5e. Fracture risk

5e.1 Background and Evidence Review

A safety notice for fracture risk with SGLT-2 inhibitors was published following increased incidence of upper and lower limb fracture in the CANVAS trial of canagliflozin (HR=1.26, 95%CI 1.04-1.52). The excess fracture risk was detected in only one of the two large subcohorts that comprise the CANVAS Program of trials (i.e. it was not apparent in CANVAS-R), for reasons that could not be explained by baseline participant demographic or protocol heterogeneity, and could conceivably represent a finding which resulted from the play of chance (37). A fracture risk with canagliflozin was not identified in the CREDENCE trial of canagliflozin, nor has it been identified in large placebo-controlled trials of other SGLT-2 inhibitors (including those inclusive of participants with low eGFR) (1-3, 8, 12, 13, 21-24, 45). Outside of trial populations, a recent systematic review of 37 large population-based studies found no association between SGLT-2 inhibitor prescription and fractures (55). A hypothesised link to hypovolaemia-related falls has not been substantiated, although the incidence of non-serious falls is often not recorded in large outcome trials (56).

Data from preclinical and phase I studies have reported short-term alterations in mineral biochemistry, including increases in phosphate, FGF-23 and parathyroid hormone (PTH) levels with SGLT-2 inhibitor use (57, 58), which were replicated in a cohort of 31 participants with type 2 DM and albuminuria treated with dapagliflozin (59). A trial of canagliflozin versus placebo in older individuals with type 2 DM demonstrated significantly reduced bone mineral density at the hip, but at no other site, with canagliflozin treatment over 104 weeks (60).

Whilst experimental data indicate that SGLT-2 inhibitors may modify bone mineral metabolism, precise mechanisms have not been elucidated. Based on current evidence, it is likely that either these bone metabolism changes do not translate into increased fracture risk, or that any small increased fracture risk with canagliflozin is outweighed by the cardiorenal benefits in populations at risk of heart failure or progressive CKD. Our recommendation therefore highlights the importance of routine CKD-mineral bone disease (MBD) monitoring and management in these individuals. More research is required into the mechanisms of SGLT-2 inhibitor-induced bone biochemical alterations, potential interactions with other drugs that modify osteoporosis risk (e.g. thiazolidinediones) and the clinical significance of this in a CKD +/- type 2 DM population already at risk of bone disease.

5e.2 Recommendations for Implementation

1. In people with CKD treated with SGLT-2 inhibitors, we suggest monitoring of bone parameters including calcium, phosphate and PTH should be performed as appropriate for CKD stage (NICE NG203) (Grade 2D).

Rationale: Whilst there has been report of an increased risk of fractures in one trial where participants were treated with canagliflozin, this has not been confirmed in any other study and may represent the play of chance. People with CKD are at increased risk of bone disease and their clinician should be monitoring them to ensure that interventions are utilised to maintain good bone health irrespective of the prescription of SGLT-2 inhibitors. NICE NG203 CKD guidance is available at https://www.nice.org.uk/guidance/ng203.

5e.3 Clinical Research Recommendations

1. Establishing any long-term impact of SGLT-2 inhibition on the development and progression of CKD-MBD
2. Establishing if SGLT-2 inhibition modifies osteoporosis risk posed by thiazolidinediones.