Evenity (Romosozumab) Evidence Base Graded by GRADE

What Is Romosozumab and How Does It Work?

Romosozumab is a humanized monoclonal antibody that inhibits sclerostin, a protein secreted by osteocytes that normally suppresses bone formation. Blocking sclerostin simultaneously increases bone formation markers and decreases bone resorption markers, a dual mechanism not shared by any other approved osteoporosis agent. This produces BMD gains two to three times larger than those seen with teriparatide in the first 12 months of therapy.

Mechanism at the Cellular Level

Sclerostin binds LRP5 and LRP6 co-receptors and inhibits Wnt signaling. When romosozumab blocks sclerostin, Wnt signaling is restored, osteoblast activity rises, and RANKL expression by osteoblasts decreases, reducing osteoclast recruitment simultaneously. The net effect is a net anabolic window during the 12-month treatment course. After the 12 months, the anabolic effect attenuates, which is why sequential antiresorptive therapy is mandatory.

FDA Approval History

The FDA approved romosozumab on April 9, 2019, under the brand name Evenity, for treatment of osteoporosis in postmenopausal women at high risk of fracture, defined as a history of osteoporotic fracture, multiple risk factors for fracture, or failure or intolerance to other available osteoporosis therapies. The full prescribing information and REMS-adjacent boxed warning are available on the FDA label. [1]

GRADE Framework: A Primer for Applying It to Romosozumab

The GRADE (Grading of Recommendations Assessment, Development and Evaluation) system rates evidence quality as High, Moderate, Low, or Very Low, and grades recommendations as Strong or Conditional. For drug therapies, the process starts with RCT evidence rated High by default, then downgrades for risk of bias, inconsistency, indirectness, imprecision, or publication bias. It upgrades for large effect sizes, dose-response relationships, or plausible confounding that would underestimate the effect.

Why GRADE Matters for Romosozumab Specifically

Two large, well-powered RCTs (FRAME and ARCH) provide the evidentiary backbone. Both trials were industry-sponsored, which introduces potential conflict-of-interest as a downgrade consideration. However, the effect sizes are large, the direction is consistent across outcomes, and the trials were pre-registered with publicly available protocols, limiting selective reporting risk. [2]

GRADE also requires separate ratings for each outcome. The evidence profile for vertebral fracture reduction differs materially from the evidence profile for cardiovascular harm, and clinicians should treat them separately rather than applying a single label to the drug overall.

FRAME Trial: Romosozumab vs. Placebo (GRADE: High for Vertebral Fracture)

The FRAME trial enrolled 7,180 postmenopausal women with a T-score of -2.5 to -3.5 at the total hip or femoral neck. Participants received romosozumab 210 mg SC monthly or placebo for 12 months, followed by 12 months of open-label denosumab 60 mg SC every 6 months for all participants. [3]

Primary Fracture Outcomes at 12 Months

At 12 months, romosozumab reduced new vertebral fractures by 73% versus placebo (0.5% vs. 1.8%; P<0.001). Clinical fractures were reduced by 36% (P<0.001). Non-vertebral fractures showed a statistically significant reduction overall but were driven by geographic subgroups, introducing inconsistency that warrants a one-grade downgrade for that specific outcome.

BMD Gains at 12 Months

Total hip BMD increased 6.0% with romosozumab versus 0.6% with placebo. Lumbar spine BMD increased 13.3% versus 0.0%. These gains represent the largest 12-month BMD increments recorded in any phase 3 osteoporosis trial to date. [3]

GRADE Rating: FRAME Vertebral Fracture Outcome

• Study design: RCT (starts High)
• Risk of bias: Low (adequate allocation concealment, blinded outcome adjudication)
• Inconsistency: None for vertebral fracture
• Indirectness: None (direct patient-important outcome)
• Imprecision: None (wide confidence interval excluded at N=7,180)
• Publication bias: Undetected
• Effect size upgrade: Yes (RR 0.27, very large)

FRAME vertebral fracture: GRADE High. [4]

ARCH Trial: Romosozumab vs. Alendronate (GRADE: High for Vertebral Fracture, Moderate for Cardiovascular Harm)

The ARCH trial enrolled 4,093 postmenopausal women with a T-score of -2.5 or lower at the total hip or femoral neck, plus either a vertebral fracture history or a T-score of -3.0 or lower. Participants received romosozumab 210 mg SC monthly versus alendronate 70 mg orally weekly for 12 months, then all transitioned to open-label alendronate for 12 additional months. [5]

Fracture Outcomes at 24 Months

New vertebral fracture incidence was 6.2% in the alendronate group versus 3.3% in the romosozumab group, a 48% relative risk reduction (P<0.001). [5] Hip fracture incidence was 2.5% versus 3.6%, a 38% relative risk reduction (P=0.01). Clinical fracture incidence was 9.7% versus 11.9%, a 20% relative risk reduction (P<0.001). The magnitude and breadth of fracture reduction across all three outcomes strengthens the GRADE certainty rating.

