The ASCO Educational Book recently published a paper by Dr. William McKean and colleagues that speaks to both the promise and challenges of treating cancer patients with immunotherapies. This well-researched article, titled “Biomarkers in Precision Cancer Immunotherapy: Promise and Challenges,” systematically reviews the effectiveness of different classes of biomarkers in predicting both beneficial treatment response and harmful side effects. Ultimately, the authors conclude that “none of these are comprehensive in predicting potential benefit.” The paper makes only a brief mention, however, of another technology that could offer a real solution: advanced imaging. At AIQ Solutions (AIQ) we are working on technology that combines PET/CT scan data with advanced algorithms to better determine which patients will have a desirable response to immunotherapy.

Dr. McKean et al. defined the problem, “[A] significant limitation behind these current treatment modalities is an irregularity in clinical response, which is especially pronounced among checkpoint inhibition. This unpredictability leads to significant side effects, financial costs, and health care burden, with unsatisfactory clinical benefit in the majority of treated patients.” Several studies have shown that immunotherapies are effective for less than 50% of patients who qualify. At the same time, immune-related adverse events (irAE) can occur in up to 60% of patients with certain cancers, according to the ASCO clinical practice guidelines for the treatment of immune-related adverse events in patients treated with immune checkpoint inhibitor therapy. To improve the effectiveness of immunotherapies, oncologists need a better way to balance the potential for therapeutic benefit with the risk of irAE.

There is a complex interplay between tumor response, irAE, and immune system activation. Establishing the patient-specific relationship among these three factors is necessary for an accurate prediction of a therapy’s benefit and risk. This may be a tall order for traditional biomarkers like those reviewed by McKean et al. Time-series PET/CT images, however, contain a wealth of information about the patient’s physiologic state. It would be a similarly tall order to manually extract and distill the critical data from this image set. Instead, AIQ is developing a software platform that uses deep learning to automatically analyze tumor response, toxicity risk in healthy organs, and overall immune system activation. AIQ’s technology will help better understand the relationship among these interdependent factors and use that to improve predictions of treatment effectiveness and risk.

Dr. McKean and his colleagues conclude their paper with recommendations for how biomarkers should be developed and validated. They identify five characteristics of the “ideal” biomarker: noninvasive, time-sensitive, multifactorial and comprehensive, therapy-specific, and (able to predict) resistance and toxicity. AIQ’s platform is designed to deliver all of these. Finally, the authors assert, “all future clinical trials must be designed to incorporate biomarkers to some degree. At a minimum, pretreatment and on-treatment information...must be collected…This is especially important for early-phase trials of novel immunotherapies.”

For pharmaceutical companies developing novel immunotherapies, AIQ can provide unique insights into treatment response and toxicity risk while collaborating on the development and validation of a technology that may ultimately improve patient outcomes. AIQ has ongoing partnerships with several research institutions and biotech companies and is actively seeking other partners. If you are interested in working with AIQ Solutions, please reach out to our team to discuss potential collaborations.

Per the US FDA, AIQ technology is indicated for use in the identification, quantification, and evaluation of Regions of Interest (ROIs) from digital medical images, including PET and CT. Any other applications of AIQ technology are for research purposes only.