Science

Advances in artificial intelligence have paved the way for leveraging hematoxylin and eosin-stained tumor slides for precision oncology. We present ENLIGHT–DeepPT, an indirect two-step approach consisting of DeepPT, a deep-learning framework that predicts genome-wide tumor mRNA expression from slides, and ENLIGHT, which predicts response to targeted and immune therapies from the inferred expression values. We show that DeepPT successfully predicts transcriptomics in all 16 The Cancer Genome Atlas cohorts tested and generalizes well to two independent datasets. ENLIGHT–DeepPT successfully predicts true responders in five independent patient cohorts involving four different treatments spanning six cancer types, with an overall odds ratio of 2.28 and a 39.5% increased response rate among predicted responders versus the baseline rate. Notably, its prediction accuracy, obtained without any training on the treatment data, is comparable to that achieved by directly predicting the response from the images, which requires specific training on the treatment evaluation cohorts.

In this study we investigated a unique subpopulation of pancreatic ductal adenocarcinoma (PDAC) patients with germline BRCA mutations (glBRCA). These tumors respond well to platinum-based chemotherapy and PARP inhibitor therapy, but resistance becomes a significant hurdle. To address this challenge, we developed model systems that replicate clinical scenarios and treatment resistance. These models will be instrumental in identifying mechanisms of resistance, ultimately leading to the development of new treatment options for glBRCA-positive PDAC patients.

This work presents Pangea’s proprietary ENLIGHT platform, a GI-based algorithm that extends and improves upon SELECT. ENLIGHT was developed by Pangea for prediction of patient response to targeted therapies and ICB in multiple scenarios. The manuscript presents a comprehensive retrospective validation for ENLIGHT’s applicability to precision oncology and clinical trial design.

This case report describes an exceptional report to checkpoint blockade therapy in a young patient with a rare liver cancer. The treatment highlighted by ENLIGHT went contrary to common checkpoint blockade biomarkers and led to complete remission.

With this work, we demonstrate for the first time the ability to leverage common histopathology images for accurate transcriptome-based prediction of response to precision oncology treatments, bypassing the need for lengthy and costly NGS. The ENLIGHT-DeepPT technology (ENLIGHT-DP for short) can have far-reaching implications, as it can make precision profiling and treatment-matching feasible in situations where NGS is not – for lack of resources, time, or sufficient biopsy material. Furthermore, it can serve as a very fast and inexpensive screening test, to determine which patients should be referred for more accurate comprehensive molecular profiling.

This work presents SELECT, a GI-based approach for response prediction in the oncology realm. This work extends upon ISLE with an emphasis on a more comprehensive retrospective validation on targeted therapies and immunotherapies

This work uses the SELECT pipeline to identify multiple potential therapy combinations for a variety of Triple Negative Breast Cancer (TNBC) sub-types. In particular, two powerful combinations targeting the BET complex and two synthetic-lethal targets of it were identified and validated.

Another demonstration of the utility of the approach for drug target identification. In this case in Pancreatic Cancer and Melanoma.

This work demonstrates the applicability of the SELECT GI-based approach to drug development. Specifically, a novel target for Uveal Melanoma was discovered here for the first time based on a computational analysis done by SELECT and was validated prospectively in in-vitro and in-vivo models. This target is currently being validated in a clinical trial.

This work presents INCISOR, an algorithm for inferring Synthetic Rescuer genes, and using them to predict resistance to targeted and immunotherapies.