Page 142 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

researchers should avoid interpreting analyses to suggest that men with certain demographic (or
other nonmodifiable) features are most likely to accept treatment and thus other men should not
receive the offer of treatment.
Further surveys of patients, their families, and their clinicians are warranted. To improve
reliability, these should be adequately powered to ensure that sufficient numbers of men were
treated with different interventions and to allow full analyses of the tested predictors. Studies
should use established methods including standardized qualitative research designs and, ideally,
validated questionnaires to elicit preferences. Studies of this sort also need to consider the overall
adequacy of discussion with patients regarding management options—and documentation of
those discussions. Adequate documentation of these discussions will surely improve the veracity
of some of these survey data.
Though not requested by the sponsors, future Key Questions of interest to be addressed by
systematic review and primary studies could include comparisons of interventions that improve
the likelihood that eligible men are offered AS, that improve acceptance of AS, and that improve
adherence with AS. Arguably, it is more important to first establish how to successfully get men
offered, accepting, and adhering to AS before determining which men are at greatest risk of
failing to receive AS. If no intervention successfully improves the likelihood that men will
adhere to AS, it may not be particularly relevant to flag those men most at risk of nonadherence.
Another issue for consideration could be when and how to discuss with patients the option of
transitioning from AS to either WW or other nontreatment protocols, for those patients who may
decide that they might no longer desire curative treatment regardless of progression.

Key Question 4. Active Surveillance Versus Immediate

Curative Treatment

An RCT with long followup would provide the best evidence to adequately assess the
differential effects between AS and immediate curative treatment. The least biased, most reliable
study design comparing two interventions is the RCT that adheres to modern methodological
standards. While the patient and his clinicians cannot be blinded to his treatment plan, outcome
assessors—particularly those who conduct psychometric testing—should be blinded. The
primary outcomes of interest should be the same as those listed above, under research needs for
Key Question 2, namely patient-centered clinical outcomes, including psychometric
measurements, adverse events, resource utilization, and costs. However, we acknowledge that
conducting and completing an adequately powered trial of sufficient duration may be challenging
and resource-intensive. The greatest difficulty is likely to be recruiting sufficient physicians and
patients who are willing to allow chance to dictate the choice between AS and immediate
treatment (note that the Active Surveillance Therapy Against Radical Treatment in Patients
Diagnosed With Favorable Risk Prostate Cancer (START) has been terminated because of not
g
meeting accrual target ). One will also need to factor in the probability of acceptance of the
scientific equipoise across these treatment options in maximizing recruitment (witness the rapid
enrollment in the Prostate Testing for Cancer and Treatment trial (ProtecT) discussed further in
the Discussion). Trials would then need to be of sufficiently long duration to collect data on the
clinically relevant outcomes.
In lieu of RCTs, adequate findings may be possible from long-term databases with
prospectively collected data. However, these studies ideally should also use AS protocols that are

g See http://clinicaltrials.gov/ct2/show/NCT00499174; last accessed September 30, 2011.

88