Extended Follow up and Immune Alterations in Patients with Classic Hodgkin Lymphoma (cHL) Treated with Frontline Single Agent Pembrolizumab Followed By AVD Chemotherapy

Background:

Hodgkin cells evade immune response through several mechanisms including overexpression of PD-L1/2, reduction of MHC expression, production of indolamine (IDO), secretion of inhibitory cytokines, and alterations to the microenvironment. We conducted a phase 2 clinical trial of sequential pembrolizumab x 3 followed by doxorubicin, vinblastine, dacarbazine (AVD) chemotherapy (4-6 cycles) for newly diagnosed early unfavorable or advanced stage cHL. Interim response to single agent anti PD-1 therapy was assessed by PET-CT and by decline in metabolic tumor volume (MTV). We report the results of changes in cytokine levels and response to PD-1 blockade and survival outcomes following extended follow up of > 5 years.

Methods:

Levels of plasma soluble cytokines and other mediators including IFN-γ, IL-10, IL-13, IL-2, IDO, LAP, PDL1, TARC (CCL17), TNF-α, and Galectin-1 were measured by Luminex using ProcartaPlex kits (ThermoFisher) according to the manufacturer’s instructions. All standards and samples were measured in duplicates pre-treatment, following single agent PEM, and after AVD. Pre-treatment diagnostic biopsy specimens were double stained for PD-L1 (E1L3N, XP Cell Signaling) and PAX5, and scored by two expert hematopathologists (QC LBS).

Response by metabolic tumor volume (MTV) was defined as complete metabolic response (CMR), ≥ 90% reduction by MTV (near CMR; nCMR), or < 90% reduction by MTV (partial response; PR). Patients were categorized as excellent responders (CMR or nCMR) versus others (PR). Changes in biomarkers were calculated as percentage change from pre-treatment to post-pembrolizumab and from post-pembrolizumab to post-chemotherapy. Percentage changes were classified as increase (≥ 25%), decrease (≤ -25%), or stable. Comparisons between patients with PR versus CMR/nCMR were calculated with Wilcoxon rank sum and Fisher’s exact tests. Additionally, we assessed levels of soluble PD-L1 as well as the relationship between baseline levels of cytokines and PDL-1 expression in Reed-Sternberg (RS) cells by H-score.

Results:

Thirty patients were enrolled from September 2017 through August 1, 2019. Response to single agent PD-1 was PR in 11 (36.7%), nCMR in 8 (26.7%), and CMR in 11 (36.7%). The updated median follow-up is 67.6 (range: 59.3-80.4) months. No patients have relapsed or died, for a PFS and OS of 100%.

Twenty-eight patients had peripheral blood available for cytokine analysis. There was no association between baseline levels of TNF‐α, IFN‐γ, TGF‐β, IL-10, IL-13, Galectin-1, IDO, TARC, or soluble PDL-1 and response to single agent pembrolizumab.

Following single agent pembrolizumab changes in cytokine levels by response (CMR/nCMR vs PR) showed that all PRs had stable or a decline in TNF‐α, IFN‐γ, TGF‐β, IL-10, IL-13, and IDO levels. In contrast, among the 18 patients with CMR/nCMR variable increase in cytokines was observed: TNF‐α increased in 1 pt (5.6%), IFN‐γ increased in 6 pts (33%), TGF‐β increased in 2 pts (11%), IL-10 increased in 3 pts (17%), and both IL-13 and IDO increased in 5 pts (28%). Following sequential chemotherapy there was no association between responses and cytokine alterations. Soluble PD-L1 was undetectable in 26/28 (93%) pts at one time point and in 18 pts at all three time points evaluated. On H-scoring there was no association with any of the cytokine’s levels at baseline and PDL-1 expression in Reed Sternberg cells.

Conclusions:

Mature data with > 5 year median follow-up confirm that sequential pembrolizumab and AVD chemotherapy remains a highly effective treatment with 100% PFS and OS. Baseline cytokine levels were not predictive of outcomes. Patients with PRs to Pembrolizumab monotherapy uniformly had a decline in cytokine levels while most with excellent responses (CMR/nCMR) had variability with most cytokines declining and some increasing. Further investigation is required to understand the implications of these findings for pembrolizumab-based therapy in cHL.