Effects of Exercise on Cardiorespiratory Fitness and Biochemical Progression in Men With Localized Prostate Cancer Under Active Surveillance: The ERASE Randomized Clinical Trial

Abstract

Importance: Men with prostate cancer who are undergoing active surveillance are at an increased risk of cardiovascular death and disease progression. Exercise has been shown to improve cardiorespiratory fitness, physical functioning, body composition, fatigue, and quality of life during and after treatment; however, to date only 1 exercise study has been conducted in this clinical setting.

Objective: To examine the effects of exercise on cardiorespiratory fitness and biochemical progression in men with prostate cancer who were undergoing active surveillance.

Design, setting, and participants: The Exercise During Active Surveillance for Prostate Cancer (ERASE) trial was a single-center, 2-group, phase 2 randomized clinical trial conducted at the University of Alberta, Edmonton, Canada. Eligible patients were recruited from July 24, 2018, to February 5, 2020. Participants were adult men who were diagnosed with localized very low risk to favorable intermediate risk prostate cancer and undergoing active surveillance. They were randomized to either the high-intensity interval training (HIIT) group or usual care group. All statistical analyses were based on the intention-to-treat principle.

Interventions: The HIIT group was asked to complete 12 weeks of thrice-weekly, supervised aerobic sessions on a treadmill at 85% to 95% of peak oxygen consumption (V̇o2). The usual care group maintained their normal exercise levels.

Main outcomes and measures: The primary outcome was peak V̇o2, which was assessed as the highest value of oxygen uptake during a graded exercise test using a modified Bruce protocol. Secondary and exploratory outcomes were indicators of biochemical progression of prostate cancer, including prostate-specific antigen (PSA) level and PSA kinetics, and growth of prostate cancer cell line LNCaP.

Results: A total of 52 male patients, with a mean (SD) age of 63.4 (7.1) years, were randomized to either the HIIT (n = 26) or usual care (n = 26) groups. Overall, 46 of 52 participants (88%) completed the postintervention peak V̇o2 assessment, and 49 of 52 participants (94%) provided blood samples. Adherence to HIIT was 96%. The primary outcome of peak V̇o2 increased by 0.9 mL/kg/min in the HIIT group and decreased by 0.5 mL/kg/min in the usual care group (adjusted between-group mean difference (1.6 mL/kg/min; 95% CI, 0.3-2.9; P = .01). Compared with the usual care group, the HIIT group experienced decreased PSA level (-1.1 μg/L; 95% CI, -2.1 to 0.0; P = .04), PSA velocity (-1.3 μg /L/y; 95% CI, -2.5 to -0.1; P = .04), and LNCaP cell growth (-0.13 optical density unit; 95% CI, -0.25 to -0.02; P = .02). No statistically significant differences were found in PSA doubling time or testosterone.

Conclusions and relevance: The ERASE trial demonstrated that HIIT increased cardiorespiratory fitness levels and decreased PSA levels, PSA velocity, and prostate cancer cell growth in men with localized prostate cancer who were under active surveillance. Larger trials are warranted to determine whether such improvement translates to better longer-term clinical outcomes in this setting.

Trial registration: ClinicalTrials.gov Identifier: NCT03203460.

Figure 1.. CONSORT Diagram

Figure 2.. Changes in Prostate-Specific Antigen (PSA), PSA Doubling Time, PSA Velocity, and Testosterone

Means are based on unadjusted data. Error bars indicate 95% CIs, and P values indicate between-group difference at the postintervention period adjusted for the baseline values of the outcome and resistance exercise behavior. HIIT indicates high-intensity interval training; PSAV, PSA velocity.

Figure 3.. Changes in LNCaP Cell Line Growth

Each bar represents the unadjusted change in LNCaP cell line growth in each participant from baseline to the postintervention period. The overall percentage of mean difference between the high-intensity interval training (HIIT) and usual care groups was statistically significant (−5.1%; P = .02). The analysis was adjusted for the baseline values and resistance exercise behavior.