American Society of Hematology, Blood, Supplement 1(136), p. 27-28, 2020
DOI: 10.1182/blood-2020-138858
Full text: Download
Background.Tumor tissue in classical Hodgkin Lymphoma (cHL) contains 1-10% malignant Hodgkin/Reed-Sternberg cells and a significant number of immune cells in the tumor microenvironment that are characterized by expression of inhibitory molecules (PD-1, CTLA-4, LAG-3, TIM-3, TIGIT). Despite overall effectiveness of anti-PD-1 treatment many patients still have relapsed or refractory (r/r) disease, therefore the search for predictive/prognostic biomarkers in patients on immunotherapy is highly demanded. Materials and methods.The study included 39 primary tumor specimens from patients with r/r cHL obtained before starting the treatment with nivolumab (primary biopsies). Specimens from 11 patients were studied before and after treatment (sequential biopsies). Treatment response was evaluated by PET-CT according to LYRIC criteria. Immunohistochemical staining for CD68, CD163, PD-1, LAG-3, TIM-3, CTLA-4, TIGIT was performed with an automated staining system (Bond III; Leica Biosystems). The slides were scanned with Aperio ScanScope XT (AperioTechnologies Inc.) and were analyzed with ImageScope Analysis software (Aperio Technologies) и Qupath (https://qupath.github.io). We explored progression-free survival (PFS) depending on the proportion of cells positive for CD68, CD163, PD-1, LAG-3, TIM-3, CTLA-4, TIGIT in the tumor microenvironment and analyzed the changes of these parameters between primary and sequential biopsies after treatment with nivolumab. Statistical analysis was performed using SPSS software (v.23). Data on sequential biopsies were analyzed with the Wilcoxon signed-rank test. PFS was calculated with the Kaplan-Meier method. The significance level was p ≤ 0.05. Results.A significant correlation was found in primary biopsies group between the value of CD163 and CTLA-4 (correlation coefficient -0,62, p < 0.05). There was no significant association between PFS and proportion of cells positive for CD68, PD-1, TIM-3, CTLA-4, TIGIT, LAG-3 in primary biopsies group. ROC analysis allowed to establish a 9% cut-off value of CD163 expression, dividing these patients into subgroups of CD163high and CD163low. In the CD163low group, the two-year PFS was 19,1% (95% CI 6%-37,7%) with a median PFS of 8,8 months (95% CI 5,7-12) and in the CD163high group - 53,8% (95% CI 28,4%-73,7%) with a median of 24,8 months (95% CI 18,8 - 39,2). In sequential biopsies, a statistically significant increase in numbers of PD-1+ and TIGIT+ T-cells and depletion of CD68+ and CD163+ cells was observed compared to corresponding cell counts in primary biopsies (median PD-1 - 3% vs 10%; median TIGIT - 10% vs 14%; median CD68 - 10% vs 7%; median CD163- 8% vs 3,5%; р <0,05). Conclusion.A comprehensive analysis of expression of CD68, CD163, LAG-3, TIGIT, CTLA4, TIM-3, PD-1 was performed in patients with r/r cHL before and after treatment with nivolumab. Significant association was found between the expression of CD163 and CTLA4. The results of the study indicate inferior PFS among patients with low expression (<9%) of CD163 in lymph node samples before immunotherapy. Biopsies taken after treatment with nivolumab showed a statistically significant increase in the number of PD-1+ and TIGIT+ cells and a decrease in the number of CD68+ and CD163+ cells compared with data from primary biopsies. The results of the study may contribute to our knowledge regarding biology of classical Hodgkin lymphoma and the mechanisms of resistance to therapy with immune checkpoints inhibitors. This study was supported by BMS research grant CA209-8EG Disclosures Ionova: Takeda:Other: principal investigator of the observational studies sponsored by Takeda;BMS:Other: principal investigator of the observational studies sponsored by BMS.