HORIZON-PFT Results in Detail: Numbers, Subgroups, and Time Course

At a glance

Parameter	Detail
N	7,765 postmenopausal women
Intervention	Zoledronic acid 5 mg IV once yearly × 3 years
Comparator	IV placebo (normal saline) once yearly
Duration	36 months
Primary endpoint	New morphometric vertebral fracture at 36 months
Key result	70% relative risk reduction (3.3% vs 10.9%; p <0.001)

Trial Design Context

HORIZON-PFT (Health Outcomes and Reduced Incidence with Zoledronic Acid Once Yearly, Key Fracture Trial) enrolled postmenopausal women aged 65, 89 with femoral neck T-score ≤ −1.5 with existing vertebral fracture, or T-score ≤ −2.5 with or without vertebral fracture. Participants were randomized 1:1 to zoledronic acid 5 mg IV or placebo, infused over at least 15 minutes once annually for three consecutive years. All participants received calcium 1,000, 1 to 500 mg/day and vitamin D 400, 1 to 200 IU/day.

Concomitant use of osteoporosis therapies was prohibited except in a stratum (approximately 20% of participants) allowed prior and ongoing use of certain medications. Spine radiographs were obtained at baseline, 12 months, 24 months, and 36 months, read centrally by a quantitative morphometry lab blinded to treatment assignment.

Primary Endpoint: Morphometric Vertebral Fracture

The primary efficacy measure was new morphometric vertebral fracture, defined as ≥20% and ≥4 mm reduction in anterior, middle, or posterior vertebral body height on lateral spine radiographs compared with baseline.

The 70% relative risk reduction at 36 months translated to an absolute risk reduction of 7.6 percentage points. The number needed to treat (NNT) to prevent one morphometric vertebral fracture over three years was 13. The effect was already pronounced at 12 months (60% RRR), suggesting rapid onset of antifracture protection.

Secondary Endpoints: Non-Vertebral and Hip Fractures

The pre-specified secondary endpoints addressed clinical fractures that matter most to patients and healthcare systems.

The 41% reduction in hip fractures represented a clinically significant finding. Prior oral bisphosphonate trials had inconsistent hip fracture data owing to smaller sample sizes or post-hoc analyses. HORIZON-PFT was powered and pre-specified for this outcome. The NNT for hip fracture prevention over three years was approximately 91, reflecting the lower baseline event rate at this skeletal site.

Wrist fractures did not reach significance. The investigators noted that wrist fractures are often related to falls in younger postmenopausal women and may involve biomechanical factors less responsive to increased bone density alone.

Bone Mineral Density Time Course

BMD changes were assessed by DXA at the lumbar spine, total hip, and femoral neck at 6, 12, 24, and 36 months. The pattern demonstrated rapid early gains that continued accruing through year three.

BMD differences from placebo were statistically significant (p <0.001) at every time point from 12 months onward. The lumbar spine showed the largest absolute gains, consistent with the predominance of trabecular bone at that site and the known affinity of bisphosphonates for metabolically active bone surfaces.

Bone turnover markers (serum C-telopeptide, bone-specific alkaline phosphatase, and N-terminal propeptide of type I collagen) showed maximal suppression at 9 to 11 days post-infusion, with partial recovery before the next annual dose. This cyclical suppression pattern is distinct from the continuous suppression seen with daily oral alendronate or weekly risedronate.

Subgroup Analyses

Pre-specified subgroup analyses examined consistency of the vertebral fracture reduction across baseline characteristics.

No subgroup interaction reached statistical significance, indicating that the treatment effect was consistent regardless of age, fracture history, baseline BMD severity, or prior bisphosphonate exposure. This broad applicability supported the subsequent approval of zoledronic acid across the full osteoporosis population rather than only severe disease.

Height Loss and Kyphosis

An often-overlooked secondary endpoint: cumulative height loss over three years was significantly less in the zoledronic acid group (mean 4.2 mm) compared with placebo (mean 6.7 mm), a between-group difference of 2.5 mm (p <0.001). While seemingly small, progressive height loss correlates with vertebral deformity burden and impaired pulmonary function. The Endocrine Society guidelines cite prevention of height loss as a meaningful clinical outcome in osteoporosis management.

Safety Signals and Adverse Events

Post-infusion reactions occurred in 31.6% of zoledronic acid patients after the first dose, compared with 6.2% for placebo. These reactions (fever, myalgia, headache, arthralgia) were mostly mild to moderate and resolved within 3 days. By the third annual infusion, incidence dropped to 6.6%.

Serious atrial fibrillation occurred in 1.3% of the zoledronic acid group versus 0.5% of placebo (p = 0.003). This signal was unexpected. Subsequent dedicated analyses, including the HORIZON Recurrent Fracture Trial and FDA review of pooled bisphosphonate data, did not confirm a consistent class-wide or drug-specific cardiac arrhythmia risk. The FDA concluded that available evidence did not support a causal relationship but recommended continued monitoring.

