Tissue and imaging biomarkers in the treatment of NSCLC

Summary The aim of this thesis was to investigate new biomarkers for immunotherapy in patients with non-small cell lung cancer (NSCLC). Although the use of immune checkpoint inhibitors in NSCLC has expanded considerably, the currently used biomarker PD-L1 expression is not sufficiently reliable. Not all patients with PD-L1-positive tumors respond to immunotherapy, while durable responses can also occur in PD-L1-negative patients. Therefore, additional biomarkers are needed to improve treatment selection. Tumor microenvironment and immune cells The first part of the thesis focused on the tumor microenvironment and its potential predictive value for nivolumab treatment. In 139 NSCLC patients, PD-L1 expression and the presence of CD3+, CD8+ and PD-1+ immune cells in tumor tissue and stroma were associated with durable clinical benefit. Higher infiltration of CD8+ and CD4+ immune cells was also associated with longer overall survival. Importantly, these associations were mainly observed in recently obtained biopsies after prior chemotherapy and not in archival biopsies. This highlights the importance of using fresh tumor tissue when investigating tissue-based biomarkers. Because T-cell priming primarily occurs in lymphoid tissue, the immune composition of tumor-draining lymph nodes (TDLNs) was also investigated. In 11 NSCLC patients, differences were observed between T-cell populations in mediastinal lymph nodes and peripheral blood. PD-1 expression, particularly on CD8+ T cells, was significantly higher in TDLNs. These findings suggest that TDLNs may provide relevant information for biomarker development and warrant further investigation as potential predictors of response or early treatment effects. PET imaging of the PD-1/PD-L1 axis The second part focused on PET imaging using tracers targeting PD-1 and PD-L1. For the PD-L1-targeting tracer ^18F-BMS-986192, kinetic modelling showed that static PET imaging approximately 50 minutes after injection was suitable for quantifying tracer uptake. In the first human study, tracer uptake correlated with PD-L1 immunohistochemistry and was higher in lesions that responded to nivolumab. The PD-1-targeting tracer ^89Zr-nivolumab also showed higher uptake in responding lesions and in tumors containing PD-1-positive immune cells. However, substantial heterogeneity in tracer uptake was observed both within and between patients, suggesting considerable variation in PD-1 and PD-L1 expression throughout the body. The first-in-human studies with ^89Zr-pembrolizumab and ^89Zr-durvalumab similarly demonstrated heterogeneous tracer uptake. Higher tumor uptake was associated with treatment response, although these associations did not reach statistical significance. For ^89Zr-pembrolizumab, tumor uptake also did not correlate with PD-1 expression measured by immunohistochemistry. Conclusion Overall, this thesis demonstrates that characteristics of the tumor immune microenvironment and PET imaging of PD-1/PD-L1 have potential as biomarkers for immunotherapy in NSCLC. PET imaging may offer an important advantage by visualizing target expression throughout the body, thereby capturing tumor heterogeneity that cannot be assessed with a single biopsy. However, larger studies are required to validate the predictive value of these approaches. Finally, PET tracers targeting CD8+ T cells may provide opportunities for early assessment of immune responses. This is particularly relevant because conventional imaging can be difficult to interpret during immunotherapy due to pseudoprogression. Future research should therefore consider not only changes in tumor size, but also changes in immune activity and metabolic response.