Clinical practice guideline for management of osteoporosis and fracture prevention in Canada: 2023 update

We were pleased to see the 2023 Osteoporosis Canada guidelines in CMAJ; this much-needed update offers useful guidance on many topics that have evolved in the field. However, we were concerned to see the retained concept of a DXA-bone density(BMD) T-score as a central factor in making treatment recommendations. For over a decade, it has been recognized that T-scores are just one part of fracture risk; this is why absolute fracture risk scoring systems such as FRAX® were developed and validated, including in Canadian populations(1)(2). As the FRAX tool already includes the T-score as a factor, it begs the question as to why T-score alone would also be a stand-alone indication.
Conflation of T-score and fracture risk leads to contradictory conclusions. For example, the guidelines recommend pharmacotherapy in post-menopausal women with 10-year major fracture risk(FRAX) ≥20% or who are older than 70 with T score < -2.5 at the spine or hip. Consider a typical 73-year-old healthy woman with BMI 18 kg/m2, no additional risk factors and femoral neck T-score of -1.8 with spine T-score -2.7. According to the guidance, drug treatment is recommended. However, the FRAX Canada 10-year absolute fracture risk estimate is just 10%, which, according to the guidelines, would also indicate no treatment recommended.
A second problem with the T-score-based recommendation is that it fails to consider Canadian population modeling data showing significant drawbacks to this approach. Using the Manitoba BMD registry, Leslie showed a T-score ≤ -2.5 based treatment paradigm would render 16% of women 50-64 and 35% of women 65 or older as treatment candidates, vs just 1.4% and 17.1% using a FRAX risk ≥ 20%. The T-score strategy more than doubles the number of women on drug therapy without any meaningful additional reduction in population fractures(3). After ranking the strategies, the authors concluded a FRAX-based 20% risk strategy was the highest ranked with no other strategy being close. Curiously, this study is cited in the guidelines, but its implications are not reflected by the treatment algorithm offered.
A different approach to the question of osteoporosis treatment is that recommended by the Canadian Task Force on Preventive Health Care 2023 Osteoporosis guidance, published earlier this year in CMAJ(4). It refrains from setting either T-score or specific absolute fracture risk intervention thresholds, preferring to focus upon shared decision making at each step of the consultation. The CTFPHC approach suggests that fracture risk estimation should start with a FRAX calculation prior to ordering DXA-BMD; if it looks like the patient might desire treatment, a BMD may be ordered to refine the numerical risk estimate. Use of the provided patient decision aid(5) permits the patient to make an autonomous decision that reflects their values and preferences. The days of advising patients to start drug therapy based on T-scores alone should be largely historical.