Evidence Based Review Article

The Journal of Informed Pharmacotherapy 2000;3:206-208.

Effects of risedronate treatment on vertebral and nonvertebral fractures in women with postmenopausal osteoporosis: Another viable option?

Download a PDF version of this article

Reviewer: Alan Low, B.Sc.(Pharm), Pharm.D.
Reviewer's email address:
Reviewer's profession/specialty:
Ambulatory Care, Neurology

Original Citation

Harris ST, Watts NB, Genant HK, et al.  Effects of risedronate treatment on vertebral and nonvertebral fractures in women
with postmenopausal osteoporosis.  JAMA 1999;282:1344-52.
  PubMed Cit

Overall Study Question

This randomized, double-blind, placebo-controlled, parallel-group study was conducted at 110 study centres across North America.  It was designed to test the efficacy of risedronate 2.5mg and 5mg compared to placebo given daily over 3 years in reducing the incidence of vertebral and other fractures in women who were at least 5 years post menopause, no older than 85 years of age, and had either 2 or more radiographically identified vertebral fractures or had one vertebral fracture with bone mineral density measurements that were at least 2 standard deviations below the mean for young adults.  The main outcome measure was the incidence of patients with new vertebral fractures.  Secondary outcome measures included the cumulative incidence of non-vertebral fractures, change in bone mineral density (BMD) and markers of bone turnover, and overall incidence of adverse events.  All subjects received calcium 1000mg/day taken with the evening meal.  For subjects with baseline levels of 25-hydroxyvitamin D less than 40nmol/L, supplementation with Vitamin D (cholecalciferol) up to 500IU/day was provided.

Are the Results of the Study Valid?

1. Was assignment of patients randomized?

Yes.  All patients were randomly and equally assigned into one of three groups and received either risedronate 2.5mg, risedronate 5mg or placebo.  Patients across all three groups had similar demographics reported.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Yes.  All patients who were randomized in the study were properly accounted for.  However, due to a protocol amendment after the first year, the risedronate 2.5mg arm of the study was terminated.  The investigators report on the disposition of the subjects in the 2.5mg group until withdrawal due to the protocol amendment.  Overall, 9,400 patients were screened and 2,258 met the study criteria.  These individuals were randomized to receive risedronate 2.5mg (n = 811), risedronate 5mg (n = 813) or placebo (n = 815).  Nineteen patients were not treated.  Ten of these patients voluntarily withdrew from the study while 9 were involved in protocol violations.  Four hundred and eighty-nine risedronate 5mg and 450 placebo recipients completed the 3-year study.

3. Were patients, their clinicians, and study personnel 'blind' to treatment?

Yes.  All patients and investigators were blind to the treatment including the radiologists who performed the vertebral fracture assessments.  As the incidence of adverse events reported was similar, this would suggest that unblinding due to adverse effects would not be an issue.

4. Were the groups similar at the start of the trial?

Yes.  All baseline demographics were similar between the three treatment arms at the start of the trial.

5. Aside from the experimental intervention, were the groups treated equally?

Yes.  All subjects received calcium 1000mg/day taken with the evening meal.  Subjects with baseline levels of 25-hydroxyvitamin D less than 40nmol/L received supplementation with Vitamin D (cholecalciferol) up to 500IU/day.  The specific dose of Vitamin D administered was not specified.

6. Overall, are the results of the study valid?


What were the Results?

1. How large was the treatment effect? 

The proportion of subjects with at least 1 incident vertebral fracture over the 3 year study period (primary endpoint) was 11.3% for the risedronate 5mg group and 16.3% for the placebo group (relative risk 0.59 [95%CI 0.43-82]), p=0.003). This represents an absolute risk reduction of 5% or a number needed to treat of 20 over three years.  The cumulative incidence of non-vertebral fractures (secondary outcome) over 3 years of treatment was 5.2% for risedronate 5mg and 8.4% for placebo (relative risk 0.6 [95%CI 0.39-0.94, p=0.02] which corresponds to an absolute risk reduction of 3.2% or a NNT of 32 over three years.