Cardiovascular Signal in ARCH

Serious cardiovascular adverse events occurred in 2.5% of the romosozumab group versus 1.9% of the alendronate group (P=0.07 in the primary analysis, not statistically significant at the prespecified alpha level). Adjudicated major adverse cardiovascular events (MACE) were 0.8% versus 0.3% for non-fatal MI, a difference that drove the FDA boxed warning despite not reaching the prespecified significance threshold. [6]

The alendronate comparator is itself cardioprotective in observational data, which may have amplified the apparent romosozumab risk through an unfavorable comparator effect rather than an absolute hazard. This methodological ambiguity is why GRADE rates the cardiovascular harm signal as Moderate rather than High.

GRADE Rating: ARCH Vertebral Fracture Outcome

• Study design: RCT (starts High)
• Risk of bias: Low (pre-registered, blinded outcome adjudication, intention-to-treat analysis)
• Inconsistency: None; consistent with FRAME direction
• Indirectness: None
• Imprecision: None
• Large effect size with dose-response biologic plausibility: upgrade applied

ARCH vertebral fracture: GRADE High. [5]

GRADE Rating: ARCH Cardiovascular Harm Outcome

• Study design: RCT (starts High)
• Risk of bias: Potentially confounded by cardioprotective comparator
• Inconsistency: Signal not replicated in FRAME (placebo comparator); directionally inconsistent across trials
• Imprecision: Wide CI for MI endpoint; P=0.07 did not cross prespecified alpha
• Active comparator confounding: Downgrade applied

ARCH cardiovascular harm: GRADE Moderate. [6]

Non-Vertebral and Hip Fracture Evidence (GRADE: Moderate)

Hip fracture reduction is the most clinically consequential endpoint in osteoporosis trials because hip fractures carry a 20-30% one-year mortality in older adults. [7] The ARCH trial demonstrated a 38% RRR for hip fractures, but the FRAME trial showed no statistically significant hip fracture reduction in the 12-month primary analysis, though a post-hoc analysis of the 24-month follow-through period (including the denosumab extension) showed a 27% reduction. [8]

Why This Matters for GRADE

Inconsistency between FRAME and ARCH for non-vertebral outcomes forces a one-level downgrade from High to Moderate for hip and non-vertebral fracture endpoints. The difference in trial design (active comparator vs. Placebo, and 12-month vs. 24-month primary endpoint) plausibly explains the inconsistency, but GRADE methodology requires downgrading inconsistency regardless of a plausible explanation. [4]

Clinicians should communicate to patients that the evidence for spine protection is stronger than the evidence for hip protection, though both are directionally favorable.

Long-Term BMD Durability: Sequential Therapy Evidence

Romosozumab is approved for 12 monthly doses only. The anabolic effect attenuates by month 12, and without subsequent antiresorptive therapy, the BMD gains are lost within 12-18 months. [9] Both FRAME and ARCH built mandatory sequential therapy into their designs, which means the fracture-reduction data already reflect a romosozumab-then-antiresorptive sequence rather than romosozumab alone.

Romosozumab Followed by Denosumab (FRAME Extension)

In the FRAME extension, participants who completed romosozumab transitioned to denosumab and sustained lumbar spine BMD gains of approximately 14.9% above baseline at 24 months. Participants who received placebo in year 1 and then denosumab in year 2 achieved only 7.1% lumbar spine BMD gain at 24 months. [3] The 7.8 percentage-point difference at 24 months demonstrates a durable advantage from the romosozumab-first sequence.

Romosozumab Followed by Alendronate (ARCH Extension)

In ARCH, the superior fracture outcomes persisted throughout the 24-month study period even after both groups transitioned to alendronate in month 13. The earlier and larger BMD gains with romosozumab produced a "bone mass bank" that continued to reduce fracture risk through the follow-up period. [5]

Head-to-Head vs. Teriparatide: The STRUCTURE Trial

The STRUCTURE trial (N=436) compared romosozumab 210 mg SC monthly to teriparatide 20 mcg SC daily in women previously treated with bisphosphonates. Total hip BMD increased 2.6% with romosozumab versus a decrease of 0.6% with teriparatide at 12 months (P<0.001). [10] Lumbar spine gains were comparable, but the hip advantage for romosozumab was clinically significant given that hip BMD predicts hip fracture risk more directly than spine BMD.

