John D. Minna, MD

John D. Minna, MD, Max L. Thomas Distinguished Chair in Molecular Pulmonary Oncology and Sarah M. and Charles E. Seay Distinguished Chair in Cancer Research at The University of Texas Southwestern Medical Center, Dallas, was enthusiastic about this approach.

“When I first heard about it, I was so impressed that I arranged for the authors to give a presentation at the National Cancer Institute monthly Lung Cancer SPORE webcast. Basically, they are taking image analysis and simple clinical variables to identify features that predict who will and who won’t respond to immune checkpoint inhibitors,” Dr. Minna said.

“The researchers have done fairly large studies at their own institution and validated them at other centers using a standard CT scan and blood and clinical characteristics [serum albumin and number of -metastatic sites] to arrive at their prediction,” he continued.

Implications for Clinical Practice

“Now the question is whether this is easily learned at different centers, and, if the algorithm were applied to thousands of patients, would it tell you the same thing?” asked Dr. Minna. “All CT scans are electronic, so they could be sent to Moffitt to predict whether the patient would do poorly or well on immune checkpoint inhibitor therapy.”

“If these results are confirmed, then every radiology report could include this information to facilitate clinical decision-making. This has a lot of promise and needs to be immediately and rapidly validated. If it holds up, it could go into clinical practice next year,” Dr. Minna noted.

When asked whether these algorithms would be available for wider use, Dr. Gillies replied in an e-mail: “These algorithms are designed for export and free use in a research setting.We are also working on a cloud based analytical tool that could be accessed. We had not planned to provide a service but would be willing to consider it, if this would iimprove clinical utility.” 

DISCLOSURE: Dr. Minna has reported no conflicts of interest.