Background Identification of patients who will sustain a fragility fracture (FF) is the main objective of the WHO Fracture Risk Assessment score (FRAX®).
Objectives To describe the FRAX scores at the time of an incident FF and the rates of subsequent fractures according to FRAX category over 4 years following this clinical FF.
Methods An ongoing prospective cohort of men and women over 50 years of age was followed up in the OPTIMUS study, a placebo-controlled intervention aimed at increasing the rate of initiation and persistence on osteoporosis treatment after an incident FF leading to an orthopaedic consultation. At year 1, 56% were then receiving an effective osteoporosis treatment. The occurrence of novel FF was obtained from patients during phone follow-ups over 4 years. Canadian-adjusted FRAX scores were calculated from baseline and follow up information, without BMD.
Results From January 2007 to June 2011, 1172 patients (961 women; Mean age 68.6 years) with FF were included. FRAX score was not calculated in 20 patients because of missing height and/or weight values. BEFORE the incident FF, 596/1152 (51.7%) were attributed a Low, 70 (6.1%) a Moderate, and 486 (42.2%) a High risk, according to FRAX. AFTER the fracture, 314 (27.3%) patients were still scored Low, 217 (18.8%) Moderate and 621 (53.9%) High risks. Prior (After) the FF, FRAX-estimated risk was estimated High in 74.2 (86.2)% of Hip, 47.0 (55.0)% of Proximal Humerus, 47.1 (64.7)% of vertebra, 34.8 (48.9)% of wrist, 20.7 (30.0)% of ankle and 36.1 (50.4)% of Other Minor sites, respectively. Over 2285 patient-years of follow up, 99 recurrent FF occurred in 86 patients. The 10-year risk was estimated as High in 57 (66.3%) episodes of recurrent FF. Odds ratios (ORs) for recurrent fractures were thus 1.77 [CI 1.01-3.09] for post-FF High vs Low risk, and 1.02 [CI 0.48-2.19] for Moderate vs Low risk patients.
Conclusions The Canadian-adjusted FRAX® High-risk category was NOT sensitive to identify patients BEFORE they sustained an incident FF: 74% of Hip but only 34% of non-Hip FF. Thus, Primary intervention based on FRAX® High Risk only is unlikely to decrease non-Hip fracture rates. Relative to High risk defined as ≥20% for Major FF only, combining estimated risk ≥3% for Hip with risk ≥20% for Major FF increased the size of the High risk category by over 60%, mostly at the expense of the Moderate category. AFTER an incident FF, the FRAX® High-risk category was statistically but not CLINICALLY useful to predict which patient would present a subsequent FF (66% of all recurrent FF patients; OR < 2 for High versus Low and Moderate risk). Even patients with incident FF at Minor sites and those classified post-FF at FRAX® Low/Moderate risk had high rates of recurrent FF. This suggests that all recent FF patients might be considered for therapy, irrespective of their FRAX® estimated 10-year risk.
Disclosure of Interest P.-M. April: None Declared, F. Cabana: None Declared, M.-C. Beaulieu: None Declared, M. Beaulieu Employee of: Merck Canada, S. Roux: None Declared, G. Boire Grant/research support from: Merck Canada, The Alliance for Better Bone Health, Novartis Canada, Amgen Canada, Warner Chilcott Canada, Servier Canada