Introduction: High-grade serous ovarian cancer (HGSC) is the most common and aggressive histotype of ovarian cancer, often associated with effusions at presentation. Despite advancements in diagnosis and treatment, survival rates for HGSC patients have not improved in recent years. Currently, there is a need for new biological markers that can reliably predict the course of the disease and response to treatment. One potential source of such markers is the effusion that often accompanies HGSC. HGSC effusions contain tumor spheroids, dissociated tumor cells, and immune cells, and by studying the features and prognostic significance of these components, we may be able to identify new markers that can aid in the diagnosis and treatment of HGSC. However, there is a lack of quantitative data about the immune cells in HGSC effusions and data on their clinical importance. Furthermore, no study has yet investigated the presence of immune cells and endothelial cells in tumor spheroids and the expression of PD-1 and PD-L1. In general, cytopathology has also been underused as a modality to identify immunophenotypic and molecular features of HGSC.
Objective: Three main objectives were established: 1) to quantify different immune cells in HGSC effusions, the expression of programmed death receptor-1 (PD-1) and its ligand (PD-L1) in both tumor and immune cells, and to calculate their correlation with patient progression-free survival (PFS) and overall survival (OS); 2) to study tumor spheroids within effusions with an emphasis on their morphological features, cell composition (tumor cells, immune cells, and endothelial cells), and PD-1 and PD-L1 expression, and 3) to evaluate the accuracy of cytopathology to identify immunophenotypic features of tumor cells and BRCA1/2 mutations from HGSC effusions.
Methods: Patients with histologically confirmed HGSC and malignant effusions were included in the study. Flow cytometry analysis was performed to detect percentages of different immune cells in the effusions: T lymphocytes (CD3+ and their subsets CD4+, CD8+, regulatory T cells (Tregs), NK T (NKT) cells, and CD103+ T lymphocytes), B lymphocytes, NK cells, macrophages, and dendritic cells (DCs), as well as their expression of PD-1 and PD-L1. The cut-off of low and high percentages of different immune cells was determined by the median of the variables. Giemsa smears were used to evaluate the size and density of tumor spheroids in effusions, and a cut-off value for low and high density of tumor spheroids that could predict better/poorer survival outcomes was determined. Cytoblock immunocytochemistry (ICC) was used to describe the cell composition of the spheroids to establish if immune cells (T and B lymphocytes, and macrophages) and endothelial cells composed tumor spheroids together with the tumor cells, as well as PD-1 and PD-L1 expression within the spheroids. Flow-cytometric results on immune cells and morphological and immunophenotypic features of spheroids were used to calculate if there is any correlation with PFS and OS. ICC staining for PAX8, WT1, P53, P16, and Ki-67 was performed on cytospins and cytoblocks and compared with the results in tissue blocks of the primary tumor. Next-generation sequencing (NGS) was used to detect germline/somatic BRCA1/2 mutations in the effusions. Both ICC and NGS results were compared with immunohistochemistry (IHC) and NGS results from tissue blocks of the primary tumor and blood samples.
Results: Flow-cytometric results showed that T lymphocytes were the predominant cells in the effusions (median 51%), while the presence of other immune cells was much lower (median 䁤10%). PD-1 was mainly expressed on T lymphocytes (median 20%), and lower expression was observed on other immune cells. PD-L1 expression was not detected in immune cells at all. Patients with high percentages of CD103+ T lymphocytes, NK cells, DCs, and PD-1+ Tregs were significantly associated with better PFS and OS than patients with low percentages of those immune cells, while high percentages of CD8+ T subsets, macrophages, and PD1+ NK cells, and low percentages of CD4+ T subsets associated with significantly better OS. The presence of small, medium, and large spheroids was observed in Giemsa smears but the size of the spheroids did not correlate with patients’ survival. A 10% cut-off for spheroid density in Giemsa smears was established, which significantly predicted the PFS and OS of the patients. The presence of spheroids was also observed in cytoblocks, but only in cases of big-sized and high-density spheroids (25/43 cytoblocks). ICC results confirmed the presence of spheroid-associated lymphocytes, macrophages, and endothelial cells in the spheroids. PD-1 was mainly expressed in spheroid-associated lymphocytes, while PD-L1 expression was seen in both spheroid-associated tumor cells and immune cells. No correlation with patient survival was observed. ICC/IHC results for PAX8, WT1, P53, and P16 showed good reliability between cytospins, cytoblocks, and tissue blocks (α > 0.75), whereas poor reliability and significant differences were observed for Ki-67 between effusions and tissue blocks (α < 0.26; p < 0.001). A 100% concordance was confirmed for the germline BRCA1/2 mutations between effusions and tissue blocks, but only 14% concordance for the somatic mutations.
Conclusions: The results from the doctoral thesis highlight the potential of HGSC effusions to describe tumor and immune cells, explore their prognostic significance and the possibility of evolving new prognostic markers, as well as the importance of cytopathology to confirm HGSC histotype and BRCA1/2 mutations from effusions as reliable as from the primary tumor tissue.
|
