Qurrat U van den Blink, Kate Garcez, Caroline C Henson, Susan E Davidson, Claire E Higham

Pelvic radiotherapy is a treatment delivered to an estimated 150,000 to 300,000 people annually across high-income countries. Fractures due to normal stresses on weakened bone due to radiotherapy are termed insufficiency fractures. Pelvic radiotherapy-related interruption of the blood supply to the hip is termed avascular necrosis and is another recognised complication. The reported incidences of insufficiency fractures are 2.7% to 89% and risk of developing avascular necrosis is 0.5%. These complications lead to significant morbidity in terms of pain, immobility and consequently risk of infections, pressure sores and mortality. To assess the effects of pharmacological interventions for preventing insufficiency fractures and avascular necrosis in adults over 18 years of age undergoing pelvic radiotherapy. We performed electronic literature searches in the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase and DARE to 19 April 2017. We also searched trial registries. Further relevant studies were identified through handsearching of citation lists of included studies. Randomised controlled trials (RCTs) or non RCTs with concurrent comparison groups including quasi-RCTs, cluster RCTs, prospective cohort studies and case series of 30 or more participants were screened. We included studies assessing the effect of pharmacological interventions in adults over 18 years of age undergoing radical pelvic radiotherapy as part of anticancer treatment for a primary pelvic malignancy. We excluded studies involving radiotherapy for bone metastases. We assessed use of pharmacological interventions at any stage before or during pelvic radiotherapy. Interventions included calcium or vitamin D (or both) supplementation, bisphosphonates, selective oestrogen receptor modulators, hormone replacement therapy (oestrogen or testosterone), denosumab and calcitonin. Two review authors independently assessed trial quality and extracted data. We contacted study authors to obtain missing data. Data were to be pooled using the random-effects model if study comparisons were similar, otherwise results were to be reported narratively. We included two RCTs (1167 participants). The first RCT compared zoledronic acid with placebo in 96 men undergoing pelvic radiotherapy for non-metastatic prostate cancer.The second RCT had four treatment arms, two of which evaluated zoledronic acid plus adjuvant androgen suppression compared with androgen suppression only in 1071 men undergoing pelvic radiotherapy for non-metastatic prostate cancer.Both studies were at a moderate to high risk of bias and all evidence was judged to be of very low certainty.The studies provided no evidence on the primary outcomes of the review and provided limited data in relation to secondary outcomes, such that meta-analyses were not possible. Both studies focused on interventions to improve bone health in relation to androgen deprivation rather than radiation-related insufficiency fractures and avascular necrosis. Few fractures were described in each study and those described were not specific to insufficiency fractures secondary to radiotherapy. Both studies reported that zoledronic acid in addition to androgen deprivation and pelvic radiotherapy led to improvements in BMD; however, the changes in BMD were measured and reported differently. There was no available evidence regarding adverse effects. The evidence relating to interventions to prevent insufficiency fractures and avascular necrosis associated with pelvic radiotherapy in adults is of very low certainty. This review highlights the need for prospective clinical trials using interventions prior to and during radiotherapy to prevent radiation-related bone morbidity, insufficiency fractures and avascular necrosis. Future trials could involve prospective assessment of bone health including BMD and bone turnover markers prior to pelvic radiotherapy. The interventions for investigation could begin as radiotherapy commences and remain ongoing for 12 to 24 months. Bone turnover markers and BMD could be used as surrogate markers for bone health in addition to radiographic imaging to report on presence of insufficiency fractures and development of avascular necrosis. Clinical assessments and patient reported outcomes would help to identify any associated adverse effects of treatment and quality of life outcomes.