Physical activity decreases the risk of cancer reclassification in patients on active surveillance: a multicenter retrospective study.

ABSTRACT

BACKGROUND: Physical activity (PA) is associated with favorable outcomes in prostate cancer (PCa) patients. We assessed its effect on the risk of PCa reclassification (PCaR) during active surveillance. METHODS: Anthropometric, demographic, and clinical data concerning men diagnosed with a low-risk PCa and initially managed with active surveillance at the two participating institutions were retrospectively collected. The Physical Activity Scale for the Elderly (PASE) was used for patients' self-assessment of their daily exercise and their consequent stratification into three groups sedentary (PASE ≤ 65), moderately active (65 < PASE < 125), active (PASE ≥ 125). Kaplan-Meier model was used to evaluate the predictive role of PA on PCaR, computed at 2, 5, 10 years after diagnosis; differences between lifestyle groups were assessed using the log-rank and uni-/multivariable Cox analyses applied to identify predictors of reclassification. RESULTS: Eighty-five patients were included in the analysis, with a median age of 66 years (IQR 59-70); 16% were active, 45% were former smokers, and 3 presented with metabolic syndrome (MetS). Prostate-specific antigen (PSA) density was 0.12 (IQR 0.07-0.15); 34 men showed a PSA doubling time <10 years. The Median PASE score was 86 (IQR 61.5-115.8) 24 patients were sedentary, 46 moderately active, and 15 active. At a median follow-up of 37 months (IQR 14-53), 25% of patients experienced PCaR. These were less physically active (PASE score 69.3 vs 87.8; p = 0.056) and presented with significantly smaller prostates (46 ml vs 50.7 ml; p = 0.001) and a higher PSAD (0.14 vs 0.10; p = 0.019). At 2 years, the risk of reclassification was 25 ± 5%, while it was 38 ± 7% at both 5 and 10 years. The risk was significantly different in the three PA groups (Log Rank p = 0.033). PASE score was the only independent predictor of PCaR (HR 0.987; 95%CI 0.977-0.998; p = 0.016). CONCLUSIONS: PA influences PCa evolution, as increasing levels are associated with a significantly reduced risk of tumor reclassification among patients undergoing active surveillance.