Over 3 years, the subjects in the 5mg risedronate group had significant increases in BMD compared to placebo for the following areas measured: lumbar spine (5.4%), femoral neck (1.6%), and femoral trochanter (3.3%).  Markers of bone turnover (i.e. bone-specific alkaline phosphatase and the deoxypyridinoline-creatinine ratio) were available for only 32% of study subjects and, in comparison with placebo, declined with risedronate treatment.  The overall incidence of adverse effects was similar between the risedronate 5mg (34%) and placebo (29%) groups.  Serious adverse events were observed in 29% and 27% of the groups, respectively.  Since bisphosphonates are known to commonly cause upper gastrointestinal tract adverse effects, the investigators also reported the incidence of any upper gastrointestinal adverse events (30% vs. 27% for risedronate 5mg and placebo groups).  The groups also appeared to have a similar incidence of specific gastrointestinal symptoms.

2. How precise was the estimate of the treatment effect?

The relative risk 95% CI for the cumulative incidence of vertebral fractures over the 3 years revealed a statistically significant effect.  In my opinion, the effect is clinically relevant and the reduction of incident vertebral fractures is clinically important.

Will the Results Help Me in Caring for My Patients?

1. Can the results be applied to my patient care?

Yes.  The information in this study further supports the premise that risedronate can effectively reduce the incidence of vertebral fractures and that this drug may also reduce non-vertebral fractures in postmenopausal women who are at high risk for fractures, including those who have previously experienced fractures.  The treatment was not associated with an increase in adverse events in the clinical trial setting.

2. Were all clinically important outcomes considered?

Yes.  Although incident vertebral fractures were the primary endpoint, a more clinically relevant endpoint is hip fractures. The current study investigated non-vertebral fractures as a secondary endpoint and the authors found a statistical significance between groups.  Subjects in the risedronate 5mg and placebo, experienced 14 vs. 22 wrist fractures, 12 vs. 15 hip and/or pelvis fractures, 4 vs. 10 humerus fractures, 4 vs. 8 leg fractures and 3 vs. 0 clavicle fractures, respectively.  An investigation of hip fractures as a primary endpoint is currently underway.

3. Are the likely treatment benefits worth the potential harms and costs?

This is difficult to determine. The cost of the drug is significantly higher than the cost of calcium and vitamin D and the reduction in vertebral fractures can mean different symptoms and effects for different people.  Some vertebral fractures occur without the patients being aware, others cause pain, and after many vertebral fractures, deformity is apparent.  However, there is emerging evidence that previous vertebral fractures are a major risk factor for more fractures, including hip and wrist fractures.  It may be a cyclical phenomenon that increases exponentially with the accumulation of more fractures and it could be interpreted that by reducing the incidence of vertebral fractures may slow or prevent this cascade.


This study involved a sufficient sample size to detect a difference in fracture rate associated with risedronate therapy for non-vertebral fractures that are clearly clinically relevant.  Unfortunately, this type of fracture was only a secondary endpoint in this trial.  Since this study was placebo-controlled and thus did not involve other alternatives currently used to reduce fractures in this population, it is difficult to determine the current role of this drug.  This study does, however, demonstrate that risedronate appears to be an effective agent for the reduction of vertebral fractures in this population and that this drug may also reduce the incidence of non-vertebral fractures.  Another advantage of this study is that subjects with previous or active gastrointestinal illnesses or concomitant use of ASA or an NSAID were included.  These patient groups have been excluded in some of the earlier trials involving  bisphosphonates.  Finally, this study shows that risedronate is certainly another option available to help reduce vertebral fractures in this study population that can be generalized to our practice population.

Copyright © 2000 by the Journal of Informed Pharmacotherapy. All rights reserved.