What was the primary endpoint of the ARCH trial?

New vertebral fractures at 24 months, assessed by vertebral radiographs. The romosozumab-to-alendronate sequence reduced this by 48% compared with alendronate alone (6.2% vs 11.9%; RR 0.52 to 95% CI 0.40, 0.66).

How quickly did fracture protection appear with romosozumab in ARCH?

Kaplan-Meier curves for clinical fracture diverged within the first 6 months. By 12 months (still in the double-blind phase), vertebral fracture rates were already 50% lower in the romosozumab group.

Did ARCH show hip fracture reduction?

Yes. Hip fractures were reduced by 38% (HR 0.62; 95% CI 0.42, 0.92; p = 0.02), one of the few anabolic agent trials to demonstrate a statistically significant hip fracture benefit.

What happened to BMD after patients switched from romosozumab to alendronate?

BMD gains were maintained or continued to increase modestly at the hip during the alendronate phase. The transition preserved the rapid bone accrual achieved during the romosozumab period.

Why does the Evenity label have a cardiovascular boxed warning?

ARCH showed a numerical (not statistically significant) excess of cardiovascular events with romosozumab vs alendronate during the 12-month double-blind period (2.5% vs 1.9%). The FDA added a boxed warning contraindicating use within one year of MI or stroke.

How does ARCH differ from the FRAME trial?

FRAME compared romosozumab to placebo in a lower-risk population. ARCH compared romosozumab to an active comparator (alendronate) in higher-risk women with prior fractures. ARCH showed hip fracture benefit; FRAME did not.

Was the ARCH trial double-blinded throughout?

Only the first 12 months were double-blinded. After month 12, all patients received open-label alendronate, though fracture adjudication remained blinded.

Who should receive romosozumab based on ARCH data?

Postmenopausal women with osteoporosis at high fracture risk (prior fracture, very low BMD) who do not have recent cardiovascular events. Guidelines from AACE and others recommend this sequence for very high-risk patients.

What is the number needed to treat from ARCH?

Approximately 18 patients treated with the romosozumab-to-alendronate sequence to prevent one additional vertebral fracture over 24 months, compared with alendronate alone.

Can romosozumab be used for more than 12 months?

The FDA label limits romosozumab to 12 monthly doses. ARCH did not test extended duration. After the 12-dose course, patients should transition to an antiresorptive agent to maintain BMD gains.

ARCH Results in Detail: Numbers, Subgroups, and Time Course

By HealthRX Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

At a glance

| Field | Detail | |---|---| | Trial name | ARCH (Active-contRolled fraCture study in postmenopausal women with osteoporosis at High risk) | | N | 4,093 postmenopausal women | | Intervention | Romosozumab 210 mg SC monthly x 12 months, then alendronate 70 mg weekly | | Comparator | Alendronate 70 mg weekly x 12 months, then continued alendronate | | Duration | 12-month double-blind period + open-label alendronate (median follow-up ~33 months) | | Primary endpoint | Incidence of new vertebral fractures at 24 months | | Key result | 48% relative risk reduction vs alendronate (p <0.001) | | Registration | NCT01631214 |

Why ARCH Matters Beyond the Abstract

Most osteoporosis trials compare active drug to placebo. ARCH tested romosozumab head-to-head against alendronate, the most widely prescribed bisphosphonate worldwide. The trial enrolled women with established osteoporosis (prior fracture history or very low BMD), making the population clinically relevant for the exact patients clinicians worry about most: those at imminent fracture risk who need the strongest first-line option available.

The trial's design also addressed a real clinical question. If a patient starts romosozumab and then transitions to a bisphosphonate (as the FDA label for Evenity recommends), does the sequence outperform starting a bisphosphonate from day one? ARCH answered that definitively.

