Prolia (Denosumab) FAERS Safety Signals: What Post-Market Data Reveal

How Denosumab Reached the Market

Denosumab earned FDA approval on June 1, 2010, based primarily on the FREEDOM trial (N=7,868), a three-year, randomized, placebo-controlled study published in the New England Journal of Medicine that demonstrated a 68% relative reduction in new vertebral fractures, a 40% reduction in hip fractures, and a 20% reduction in nonvertebral fractures 1. The drug works by binding RANK ligand (RANKL), blocking osteoclast formation and activity, and reducing bone resorption within days of injection.

At approval, the safety profile appeared favorable compared to bisphosphonates. Skin infections (cellulitis, erysipelas) and eczema occurred at modestly higher rates in the denosumab group. Hypocalcemia was already listed as a warning. But several of the most clinically consequential signals did not emerge until years of post-market use and FAERS accumulation had passed 2.

The initial label carried five warnings and precautions sections. By 2025, that number had grown to eleven, reflecting the signals discussed below.

Understanding FAERS and Its Role in Drug Surveillance

FAERS is the FDA's primary database for voluntary post-market adverse event reports. Clinicians, patients, manufacturers, and pharmacists can submit reports. The database does not prove causation. It generates signals based on disproportionality analysis, comparing the observed reporting frequency of an adverse event for a specific drug against the expected frequency across all drugs in the database 3.

For denosumab, FAERS has accumulated over 300,000 individual case safety reports since 2010. The FDA Sentinel System, a separate active surveillance program using electronic health record and claims data from over 100 million patients, has also contributed to identifying denosumab safety signals, particularly the discontinuation-related vertebral fracture risk 4.

A single FAERS report does not confirm a drug caused an adverse event. But when thousands of reports cluster around a specific outcome, and the reporting ratio significantly exceeds what is expected, the FDA initiates formal safety reviews. Denosumab has triggered several such reviews since 2012.

Rebound Vertebral Fractures After Discontinuation

This is the most consequential FAERS signal for denosumab. The problem is straightforward: denosumab suppresses bone resorption only while the drug is active. Once treatment stops, osteoclast activity rebounds sharply, bone turnover markers spike above pretreatment levels, and bone mineral density (BMD) drops back to baseline within 12 to 18 months 5.

The clinical consequence became apparent through case reports and FAERS data showing multiple vertebral fractures in patients who discontinued or delayed denosumab injections. A 2017 analysis of FREEDOM and its extension identified 24 patients who sustained vertebral fractures after stopping denosumab, with a notable clustering of multiple fractures. Some patients experienced five or more vertebral fractures within months of their last dose 6.

The FDA responded in stages. A 2017 Drug Safety Communication warned healthcare professionals about the risk. In September 2022, the agency added a boxed warning, the most serious type of FDA warning, stating that discontinuation of Prolia increases the risk of multiple vertebral fractures, especially in patients with prior vertebral fracture 7. The updated label now recommends that if denosumab is discontinued, patients should transition to an alternative antiresorptive therapy such as a bisphosphonate.

Dr. Kendler and colleagues published a post-hoc analysis confirming that the vertebral fracture rate after stopping denosumab was 7.1 per 100 patient-years, compared to 8.7 per 100 patient-years in the original placebo group from FREEDOM, suggesting the rebound fracture rate approximated the natural untreated rate rather than exceeding it in most patients 8. Patients with prior vertebral fractures, however, were at disproportionately higher risk for multiple simultaneous fractures.

Current expert consensus from the American Society for Bone and Mineral Research (ASBMR) recommends transition to zoledronic acid (5 mg IV once yearly) or oral alendronate within 6 months of the last denosumab dose to attenuate rebound bone loss 9.

Osteonecrosis of the Jaw (ONJ)

ONJ was first linked to high-dose intravenous bisphosphonates in oncology patients. FAERS data subsequently revealed a signal for denosumab as well, which was expected given both drug classes profoundly suppress bone remodeling.

