Modeling Tumor Immuno-Dynamics to Predict Patient Survival & Immunotherapy Efficacy

Identification: 1002


Modeling Tumor Immuno-Dynamics to Predict Patient Survival & Immunotherapy Efficacy

Nicholas K. Akers, Eric Schadt, Bojan Losic

Department of Genetics, Icahn Institute for Genomics and Multiscale Biology, Icahn School of Medicine at Mount Sinai, NY, NY 10029, USA

State of the art models of cancer survival are increasingly utilizing molecular data that feature the number of non-synonymous mutations as significant predictors.It is widely hypothesized that this association between mutations and survival is a result of neo-epitopes that induce a robust anti-tumor immune response, however this has generally been difficult to prove.Indeed, most neoepitope prediction pipelines explain roughly the same survival variance as mutations.

Reasoning that integrating novel neo-epitope measures and immune response will lead to superior performance, we built a statistical model to quantify the influence of tumor immune-dynamics on patient survival. Neoepitopes (MHC-I/II) were predicted from mutations and filtered on a self-ligandome. The clonal structure of these mutations and functions of the distribution of resulting epitopes were assessed as predictive features along with others such as HLA and mutation expression.The immune response was inferred using VDJ sequencing.

Leveraging the cancer genome atlas (TCGA) for 9 major cancer types including bladder, breast, colorectal, glioblastoma, liver, lung, melanoma, pancreatic, and uterine cancer, we obtained omics data for a total of 2,866 patients.Random forest based recursive feature elimination was used to determine predictive features while overfitting was controlled with k-fold cross validation.A cox proportional hazard model of survival as a function of mutation burden was compared to that with a distributional function of epitope and immune features. Our results confirm that in melanoma these distributional predictors for survival outperform mutation burden alone and simultaneously suggest a quantitative classification of immunotherapy efficacy across cancer subtypes. Similar results hold across other types, suggesting distributional epitope measures and immune response capture important novel survival correlates.


Credits: None available.

You must be logged in and own this product in order to post comments.