The PD1/PD-L1 Axis in the Classical Hodgkin Lymphoma Microenvironment: PD-1 Is Rarely Expressed but Identifies Patients with High-Risk Disease

Abstract 1560 Introduction: Classical Hodgkin lymphoma (CHL) is histologically unique: the malignant Hodgkin Reed-Sternberg cell (HRS) represents a fraction of tumour bulk, the majority being inflammatory cells: a combination of inadequate immune response and tumor-supportive microenvironment. PD-1 is a T-cell expressed CD28-analogue that inhibits activity of the expressing cell by interactions with ligands including the B7 analogue PD-L1. PD-L1 is upregulated in a range of immune and non-immune cells including malignant cells. Immunosuppression through this axis may contribute to failed immune response and adverse outcome in some tumors (Keir et al. Annu Rev Immunol, 2008; 26: 677–704) and there are reports regarding expression levels and pathway activity (Chemnitz et al. Blood, 2007; 110: 3226–33) as well as biological importance in CHL (Yamamoto et al. Blood, 2008; 111: 3220–3224). PD-L1 is overexpressed by HRS cells and encoded by a chromosomal region showing recurrent copy number gains in CHL (Steidl et al. Blood, 2010; 116: 418- 427). PD1/PD-L1 interactions are also essential to the survival of germinal centre B cells (Good-Jacobson et al. Nature Imm; 2010; 11: 535–42) the normal counterparts of the HRS, implying a potential direct supportive role for this pathway independent of the immune response. While the mechanistic importance of this axis in CHL is widely stated, little published evidence exists to substantiate it. This immunohistochemical (IHC) study set out to determine levels and variability of expression of PD-1 and PD-L1 at diagnosis in CHL and determine prognostic importance. Methods: IHC analysis was performed on tissue microarrays (TMAs) from 122 previously untreated CHL patients with known clinical outcome: median age 30 (range 18–80), 35% female, 71% advanced stage, median follow up 16.5 (range 2–40) years. Triplicate cores were made from areas of high cellularity avoiding fibrotic or acellular portions, arrayed and stained for PD-1 or PD-L1. Positive cells were counted across all cores using an automated image analysis system (Ariol), output confirmed by expert histopathologists, and means calculated corrected to % infiltration per 1mm2. 105 cases (86%) had tissue of sufficient quality for full IHC assessment. Groups were determined based on population distribution of infiltration levels and confirmed using the test/validation set method. Outcomes of freedom from first-line treatment failure (FFTF), disease-specific survival (DSS) and overall survival (OS) were assessed using the Kaplan Meier method with statistically significant differences between groups calculated using the log-rank test. Results: PD-1 expression was strikingly low or absent in the microenvironment of the majority of cases: <1% of cells in 40% of cases and <5% of cells in 80%. Median expression was 0.14% of all nucleated cells (range 0–8.9%). When present, this was predominantly as a low-level background infiltrate with frequent clusters/nodules of PD-1 positive lymphoid cells aggregating around HRS cells. Notably, the rare patients with high expression of PD-1 had adverse outcome. The 20% of patients with infiltration >0.5% cells had DSS 61% vs 89% at both 5 and 10 years (p=0.03). In contrast, PD-L1 was expressed at a high level in both HRS and the microenvironment in the majority of cases: >10% cells in 38% of cases, and <1% in only 5%. Median expression was 6.1% (range 0.2%–40%). Level of PD-L1 expression in the microenvironment was not associated with any measure of clinical outcome. Conclusions: PD-1 may have biological significance, as measured by impact on prognosis in a minority of patients with CHL, where overexpression is associated with an increased chance of CHL-related death. This may relate to immunosuppression, but the very low overall levels of expression and tendency to cluster around the malignant cells (a pattern seen more often in nodular lymphocyte predominant HL) imply a more localised HRS interactive role perhaps equivalent to the pathway's activity in a normal germinal centre. Its ligand PD-L1 is highly expressed in the microenvironment as well as in the HRS, which may relate to an activated rather than suppressive microenvironment, but without association with prognosis. We conclude that for the majority of patients with CHL, the definitive role for this pathway remains to be demonstrated conclusively. Disclosures: Gribben: Roche: Honoraria; Celgene: Honoraria; GSK: Honoraria; Mundipharma: Honoraria; Gilead: Honoraria; Pharmacyclics: Honoraria.