Matthew B. Rettig, MD Awarded $6.4M from Department of Veterans Affairs

A team of UCLA researchers at the West Los Angeles VA, led by Matthew Rettig, MD, Professor-in-Residence in the Department of Medicine and Chief of Hematology-Oncology at the West Los Angeles VA, has been awarded a $6.4 million grant from the Department of Veterans Affairs to develop artificial intelligence algorithms to accurately predict metastatic recurrence amongst Veterans with high risk, localized prostate cancer. Prostate cancer is the most commonly diagnosed malignancy other than non-melanoma skin cancer in the United States, and high risk, localized prostate cancer represents 20-25% of the ~250,000 incident cases of prostate cancer in the US. Outcomes of high risk, localized prostate cancer are quite variable, with some patients remaining in remission and others suffering from metastatic progression and death. Our ability to discriminate between patients who will fare well following curative-intent treatment versus those destined for lethal metastatic progression remains poor. 

The VA/UCLA multidisciplinary investigative team is represented by urology (Isla Garraway), radiation oncology (Nicholas Nickols) and external experts in epidemiology, artificial intelligence, radiology, and pathology. This group has generated a searchable database of 1.2 million Veterans with prostate cancer that will be leveraged to collect the following three sources of data:

  • High resolution digital pathology images of diagnostic prostate needle biopsies,
  • Prostate MRI, and
  • Area Deprivation Index (a composite of ~20 factors related to social determinants of health outcomes) 

Artificial intelligence, including computer vision and machine learning approaches, will be applied to generate prognostic models for each of the three data sources. Once an optimal model for each individual data source has been developed, the models will then be combined in all possible permutations to identify a simple, low cost “super classifier” artificial intelligence model to predict metastatic recurrence. The resulting predictive model will be made publicly available to assist in clinical decision making and to test treatment intensification and deintensification strategies in prospective clinical trials. 

Related Posts