March 2010

Prostate Cancer Assessment Predicts Disease Progression

Matthew R. Cooperberg, M.D., M.P.H.
Prostate cancer specialist
UCSF Helen Diller Family Comprehensive Cancer Center

A new test developed at UCSF known as the Cancer of the Prostate Risk Assessment (CAPRA) predicts bone metastasis, prostate cancer-specific mortality and all-cause mortality when localized prostate cancer is first diagnosed.

Prostate cancer is the most common cancer diagnosed among American men, and causes more deaths annually among men than any other tumor except lung cancer. However, only a small proportion of prostate cancers are ultimately lethal. The majority of men with prostate cancer die of other causes, most commonly cardiovascular disease.

Treatments for prostate cancer — surgery, radiation therapy, hormonal therapy and others — carry a risk of complications, side effects and other impacts to a patient's long-term quality of life. Before making treatment decisions, it is critical to estimate the likelihood that a tumor will recur after treatment, progress or pose a threat to life.

The goal of risk assessment is to detect patients at high risk of mortality and treat them aggressively. It is also useful for moderate- to low-risk patients when guiding their treatment or planning active surveillance.

More than 100 risk assessment tests have been developed in recent years, such as the D'Amico classification and various nomograms. Most tests are unable to predict long-term outcomes, apply to just one form of treatment, are not well-validated and do not apply to many patients treated in the community. Many tests are cumbersome to calculate and may be confusing to patients.

In a recent UCSF study, reported in the June 9, 2009 edition of the Journal of the National Cancer Insitute, the CAPRA score accurately predicted mortality across multiple forms of treatment and helped patients and clinicians decide which tumors need to be treated and how aggressively. For research purposes, the test classified men into low, intermediate and high risk groups.

Moreover, the CAPRA test was independently validated in three studies as accurate and consistent in predicting pathological and biochemical outcomes after radical prostatectomy.


The test provides a straightforward 0 to 10 score. It is nearly as easy to calculate as the D'Amico classification, yet with accuracy comparable to the best nomograms. The score is calculated using points assigned to:

  • Age at diagnosis
  • Gleason score, a measure of how aggressive the cancer cells appear under the microscope
  • PSA score, the prostate-specific antigen level in the blood
  • Percentage of biopsy scores that test positive for cancer
  • Clinical tumor stage based on digital exam of the prostate and ultrasound

To calculate your patient's CAPRA score, use our online chart.

Study Results

To determine the accuracy of CAPRA, 10,627 men — with clinically localized prostate cancer who underwent primary radical prostatectomy, external beam radiation therapy, brachytherapy, androgen deprivation monotherapy or active surveillance — were studied.

Age at diagnosis, clinical tumor stage, percentage of biopsy cores positive for cancer, Gleason score, and prostate-specific antigen (PSA) level were used to calculate CAPRA scores at diagnosis and the patients were followed up for a median of 71.3 months.

The average CAPRA score was 3.1. Based on score ranges of 0—2, 3—5 and 6—10, 48.7 percent of patients were categorized as low risk, 38.0 percent as intermediate risk, and 13.3 percent as high risk, respectively.

During follow up, 2.9 percent of patients developed bone metastases, 2.4 percent died of prostate cancer, and 14.9 percent died of "any cause."

Prostate cancer-specific survival and overall survival at 10 years ranged from 96.7 percent and 76.7 percent, respectively, for patients with a CAPRA score of 1, to 78.9 percent and 41.5 percent for those with a score of 8—10.

Each point increase in CAPRA score was associated with a 47 percent relative increase in likelihood of bone metastases, 39 percent relative increase in likelihood of cancer-specific mortality, and 13 percent relative increase in likelihood all-cause mortality. The accuracy of the CAPRA score for predicting metastases, cancer-specific mortality, and all-cause mortality, was estimated at 78 percent, 80 percent, and 71 percent, respectively.

Other Resources

For more information, contact the Physician Referral Service at UCSF Medical Center:

Phone (888) 689-UCSF or (888) 689-8273