Primary Endpoint: New Vertebral Fractures at 24 Months

The primary analysis used a modified intention-to-treat population. At 24 months, the cumulative incidence of new vertebral fractures was:

| Outcome | Romosozumab → Alendronate | Alendronate → Alendronate | Relative Risk (95% CI) | p-value | |---|---|---|---|---| | New vertebral fracture (24 mo) | 6.2% (127/2,046) | 11.9% (243/2,047) | 0.52 (0.40, 0.66) | <0.001 |

The 48% relative risk reduction translates to an absolute risk difference of 5.7 percentage points. The number needed to treat (NNT) to prevent one vertebral fracture over 24 months was approximately 18, a clinically meaningful figure for a population with baseline vertebral fracture rates this high.

At the 12-month mark, before the open-label transition, the vertebral fracture rate was already lower in the romosozumab group (4.0% vs 8.0%), representing a 50% relative risk reduction (RR 0.50; 95% CI 0.37, 0.67). This confirms that most of the vertebral fracture benefit accrued during the initial romosozumab treatment period itself, not simply from the subsequent alendronate phase.

Secondary Endpoints: Clinical, Nonvertebral, and Hip Fractures

ARCH pre-specified several secondary fracture endpoints analyzed at the time of the primary analysis (event-driven, median ~33 months of follow-up):

| Endpoint | Romosozumab → Alen | Alendronate → Alen | Hazard Ratio (95% CI) | p-value | |---|---|---|---|---| | Clinical fracture | 9.7% | 13.0% | 0.73 (0.61, 0.86) | <0.001 | | Nonvertebral fracture | 8.7% | 10.6% | 0.81 (0.66, 0.99) | 0.04 | | Hip fracture | 2.0% | 3.2% | 0.62 (0.42, 0.92) | 0.02 |

The hip fracture reduction (38%) is particularly notable. Hip fractures carry the highest morbidity and mortality of any osteoporotic fracture, and few anabolic agents have demonstrated statistically significant hip fracture reduction in a single trial. The FRAME trial (romosozumab vs placebo) did not show a significant hip fracture benefit, likely because the placebo-controlled population had lower baseline fracture risk.

Clinical fractures (a composite of symptomatic vertebral and all nonvertebral fractures) showed a 27% reduction. This endpoint matters because it captures the fractures patients actually present with, not just those found on protocol-mandated radiographs.

Time-Course Pattern: When the Curves Separate

The Kaplan-Meier curves for clinical fracture diverged within the first 6 months. By month 12, when patients in the romosozumab arm transitioned to alendronate, the gap was already established. During the open-label period (months 12, 33+), the curves continued to separate, though at a slower rate.

This pattern has two clinical implications. First, romosozumab's bone-forming mechanism provides rapid fracture protection that bisphosphonates alone cannot match in the early treatment window. Second, transitioning to alendronate after romosozumab preserves and extends the gains rather than allowing regression. The American Association of Clinical Endocrinologists (AACE) 2020 guidelines cite this sequential strategy as a preferred approach for very high-risk patients.

For nonvertebral fractures, curve separation was more gradual, becoming statistically significant only at the final analysis. Hip fracture curves separated around month 12, with the hazard ratio estimate carrying wider confidence intervals (0.42, 0.92) reflecting the smaller event count.

BMD Gains: Magnitude and Trajectory

While BMD was not the primary endpoint, the density changes contextualize the fracture reductions.

| Site | Romosozumab (12 mo) | Alendronate (12 mo) | Difference | |---|---|---|---| | Lumbar spine | +13.7% | +5.0% | +8.7 percentage points | | Total hip | +6.2% | +2.8% | +3.4 percentage points | | Femoral neck | +4.9% | +1.7% | +3.2 percentage points |

At 12 months, romosozumab produced nearly triple the lumbar spine BMD gain and more than double the hip BMD gain versus alendronate. After transition to alendronate, BMD in the romosozumab group continued to increase modestly at the hip and was maintained at the spine, while the alendronate-only group continued its slower upward trajectory.

The rapid BMD accrual explains how fracture protection appears so early. Bone turnover markers confirmed this: P1NP (a formation marker) spiked within the first month on romosozumab and returned to baseline by month 12, while CTX (a resorption marker) dropped. This dual mechanism (simultaneous formation increase and resorption decrease) is unique to sclerostin inhibition and distinguishes romosozumab from both pure anabolics like teriparatide and pure antiresorptives like alendronate.