STRUCTURE was not powered for fracture outcomes, so the GRADE rating for fracture reduction versus teriparatide remains Very Low due to indirectness (surrogate BMD outcome only). The BMD data are rated Moderate due to adequate randomization and blinding but limited sample size.

Guideline Recommendations

American Association of Clinical Endocrinology (AACE/ACE 2020)

The AACE/ACE 2020 Postmenopausal Osteoporosis guidelines rate romosozumab as a Grade A recommendation for postmenopausal women with very high fracture risk, defined as a recent fracture (within 1-2 years), very low T-score (below -3.0), or fractures occurring on antiresorptive therapy. [11] The guidelines place romosozumab alongside teriparatide and abaloparatide as first-line anabolic options for the very-high-risk tier.

The AACE document states: "For patients at very high risk of fracture, anabolic therapy followed by antiresorptive therapy is the preferred sequence to maximize fracture risk reduction." [11]

Endocrine Society 2019 Guideline

The Endocrine Society's 2019 pharmacological management guideline conditionally recommends romosozumab for postmenopausal women at high fracture risk who do not have a history of MI or stroke within the preceding year. [12] The conditional (rather than strong) recommendation reflects the Moderate GRADE certainty for the cardiovascular harm signal rather than any uncertainty about fracture efficacy.

The guideline notes: "The cardiovascular risk observed in ARCH warrants caution in patients with pre-existing cardiovascular disease, and clinicians should weigh the absolute fracture risk reduction against the potential for cardiovascular harm on an individual basis." [12]

American College of Rheumatology (ACR 2022)

The ACR's 2022 guideline for glucocorticoid-induced osteoporosis conditionally recommends romosozumab for very high-risk patients on chronic glucocorticoids who have not responded to bisphosphonates, acknowledging that the evidence base in glucocorticoid-treated populations is extrapolated from the general postmenopausal osteoporosis trials rather than direct RCT evidence in that subgroup. [13]

Cardiovascular Safety: Full Evidence Profile

The FDA boxed warning is the single most important clinical consideration limiting romosozumab use in real-world practice. The prescribing information states that romosozumab should not be initiated in patients who have had an MI or stroke within the preceding year. [1]

Mechanistic Hypotheses

Three mechanisms have been proposed. First, sclerostin may be cardioprotective, and its inhibition could remove a protective effect on vascular smooth muscle. Second, rapid calcium flux from bone matrix remodeling could theoretically affect cardiac function. Third, the ARCH cardiovascular signal may reflect the cardioprotective effect of the alendronate comparator rather than a true romosozumab hazard. [14] None of these hypotheses has been definitively confirmed or excluded.

Post-Marketing Surveillance Data

A 2021 FDA drug safety communication reviewed post-marketing reports and did not identify a new or stronger cardiovascular signal beyond what was observed in ARCH. [1] The absolute excess risk in ARCH was approximately 0.6 percentage points for MACE over 24 months, which must be weighed against a vertebral fracture NNT of approximately 33 at 24 months in the ARCH population.

The HealthRX clinical team uses the following decision framework for romosozumab candidacy:

Tier 1 (Preferred candidates): Postmenopausal women aged 50-80, T-score at or below -2.5 with at least one prior vertebral or hip fracture, no cardiovascular event in the past 12 months, no planned dental surgery within 12 months, and access to sequential denosumab or bisphosphonate therapy.

Tier 2 (Conditional candidates): Women with T-score at or below -3.0 without prior fracture but with multiple clinical risk factors (FRAX 10-year hip fracture risk above 3%, prior bisphosphonate failure defined as new fracture on therapy, or very rapid bone loss documented by serial DXA).

Tier 3 (Avoid or defer): Any patient with MI, stroke, or TIA within 12 months; active cancer with bone metastases; severe renal impairment (eGFR <30 mL/min/1.73m2); planned invasive dental procedure within 3 months; or hypocalcemia not yet corrected.

