When do bisphosphonates make the most sense?
A Cochrane Musculoskeletal Group review
Should you prescribe bisphosphonates for postmenopausal patients for primary as well as secondary prevention of osteoporotic fractures? Here’s what the evidence tells us.
TABLE 1
Fracture risk reduction: How the bisphosphonates compare*
| Study | Vertebral fracture RR (95% CI) | Nonvertebral fracture (hip, wrist, others) RR (95% CI) | Hip fracture RR (95% CI) | Wrist fracture RR (95% CI) |
|---|---|---|---|---|
| Primary prevention | ||||
| Alendronate3 | 0.55 (0.38-0.80) | 0.89 (0.76-1.04) | 0.79 (0.44-1.44) | 1.19 (0.87-1.62) |
| Risedronate4 | 0.97 (0.42-2.25) | 0.81 (0.25-2.58) | N/A | N/A |
| Secondary prevention | ||||
| Alendronate3 | 0.55 (0.43-0.69) | 0.77 (0.64-0.92) | 0.47 (0.26-0.85) | 0.50 (0.34-0.73) |
| Risedronate4 | 0.61 (0.50-0.76) | 0.80 (0.72-0.90) | 0.74 (0.59-0.94) | 0.67 (0.42-1.07)† |
| Ibandronate15,16 Oral daily Oral intermittent | 0.62 (0.42-0.75) 0.50 (0.26-0.66) | No effect)‡ No effect | N/A N/A | N/A N/A |
| Zoledronic acid22 | 0.30 (0.2-0.38) | 0.75)§ (N/A) | 0.59¶ (0.42-0.83) | N/A |
| CI, confidence interval; N/A, not available; RR, relative risk. *Bold type indicates statistical significance (P<.05). †P=.10. ‡RR of nonvertebral fracture was 0.69 (P=.013) for daily oral ibandronate in the subgroup with femoral neck BMD T-score <–3.0. §P<.001. ¶Hazard ratio. | ||||
What are the absolute benefits? A look at number needed to treat
In addition to looking at the RR, the authors of both the alendronate and risedronate reviews calculated the number needed to treat (NNT) to prevent one fracture (TABLE 2) in the trial participants;3,4 they focused on the secondary prevention outcomes, as these were statistically significant. The reviewers also estimated what the NNT would be if the risk reductions achieved with alendronate and risedronate in the reviews occurred when treating community-based samples of women at moderate compared with high fracture risk.
The biggest differences involved hip fracture: For alendronate, if a community-based sample of women at moderate risk of fracture were treated with the drug and the reduction in RR seen in the secondary prevention trials applied, the NNT would be 100. Thus, for every 100 women treated for 5 years with alendronate, 1 hip fracture would be prevented. However, if this same RR reduction were applied to women at high risk of fracture, the NNT would be only 22.3 For risedronate, the estimated NNT to prevent one hip fracture in women at moderate risk was 203, compared with only 45 for women at high risk.4 These estimates indicate that the benefits of bisphosphonate therapy in preventing fractures are greatest in women with a high underlying fracture risk.
TABLE 2
NNT analysis: Women at higher risk are most likely to benefit3,4
| NNT | |||
|---|---|---|---|
| Observed in secondary prevention trials in reviews | Estimated for community-based sample of women with | ||
| High fracture risk* | Moderate fracture risk* | ||
| Alendronate (10 mg/d) | |||
| Vertebral fracture | 19 | 20 | 42 |
| Nonvertebral fracture | 47 | 16 | 27 |
| Hip fracture | 146 | 22 | 100 |
| Wrist fracture | 69 | N/A | N/A |
| Risedronate (5 mg/d) | |||
| Vertebral fracture | 19 | 23 | 49 |
| Nonvertebral fracture | 49 | 19 | 31 |
| Hip fracture | 138 | 45 | 203 |
| Wrist fracture | N/A | N/A | N/A |
| N/A, not available; NNT, number needed to treat. *NNT calculated by applying the relative risk reduction observed in the reviews to published estimates of 5-year fracture risk in a community-based sample of women >50 years of age at moderate and high risk. | |||
Adverse effects do not increase with longer-term treatment
In both the alendronate and risedronate reviews, adverse effects and the risk of discontinuing treatment due to adverse events were similar in the intervention and control groups.3,4 Postmarketing data suggest that there is potential for upper gastrointestinal (GI) problems, however;7 osteonecrosis of the jaw has also been reported infrequently.8,9 More recently, there have been reports of a possible link between bisphosphonates and atypical femoral fractures, which we’ll say more about in a bit.
Some potential adverse events—eg, osteonecrosis of the jaw and atypical femoral fractures—may be related to treatment duration. The maximum duration of the trials included in these meta-analyses was 4 years for alendronate and 3 years for risedronate. However, additional published data do not appear to support a relation between adverse events and treatment duration.
For alendronate, researchers extended the Fracture Incidence Trial (FIT) for a 10-year follow-up,10,11 comparing women who took the drug for the first 5 years with women who took it for 10 years. Adverse effects were similar in both groups.
For risedronate, researchers followed a small subsample (n=164) of the participants in the Vertical Efficacy with Risedronate Therapy (VERT) Study Group for up to 7 years.12,13 For the first 5 years, half of the participants took 5 mg/d risedronate, while the other half took a placebo. During the final 2 years, all participants received 5 mg/d risedronate. The incidence of adverse events among those who took the drug for 7 years was similar to that reported in the first 3 years of the original trial.13
Ibandronate studies focus on dose
Nonvertebral fracture. The Cochrane systematic review examining ibandronate for postmenopausal osteoporosis is not yet completed.5 However, Cranney et al performed a pooled analysis of individual patient data from 8 RCTs to examine the efficacy of different doses of the drug for the secondary prevention of nonvertebral fracture.14 (No studies of the drug for primary prevention have been done.) After 2 years of treatment at higher doses of ibandronate (annual cumulative exposure ≥10.8 mg, equivalent to 150 mg orally/month, 3 mg IV quarterly, or 2 mg IV every 2 months), the hazard ratio was 0.62 (95% CI, 0.396-0.974), compared with those taking lower doses (annual cumulative exposure of 5.5 mg). The individual results of the 2 largest trials did not demonstrate an effect on nonvertebral fracture, except in the subgroup of women with very low femoral neck bone mineral density (BMD) (T-scores <–3.0). 15-17
