Establishment of Ex-vivo model from NSCLC Pleural Effusions to Assess Checkpoint Inhibitor Response
​
Iris, Kamer1, Yael, Alpan1, Oranit, Zadok1, Inbal, Daniel-Meshulam1, Michael, Peled2, Sumit, Chatterji2, Efrat, Ofek3, Amir, Onn2, Jair, Bar1
​
1 Jusidman Cancer Center, Chaim Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel.
2 Pulmonology department, Chaim Sheba Medical Center, Tel Hashomer, Ramat Gan, Israel.
3 Pathology Institute, Sheba Medical Center, Ramat-Gan, Israel.
​
Background: A major challenge in cancer immunotherapy is the lack of robust, accessible experimental models to study treatment response. Malignant pleural effusions (MPE) occurring in 20% of NSCLC patients and associated with poor prognosis. This study aimed to develop an ex-vivo culture model using MPE samples from NSCLC patients to investigate the mechanisms of response to immune checkpoint inhibitors (ICIs) and to study the cellular and molecular pathways involved.
Methods:MPE samples were collected from metastatic NSCLC patients treated at Sheba Medical Center. Cell populations within the effusions were characterized by flow cytometry and immunohistochemistry to identify lymphocyte subsets and cancer epithelial cells. Cultures were established from these samples and treated with the Durvalumab (anti-PD-L1), either alone or in combination with Tremelimumab (anti-CTLA4). The response to ICIs was assessed by measuring IFN-gamma protein secretion using ELISA and gene expression using RT-PCR. To explore mechanisms underlying differential responses, RNA seq was performed on sorted CD3+ T cells and B cells (CD19+/CD20+) from both responder and non-responder groups.
Results: A total of 46 MPE samples from 32 metastatic NSCLC patients were analyzed. Fresh samples predominantly contained lymphocytes, including both CD4+ and CD8+ cells, as well as variable numbers of cancer epithelial cells. Of the 15 ex-vivo cultures treated with ICIs, seven (47%) demonstrated a significant induction of IFN-gamma gene expression or protein secretion, indicating an immune response. RNA-seq analysis revealed 27 differentially expressed genes in B cells between responders and non-responders, while no significant differences were observed in T cells.
Conclusions: This ex-vivo model derived from NSCLC MPE provides a valuable platform to investigate the mechanisms of response to ICIs and to study the cellular and molecular pathways involved. The findings highlight the potential importance of B cells in mediating response to ICI treatment, warranting further investigation into their role in immunotherapy outcomes.