GRAPHICAL ABSTRACT

INTRODUCTION

Romosozumab, an anti-sclerostin monoclonal antibody, rapidly increases bone mineral density (BMD) and reduces fracture risk in postmenopausal women with osteoporosis [1–3]. Treatment with romosozumab following alendronate has shown superior efficacy in fracture risk reduction compared with teriparatide [4]. Because of its substantial therapeutic benefits, romosozumab is considered one of the most effective treatment options for osteoporosis, particularly for patients at very high fracture risk [5–7].

However, concerns regarding cardiovascular safety emerged from the Active-Controlled Fracture Study in Postmenopausal Women with Osteoporosis at High Risk (ARCH) trial, which reported a numerically higher incidence of adjudicated serious cardiovascular adverse events (CVAEs) within 1 year among postmenopausal women treated with romosozumab compared with those treated with alendronate, although the difference did not reach statistical significance [2]. A subsequent post hoc analysis of the ARCH trial reported a significant increase in the risk of major adverse cardiovascular events (MACE) among patients receiving romosozumab [8]. Consequently, the U.S. Food and Drug Administration issued a prescribing warning recommending against romosozumab use in patients with a history of myocardial infarction or stroke within the preceding year. However, no increased risk of cardiovascular disease was observed in the FRActure study in postmenopausal woMen with ostEoporosis (FRAME) trial comparing romosozumab with placebo [1].

To date, only a few observational studies have evaluated the cardiovascular risks associated with romosozumab treatment [9–11]. The present study aimed to compare the risk of cardiovascular events among patients with osteoporosis initiating romosozumab with that among patients initiating denosumab using a multicenter retrospective cohort derived from a standardized electronic health record data network. Denosumab was selected as the active comparator to achieve clinical equipoise because it is currently considered cardiovascularly neutral [12] and demonstrates potent anti-osteoporotic effects comparable to those of romosozumab.

METHODS

Study design and data sources

We conducted an observational comparative cohort study using the Osteoporosis Analysis and Surveillance Initiative using Standardized data (OASIS) network, a distributed research network based on the Observational Medical Outcomes Partnership Common Data Model (OMOP-CDM). The OASIS network includes standardized electronic health records from Yonsei University Health System (YUHS; approximately 6.27 million individuals, spanning January 2006 to August 2023), Asan Medical Center (AMC; approximately 5 million individuals, from May 1989 to June 2024), and Korea University Medical Center (KUMC; approximately 2.18 million individuals, from January 2009 to June 2021) [13]. The Institutional Review Boards (IRBs) of the three hospitals approved this study and granted exemptions from the requirement for individual informed consent (IRB numbers: 2024AN0175, 2023-0860, 2023-06-007, and 4-2023-0492).

Study population and exposure

The study included patients aged ≥50 years who were newly prescribed romosozumab or denosumab between January 1, 2020, and the end of each hospital’s data collection period. The romosozumab group served as the target cohort, and the denosumab group served as the comparator cohort. The index date was defined as the date of the first prescription of romosozumab for the target cohort and denosumab for the comparator cohort. The detailed definition of exposure is provided in Supplemental Table S1. The cohort construction process is illustrated in Supplemental Fig. S1.

Study outcomes

Using all data sources and pairwise exposure cohorts, we evaluated the relative risks of two primary cardiovascular outcomes. The two primary outcomes of interest were: (1) 3-point MACE, including acute myocardial infarction, ischemic stroke, and sudden cardiac death; and (2) all CVAEs, which additionally included heart failure, peripheral artery disease, and non-coronary revascularization, as detailed in the concept sets in Supplemental Table S2. To compare outcome risks between the target and comparator cohorts, we defined two time-at-risk periods from the index date to 1 year and 3 years.

Statistical analysis

We used a systematic federated analytical framework incorporating large-scale propensity score (PS) adjustment, bias diagnostics, and meta-analysis from the Observational Health Data Sciences and Informatics (OHDSI) to address residual confounding, publication bias, and p-hacking. By performing 1:1 large-scale PS matching between the target and comparator groups, we sought to ensure that the characteristics of the two groups were similar and to reduce selection bias. The PS was estimated using L1-regularized logistic regression tuned by 10-fold cross-validation, and matching was performed with a caliper of 0.2 on the logit scale. Covariates included demographic characteristics (such as age and sex), comorbidities, medication history within 1 year before the index date, and risk scores (Charlson index, diabetes complications severity index [DCSI], CHADS2, and CHA2DS2-VASc score).

To assess study reliability, we evaluated study diagnostics and quantified empirical equipoise, defined as sufficient overlap between preference score distributions, where the preference score is a transformation of the PS that accounts for differences in the prevalence of each treatment between the target and comparator cohorts [14]. We considered empirical equipoise to be achieved if 25% of patients in both groups had preference scores between 0.3 and 0.7. In addition, absolute standardized mean differences (aSMDs) were calculated for each covariate between treatment groups, with values <0.25 considered well balanced [15]. Because adequate covariate balance is required for valid pooling of effect estimates, cohorts that failed to meet this criterion after matching were excluded from the meta-analysis and analyzed separately. This study included two outcomes, three databases, one meta-analysis, two time-at-risk definitions, and one PS adjustment, yielding a total of 12 individual effect estimates. Because each estimate was generated independently, prior events relevant to the outcome were excluded on an analysis-by-analysis basis, leading to minor differences in the study populations.

