Active surveillance for prostate cancer: a systematic review of the literature

Context: Prostate cancer (PCa) remains an increasingly common malignancy worldwide. The optimal management of clinically localized, early-stage disease remains unknown, and profound quality of life issues surround PCa interventions.

Objective: To systematically summarize the current literature on the management of low-risk PCa with active surveillance (AS), with a focus on patient selection, outcomes, and future research needs.

Evidence acquisition: A comprehensive search of the PubMed and Embase databases from 1980 to 2011 was performed to identify studies pertaining to AS for PCa. The search terms used included prostate cancer and active surveillance or conservative management or watchful waiting or expectant management. Selected studies for outcomes analysis had to provide a comprehensive description of entry characteristics, criteria for surveillance, and indicators for further intervention.

Evidence synthesis: Data from seven large AS series were reviewed. Inclusion criteria for surveillance vary among studies, and eligibility therefore varies considerably (4-82%). PCa-specific mortality remains low (0-1%), with the longest published median follow-up being 6.8 yr. Up to one-third of patients receive secondary therapy after a median of about 2.5 yr of surveillance. Surveillance protocols and triggers for intervention vary among institutions. Most patients are treated for histologic reclassification (27-100%) or prostate-specific antigen doubling time <3 yr (13-48%), while 7-13% are treated with no evidence of progression. Repeat prostate biopsy with a minimum of 12 cores appears to be important for monitoring patients for changes in tumor histology over time.

Conclusions: AS for PCa offers an opportunity to limit intervention to patients who will likely benefit the most from radical treatment. This approach confers a low risk of disease-specific mortality in the short to intermediate term. An early, confirmatory biopsy is essential for limiting the risk of underestimating tumor grade and amount.