麻豆传媒

For low-risk prostate cancer patients, seven clinical characteristics significantly predicted risk of disease progression and could be used to identify those who may benefit from scaled-back active surveillance, researchers reported.

Multivariable analysis of two cohorts totalling nearly 1,400 men on active surveillance showed that several factors independently predicted risk for disease reclassification, including organ size, time since initial cancer diagnosis, extent of biopsy core involvement, and certain measures of prostate-specific antigen (PSA), according to the study led by Matthew Cooperberg, MD, MPH, of the Helen Diller Family Comprehensive Cancer Center at University of California San Francisco (UCSF).

"Where we're trying to go now is getting more granular with risk stratification, trying to figure out which of the low-risk cancers truly have indolent course and would not progress even with minimal observation versus those which eventually over the years and decades may become higher risk," Cooperberg said in a that accompanied the study's publication in .

Men on active surveillance for prostate cancer require regular tumor assessments, he noted, with most guidelines now recommending biopsies every 1 to 2 years.

"That timing is what we are now really trying to tailor," said Cooperberg. "The frequency of the biopsies in particular in active surveillance is a concern. Every time we do a biopsy of the prostate there is discomfort, there is risk of bleeding, and there's a low risk of a significant infection."

For their study, the researchers examined outcomes in 850 men with low-risk disease (Gleason grade group 1) undergoing surveillance at nine U.S. centers involved in the Canary Prostate Active Surveillance Study (PASS) to create a risk classifier model, which was then validated in 533 men meeting the same criteria on a surveillance program at UCSF. All men included in the analysis had at least one confirmatory biopsy of their disease status.

At 4 years, the area under the receiver operating curve for predicting no change in disease classification was 0.70 for the PASS and validation cohorts (P<0.001 for both). For men in the bottom 10th and 25th percentiles of risk, the model showed a negative predictive value of 0.95 (95% CI 0.89-1.00) and 0.88 (95% CI 0.83-0.94), respectively.

"The model which we developed really can be calculated at any point in surveillance," Cooperberg said. "The more assessments we have -- so the more PSA history, the more prior biopsies, etc. -- the more accurate the model is going to be."

Factors that ultimately went into the predictive model included PSA levels at diagnosis (HR 1.51, 95% CI 1.15-1.98, P=0.003), PSA kinetics (HR 1.46, 95% CI 1.23-1.73, P<0.001), time since the initial prostate cancer diagnosis (HR 1.62, 95% CI 1.28-2.05, P<0.001), maximum percent of positive cores on biopsy (HR 1.30, 95% CI 1.09-1.56, P=0.004), size of the prostate (HR 0.40, 95% CI 0.25-0.62, P<0.001), and patients' body mass index (HR 1.08, 95% CI 1.05-1.12, P<0.001).

Any history of a negative biopsy following the initial cancer diagnosis also predicted a lower risk of disease change (P<0.001 for both):

• 1 vs 0: HR 0.52 (95% CI 0.38-0.71)
• ≥2 vs 0: HR 0.18 (95% CI 0.08-0.40)

The model is available as an to predict patients' risk of reclassification.

Cooperberg pointed out that less-invasive methods for assessing disease change -- such as MRI or biomarkers -- have yet to be shown to be able to replace standard tissue biopsies, which can be a barrier to active surveillance.

"It's a combination of the desire to avoid multiple biopsies and the uncertainty associated with active surveillance that still discourages some men -- and I think some clinicians in particular -- from embracing active surveillance more universally."

Patients in the PASS cohort, which was initiated in 2008, had a median age of 64 years, and the vast majority were white (91%). Patients in the UCSF cohort were slightly younger (median age 61) and less white (79%). Men who went on to receive treatment within 6 months of their diagnosis were excluded from the present analysis.

Comment