Daniel J George, Karim Fizazi, Alicia K Morgans, Hannah D McManus, Yen Yen Yip, Shankar Srinivasan, Maxime Duflot, Alexander Upton, Fred Saad
In nmCRPC, darolutamide had a higher NNH (lower risk) than apalutamide and enzalutamide for the all-grade AEs hot flash (100.0/not reported [NR])/20.0), fatigue (29.4/10.8/5.3), arthralgia (negative [NEG]/11.9/100.0), cognitive impairment (≥ 250/47.6/33.3), hypertension (71.4/20.0/14.3), and falls (NEG/15.2/14.3) in the primary analysis, a trend generally maintained in the final analysis. A similar trend was reported for most grade ≥ 3 AEs. In mHSPC, darolutamide also had a higher NNH for most all-grade/grade ≥ 3 AEs. Exceptions included all-grade constipation (higher NNH for apalutamide and enzalutamide) and anemia (higher NNH for apalutamide). NNH comparisons across trials should be interpreted cautiously because of heterogeneity in trial designs and study populations.
INTRODUCTION: Androgen receptor inhibitors (ARIs) are standard of care in prostate cancer. Patient-centered treatment goals include optimizing efficacy, safety, tolerability, and quality of life, and prolonging time to treatment change. Number needed to harm (NNH) analyses can help contextualize adverse event (AE) risk and support treatment decisions. This exploratory study assessed NNH for AEs of interest with ARIs in nonmetastatic castration-resistant prostate cancer (nmCRPC) and metastatic hormone-sensitive prostate cancer (mHSPC).
METHODS: NNH was calculated using AE data from pivotal trials comparing ARIs plus androgen deprivation therapy (ADT) versus ADT and derived using the reciprocal of the absolute risk increase, then compared between ARIs.
RESULTS: In nmCRPC, darolutamide had a higher NNH (lower risk) than apalutamide and enzalutamide for the all-grade AEs hot flash (100.0/not reported [NR])/20.0), fatigue (29.4/10.8/5.3), arthralgia (negative [NEG]/11.9/100.0), cognitive impairment (≥ 250/47.6/33.3), hypertension (71.4/20.0/14.3), and falls (NEG/15.2/14.3) in the primary analysis, a trend generally maintained in the final analysis. A similar trend was reported for most grade ≥ 3 AEs. In mHSPC, darolutamide also had a higher NNH for most all-grade/grade ≥ 3 AEs. Exceptions included all-grade constipation (higher NNH for apalutamide and enzalutamide) and anemia (higher NNH for apalutamide). NNH comparisons across trials should be interpreted cautiously because of heterogeneity in trial designs and study populations.
DISCUSSION: Similar NNH trends observed in nmCRPC and mHSPC patient populations suggest a lower risk of incremental harm for most AEs with darolutamide versus apalutamide and enzalutamide. Understanding ARI risk profiles is key to optimizing treatment selection and improving tolerability in patients.