Renal safety: transient increases in serum creatinine >0.5 mg/dL occurred in 1.2% of zoledronic acid patients versus 0.4% of placebo within 9 to 11 days post-infusion. Most resolved within 30 days. Patients with creatinine clearance <35 mL/min were excluded from the trial.

Limitations Worth Noting

The investigators acknowledged several constraints. First, the trial population was exclusively postmenopausal women; male osteoporosis data came from a separate, smaller study (HORIZON-RFT included men but was a different design). Second, 36 months is insufficient to evaluate rare long-term events like atypical femoral fractures or osteonecrosis of the jaw. Third, approximately 24% of participants did not complete the full three-year follow-up, though sensitivity analyses using worst-case imputation still showed significant benefit. Fourth, the 20% of participants in the "concurrent therapy" stratum complicated interpretation, though pre-specified analyses showed the primary result held in both strata.

Clinical Translation

The 70% vertebral fracture reduction and 41% hip fracture reduction established once-yearly IV zoledronic acid as a first-line option for patients who cannot tolerate oral bisphosphonates, have adherence concerns, or have upper GI contraindications. The FDA-approved prescribing information for Reclast reflects these data directly. Real-world adherence studies have since shown that the IV annual schedule achieves higher persistence rates (approximately 60% at 2 years) compared with oral weekly bisphosphonates (approximately 30 to 40% at 2 years), a finding that matters because fracture reduction disappears when patients stop treatment.

Frequently asked questions

›What was the absolute risk reduction for vertebral fractures in HORIZON-PFT?

The absolute risk reduction was 7.6 percentage points (10.9% placebo vs 3.3% zoledronic acid) over three years, yielding a number needed to treat of 13 to prevent one morphometric vertebral fracture.

›How quickly did zoledronic acid show fracture reduction benefits?

At the 12-month assessment, morphometric vertebral fractures were already reduced by 60% (RR 0.40, 95% CI 0.27, 0.58). BMD differences from placebo were significant at all time points from 12 months onward.

›Did HORIZON-PFT demonstrate hip fracture reduction?

Yes. Hip fractures were reduced by 41% (HR 0.59, 95% CI 0.42, 0.83, p = 0.002). This was a pre-specified secondary endpoint with adequate statistical power.

›Were the results consistent across different patient subgroups?

Pre-specified subgroup analyses showed consistent vertebral fracture reduction regardless of age, baseline T-score, prior fracture history, or previous bisphosphonate use. No interaction tests reached significance.

›What were the most common side effects of annual IV zoledronic acid?

Post-infusion reactions (fever, myalgia, headache) occurred in about 32% of patients after the first dose but dropped to under 7% by the third infusion. Most symptoms resolved within 72 hours.

›Was the atrial fibrillation signal in HORIZON-PFT confirmed in later studies?

No. The FDA reviewed pooled data from multiple bisphosphonate trials and concluded that evidence did not support a causal link between zoledronic acid and serious atrial fibrillation, though monitoring was recommended.

›How does HORIZON-PFT compare with oral bisphosphonate fracture trials?

The 70% vertebral fracture reduction is numerically larger than the approximately 50% reduction seen in the FIT trial with alendronate, though cross-trial comparisons are limited by different populations and methodologies.

›What was the dropout rate in HORIZON-PFT?

Approximately 24% of participants did not complete the full 36-month follow-up. Sensitivity analyses using conservative imputation methods confirmed that the primary endpoint remained statistically significant.

›Who was excluded from HORIZON-PFT?

Key exclusions included creatinine clearance <35 mL/min, prior sensitivity to bisphosphonates, active cancer, and metabolic bone disease other than osteoporosis. The minimum age was 65.

›How much bone density did patients gain on zoledronic acid over three years?

Lumbar spine BMD increased by 6.7% and total hip by 3.6% from baseline. The treatment difference versus placebo was approximately 5 to 6% at both sites by 36 months.

References

1. Black DM, Delmas PD, Eastell R, et al. Once-yearly zoledronic acid for treatment of postmenopausal osteoporosis. N Engl J Med. 2007;356(18):1809-1822. https://pubmed.ncbi.nlm.nih.gov/17476007/
2. Reclast (zoledronic acid) prescribing information. FDA. https://www.accessdata.fda.gov/drugsatfda_docs/label/2022/021817s030lbl.pdf
3. Shoback D, Rosen CJ, Black DM, et al. Pharmacological management of osteoporosis in postmenopausal women: an Endocrine Society guideline update. J Clin Endocrinol Metab. 2020;105(3):587-594. https://pubmed.ncbi.nlm.nih.gov/31074826/
4. Sharma A, Chatterjee S, Arbab-Zadeh A, et al. Risk of serious atrial fibrillation and stroke with use of bisphosphonates: evidence from a meta-analysis. Chest. 2013;144(4):1311-1322. https://pubmed.ncbi.nlm.nih.gov/22147950/
5. Fosamax (alendronate sodium) prescribing information. FDA. https://www.accessdata.fda.gov/drugsatfda_docs/label/2012/021575s017lbl.pdf