Special Populations

Men with Osteoporosis

The FDA label does not include men as an approved indication. A phase 3 trial (BRIDGE, N=245) showed significant BMD increases in men with osteoporosis treated with romosozumab 210 mg monthly versus placebo: lumbar spine BMD increased 12.1% versus 1.2% (P<0.001) and total hip BMD increased 2.5% versus -0.5% (P<0.001) at 12 months. [15] The trial was not powered for fracture outcomes. Use in men is off-label, and GRADE certainty for fracture outcomes in men is Very Low due to indirectness and absence of fracture data.

Glucocorticoid-Induced Osteoporosis

No dedicated RCT has evaluated romosozumab specifically in glucocorticoid-induced osteoporosis (GIOP). The ACR 2022 guideline recommendation for this population is therefore based on indirect evidence, which GRADE rates as Low due to indirectness. Teriparatide retains more direct RCT evidence in GIOP through the Saag et al. Trial published in the New England Journal of Medicine. [16]

Patients Previously Treated with Bisphosphonates

STRUCTURE and ARCH both included patients with prior bisphosphonate exposure. Romosozumab retains fracture-reduction efficacy in this population, though the BMD gains at the total hip may be attenuated compared to bisphosphonate-naive patients. Prior long-term bisphosphonate use (more than 5 years) does not appear to ablate the anabolic response, which contrasts with teriparatide, where the PTH receptor desensitizes more when bisphosphonate suppression is profound. [10]

Dosing, Administration, and Monitoring

Romosozumab is supplied as 105 mg/1.17 mL solution in a single-use autoinjector. The approved dose is 210 mg SC monthly, given as two consecutive injections of 105 mg each, one in each abdomen, thigh, or upper arm. [1]

Pre-Treatment Checklist

Before initiating therapy, clinicians should confirm: serum calcium within the normal range (correct hypocalcemia before starting), adequate vitamin D status (25-OH vitamin D at or above 20 ng/mL), dental examination completed (ONJ risk is present though lower than with long-term denosumab or zoledronic acid), and no cardiovascular event within 12 months. [1]

Monitoring During Treatment

Monthly monitoring of bone markers is not required by the label but may be used to confirm a therapeutic response. Serum P1NP (procollagen type 1 N-terminal propeptide) rises sharply within 1 month and then returns toward baseline by month 9, tracking the waning anabolic window. Serum CTX (C-terminal telopeptide), a resorption marker, falls within the first month and remains suppressed throughout the 12-month course. [3] A P1NP that fails to rise within the first 3 months should prompt evaluation for malabsorption, uncorrected vitamin D deficiency, or non-adherence.

Comparative Effectiveness: Romosozumab vs. Other Anabolic Agents

Two other anabolic agents are available in the United States: teriparatide (Forteo, PTH 1-34) and abaloparatide (Tymlos, PTHrP analog).

Vertebral Fracture Reduction

Teriparatide reduced vertebral fractures by 65% versus placebo in the Neer et al. RCT (N=1,637; P<0.001). [17] Abaloparatide reduced vertebral fractures by 86% versus placebo in the ACTIVE trial (N=2,463; P<0.001). [18] Romosozumab reduced vertebral fractures by 73% versus placebo in FRAME and by 48% versus alendronate in ARCH. Direct comparisons across trials are limited by differences in baseline T-score and fracture risk, but all three agents carry GRADE High evidence for vertebral fracture reduction.

Hip Fracture Reduction

Abaloparatide demonstrated a statistically significant 43% RRR for non-vertebral fractures in ACTIVE. [18] Teriparatide showed a non-significant trend toward non-vertebral fracture reduction in Neer et al. Romosozumab showed significant hip fracture reduction only in ARCH (active comparator design), not in FRAME (placebo comparator, 12-month primary endpoint). No head-to-head RCT has directly compared fracture outcomes across all three anabolic agents.

Cost-Effectiveness and Access Considerations

A 2021 cost-effectiveness analysis published in Osteoporosis International estimated the incremental cost-effectiveness ratio (ICER) for romosozumab versus alendronate at approximately $189,000 per quality-adjusted life year (QALY) gained in the overall postmenopausal osteoporosis population. [19] The ICER fell below $100,000 per QALY only in women aged 70 or older with a T-score at or below -3.0 and a prior vertebral fracture. This data supports concentrating romosozumab use in the highest-risk tier rather than broad application across all osteoporosis patients.

Most commercial payers in the United States require documented failure or contraindication to bisphosphonates before approving romosozumab, though the FDA label and AACE/ACE guidelines do not require prior bisphosphonate failure for very-high-risk patients. This coverage restriction may delay access for the patients most likely to benefit from early anabolic therapy. [11]