Abstract 355P
Background
Immune checkpoint inhibitors (ICIs) has been broadly integrated in the treatment of advanced non-small cell lung cancer (NSCLC). However, only 20% of patients experience a prolonged response.
Methods
42 advanced NSCLC patients undergoing ICIs were enrolled. They were divided between those still responding after 1 year of treatment (long responders - LR[AP1]) and those who progressed within the first 3 months (fast progressors - FP). Immune populations and soluble molecules were evaluated, with a total of 200 blood samples. Different therapeutic protocols including anti-PD-1/PD-L1/CTLA4, alone or in combination with chemotherapy (CT) were considered. Peripheral immunoscore of LR was also monitored over time to understand whether prolonged therapy could lead to a different immunological asset.
Results
Peripheral blood immune content, including blood count, percentage of lymphocyte and monocyte subclasses together with HLADR and CD169 expression on monocytes and CD86 expression on B cells permitted to distinguish patients LR from FP. The same combination of indexes was not able to discriminate LR according to therapy. Monocytes were higher in ICIs monotherapy compared to ICI combined with CT that resulted in an increase in neutrophils. Monitoring LR over time ICIs monotherapy resulted in an increase in monocytes instead CT + ICIs therapy registered a decrease in lymphocytes and monocytes and a neutrophils increase. T cell percentage decreased over time for both treatments although with an inverted T CD4+/CD8+ ratio. T cell response was also shaped by gender in ICI monotherapy. TGFB1 and CXCL10 were higher in FP than LR. Surprisingly, only FP autoantibodies inhibited 3D-growth of NSCLC cells derived from a tumor grown under ICI selection pressure.
Conclusions
These results showed the diversification of immune response in different therapeutic regimens and the presence of an active anti-tumor response even in FP. Implementation with AI, radiology/radiomics will allow to identify signature for the monitoring of therapeutic failure. This will improve the understanding of disease biology and will allow the development of prognostic models to be validated prospectively.
Funding
Ricerca Corrente.
Disclosure
All authors have declared no conflicts of interest.