Dr. Etzioni and her colleagues used data from several sources. For their investigation, times from PSA-recurrence to metastasis were estimated from 441 RP patients treated at Johns Hopkins University in Baltimore, MD, who did not receive salvage therapy at PSA-recurrence. Estimated life expectancy was based on data from patients from the Cancer of the Prostate Strategic Urologic Research Endeavor (CaPSURE) database, and data from the Surveillance, Epidemiology, and End Results (SEER) registry.

After mining these datasets, the researchers created 1 million virtual patients with different age and disease characteristics. They computed the fraction of patients with PSA recurrence for which the time from PSA recurrence to another cause of death was less than the time from PSA recurrence to metastasis in the absence of salvage therapy.

Continue Reading

“There are no clinical trials of the benefit of early treatment of PSA recurrence, at PSA rise versus later, at the time that metastases become apparent,” said Dr. Etzioni. “There are some observational studies that have happened to treat people earlier versus later and they try to look at the question of whether the people who were treated earlier did better. But there are many ways these studies can be biased and in my opinion they all are. We are working on statistical models that will address this bias.”

RELATED: ASCO Endorses AUA/ASTRO Treatment Guideline Regarding Radiotherapy After Prostatectomy

She said clinicians should be cautious before rushing to treatment, and assess the need based on disease risk and patient characteristics such as age and comorbidities. However, Mark Garzotto, MD, of Oregon Health & Science University in Portland, OR, said these numbers are only based on a simulation from several different datasets and must be viewed cautiously. He said there were significant limitations in the datasets used to create the modeling for this investigation

“One must remember this is a computer simulation based on merging of independent datasets. Thus, these are predicted and not actual patient events. It takes a leap of faith to put enough trust in this information to the degree that you would actually use it for clinical decision-making,” Dr. Garzotto told Cancer Therapy Advisor

“What makes this model difficult to accept is that we have data from large trials showing that adjuvant treatment lowers the rate of metastases and death.  In the model created here, the authors here used data from a study that lowered the rate of metastases by 80%.”

He said another major limitation is that the follow-up (4 to 7 years) is relatively short on the patient datasets used for this investigation. CaPSURE is a longitudinal, observational study of approximately 15,000 men that started in 1995.2

Patients involved have all stages of biopsy-proven prostate cancer and were enrolled at 43 community urology practices, academic medical centers, and Veterans Affairs hospitals throughout the United States. The majority of men are diagnosed with stage b cancer and they receive a variety of treatments (surgery, irradiation, hormonal therapy, and watchful waiting).


  1. Xia J, Trock BJ, Gulati R, et al. Overdetection of recurrence after radical prostatectomy: estimates based on patient and tumor characteristics. Clin Cancer Res. 2014;20(20):5302-5310.
  2. University of California, San Francisco. Department of Urology. CaPSURE: UCSF Cancer of the Prostate Strategic Urologic Research Endeavor. https://urology.ucsf.edu/research/cancer/capsure. Accessed November 26, 2014.