Subgroup Analyses

Pre-specified subgroup analyses examined consistency across baseline characteristics:

Prior vertebral fracture status: Patients with prior vertebral fractures (who comprised the majority) showed consistent benefit. Those with multiple prior fractures had higher absolute risk reductions.
Age strata (<75 vs ≥75 years): The relative risk reduction was consistent across age groups. Older patients had higher absolute event rates in both arms, yielding smaller NNTs.
Baseline T-score: Patients with lower baseline BMD (T-score ≤ −3.0 at any site) showed similar relative reductions. No interaction was detected between baseline BMD severity and treatment effect.
Geographic region: Results were consistent across North America, Europe, and Latin America enrollment sites.

No subgroup showed a statistically significant interaction with treatment, meaning the benefit appeared broadly applicable across the enrolled population rather than driven by a specific subset.

Cardiovascular Safety Signal

ARCH reported a numerical imbalance in adjudicated serious cardiovascular events during the 12-month double-blind period: 50 events in the romosozumab group (2.5%) vs 38 in the alendronate group (1.9%). This difference was not statistically significant, but it triggered regulatory scrutiny.

The FDA's 2019 approval of Evenity included a boxed warning about potential cardiovascular risk, contraindicating use in patients who have had a myocardial infarction or stroke within the preceding year. The FRAME trial (placebo-controlled) did not show a similar imbalance, raising questions about whether the ARCH signal reflects a true romosozumab risk or a cardioprotective effect of alendronate that made the comparator arm look relatively better.

Post-hoc analyses noted that the cardiovascular events in ARCH were not clustered at any particular time point and that baseline cardiovascular risk factors were balanced between groups. The European Medicines Agency also approved romosozumab but with similar cardiovascular contraindications.

Limitations the Authors Acknowledged

Several limitations deserve mention for clinicians interpreting ARCH:

Population specificity. All participants were postmenopausal women with osteoporosis and prior fracture history. Results may not generalize directly to men, premenopausal women, or patients without prior fractures.
Open-label second phase. After month 12, both investigators and patients knew treatment assignment. While fracture adjudication was blinded, reporting bias for symptomatic endpoints cannot be fully excluded.
No pure placebo arm. ARCH cannot quantify the absolute fracture risk reduction of either drug. It can only establish relative superiority of the sequence strategy over bisphosphonate monotherapy.
Cardiovascular event adjudication. The trial was not powered to detect differences in cardiovascular outcomes. The boxed warning stems from a signal that remains statistically inconclusive within ARCH itself.
Finite treatment duration. Romosozumab's label limits use to 12 monthly doses. ARCH does not address what happens if patients receive no follow-on antiresorptive therapy after romosozumab.

Putting the Numbers in Clinical Context

The ARCH results established a treatment hierarchy for severe osteoporosis: starting with a bone-forming agent and transitioning to an antiresorptive outperforms starting with an antiresorptive alone. The 48% vertebral fracture reduction, 38% hip fracture reduction, and rapid time-to-benefit position this sequence as the strongest evidence-based option for patients at imminent fracture risk, with the cardiovascular contraindication as the primary gating factor for patient selection.

Frequently asked questions

›

References

Saag KG, Petersen J, Brandi ML, et al. Romosozumab or alendronate for fracture prevention in women with osteoporosis. N Engl J Med. 2017;377(15):1417-1427. PubMed
Cosman F, Crittenden DB, Adachi JD, et al. Romosozumab treatment in postmenopausal women with osteoporosis (FRAME trial). N Engl J Med. 2016;375(16):1532-1543. PubMed
Camacho PM, Petak SM, Binkley N, et al. American Association of Clinical Endocrinologists/American College of Endocrinology clinical practice guidelines for the diagnosis and treatment of postmenopausal osteoporosis, 2020 update. Endocr Pract. 2020;26(Suppl 1):1-46. PubMed
Evenity (romosozumab-aqqg) prescribing information. Amgen Inc. 2019. FDA Label
Fosamax (alendronate sodium) prescribing information. Merck & Co. 2012. FDA Label