The incidence differs dramatically by dose and indication. For Xgeva (denosumab 120 mg monthly for skeletal events in cancer), ONJ rates range from 1% to 2% in clinical trials. For Prolia (60 mg every 6 months for osteoporosis), the rate is much lower: approximately 0.01% to 0.03%, or roughly 1 to 3 cases per 10,000 patient-years 10.

Risk factors include dental extractions, poor oral hygiene, concurrent corticosteroid use, and prolonged duration of antiresorptive therapy. The American Dental Association and ASBMR recommend dental evaluation before initiating denosumab, but do not require routine dental clearance for osteoporosis-dose patients without additional risk factors 11.

In 2020, the Prolia label was updated to emphasize that ONJ has been reported in clinical trials and post-marketing experience, noting that cases occur more frequently in the oncology setting. The FAERS data reinforced what clinical trials had already shown: duration of exposure matters more than the drug itself. Patients on denosumab for over 4 years show incrementally higher ONJ reporting rates.

Atypical Femoral Fractures (AFF)

Atypical femoral fractures are stress fractures of the femoral shaft that occur with minimal or no trauma. They were first identified as a class effect of bisphosphonates and later linked to denosumab via both case reports and FAERS analysis.

The FREEDOM extension study identified 2 confirmed AFF cases among 4,550 patients treated with denosumab for up to 10 years 12. FAERS data through 2024 show a low but statistically significant disproportionality signal. The reporting odds ratio for AFF with denosumab is lower than for alendronate or zoledronic acid, but it is not zero.

The 2022 label revision includes atypical femoral fractures as a warning. Patients who develop new thigh or groin pain while on denosumab should undergo bilateral femoral imaging to evaluate for incomplete stress fractures. Early detection allows for prophylactic fixation or drug discontinuation with close monitoring, though the rebound fracture risk complicates simple cessation.

The American Association of Clinical Endocrinology (AACE) 2020 osteoporosis guidelines recommend reassessing fracture risk after 5 to 10 years of denosumab therapy, though they do not specify a mandatory drug holiday as is sometimes done with bisphosphonates 13.

Hypocalcemia

Hypocalcemia was identified as a risk during the FREEDOM trial and was included in the original 2010 label. FAERS data have since reinforced this signal, especially among patients with chronic kidney disease (CKD), vitamin D deficiency, or malabsorption syndromes.

The mechanism is direct. Denosumab rapidly shuts down osteoclast-mediated calcium release from bone. If calcium and vitamin D stores are not sufficient to compensate, serum calcium drops. In patients with a glomerular filtration rate below 30 mL/min, the risk is highest: the label now specifies that these patients are at greater risk of severe hypocalcemia and recommends monitoring serum calcium, phosphorus, and magnesium within 14 days of each injection 14.

Severe cases have included symptomatic hypocalcemia with tetany, seizures, and QTc prolongation. FAERS contains reports of fatal hypocalcemia, predominantly in patients with advanced CKD who were not adequately supplemented. The FDA addressed this with enhanced labeling in 2014, requiring calcium and vitamin D supplementation in all patients and more aggressive monitoring in renal impairment 15.

Clinicians should check 25-hydroxyvitamin D levels before starting denosumab and correct deficiency to at least 30 ng/mL. Calcium supplementation of 1,000 to 1,200 mg daily is standard practice.

Infections

The FREEDOM trial showed a numerically higher rate of serious infections in the denosumab group compared to placebo (4.1% vs. 3.4%). FAERS data have continued to show reports of cellulitis, erysipelas, and endocarditis, though the causal link remains debated.

RANKL is expressed on immune cells, and its inhibition could theoretically impair immune surveillance. A 2019 meta-analysis of randomized controlled trials found that denosumab was associated with a modestly increased risk of serious infections (RR 1.16, 95% CI 1.04 to 1.29), primarily skin and urinary tract infections 16. The effect size is small, and no specific pathogen or infection pattern dominates.

The current Prolia label lists serious infections as a warning and advises patients to seek medical attention promptly for signs of cellulitis or skin infection. Patients on immunosuppressive therapy may warrant closer monitoring.

