Digital Pathology and AI May be ‘Valuable Tools’ in Prostate Cancer

Commentary
Video

Artificial intelligence models may be “seamlessly incorporated” into clinical workflow in the management of prostate cancer, says Eric Li, MD.

Novel tools such as digital pathology and artificial intelligence (AI) may prove to be “valuable” and lead to more personalized treatment of patients with prostate cancer, according to Eric Li, MD.

In a conversation with CancerNetwork® during the 2023 Society of Urologic Oncology (SUO) Annual Meeting, Li, a urology resident at Northwestern University Feinberg School of Medicine in Chicago, Illinois, discussed his key takeaways from a presentation on the 15-year follow-up findings from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening study. This study examined the capability of an AI-derived digital pathology biomarker in predicting prostate cancer–specific survival in a population of patients undergoing radical prostatectomy.

Specifically, Li indicated that the AI technology is still in its early days, and that the study was the first notable assessment of its efficacy. He stated that larger cohort studies are still needed to further confirm the model’s benefit, although it’s possible that the technology could be “seamlessly incorporated” into clinical workflow.

Transcript:

This [research] was just like a first dive—just the tip of the iceberg. We could further develop artificial intelligence models for patients undergoing surgery in larger cohorts. But really, more healthcare systems have started incorporating digitization of their slides. These models can be seamlessly incorporated into clinical workflow and help inform patient decision-making, specifically regarding whether they need postoperative androgen deprivation therapy or postoperative radiation.

The other thing that I would just like to mention is that the PLCO [study] was a historic cohort that was started in the 1990s. We haven't really had modern tools, just prostate-specific membrane antigen [PSMA] imaging or some of the systemic therapies that we could give nowadays. The next step is going to be trying to incorporate digital pathology biomarkers into clinical trial designs, just to see how it can affect contemporary survival.

Among the biomarkers, digital pathology and artificial intelligence can be really valuable tools for clinicians to help personalize management for patients.

Reference

Li E. A.I. derived digital pathology biomarker predicts prostate cancer specific survival for patients undergoing radical prostatectomy from PLCO trial with over 15 years of follow-up. 2023 Society of Urologic Oncology (SUO) Annual Meeting; November 28-December 1, 2023; Washington, DC.

Related Videos
The use of a single-port robot may allow for surgically treating more patients with head and neck cancer in a more timely manner, according to Hilary McCrary, MD, MPH.
Rohit Gosain, MD; Rahul Gosain, MD; and Rana R. McKay, MD, presenting slides
Rohit Gosain, MD; Rahul Gosain, MD; and Rana R. McKay, MD, presenting slides
Rohit Gosain, MD; Rahul Gosain, MD; and Rana R. McKay, MD, presenting slides
Rohit Gosain, MD; Rahul Gosain, MD; and Rana R. McKay, MD, presenting slides
The AACR Health Disparities Report highlights the changes needed to achieve health equity for patients with cancer.
Robert A. Winn, MD, gives insight into how institutions and community practices can improve access to cancer care.
Patients with KRAS G12C-mutated non–small cell lung cancer who have brain metastases or intolerability of intravenous infusion may be more suitable to receive a small molecule inhibitor compared with chemotherapy, says Sandip P. Patel, MD.
Treatment with sotorasib or adagrasib appears to be more tolerable among patients with KRAS G12C-mutated non–small cell lung cancer compared with docetaxel, according to Sandip Patel, MD.
Notions of blame around smoking may be a prominent factor in patients with lung cancer feeling stigmatized for developing their disease, according to Lisa Carter-Bawa PhD, MPH, APRN, ANP-C, FAAN.
Related Content