Drug Holidays from Bisphosphonates

A bisphosphonate drug holiday is a break from medication therapy following long-term treatment. Bisphosphonates are effective at reducing the risk of vertebral, non-vertebral and hip fractures in postmenopausal women and adult men. Though there are many benefits of treatment, some rare but serious side effects have been associated with the use of bisphosphonates including atypical fractures. Drug holidays may reduce the risk of serious adverse effects. Bisphosphates are long-acting medications, and therefore adequate efficacy may be maintained after discontinuation. Therefore, drug holidays may be a great option for patients to reduce the risks of treatment.

Bisphosphonate therapy has been shown superior to placebo at reducing the rate of fractures in randomised controlled trials. In one review they found therapy to be associated with a 50% reduction in hip fracture risk, 40 to 50% reduction in vertebral fractures and 20 to 30% reduction in non-spine fractures.

In the FIT study, over 2000 women were randomised to alendronate or placebo for up to five years. Many benefits of alendronate therapy were identified in the trial including a 50% reduction of vertebral fractures with five years of treatment. Also, there was significantly less non-vertebral fractures.

In another study, the HORIZON-PFT, patients received three years of annual zoledronic acid infusions. Zoledronic acid therapy was found to reduce the risk of hip fracture by 41%, reduce the risk of non-vertebral fractures by 25% and increase bone mineral density (BMD).

Adverse effects

Atypical femoral fractures (AFF) and osteonecrosis of the jaw are the two rare but potentially serious adverse effects that are associated with bisphosphonate therapy. These side effects are so rare that many trials of these medications have no cases in the study population. Of note, these serious side effects are more common in cancer patients who receive high dose bisphosphonates but still linked to patients who are using bisphosphonates for fracture prevention.

Atypical femoral fractures

AFF is often only on one side, requires surgical repair and can occur with minimal trauma. Though the cause is unknown, bisphosphonates are anti-remodelling drugs and it is thought that AFFs are exacerbated by impaired bone repair.

In a study of almost 200,000 women, 277 AFF incidents occurred. The risk of AFF was higher with increased duration of therapy with a bisphosphonate. The hazard ratio of 8.86 for patients who used bisphosphonates for 3 to 5 years (when compared with use of bisphosphate under 3 months) and increased up to 43.51 for patients who received 8 years of therapy or more. Post discontinuation of therapy, the risk of AFF fell rapidly.

Osteonecrosis of the jaw

Osteonecrosis of the jaw is bone that is exposed in the facial region. To meet the criteria of medication-related osteonecrosis of the jaw, patients must have been exposed to an antiresorptive drug, antiangiogenic drug or romosozumab. They must also have no history of facial radiation therapy, and the exposed bone must have not healed over a period of greater than 8 weeks.

The condition is rare, reported in less than 1 case in 10,000 patient years in patients using bisphosphonates for fracture prevention. Whilst some serious cases require extensive jaw surgery, most cases are mild.

Drug holidays

Bisphosphonates have been shown in pharmacokinetic studies to remain in the bone for years after cessation. For example, alendronate has a similar half-life to bone mineral, about 10.5 years. So, a patient having a drug holiday may continue to benefit from past treatment for years. Another benefit of stopping bisphosphate therapy is the theoretical increase of bone turnover.

The FLEX trial (an extension of the FIT trial) included over a thousand women who had received on average five years of alendronate therapy. Patients were randomised to an additional five years of therapy with alendronate or placebo. Patients in the alendronate group maintained BMD (bone mineral density) at the total hip and had a greater increase in lumbar spine BMD and total body BMD than the placebo subjects.

Fracture rates between groups were similar. However, the rate of clinical vertebral fractures was significantly different between groups with 5.3% in placebo group and 2.4% in alendronate group.

Overall, the authors concluded that for many women, a drug holiday of up to five years after five years of treatment of alendronate was unlikely to increase fracture risk. However, continued treatment without a drug holiday may be appropriate for patients with very low BMD or for patients at high risk of vertebral fractures.

In another follow up study of the HORIZON-PFT trial, 921 women who had received three annual doses of zoledronic acid were randomised to continue for another three years or receive placebo. The primary outcome, change in femoral neck BMD, was significantly different, +0.24% in zoledronic acid group compared to -0.80% in placebo group. The difference in lumbar spine was also significantly different.

There was no significant increased risk for nonvertebral, hip or all clinical fractures between groups. However, the risk of vertebral fracture was significantly lower in the treatment group.

The authors concluded that patients who had a drug holiday had residual benefits from therapy. Patients who continued therapy were more likely to maintain gained BMD. This study also recommended that patients that are deemed high risk continue bisphosphonate therapy.

There was also a three-year extension for women from that trial who had already received 6 years of zoledronic acid, 190 patients were randomised to receive a further three years of treatment or placebo. Due to the small subject population, there were few fractures and no significant difference between groups. From this trial, it is likely that patients who receive zoledronic acid therapy for six years can have a drug holiday of up to three years.

The FLEX and HORIZON studies were important in establishing the benefits and risks of drug holidays. The risk of vertebral fractures was approximately 50% higher in patients who were in the placebo groups in these trials than the patients in the treatment groups. No significant differences were found between the risk of hip fracture or all non-spine fractures. There are concerns that these outcomes were not able to be reviewed as there were not enough patients in either study to identify a significant difference in risk of these outcomes.

There are no randomised controlled trials for drug holidays from risedronate therapy. There is a cohort analysis of over 50,000 subjects comparing subjects who had a drug holiday from either risedronate or alendronate. Subjects included had used bisphosphonate therapy for over three years and had a break of over 120 days. Follow up was up to 3 years.

Overall, patients who used risedronate prior to a drug holiday were found to be at an 18% increased relative risk for hip fracture over 3 years compared with patients who used alendronate prior to a drug holiday. This effect was more pronounced after two years of drug holiday. Of note, there were limitations of this cohort study including that not all relevant data for subjects was available (such as BMD) and therefore the investigators couldn’t rule out significant differences between subject groups. However, the trial indicates that a risedronate drug holiday has increased risks compared with an alendronate drug holiday.

Of note, if patients do have drug holidays from risedronate, this should be reviewed at the end of two years due to the potential for increased risk of hip fracture.

The ideal candidates for drug holidays are patients who do not have very low BMD, have no fracture history and patients who are younger. Contraindications of a drug holiday include history of hip fracture, vertebral fractures or if a patient has had a fracture while receiving treatment.

For patients who are in the middle ground, patients need to be advised of the risks versus benefits and allowed to make an informed decision.

Once a drug holiday has been commenced, monitoring BMD is recommended. If BMD levels drop too low, or the patient has a fracture, bisphosphonate therapy should probably be restarted. At the end of a proposed drug holiday, unfortunately there is no clear evidence to guide us. Restarting therapy with oral or IV bisphosphonate may be appropriate, or switching to denosumab might be another option.

Drug holidays may be appropriate for patients who have received five years of treatment with oral bisphosphonates or three to six years of therapy with IV bisphosphonates. Treatment with bisphosphonates needs to be regularly reviewed to assess the risks and benefits of treatment. The potential for serious adverse effects (especially AFF) that increase with increased length of treatment must be considered as well as the potential for maintained benefit during a drug holiday. It is essential for patients to be informed on the risks and benefits.

Hopefully there will be more studies in the future to guide clinicians on the best timing for drug holidays. Also, hopefully there will be studies on how to recommence patients on these treatments following a drug holiday.