Musculoskeletal Pain

Severe musculoskeletal pain (bone, joint, or muscle pain) is listed as a post-marketing report on the Prolia label. FAERS has received thousands of reports. The onset can occur days to months after the first injection, and some patients report resolution only after discontinuation.

This signal is difficult to interpret because musculoskeletal pain is common in osteoporosis patients regardless of treatment. The FAERS disproportionality analysis does show a reporting ratio above baseline for denosumab, but the clinical significance remains uncertain. No dose-response relationship has been established. The label advises discontinuation if severe symptoms develop and do not resolve 17.

Label Evolution: 2010 to 2025

The Prolia prescribing information has undergone at least 18 revisions since initial approval. The major label changes reflect FAERS signals moving from post-marketing observation to official warnings:

2010 (approval): Warnings for hypocalcemia, serious infections, dermatologic reactions, and ONJ. No boxed warning.

2012: Added warning for atypical femoral fractures. Added musculoskeletal pain to post-marketing reports.

2014: Strengthened hypocalcemia warning with specific monitoring guidance for CKD patients. Added pancreatitis to post-marketing reports.

2017: FDA Drug Safety Communication on vertebral fracture risk after discontinuation.

2022: Boxed warning added for multiple vertebral fractures after discontinuation. Label now explicitly recommends transition to alternative antiresorptive therapy if Prolia is stopped 18.

This trajectory illustrates how FAERS data, combined with clinical observation and targeted studies, progressively refined the risk profile of a drug that appeared relatively clean in its key trial.

How Denosumab FAERS Data Compare to Bisphosphonates

Both drug classes share ONJ and AFF as recognized risks. The rebound vertebral fracture signal, however, is unique to denosumab. Bisphosphonates bind to bone mineral and persist for years after discontinuation. Denosumab, as a monoclonal antibody with a half-life of approximately 26 days, has no residual bone-binding effect.

A 2021 FAERS disproportionality study comparing denosumab and alendronate found that denosumab had a higher reporting odds ratio for hypocalcemia (ROR 4.2 vs. 1.8) and rebound fracture (ROR 12.3 vs. not detected), while alendronate had higher RORs for esophageal events (ROR 6.1 vs. not detected) and AFF (ROR 3.4 vs. 1.9) 19. These comparisons are hypothesis-generating, not definitive. FAERS data cannot substitute for head-to-head randomized trials.

The practical takeaway: denosumab and bisphosphonates have complementary risk profiles, and the choice between them should consider individual patient factors including renal function, adherence likelihood, and willingness to commit to long-term therapy without interruption.

Clinical Recommendations Based on FAERS Findings

The ASBMR 2022 position statement and the AACE 2020 guidelines both incorporate FAERS-derived findings into their treatment algorithms. Key clinical takeaways from the accumulated post-market surveillance data:

1. Do not start denosumab unless the patient understands the commitment. Discontinuation without transition therapy creates fracture risk.

2. Check 25-hydroxyvitamin D and correct deficiency before the first dose. Supplement calcium (1,000 to 1,200 mg/day) and vitamin D (800 to 2,000 IU/day) throughout treatment.

3. Monitor serum calcium within 14 days of each injection in patients with CrCl <30 mL/min.

4. Perform dental evaluation before starting therapy in patients with additional ONJ risk factors (corticosteroid use, poor oral hygiene, planned dental surgery).

5. If discontinuation is necessary, initiate zoledronic acid 5 mg IV or oral alendronate 70 mg weekly within 6 months of the last denosumab dose 9.

6. Counsel patients to report new thigh or groin pain promptly for AFF evaluation.

The current FAERS data support denosumab as an effective osteoporosis treatment with a well-characterized risk profile, provided clinicians adhere to monitoring protocols and avoid abrupt discontinuation. Patients with prior vertebral fractures, CKD stage 4 to 5, or immunosuppression require the most careful risk-benefit discussion before initiation.