• CONTEXT
    • Bisphosphonates have been widely used in the treatment of osteoporosis. Uncommon side effects have emerged in postapproval use. Because bisphosphonates accumulate in bone and are released for months or years after treatment is stopped, it is reasonable to consider the clinical question of how long to treat.
  • OBJECTIVE
    • In this personal perspective, we review the pharmacology and mechanism of action of bisphosphonates and the clinical studies that support their efficacy. We then review the literature for longer-term studies and reports of possible side effects that were not seen in clinical trials.
  • RESULTS
    • Bisphosphonates have demonstrated antifracture efficacy in randomized, placebo-controlled trials of 3 and 4 yr duration and have been widely used since the initial release of alendronate in 1995. For zoledronic acid and risedronate, an early effect (fractures reduced within 6-12 months of starting therapy) has been shown. A sustained effect for risedronate has been shown through 5 yr and suggested through 7 yr. Ten-year data with alendronate and 8 yr data with risedronate indicated good tolerability and safety; it is unlikely that longer-term studies will be done. Side effects that emerged in clinical trials include esophageal irritation with oral administration and acute phase response with iv treatment or high-dose oral therapy. Uncommon side effects that have been noted with wide clinical use include osteonecrosis of the jaw, musculoskeletal complaints, and atypical fractures. The numbers of events are small, and a clear cause-and-effect relationship between these events and bisphosphonate treatment has not been established. Because bisphosphonates accumulate in bone, they create a reservoir leading to continued release from bone for months or years after treatment is stopped. Studies with risedronate and alendronate suggest that if treatment is stopped after 3-5 yr, there is persisting antifracture efficacy, at least for 1-2 yr.
  • CONCLUSIONS
    • Bisphosphonates are popular and effective for treatment of osteoporosis. Because they accumulate in bone and provide some residual antifracture reduction when treatment is stopped, we recommend a drug holiday after 5-10 yr of bisphosphonate treatment. The duration of treatment and length of the holiday are based on fracture risk and pharmacokinetics of the bisphosphonate used. Patients at mild risk might stop treatment after 5 yr and remain off as long as bone mineral density is stable and no fractures occur. Higher risk patients should be treated for 10 yr, have a holiday of no more than a year or two, and perhaps be on a nonbisphosphonate treatment during that time.