Using AI to Improve Prostate Cancer Clinical Trial Imaging: Part 1

PSMA-PET Imaging: Looking Back

PSMA-PET imaging is revolutionizing prostate cancer drug development by offering unparalleled image quality and disease detection / monitoring. In recent years there have been an increasing number of clinical trials investigating prostate cancer using PET imaging with dedicated PSMA targeted radioligands.

Challenges of PSMA-PET Imaging

In 2018, Perceptive became the first, and to date, only imaging core lab to provide clinical trial imaging that supported an FDA submission of quantitative PET data using AI. The data was collected from the VISION trial, the pivotal Phase 3 study evaluating 177Lu-PSMA-617 (now marketed as Pluvicto™) for the treatment of metastatic, PSMA-positive prostate cancer (mCRPC).

Patient enrollment in the VISION trial relied on the identification and accurate measurement of metastatic lesions for determination of PSMA status, thus requiring high-quality anatomic imaging and careful reads by imagers. Historically, the imaging workflow for these types of trials relied solely on time-consuming visual reads, but this trial would be the first to combine quantitative outputs to PSMA reads to produce stronger data for the sponsor to assure patients and the public of drug efficacy.

Introducing the Automated PSMA Segmentation Tool

In conjunction, Perceptive’s clinical imaging core lab had developed a novel PSMA segmentation methodology based on an AI algorithm that used information from both PSMA-PET imaging as well as CT scans to mask the skeleton and reliably delineate bone lesions, which is often the organ system with the highest burden of prostate cancer metastases.

Coupled with a highly trained group of image analysts who quickly learned how to perform software-assisted PSMA segmentation, Perceptive’s methodology provided a nuclear radiologist PSMA-PET expert high-quality draft segmentations to modify as necessary for final production. Successful performance of that FDA-mandated sub-study (led by Phillip Kuo MD, PhD, FACR) helped the sponsor demonstrate that the whole-body SUV mean on PSMA-PET was predictive of survival. So, for approximately every unit of increase in whole body SUV there’s a 10% decrease in the risk of death.

Paving the Way for Theranostic Trials

This was a pivotal result for the FDA submission process and the trial’s success and ultimate approval of Pluvicto opened the doors for the utilization of PSMA-PET and molecular imaging for future theranostic trials.

Perceptive’s methodology has since been used in ten additional trials, including post-marketing studies on PSMA tracers and in April 2025 Version 2 of the Automated PSMA Segmentation Tool was launched.

In the next installment to this series, we learn about V2 of Perceptive’s Automated PSMA Segmentation Tool and get perspective from Dr. Phillip Kuo on how it will advance prostate cancer clinical trials and what he envisions as the future of AI-enabled image analysis.

Interested in speaking to our Scientific & Medical Services team? Get in touch today.