Spatial clustering of recently activated cytotoxic lymphocytes improve association with overall survival in women with high grade serous ovarian cancer

BACKGROUND: A previous study found adding spatial contexture of tumor infiltrating lymphocyte (TILs) and cytotoxic T lymphocyte (CTLs) to abundance improved associations with overall survival of non-Hispanic Black women with high-grade serous ovarian cancer (HGSOC). This study set out to replicate previous findings in predominantly non-Hispanic White women HGSOC cohorts. METHODS: Multiplex immunofluorescence was used to characterize TILs in HGSOC tumors from women enrolled in three epidemiological studies (N = 433). Spatial clustering of TILs was derived using a permutation approach for Ripley's K. Cox proportional hazards models were used for associations of spatial clustering and abundance with overall survival. RESULTS: In models assessing differences in spatial clustering within tumors with a high abundance of TILs (>1%), we found that spatial information significantly improved the model fit for recently activated CTLs (LRT P = 0.008), where low spatial clustering in high abundance was associated with decreased risk of mortality (HR = 0.31; 95% CI 0.14-0.70; P = 0.004) compared to when recently activate CTLs were highly clustered. CONCLUSIONS: In this study, we replicated the significantly improved association with overall survival in mostly White women with HGSOC by including spatial information for recently activated CTLs. Further research is needed to understand mechanisms which recently activated spatial architecture impact survival from HGSOC. IMPACT: This study validates our previous findings that adding spatial contexture to abundance, especially CTLs or recently activated CTLs, when performing survival analyses improves the fit of the models.