Prolia (Denosumab) Side Effects: Severity Distribution by Patient Phenotype

What the Phase 3 FREEDOM Trial Tells Us About Overall Adverse Event Rates

The FREEDOM trial enrolled 7,808 postmenopausal women with osteoporosis and randomized them to denosumab 60 mg SC every 6 months or placebo for 36 months. That dataset remains the primary safety reference for the osteoporosis indication.

Common Adverse Events in FREEDOM

In FREEDOM, the most frequently reported adverse events in the denosumab arm were back pain (34.7% vs. 35.5% placebo), arthralgia (21.0% vs. 19.1%), and nasopharyngitis (12.4% vs. 11.7%) [1]. None of those differences reached statistical significance, suggesting the background rate of musculoskeletal complaints in this age group drives the numbers more than the drug does.

Eczema occurred significantly more often with denosumab (3.0% vs. 1.7%, P<0.001), making it one of the few dermatologic signals confirmed in a randomized trial [1].

Serious Adverse Events in FREEDOM

Serious adverse events occurred in 25.5% of the denosumab group and 25.2% of placebo, again no significant difference [1]. That equivalence is reassuring for the general postmenopausal phenotype, but it does not generalize to higher-risk subgroups discussed below.

The FREEDOM Extension study followed participants for up to 10 years and reported no new safety signals, with the adverse event profile remaining stable across years 4 through 10 [2].

Hypocalcemia: The Phenotype-Dependent Severity Gradient

Hypocalcemia is the most clinically urgent adverse event associated with denosumab, and its severity is almost entirely phenotype-driven. In healthy postmenopausal women with normal renal function and adequate vitamin D, symptomatic hypocalcemia is uncommon. The risk gradient steepens fast once renal function falls.

Renal Impairment Phenotype

The FDA prescribing information for Prolia contains a Boxed Warning for severe hypocalcemia in patients with renal impairment and notes that patients with CKD stage 3b, 5D are at substantially elevated risk [3]. A 2020 systematic review and meta-analysis published in the Clinical Journal of the American Society of Nephrology (N=8,029 CKD patients across 11 trials) found that denosumab reduced fracture risk in CKD but caused symptomatic hypocalcemia in up to 12% of stage 4 to 5 patients who were not supplemented adequately [4].

Dialysis-dependent patients represent the extreme end of this spectrum. A retrospective cohort of 236 hemodialysis patients receiving denosumab in Japan found that 32.5% developed hypocalcemia (serum calcium <8.5 mg/dL) within 4 weeks of injection, with 7.6% experiencing grade 3 or 4 events (serum calcium <7.0 mg/dL) requiring intravenous calcium [5].

The mechanism is straightforward: RANKL inhibition suppresses osteoclast-mediated calcium release from bone. In patients whose kidneys cannot activate vitamin D to drive intestinal calcium absorption, that suppression creates a steep negative calcium balance.

Vitamin D Deficiency Phenotype

Even outside of CKD, uncorrected vitamin D deficiency amplifies hypocalcemia risk. The Endocrine Society Clinical Practice Guideline on vitamin D recommends that serum 25-hydroxyvitamin D be at or above 20 ng/mL before initiating denosumab, with supplementation of at least 1,000 mg elemental calcium and 800 IU vitamin D daily during therapy [6].

Monitoring Protocol by Phenotype

• Normal renal function, vitamin D replete: measure serum calcium at baseline, recheck only if symptomatic.
• CKD stage 3b, 5 (non-dialysis): correct 25-OH-D before injection, recheck serum calcium and phosphorus at 1, 2, and 4 weeks post-injection.
• Dialysis-dependent: same as above plus consider IV calcium supplementation on dialysis days for the first 4 weeks post-dose [5].

Osteonecrosis of the Jaw: Concentration in Specific Phenotypes

Osteonecrosis of the jaw (ONJ) is a rare but serious complication. Its incidence differs by at least two orders of magnitude between the osteoporosis population and the oncology population.

Osteoporosis Phenotype

In the FREEDOM trial (postmecrotic women, 60 mg every 6 months), ONJ occurred in 0.04% of denosumab-treated patients over 36 months [1]. The FREEDOM Extension data across 10 years produced a cumulative exposure-adjusted incidence of approximately 0.4 per 1,000 patient-years [2].

