LAUSR

Purpose Bone-modifying agents (BMAs) for bone metastases are commonly prescribed for many years even though randomized clinical trials are only 1–2 years in duration. A systematic review on the risk-benefit of BMA use for > 2 years in breast cancer or castrate-resistant prostate cancer was conducted. Methods MEDLINE, Embase, and Cochrane databases were searched (1970–February 2019) for randomized and observational studies, and case series reporting on BMA efficacy (skeletal-related events and quality of life) and toxicity (osteonecrosis of the jaw, renal impairment, hypocalcemia, and atypical femoral fractures) beyond 2 years. Results Of 2107 citations, 64 studies were identified. Three prospective and 9 retrospective studies were eligible. Data beyond 2 years was limited to subgroup analyses in all studies. Only one study ( n  = 181) reported skeletal-related event rates based on bisphosphonate exposure, with decreased rates from 27.6% (0–24 months) to 15.5% (> 24 months). None reported on quality of life. All 12 studies (denosumab ( n  = 948), zoledronate ( n  = 1036), pamidronate ( n  = 163), pamidronate-zoledronate ( n  = 522), ibandronate ( n  = 118)) reported ≥ 1 toxicity outcome. Seven bisphosphonate studies ( n  = 1077) and one denosumab study ( n  = 948) reported on osteonecrosis of the jaw. Across three studies ( n  = 1236), osteonecrosis of the jaw incidence ranged from 1 to 4% in the first 2 years to 3.8–18% after 2 years. Clinically significant hypocalcemia ranged from 1 to 2%. Severe renal function decline was ≤ 3%. Atypical femoral fractures were rare. Conclusions Evidence informing the use of BMA beyond 2 years is heterogeneous and based on retrospective analysis. Prospective randomized studies with greater emphasis on quality of life are needed. PROSPERO registration number CRD42019126813