STOCKHOLM — The longest follow-up to date of active surveillance in patients with favorable or intermediate-risk prostate cancer shows that it is a safe and feasible approach for as long as 20 years after diagnosis.
Men in the study cohort had early-stage disease and were managed with surveillance; they were treated only if there were signs of disease progression. Up to 20 years after diagnosis, 1.5% of the 993 men had died, and 3.1% had developed metastatic disease.
In addition, death was 10 times more likely from other causes than from prostate cancer, reported Laurie Klotz, MD, from the Sunnybrook Research Institute in Toronto.
"This is the longest, most mature follow-up with the explicit strategy of conservative management and selective delayed intervention," he toldMedscape Medical News. "Even at 15 to 20 years, the prostate cancer mortality rate is really very low, you've avoided treatment in the majority of patients, and even the ones treated late still had a long period of normal quality of life before they had treatment."
Dr. Klotz presented the results here at the European Association of Urology 29th Annual Congress.
At 5, 10,15, and 20 years after diagnosis, 75.7%, 63.5%, 55.0%, and 55.0% of patients, respectively, remain untreated and on active surveillance. During the follow-up period, 15 died from prostate cancer and 7 developed metastatic disease.
Results Are Now "Slightly More Sobering"
The prospective single-group cohort study enrolled men (median age, 69 years) with histologically confirmed prostate adenocarcinoma who had undergone no previous treatment. Definitive intervention was offered to patients if they had a prostate-specific antigen (PSA) doubling time of less than 3 years, Gleason score progression (to 4 + 3 or greater), or unequivocal evidence (including MRI) of clinical progression. The remaining 993 patients were followed with active surveillance.
In an earlier follow-up of the patients, at a median of 6.8 years after enrollment, only 5 subjects had died of prostate cancer — all of them fairly early into their surveillance (J Clin Oncol. 2010;28:126-131). In all 5, it looked like their disease "wasn't preventable with earlier treatment," said Dr. Klotz. In the the latest analysis, median follow-up is 8.1 years. "But now it's slightly more sobering. We're starting to see some deaths in patients who have actually developed their metastases late. Perhaps if they'd been treated a decade before, they might have been cured."
"The devil is really going to be in the details about these men who progressed," said session chair Matthew Cooperberg, MD, MPH, associate professor of urology at the University of California, San Francisco.
"The question is how many men died of prostate cancer because they went on active surveillance instead of getting immediate treatment. The answer to that question is always going to be greater than 0," Dr. Cooperberg explained. "However, we all think that it's very, very, very low; it's far lower, most likely, than the number of men who are harmed or potentially harmed from overtreatment for low-risk disease."
A similar sentiment was expressed by Chris Bangma, MD, PhD, professor and chair of urology at Erasmus University in Rotterdam, the Netherlands. Dr. Bangma is a principal investigator of the ongoing Prostate Cancer Research International: Active Surveillance (PRIAS) study, the largest worldwide prospective active surveillance study.
"How do these survival results compare with the curative treatment that patients would have otherwise received if active surveillance did not exist? We do not know yet whether the active surveillance mortality numbers in the study by Dr. Klotz's team are comparable with those in radical prostatectomy studies," Dr. Bangma told Medscape Medical News. However, "it's my expectation that the study numbers will be far below those in the radical prostatectomy series, because the patients have been selected as having very low risk."
At this stage, active surveillance "is a treatment approach in evolution," said Dr. Klotz. Research continues into how to refine the parameters for triggering intervention.
"We've shifted to a more MRI-based evaluation of patients, which I think in the long run is going to change the shape of the time-to-treatment curve," he said. "When something doesn't seem right — either a patient has a short [PSA] doubling time or some small amount of Gleason pattern 4 — we now go to MRI; before, we treated them."