Newly developed prediction tool to help understanding if patients will respond to Checkpoint Inhibitors

IMPRES, which stands for immuno-predictive score, is a newly created predictor of Immune Checkpoint Blockade (ICB) response for melanoma patients.

ICB has been very successful in treatment of melanoma patients, however, only in a percentage of them. It is therefore important to understand who is likely to respond to ICB.

The study, “Robust prediction of response to immune checkpoint blockade therapy in metastatic melanoma,” was published yesterday in Nature Medicine (August 20th2018).

To develop their tool, Auslander et al. have analyzed the transcriptomic data of 108 neuroblastoma patients, including both high risk (i.e. no spontaneous regression) and low risk (no tumor progression). From this transcriptomic data, they focused on 28 literature-based immune checkpoint genes. Based on this analysis, they were able to identify the 15 most predictive features. The scoring system is simply made by counting the number of features that are found in the analyzed sample. The higher the scoring, the higher the likelihood of cancer regression and better overall survival of the patient.

This tool was already validated in an independent dataset of melanoma patients and it was able to recognize patients who responded to treatment, with an overall accuracy of 77-96%.

Further studies will be needed to validate this approach for other cancer types that have been already approved for ICB.