Page 131 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

retrospective cohort studies 207-209,218,223,224,229-232 were included. The two RCTs reported updated
results from the SPCG-4 and UMEA1 trials that had been included in the previously mentioned
208
9
systematic review. Among the 13 cohort studies, sample size ranged from 113 226 to 44,630
226
and followup duration from 12 months to more than 13 years. 209 Methodological quality of the
studies was rated as A in one study, 213 B in nine studies, 207-209,214,218,223,224,229,230 and C in five
studies. 226-228,231,232

d
Comparison Between Observational Management Strategies and Radical Prostatectomy
Appendix Table 4.2 presents the detailed results from the one RCT and its ancillary
investigation, as well as 14 cohort studies that compared observational management strategies
with RP.

Randomized controlled trials. The SPCG-4 RCT followed 695 men with localized prostate
cancer for a median of 12.8 years. 213 Compared with men on WW, men treated with RP had
significantly lower prostate cancer-specific mortality (RR 0.62; 95 percent CI 0.44, 0.87; P =
0.01), all-cause mortality (RR 0.75; 95 percent CI 0.61, 0.92; P = 0.007), and incidence of distant
metastases (RR 0.59; 95 percent CI 0.45, 0.79; P <0.001).
Subgroup analyses found no significant modification of the treatment effect on mortality by
PSA (< 10 vs. ≥ 10 ng/mL, interaction P = 0.72 and P = 0.30, for overall and prostate cancer-
specific mortality, respectively) or Gleason score (<7 vs. ≥7, interaction P = 0.36 and P = 0.52,
for overall and prostate cancer-specific mortality, respectively). However, age was found to be a
modifier of the treatment effect (interaction P = 0.003 when age was dichotomized at 65 years of
age; P = 0.001 when treating age as a continuous variable). The favorable effects of RP on
overall mortality were present among men younger than 65 years (HR = 0.52; 95 percent CI
0.37, 0.73; P < 0.001), but not men older than 65 years (HR = 0.98; 95 percent CI 0.75, 1.28;
P = 0.89). Effect modification by age did not reach statistical significance for prostate cancer-
specific mortality (interaction P = 0.16; HR = 0.49; 95 percent CI 0.31, 0.79; and HR = 0.76; 95
percent CI, 0.25, 2.32; comparing men younger vs. older than 65 years, respectively). The
authors reported that none of the subgroup analyses performed were specified in the main study
protocol but were determined “before any data were seen.” 213 Thus, the study may not have had
adequate power to detect effect modification; the absence of statistically significant interactions
does not indicate that clinically meaningful differences do not exist between the investigated
subgroups.
An additional subgroup analysis limited to low-risk prostate cancer patients (124 in the RP
group and 139 in the WW group) was also reported. Low-risk cancers were defined as PSA < 10
ng/mL and Gleason score < 7 or WHO grade of 1 (for tumors diagnosed only by cytologic
assessment). In this subgroup, the HRs comparing RP versus WW were 0.62 (95 percent CI 0.42,
0.92; P=0.02), 0.53 (95 percent CI 0.24, 1.14; P=0.14) and 0.43 (95 percent CI 0.23, 0.79;
P=0.008), for overall survival, prostate-cancer specific mortality and development of metastatic
disease, respectively. The study reported that the most common non-fatal complications in the
RP group within 1year of surgery were impotence and urinary leakage, with cumulative one-year
incidence proportions of 58.1 percent and 32.2 percent, respectively. Urinary obstruction,

d During the conduct of our review, preliminary results from the Prostate cancer Intervention Versus Observation
Trial (PIVOT, NCT00007644), a randomized trial of RP versus WW with palliative intervention, were presented at
the 2011 annual meeting of the American Urological Association (AUA). We did not include this study in our
evidence tables because no full text publication was available.

77