Immune Signature of Pembrolizumab Plus Gemcitabine, Vinorelbine, and Liposomal Doxorubicin As Second-Line Therapy for Relapsed or Refractory Classical Hodgkin Lymphoma

Abstract

Introduction: In our phase II study (Moskowitz AJ et al, JCO 2021), pembrolizumab plus gemcitabine, vinorelbine, and liposomal doxorubicin (P-GVD) was found to be highly effective as salvage for patients (pts) with relapsed or refractory (R/R) classic Hodgkin lymphoma (cHL) proceeding to high dose therapy and autologous hematopoietic cell transplantation (AHCT). We present analyses assessing immunologic changes associated with P-GVD treatment as well as updated clinical outcome. Methods: Transplant eligible pts with R/R cHL following failure of 1-line of therapy were eligible. Treatment consisted of 2 to 4 cycles of P-GVD followed by AHCT. Peripheral blood samples were collected at baseline, after 2 cycles of P-GVD, before AHCT, 3 and 9 months after AHCT. Peripheral blood was analyzed for serum thymus and activation-regulated chemokine (TARC) and immune cell subsets. TARC was measured by enzyme-linked immunosorbent assay. Immune cell subsets, including T-cells and myeloid-derived suppressor cells (MDSC), were measured by multi-parametric flow cytometry (FC). Metabolic tumor volume (MTV) was evaluated at baseline. Changes in MDSC and TARC levels were assessed using Wilcoxon rank sum and Friedman tests, respectively. Correlations between MTV and either TARC or MDSC were assessed using Pearson's R. Results: Of 38 evaluable pts, the complete response (CR) rate after 2 or 4 cycles of P-GVD was 95%. Two pts declined AHCT and 36 proceeded to AHCT. After a median follow-up of 30 months (range: 2-43), only 1 pt experienced disease progression 23 months after AHCT. The estimated 30-month progression-free survival (PFS) is 96%. The median baseline MTV value was 37 cm3 (range 4- 845). 36/38 and 27/38 pts had sufficient correlative samples to perform analysis for TARC and MDSC analysis, respectively. The median baseline TARC level was 526 pg/mL (range: 12-4017). TARC levels decreased significantly after 2 cycles of P-GVD (p < 0.001) and increased post-AHCT. Circulating MDSCs were 0.05% of total mononuclear cells (range: 0.003-0.59%) and significantly increased after 2 cycles of P-GVD (p<0.001). Higher baseline TARC and circulating MDSC levels correlated with higher baseline MTV (r=0.45, 95% CI: 0.11-0.70, p<0.01 and r=0.46, 95%CI 0.03-0.76, p=0.03, respectively). High-dimensional analysis of FC data revealed several distinct immune clusters (Figure 1), of which non-proliferating PD-1+ CD4+ T-cells decreased most significantly following treatment (Figure 2). Additional markers (ICOS, CTLA-4, LAG-3, FOXP3) of immune exhaustion showed only a mild decrease over the treatment course. Conclusions: After a median of 30-months follow-up, second-line P-GVD continues to be associated with durable remissions after AHCT. These analyses provide insights into key immune changes associated with chemoimmunotherapy in HL. Baseline TARC correlated with baseline MTV and was significantly lower after P-GVD, illustrating its role as a marker of tumor bulk. The subsequent rise in serum TARC post AHCT suggests immune reconstitution of TARC-secreting cells. Higher circulating levels of MDSCs correlated with baseline MTV, which may reflect an association between immune suppression and tumor bulk. MDSC levels unexpectedly increased after P-GVD, indicating that a rise in MDSCs after treatment may not necessarily predict for poor outcomes. Ongoing single-cell RNA sequencing analysis of peripheral blood mononuclear cells (PBMC) may provide further understanding into immune dynamics associated with P-GVD treatment. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal