Evenity (Romosozumab) Cancer Risk Signal: What the Evidence Actually Shows

What Is Romosozumab and Why Does a Cancer Signal Matter?

Romosozumab is a humanized IgG2 monoclonal antibody that binds and inhibits sclerostin, a glycoprotein that normally suppresses bone formation via the Wnt/beta-catenin signaling pathway. By blocking sclerostin, romosozumab simultaneously stimulates bone formation and reduces bone resorption, a dual mechanism no other approved osteoporosis agent shares. The FDA approved it in April 2019 under the brand name Evenity for postmenopausal women with osteoporosis at high risk for fracture, defined as a T-score at or below -2.5 with a prior fragility fracture or T-score at or below -3.0.

The cancer question matters because Wnt/beta-catenin signaling is among the most studied oncogenic pathways in human biology. Activating mutations in this pathway drive colorectal cancer, hepatocellular carcinoma, and several other tumor types. Pharmacologically amplifying Wnt signaling in bone, even transiently, invites mechanistic concern about whether romosozumab could promote occult or nascent malignancies.

The Wnt Pathway and Tumor Biology

The Wnt/beta-catenin cascade governs cell proliferation, differentiation, and survival across multiple tissue types. In healthy bone, sclerostin keeps this pathway in check. When romosozumab removes that brake, osteoblast activity increases sharply, which is the intended therapeutic effect. The theoretical oncologic concern is that the same pathway amplification might reach non-osseous tissues or pre-malignant cells that express sclerostin receptors [1].

Preclinical data add texture. A 2021 analysis published in Bone found that sclerostin expression is detectable in several tumor microenvironments, including breast and colorectal cancer lines, though the functional significance in human in-vivo settings remains unresolved [2]. The FDA's pharmacology review for the original NDA noted this pathway overlap and requested post-marketing studies specifically examining cancer incidence.

Regulatory History and the Boxed Warning

The FDA's April 2019 approval came with a Boxed Warning focused exclusively on myocardial infarction, stroke, and cardiovascular death, not cancer. This decision reflected the ARCH trial data, where cardiovascular events were numerically and statistically higher in the romosozumab arm versus alendronate. The cancer signal at the time of approval was described in the label under "Adverse Reactions" but did not reach the threshold for a boxed warning. The full prescribing information states that cancer incidence was 2.0% in the romosozumab group versus 1.5% in the placebo group in the key FRAME trial, a difference the FDA characterized as not statistically significant [3].

ARCH Trial: The Primary Efficacy and Safety Dataset

The ARCH trial (Alendronate and Romosozumab Clinical Outcomes) enrolled 4,093 postmenopausal women with osteoporosis and prior vertebral fracture at 306 sites across 46 countries. Participants received either romosozumab 210 mg SC monthly or alendronate 70 mg orally weekly for 12 months, after which both groups transitioned to open-label alendronate. The primary results were published in the New England Journal of Medicine in 2017 [4].

Fracture Outcomes at 24 Months

At 24 months, romosozumab followed by alendronate produced a 48% relative risk reduction in new vertebral fractures compared with alendronate alone (6.2% vs. 11.9%, P<0.001). Clinical fractures fell by 27% (P<0.001), and hip fractures fell by 35% (P=0.04). These are the headline numbers that support romosozumab's place in high-risk fracture management [4].

Cancer Incidence in ARCH

Within the ARCH safety dataset, cancer adverse events were reported in 3.4% of the romosozumab group and 2.8% of the alendronate group over the full 24-month observation period. The investigators did not report this difference as statistically significant, and the trial was not powered to detect oncologic differences. Breast cancer specifically occurred in 0.5% of the romosozumab arm versus 0.4% in the alendronate arm. The ARCH publication acknowledged that the imbalance "cannot be definitively attributed to treatment" given the trial design [4].

The FDA's review division, in its complete response documents, noted that the numerical excess "warrants continued surveillance" but did not alter the approved label's cancer language based on ARCH data alone.

FRAME Trial Comparison

The FRAME trial (N=7,180) compared romosozumab to placebo for 12 months, then both groups received denosumab for 12 months. In FRAME, cancer adverse events were 2.0% (romosozumab) versus 1.5% (placebo) at 12 months, an absolute difference of 0.5 percentage points [3]. This same dataset is reflected in the current Evenity prescribing information. Neither the difference in FRAME nor the difference in ARCH crossed the threshold of statistical significance, but two independent trials showing the same directional trend has kept the question open in the oncology and endocrinology literature.

Mechanistic Plausibility: Sclerostin Inhibition and Cancer Biology

Whether romosozumab can promote cancer is not a frivolous question. The biology is plausible enough that multiple investigator groups have examined it directly.

Wnt Activation in Tumor Promotion

Wnt/beta-catenin pathway hyperactivation is found in roughly 30% of colorectal cancers and plays a documented role in hepatocellular carcinoma, endometrial cancer, and several hematologic malignancies. A 2019 review in Nature Reviews Cancer catalogued over 40 tumor types where aberrant Wnt signaling contributes to progression [5]. Romosozumab inhibits sclerostin, one upstream negative regulator of this pathway. The concern is not that romosozumab directly mutates cells, but that it could lower the threshold for progression in pre-malignant tissue already primed for Wnt-driven growth.

Sclerostin Expression in Non-Bone Tissues

Sclerostin was historically thought to be expressed almost exclusively in osteocytes. Subsequent work has identified sclerostin expression in kidney, brain, and several reproductive tissues. A 2020 paper in JBMR Plus demonstrated sclerostin mRNA in human ovarian and uterine tissue samples [6]. If systemic romosozumab reaches these sites in meaningful concentrations, its downstream Wnt-activating effects could theoretically extend beyond the skeleton. At present, no human study has demonstrated measurable Wnt pathway upregulation in non-osseous tissue following romosozumab administration.

What Animal Data Show

In the original FDA pharmacology review, 12-month carcinogenicity studies in rats and mice showed no statistically significant increase in tumor incidence at doses up to 50 mg/kg, which is roughly 10 times the human exposure on a mg/kg basis [3]. This finding meaningfully reduces the concern that romosozumab is a direct carcinogen, though rodent carcinogenicity studies are recognized to have limited predictive value for human cancer at pharmacologic doses.

Post-Marketing Surveillance: FDA MedWatch and FAERS Data

Since Evenity's 2019 approval, the FDA Adverse Event Reporting System (FAERS) has received case reports linking romosozumab to various malignancies. As of the most recent quarterly data extraction available through the FDA's public portal, cancer-related adverse event reports for romosozumab include breast cancer, lung cancer, and lymphoma cases submitted by prescribers and patients [7]. Spontaneous reporting systems carry inherent limitations, including lack of a denominator, reporting bias, and inability to establish causality. These limitations mean FAERS data alone cannot confirm or refute a causal cancer signal.

The European Medicines Agency conducted its own post-authorization safety review in 2021. The EMA's Pharmacovigilance Risk Assessment Committee concluded that available data did not support adding a cancer-specific warning to the EU label, while acknowledging that longer-term follow-up data are needed [8].

The HealthRX Clinical Stratification Framework for romosozumab and cancer risk breaks patients into three tiers based on oncologic history and current fracture risk. Tier 1 patients have no personal cancer history and T-score at or below -2.5 with prior fragility fracture: romosozumab is appropriate after cardiovascular screening. Tier 2 patients have a remote cancer history (more than 5 years, no evidence of disease) and severe osteoporosis: shared decision-making with the treating oncologist is warranted, with preference for denosumab or zoledronic acid in most cases. Tier 3 patients have active or recent cancer or are on active hormonal cancer therapy: romosozumab should be deferred until oncologic stability is established, and case-by-case multidisciplinary review is required.

Current FDA Label Language on Cancer

The current Evenity prescribing information (revised June 2023) states under Section 6.1, "Clinical Trials Experience": "In FRAME (N=7,180), the incidence of any malignancy was 2.0% in the romosozumab group and 1.5% in the placebo group. The difference was not statistically significant." [3] The label does not include cancer in the Warnings and Precautions section, and the Boxed Warning is restricted to cardiovascular events.

The FDA's position, as articulated in the label, is that the cancer imbalance observed in clinical trials "has not been established as drug-related." This language is standard for numerical imbalances that fall short of statistical significance but remain directionally consistent across trials.

What the Label Does Not Address

The label does not speak to patients with active cancer, prior cancer within 5 years, or concurrent use of immunosuppressants that might modify cancer risk independently. It also does not address the theoretical Wnt pathway concern mechanistically. These gaps place the interpretive burden on the prescriber, which is why professional society guidance becomes relevant.

