Outcomes, Toxicities, and Patterns of Use of Programmed Death-1 (PD-1) Inhibitors in Hodgkin Lymphoma Patients in the United States: A Multicenter Retrospective Study

Introduction

Greater than 8,000 patients (pts) are diagnosed with classical Hodgkin lymphoma (cHL) in the United States each year. While cHL is a highly curable malignancy, 20% will not be cured with conventional treatments. PD-1 inhibition has resulted in impressive activity in the relapsed/refractory (R/R) setting, thus leading to FDA approval of nivolumab in 2016 and pembrolizumab in 2017. There is limited information on "real world" experience with these agents in cHL treated outside of clinical trials.

Methods

We performed a multicenter, retrospective analysis of R/R cHL pts treated with programmed death-1 (PD-1) inhibitors (PD-1i) in the non-trial setting. The primary objective of the study was to describe progression free survival (PFS) and overall survival (OS) in this population. PFS and OS were calculated from the date of PD-1i initiation to the date of disease progression or death, respectively, or the date of last follow-up. Secondary objectives included overall response rates (ORR) and complete response rates (CR), toxicities, discontinuation patterns, and post-PD-1i therapies. OS and PFS were calculated using the Kaplan-Meier method.

Results

Forty two pts from 7 centers in the United States were identified. Pt characteristics include: median age 34 years (range 20-67), 62% male, 60% stage III/IV at diagnosis, 71% NS type, 52% with prior auto stem cell transplant (SCT) and 7% with prior allo SCT. Pts received a median of 4 lines of therapy prior to PD-1i (range 1-10); 85% had ABVD/AVD as frontline therapy and 95% had brentuximab (BV) prior to PD-1i.

Nivolumab was the initial PD-1i used in all 42 pts; 78% received the 3 mg/kg every 2 weeks dosing schedule. Concurrent therapies with PD-1i were used at some point in 19% of pts (RT in 4, BV in 4). The ORR to nivolumab as assessed by the treating physician interpretation was 62% with CR 45%. The median time to best response was 3 months (mo) (range 1-10). At median follow-up of 13 mo, the median PFS and OS have not been reached (Figure 1). The 12-mo OS and PFS were 96% and 65%, respectively. The median duration of therapy was 9 mo (range 1-26). Median duration of response in pts with CR and partial response (PR), respectively was 13 mo (range 3-28) and 9 mo (range 7-25). Dose delays occurred in 33% of pts (N=14), most commonly due to toxicity in 43% (N=6), non-adherence in 29% (N=4), and pt preference in 14% (N=2). Dose delays did not impact PFS (p = 0.85). Original PD-1i was discontinued in 52% of pts (N=21). The cause for discontinuation was progression in 52% (N=11), patient preference/non-compliance in 14% (N=3) and toxicity in 14% (N=3). Four of 42 pts (10%) died, all due to disease progression.

In 40 pts with available toxicity data, 43% (N=17) experienced toxicity attributed to PD-1i. The most common toxicities were hypothyroidism in 10% (N=4), rash in 10% (N=4), pneumonitis/dyspnea in 7% (N=3), and colitis in 7% (N=3). Toxicities resulting in drug discontinuation included encephalitis (N=1), colitis (N=1) and skin GVHD (N=1). Toxicity management included systemic steroids in 10% (N=4), thyroid replacement in 10% (N=4), drug discontinuation in 7% (N=3), dose delay in 5% (N=2), or local steroid therapy in 5% (N=2). There were no treatment-related deaths. In 21 pts who discontinued PD-1i, 67% (N=14) pts had subsequent systemic therapies and 12 were evaluable for response (Table 1). Clinical benefit of systemic therapy immediately after PD-1i was seen in 75% (1 CR, 5 PR, 3 SD) with 5 pts responding to cytotoxic chemotherapies. Four pts underwent SCT after PD-1i (3 auto, 1 allo). Four pts were treated with pembrolizumab containing regimens after nivolumab of which 1 pt had a clinical response.

Conclusions

To our knowledge, this multi-center retrospective analysis is the first to describe the "real world" experience with PD-1i in cHL pts in the United States. ORR, median PFS and OS, and toxicities are similar to those observed in published clinical trials. The treating physician assessment of CR rate was higher than previously reported although independent confirmation of response was not performed. Patterns of discontinuation and dose delays are different from published studies, but do not affect outcomes. Systemic therapies after PD-1i discontinuation appear to be active. Our results support the important role of PD-1i therapy in R/R cHL and provide insight into practice patterns.

View largeDownload slide

View largeDownload slide

Close modal