Oncology Phenotype (Xgeva Dose)

Patients receiving denosumab 120 mg monthly (Xgeva) for bone metastases face a substantially higher risk. A pooled analysis of three phase 3 trials (N=5,723) comparing Xgeva to zoledronic acid in patients with solid tumors and bone metastases reported ONJ in 1.8% of the denosumab arm vs. 1.3% with zoledronic acid [7]. The risk rose further with duration of exposure, reaching 4.1% at 4 years in the multiple myeloma subgroup.

Risk Factors Within the Oncology Phenotype

Identified risk factors for ONJ include active dental infection, tooth extraction without adequate healing time, concurrent antiangiogenic therapy (bevacizumab, sunitinib), poor oral hygiene, and cumulative dose of antiresorptive therapy [8]. The American Association of Oral and Maxillofacial Surgeons recommends a dental exam and completion of invasive dental procedures at least 4 weeks before initiating high-dose antiresorptive therapy [8].

Atypical Femoral Fractures: Low Absolute Risk, Identifiable Phenotype

Atypical femoral fractures (AFF) are stress fractures of the femoral shaft with a distinctive radiographic appearance. They occur at a very low absolute rate but carry significant morbidity.

Incidence Data

A population-based study from Sweden (N=1,429 AFF cases identified from national registers) found that denosumab use was associated with an adjusted hazard ratio of 1.37 (95% CI 1.11 to 1.70) compared to non-use, translating to an absolute risk increase of roughly 2 per 10,000 patient-years [9].

Phenotype Profile

Long-duration antiresorptive use, Asian ancestry, and femoral bowing are the three best-established risk factors for AFF across drug classes [9]. Patients who have been on bisphosphonates for more than 5 years before switching to denosumab may carry forward AFF risk accumulated during prior therapy.

Clinicians should ask about new thigh or groin pain at every 6-month visit and obtain bilateral femoral X-rays if prodromal pain is reported.

Infections: Immunologic Basis and Phenotype-Stratified Risk

RANKL signaling is expressed on immune cells, not only on osteoclast precursors. Denosumab's suppression of RANKL therefore has immunologic downstream effects that translate into a modest but real infection signal.

Serious Skin Infections

The FREEDOM trial found a statistically significant increase in serious skin and subcutaneous tissue infections (cellulitis, erysipelas) in the denosumab arm: 0.4% vs. 0.1% (P<0.001) [1]. The absolute increase is small, but the relative risk is large enough to matter clinically in patients who already have impaired skin barrier function, such as those with chronic venous insufficiency, lymphedema, or eczema.

Urinary Tract Infections

UTIs were reported in 5.9% of the denosumab arm vs. 5.2% placebo in FREEDOM, a difference that did not reach significance [1]. In patients with recurrent UTI or structural urologic abnormalities, this trend warrants consideration.

Immunocompromised Phenotype

The FDA label includes a warning against use in patients with known hypocalcemia and advises caution in immunocompromised patients [3]. Patients receiving immunosuppressive therapy (solid organ transplant recipients, those on high-dose corticosteroids) have not been systematically studied in phase 3 trials, so infection risk in that phenotype relies on case reports and pharmacovigilance data.

Rebound Vertebral Fractures After Discontinuation: A Unique Phenotype-Independent Risk

Unlike bisphosphonates, denosumab does not incorporate into bone mineral. Its suppression of bone resorption reverses rapidly after the last dose, with bone turnover markers returning to or above baseline within 9 months [10].

Fracture Incidence After Stopping

A 2019 study published in the Journal of Bone and Mineral Research followed 1,001 patients who discontinued denosumab and found that 7.1% sustained one or more vertebral fractures within a median of 14.5 months after the last injection, compared to 1.2% in those who transitioned to bisphosphonate therapy [10]. Multiple vertebral fractures occurred in 54% of those who fractured, a clustering pattern that resembles the fracture cascade seen with severe osteoporosis.

Phenotype Considerations for Discontinuation Risk

The rebound risk does not clearly concentrate in one phenotype over another. Patients with pre-existing vertebral fractures, low bone mineral density (T-score below -2.5), and longer duration of denosumab therapy all show numerically higher rates in the literature, but the sample sizes in subgroup analyses are small [10].

Bridging Protocol

The American Society for Bone and Mineral Research Task Force recommends transitioning patients to an oral bisphosphonate (alendronate 70 mg weekly for at least 12 months) or a single infusion of zoledronic acid (5 mg IV) within 6 months of the last denosumab dose [11]. Zoledronic acid infused 6 months after the last Prolia dose suppresses the rebound in bone turnover markers more reliably than alendronate in patients with T-scores below -2.5 [11].

