Zoltan Szallasi, MD

2024 Collaborative Research Development Grant – Microsoft AI For Health

Boston Children’s Hospital
Machine learning based approaches to predict and overcome treatment resistance in ovarian cancer

Project Summary
We will use Artificial Intelligence (AI)/deep learning methods to analyze the rich data bases of genomic, proteomic profiles and histological images of ovarian cancer to achieve the following clinical goals. We will create an accurate predictor of response to first-line platinum-based therapy. Patients not responding to this therapy have an inferior clinical outcome and using our predictor they could be offered alternative therapies at an earlies stage. We will also identify potential therapeutic vulnerabilities of ovarian cancer that are resistant to first-line or maintenance therapy.

Bio
Zoltan Szallasi MD is a senior research scientist at Boston Children’s Hospital and an associate professor at Harvard Medical School. His group is interested in the computational analysis of cancer genomics with a special emphasis on synthetic lethality strategy driven therapeutic targeting of DNA repair deficiencies in solid tumors, such as ovarian and breast cancer. His group published the first component of an FDA approved diagnostic genomic scar signature of homologous recombination deficiency, which is widely used for the prioritization of ovarian cancer patients for PARP inhibitor therapy. His group has also developed diagnostic signatures for nucleotide excision repair deficiency, which is currently evaluated in the context of predicting platinum sensitivity. Recently, Dr. Szallasi has started to apply artificial intelligence/deep learning-based approaches to analyze genomic data of high grade serous ovarian cancer cases in order to identify novel synthetic lethality-based therapies on cases that progress on PARP inhibitor therapy.