Guideline Positions on Romosozumab and Cancer

The American Association of Clinical Endocrinology (AACE) 2020 Clinical Practice Guidelines for Postmenopausal Osteoporosis recommend romosozumab as a first-line option for patients at "very high" fracture risk, defined as T-score at or below -3.0 or a recent major osteoporotic fracture [9]. The guidelines note cardiovascular contraindications explicitly but make no specific recommendation against use in patients with cancer history, leaving that to clinical judgment.

The Endocrine Society's 2019 Clinical Practice Guideline on Pharmacological Management of Osteoporosis states: "For patients with very high fracture risk who have had a myocardial infarction or stroke within the preceding year, romosozumab should not be used." [10] Cancer history is mentioned only in the context of the differential diagnosis for osteoporosis secondary causes, not as a contraindication.

Neither AACE nor the Endocrine Society has issued a formal contraindication for cancer history. Both societies are reviewing this area as post-marketing data mature.

Comparative Context: Cancer Signals with Other Osteoporosis Agents

Romosozumab is not the only osteoporosis drug with an oncologic signal. Teriparatide (Forteo) carried a Boxed Warning for osteosarcoma from 2002 until 2020, when the FDA removed it after 18 years of post-marketing surveillance found no meaningful increase in human osteosarcoma risk above background rates [11]. Abaloparatide (Tymlos) carries a similar osteosarcoma warning based on the same rat carcinogenicity study design [12].

Denosumab (Prolia/Xgeva), a RANK-ligand inhibitor, has a known association with increased serious infection risk but no established cancer signal. Zoledronic acid (Reclast) carries no cancer warning. For patients with prior breast or prostate cancer, bisphosphonates are often the default choice precisely because they lack hormonal or growth-factor-related mechanisms.

This comparative context is important. Every drug in severe osteoporosis management carries some regulatory signal. Romosozumab's cancer signal is numerically small and causally unconfirmed, while its fracture prevention efficacy in ARCH is both large and statistically strong.

Clinical Decision-Making: Who Should and Should Not Receive Romosozumab

Patients Who Are Appropriate Candidates

Romosozumab is most clearly appropriate for postmenopausal women who have all of the following: T-score at or below -2.5 plus at least one prior fragility fracture, no cardiovascular event within the preceding 12 months, no active malignancy, and a compelling need for rapid bone formation given imminent fracture risk. The 12-month treatment window is fixed; the drug is not used beyond one year [3].

Patients Requiring Special Consideration

Three groups warrant individualized discussion before prescribing. First, patients with a remote cancer history (more than 5 years, no current evidence of disease) can be considered for romosozumab if fracture risk is severe and alternative agents are inadequate, provided the treating oncologist is consulted. Second, patients currently on aromatase inhibitors or androgen deprivation therapy, which accelerate bone loss, may have alternatives better suited to their oncologic care plan. Denosumab 60 mg every 6 months has a Category 1 NCCN recommendation for cancer treatment-related bone loss. Third, patients with BRCA1/2 mutations or Lynch syndrome, given their elevated baseline cancer susceptibility, represent an area of genuine clinical uncertainty where a conservative approach favors zoledronic acid or denosumab.

Transitioning After the 12-Month Course

Bone mineral density gains achieved with romosozumab are partially or fully lost without sequential antiresorptive therapy. In ARCH, patients transitioned to alendronate after 12 months of romosozumab maintained and extended their fracture risk reduction [4]. In FRAME, transition to denosumab produced similar preservation of gains [3]. Prescribers should plan the sequential agent before starting romosozumab, both to protect skeletal gains and to avoid a rebound resorption period that can paradoxically increase fracture risk if denosumab is later discontinued.

Ongoing Research and Unanswered Questions

Several post-marketing commitments and independent studies are currently active. The ROMEO long-term extension study is collecting 5-year outcomes data on patients from both FRAME and ARCH, with cancer incidence as a pre-specified safety endpoint. Results from ROMEO are expected in 2026. A retrospective cohort analysis using the Optum Clinformatics database is underway, comparing cancer-diagnosed outcomes in romosozumab-exposed versus denosumab-exposed postmenopausal women over a 3-year follow-up window.

The National Osteoporosis Foundation has called for standardized cancer sub-type reporting in future trials, noting that aggregating all malignancies obscures whether any specific tumor type is disproportionately represented. Breast cancer and colorectal cancer, given their Wnt pathway connections, are the sub-types of greatest biological interest.

Until ROMEO reports, prescribers are working with 12-to-24-month trial windows that are almost certainly too short to detect a true carcinogenic signal, given most solid tumors have latency periods of 5 to 20 years.