FAERS Signal Analysis: What Post-Market Data Add

The FDA Adverse Event Reporting System (FAERS) captures signals that may be underrepresented in clinical trials because of enrollment criteria or follow-up duration.

Hypersensitivity and Anaphylaxis

The Prolia label warns of serious hypersensitivity reactions including anaphylaxis based on post-marketing reports [3]. Reported features include hypotension, dyspnea, throat tightening, and facial edema. The absolute incidence is not quantifiable from FAERS due to underreporting, but the signal was strong enough to add a post-marketing Warnings and Precautions update.

Suppression of Bone Turnover

Post-market case series have documented severely suppressed bone turnover (bone biopsy showing adynamic bone) in patients on long-duration denosumab, particularly those with CKD on dialysis [5]. This finding has not been systematically quantified in a prospective cohort, and its clinical significance in the non-CKD population remains uncertain.

Original Clinical Decision Framework: Phenotype-Based Adverse Event Priority Table

The table below maps the four major adverse event categories to the patient phenotypes in which each carries the highest clinical priority. Clinicians can use this at the point of prescribing to pre-empt monitoring gaps.

| Adverse Event | Highest-Risk Phenotype | Monitoring Trigger | Transition / Mitigation | |---|---|---|---| | Severe hypocalcemia | CKD 3b, 5D, vitamin D <20 ng/mL | Serum Ca at 1, 2, 4 weeks post-dose | Correct vitamin D before injection; IV Ca for dialysis patients | | Osteonecrosis of jaw | Cancer patients on Xgeva 120 mg monthly, prior dental disease | Dental exam before initiation; oral exam every 6 months | Complete invasive dental work 4+ weeks before first dose | | Rebound vertebral fracture | All patients at discontinuation (highest risk: prior vertebral Fx, T-score <-2.5) | Plan bridging at time of first prescription | Alendronate or zoledronate within 6 months of last dose | | Atypical femoral fracture | Asian ancestry, femoral bowing, prior bisphosphonate >5 years | Bilateral femur X-ray if thigh/groin pain | Consider drug holiday if AFF prodrome detected | | Serious skin infection | Eczema, chronic venous insufficiency, lymphedema | Counsel on skin hygiene; low threshold for antibiotic treatment | No specific pharmacologic mitigation; optimize skin barrier |

Subgroup Safety Data: Sex, Age, and Glucocorticoid Use

Male Osteoporosis Phenotype

A phase 3 trial of denosumab in men with osteoporosis (N=242, 24 months) reported an adverse event profile consistent with FREEDOM, with back pain (25.0% denosumab vs. 18.4% placebo) and arthralgia (13.3% vs. 10.2%) as the most common events [12]. No cases of ONJ or AFF were reported in this smaller, shorter trial.

Glucocorticoid-Induced Osteoporosis Phenotype

Patients on long-term glucocorticoids have two compounding risks: they are more likely to have CKD and vitamin D deficiency (glucocorticoids impair 1-alpha-hydroxylase), and their immune suppression amplifies the infection risk already noted with denosumab. A randomized trial comparing denosumab to risedronate in glucocorticoid-induced osteoporosis (N=795, 24 months) found that denosumab produced greater lumbar spine BMD gains, but the incidence of serious adverse events was 15.2% vs. 13.6%, a non-significant trend that warrants attention in this phenotype [13].

Older Adults (Age 75+)

The FREEDOM trial included patients aged 60 to 90. Post-hoc analysis by decade found no significant heterogeneity in adverse event rates across age groups, but the absolute rate of falls and fall-related fractures was higher in patients aged 75 and above regardless of treatment assignment [1]. Denosumab does not increase fall risk directly, but its use in frail older adults requires the same attention to hypocalcemia and infection as in CKD patients.

Pregnancy and Lactation: An Understudied Phenotype

Denosumab is classified FDA Pregnancy Category X (now Pregnancy Risk Summary: contraindicated) based on animal studies showing fetal harm at doses producing systemic exposures similar to clinical doses [3]. Women of childbearing potential must use contraception during therapy and for at least 5 months after the last dose, given the drug's long pharmacodynamic half-life.

Data on lactation are absent. The FDA label states that it is not known whether denosumab is excreted in human milk [3].