Page 45 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

Experimental Studies
For this category, we included studies of any design that evaluated any tool (such as a
decision aid) or other intervention designed specifically to have an impact on acceptance of AS.

Key Question 4 (Comparative Effectiveness of Observational
Management Strategies and Active Treatment)
We considered studies that fulfilled the following criteria:

Population. Men with clinically localized prostate cancer (T1-T2), without (or unable to assess)
either regional lymph nodes involvement (N0-X) or metastases (M0-X), regardless of age,
histologic grade, Gleason score, or prostate-specific antigen (PSA) concentration. Studies that
enrolled mixed populations of clinically localized and more advanced disease were included only
if men with more advanced disease stages represented less than 20 percent of the study sample,
or if they reported separate treatment effect estimates for the clinically localized subgroup of
patients.

Intervention. Observational (no immediate active treatment) management strategies, including
both WW and AS strategies. We excluded studies where the observational management group
was combined with the group of patients receiving either medical or surgical ADT (i.e.,
orchiectomy). We also excluded studies defining observational management strategies as the
lack of surgical treatment or radiotherapy, without providing information on how other
treatments (such as ADT) were handled in the analyses. However, studies that provided no
definition of the observational strategy employed (e.g., “expectant management” with no other
information on how treatment groups were defined) were included.

Comparators. Radical prostatectomy (RP), external beam radiation therapy (EBRT), or
brachytherapy (BT), all with or without ADT. Based on input from the TEP and AHRQ
representatives, ADT was not considered a potentially curative treatment and studies that
compared observational management strategies only with ADT monotherapy were excluded.

Outcomes. Outcomes of interest included prostate cancer-specific mortality, all-cause mortality,
morbidity of primary treatment (including the frequency of procedures to address treatment-
related morbidity), development of metastatic disease, quality of life (QoL, including satisfaction
with treatment), and costs. Eligible studies had to report or provide sufficient data to allow the
estimation of the treatment effect (e.g., hazard ratios, odds ratios, risk differences, or risk ratios
along with sufficient statistics to calculate the uncertainty around these estimates) or provide the
P value from a test of association of the treatments examined with the outcomes of interest. For
treatment-related costs, we considered estimates from primary studies of actual patient costs (i.e.,
based on data collected during patient followup) of any observational management strategy (AS
and WW).

Study design and analysis. We considered both randomized controlled trials, and prospective or
retrospective nonrandomized comparative studies with longitudinal followup from any country.
We excluded cross-sectional and case-control studies. Eligible observational studies had to be
conducted in a multicenter setting in any country or to have utilized databases sourced from the
U.S. population (such as SEER, CaPSURE, the Prostate Cancer Outcomes Study [PCOS], the

10