Patient preferences for life expectancy cutoffs for aggressive treatment in clinically localized prostate cancer.

Abstract

Background: Guidelines for prostate cancer treatment in men with limited life expectancy are based on expert opinion. Patient preferences for when to defer treatment based on longevity are unknown. We sought to define life expectancy thresholds at which men are more likely to choose conservative management in the context of varying risks of cancer death and treatment-related side effects.

Materials and methods: We crowdsourced a conjoint analysis exercise to 2,046 men sociodemographically matched to a US prostate cancer population. Subjects were given a longevity estimate based on their age and comorbidity. They then chose between treatment and conservative management across scenarios with varying risks of cancer death at 5, 10, and 15 years, erectile dysfunction, urinary incontinence, and irritative urinary symptoms. Multivariable multinomial logistic regression identified the life expectancy threshold when men were more likely to choose conservative management over treatment.

Results: Across all men, there was a significant interaction between longevity and treatment choice (P < 0.001), with probability of treatment decreasing 15% for every 5-year decrease in life expectancy (OR0.85, 95% CI0.82-0.89). Across all tumor risk subtypes, men were significantly more likely to choose conservative management at life expectancy<10 years(OR<1, P < 0.05). For low-, favorable-intermediate-, unfavorable-intermediate-, and high-risk cancers, men were more likely to choose conservative management at life expectancy thresholds of ≤15, ≤10, ≤9, and ≤7 years, respectively (P < 0.05).

Conclusions: Preferences for when to consider conservative management of prostate cancer based on longevity align with current guidelines recommendations, except for low-risk disease, for which men are likely to consider conservative management at even higher life expectancy thresholds.