients who received PD-1 blockade prior to allo-HSCT compared to those patients who did not (78% vs. 53% respectively; p=0.066). For patients in complete response (CR) and partial response at the time of transplantation, the two-year cumulative incidence of relapse was 13 percent. A non–statistically significant higher incidence of relapse was seen in patients who did not receive PD-1 blockade prior to allo-HSCT as compared to patients who were treated with this class of agent (22% vs. 4%; p=0.098). Considering only patients in CR at the time of transplantation, however, the two-year cumulative incidence of relapse was 0 percent for those who received PD-1 blockade prior to allo-HSCT (n=22) and 20 percent for patients who did not receive PD-1 blockade (n=18). Two-year nonrelapse mortality rate was 18 percent for all patients, and no difference was seen between patients who received PD-1 blockade before allo-HSCT and those who did not (15% vs. 21%; p=0.578). Finally, multivariable analysis was used to evaluate the effect on survival of PD-1 blockade before allo-HSCT and remission status at the time of transplantation. PD-1 blockade before transplantation was associated with superior PFS, but not OS (HR, 0.23; 95% CI, 0.07-0.76; p=0.015). Both univariate and multivariate analysis identified remission status at the time of transplantation as being associated with long-term outcomes: The two-year OS for patients in CR was 81 percent compared to 75 percent in patients with partial remission and 20 percent in patients with stable disease or progressive disease (p≤0.001). Stable or progressive disease was an independent negative prognostic factor for OS and PFS (HR, 14.3; 95% CI, 3.49-58.95; p<0.001).
In Brief
In summary, Dr. De Philippis and colleagues have provided toxicity and outcomes data for their cohort of adult patients with R/R cHL who underwent haploidentical transplantation with PTCy as part of GVHD prophylaxis and compared data from patients who had received PD-1 blockade prior to transplantation to that of patients who did not. Prior exposure to PD-1 blockade did not lead to excess toxicity in the post-transplant course, and most likely, the incorporation of PTCy as GVHD prophylaxis can ameliorate the risk of moderate to severe chronic GVHD. Whether PD-blockade prior to allo-HSCT improves relapse rates remains to be determined, but it is clear that remission status at time of transplantation is an important predictor of long-term outcome. Prospective trials specifically designed to address the optimal transplant strategy are needed. Finally, ongoing follow-up of the Checkpoint 205 trial has identified a subset of patients with cHL who received PD-1 blockade as salvage therapy after autologous BMT relapse and have remained in CR for more than three years without further therapy.15 Thus, future studies will likely investigate whether a subset of patients can be cured with checkpoint blockade alone and do not need to proceed to allo-HSCT.
References
Competing Interests
Dr. Alpert and Dr. O'Dwyer indicated no relevant conflicts of interest.
