Dr. Ashesh B. Jani
The findings were reported in a plenary session at the American Society for Radiation Oncology Annual Meeting 2020.
“Quite frankly, this is an area where we are shooting in the dark with conventional imaging, and that’s where we think molecular imaging has a potential role,” noted coprincipal investigator Ashesh B. Jani, MD, of the Winship Cancer Institute of Emory University, Atlanta.
“We hypothesized that radiotherapy outcomes can be improved upon by PET by excluding patients with extrapelvic disease and also by improving treatment field decisions and target definition,” Dr. Jani added.
Patients with prostate cancer were eligible for EMPIRE-1 if they had undergone prostatectomy and had a detectable prostate-specific antigen (PSA) level but negative findings on conventional imaging (a bone scan plus abdominopelvic CT and/or MRI).
A total of 165 patients were randomized to RT guided by the conventional imaging alone or combined with PET imaging using the radiotracer fluciclovine (18F). Treatment decisions in the latter group were strictly based on where uptake was seen.
Study results
The trial’s primary endpoint was treatment failure, defined as a PSA level exceeding 0.2 ng/mL from nadir followed by another rise, a continued PSA rise despite RT, progression on imaging or digital rectal exam, or initiation of systemic therapy.
“Most imaging studies tend to focus on diagnostic accuracy, pathologic correlation, and decision changes. It’s a very high bar for an imaging study to influence failure rates,” Dr. Jani pointed out.
Adding 18F-PET to conventional imaging altered the treatment decision for 35.4% of patients in that group (P < .001). It also significantly altered a range of volumetric and dosimetric parameters.
At a median follow-up of 2.48 years, the 3-year rate of failure-free survival was 63.0% with conventional imaging alone and 75.5% with the addition of 18F-PET (P = .003). The corresponding 4-year rate was 51.2% and 75.5%, respectively (P < .001).
In multivariate analysis, the conventional imaging group had double the risk of failure events relative to the PET group (hazard ratio, 2.04; P = .033).
Provider-reported data showed no significant difference between imaging groups in maximum acute or late genitourinary toxicity and gastrointestinal toxicity. An analysis of patient-reported toxicity data is pending.
“Randomized trials of imaging tests with a primary cancer control endpoint are important but uncommonly done,” Dr. Jani commented. “This is the first such trial of PET over conventional imaging in the postprostatectomy radiotherapy setting.”
“Inclusion of fluciclovine resulted in a significant improvement in failure rate at 3 years. This warrants further investigation,” he maintained.
To that end, the investigators have launched the EMPIRE-2 trial, which is comparing RT guided by 18F-PET with PET using another radiotracer that is not yet approved by the Food and Drug Administration, gallium-68 prostate-specific membrane antigen.
Findings in context
“There are several remarkable aspects of the EMPIRE-1 trial worth noting,” said invited discussant Neha Vapiwala, MD, of the University of Pennsylvania, Philadelphia.
Dr. Neha Vapiwala
She commended the trial’s randomization, given a bias that more imaging is better, and the diversity of its participants that better reflects the general population of prostate cancer patients.
“The study procedures appear to be well tolerated despite a net overall increase in the radiation volume treated in the patients who underwent PET, although we do still await patient-reported toxicity,” Dr. Vapiwala noted. “Finally, a high bar was set, with a clinically meaningful primary endpoint for an imaging study.
“This study ultimately demonstrated that, in the PET arm, better selection with PET was able to result in better patient outcomes,” she maintained.
At the same time, Dr. Vapiwala recommended caution when reducing or withholding definitive local therapy based on PET results, as occurred in 14 patients.
“We must always be able to see the forest from the trees, and when evaluating our patients with PET scans, what we see and what we don’t see is just one piece of the puzzle. Existing level 1 evidence and oncologic principles must still apply,” she said. “While PET can help paint a more complete picture, it should not define the picture itself.”
The study was supported by the National Institutes of Health. Dr. Jani disclosed advisory board service for Blue Earth Diagnostics. Dr. Vapiwala disclosed no relevant conflicts of interest.
SOURCE: Jani A et al. ASTRO 2020, Abstract LBA1.