The incidence rate was calculated by dividing the number of outcome events by the number of person-years at risk and multiplying the result by 1,000. We used Cox proportional hazards models to estimate hazard ratios (HRs) with 95% confidence intervals (CIs) for all outcomes. We then performed a random-effects meta-analysis to pool HR estimates across databases. All tests were two-sided, and P<0.05 was considered statistically significant. Distributed network analyses were conducted by executing the analysis package locally within the firewall at each database site. The prespecified statistical results, without patient-level information, were then shared for interpretation and database-level meta-analysis. To ensure transparency and reproducibility, the end-to-end executable study package and protocol are available online (https://github.com/dr-you-group/CORDIS), and all results are displayed in a publicly available R Shiny app (https://djboo.shinyapps.io/CORDIS).

RESULTS

Study population

Given the imbalances in baseline characteristics across study centers, outcome-specific changes in sample size, and the results of 1:1 large-scale PS matching, separate analysis sets were constructed for the MACE and CVAE outcomes. Because baseline characteristic imbalances persisted after PS matching among KUMC participants, only YUHS and AMC data were included in the final meta-analysis. For the MACE outcome, among a total of 26,720 patients (MACE set: romosozumab, 2,615; denosumab, 24,105), 4,896 patients remained in the final meta-analysis set after 1:1 PS matching (2,448 in each group) (Fig. 1). For the CVAE outcome, among a total of 25,901 patients (CVAE set: romosozumab, 2,530; denosumab, 23,371), 4,758 patients were included in the final analysis set (2,379 in each group) (Supplemental Fig. S2). Institution-specific patient selection flows are presented in Supplemental Figs. S3, S4.

Clinical characteristics

In both PS-matched datasets (the MACE and CVAE sets), no imbalances exceeding the prespecified aSMD threshold were observed in baseline characteristics, including age group, proportion of women, prior medical history (e.g., fractures, diabetes mellitus, hyperlipidemia, hypertension, cardiovascular disease, and neoplasms), and prior medication use (e.g., antidiabetic, antihyperlipidemic, and antihypertensive agents and calcium or vitamin D supplements; all aSMDs <0.25) (Table 1, Supplemental Table S3). Most patients were women in their 60s or 70s. Approximately 30% had a history of fractures, 5% had diabetes, 16% had hyperlipidemia, and 18% had hypertension. Institution-specific PS matching results showed a similar balance pattern, as presented in Supplemental Tables S4, S5. Cohort balance before and after PS matching is presented in Supplemental Fig. S5, which shows that all matching-variable aSMDs were <0.25 in YUHS and AMC, whereas several aSMDs exceeded 0.25 in KUMC. Accordingly, YUHS and AMC were included in the meta-analysis, whereas KUMC results were presented as cohort-specific analyses. The preference score distributions by treatment group within each institution are shown in Supplemental Fig. S6, indicating that all institutions met the empirical equipoise criterion, although overlap was relatively lower in KUMC.

Cardiovascular risk of romosozumab compared with denosumab

After large-scale PS matching, we compared the risks of MACE and all CVAEs between patients with osteoporosis treated with romosozumab and those treated with denosumab at 1- and 3-year follow-up (Figs. 2, 3). In the meta-analysis including data from YUHS and AMC, no significant differences in MACE risk were observed between the romosozumab and denosumab groups at 1 year (adjusted hazard ratio [aHR], 1.42; 95% CI, 0.64 to 3.19; P=0.351) or 3 years (aHR, 1.51; 95% CI, 0.79 to 2.88; P=0.394) (Fig. 2A). Likewise, no significant differences were observed in the risk of all CVAEs at 1 year (aHR, 1.05; 95% CI, 0.62 to 1.78; P=0.752) or at 3-year follow-up (aHR, 1.04; 95% CI, 0.69 to 1.58; P=0.642) between the two groups (Fig. 2B). In separate cohort-specific analyses, Kaplan–Meier curves from YUHS and AMC (Fig. 3) and KUMC (Supplemental Fig. S7) showed no apparent differences between treatment groups for either MACE or all CVAEs.

Subgroup analysis by sex and age

We further examined the risk of cardiovascular events among patients with osteoporosis treated with romosozumab compared with those treated with denosumab across subgroups defined by sex and age (Fig. 4). In the meta-analysis including data from YUHS and AMC, no significant differences in MACE risk were observed between the romosozumab and denosumab groups across sex subgroups (men: aHR, 1.95; 95% CI, 0.19 to 41.91 at 1 year, and aHR, 2.51; 95% CI, 0.58 to 10.86 at 3 years) (women: aHR, 0.90; 95% CI, 0.37 to 2.19 at 1 year, and aHR, 1.08; 95% CI, 0.51 to 2.27 at 3 years) or age subgroups (<65 years: aHR, 1.04; 95% CI, 0.12 to 8.66 at 1 year, and aHR, 0.69; 95% CI, 0.09 to 4.18 at 3 years) (≥65 years: aHR, 1.28; 95% CI, 0.59 to 2.80 at 1 year, and aHR, 1.19; 95% CI, 0.63 to 2.27 at 3 years) at either follow-up period (Fig. 4A). Likewise, no significant differences in all CVAEs risk were observed between the two groups across sex subgroups (men: aHR, 1.45; 95% CI, 0.43 to 4.90 at 1 year, and HR, 1.56; 95% CI, 0.52 to 4.62 at 3 years) (women: aHR, 0.78; 95% CI, 0.43 to 1.41 at 1 year, and aHR, 0.83; 95% CI, 0.52 to 1.32 at 3 years) or age subgroups (<65 years: aHR, 1.33; 95% CI, 0.29 to 6.77 at 1 year, and aHR, 1.39; 95% CI, 0.46 to 4.22 at 3 years) (≥65 years: aHR, 1.13; 95% CI, 0.47 to 2.69 at 1 year, and aHR, 1.09; 95% CI, 0.62 to 1.94 at 3 years) at either follow-up period (Fig. 4B).

DISCUSSION

In this large-scale PS-matched multicenter study using data from three Korean tertiary-level institutions, we compared cardiovascular risk during follow-up between patients with osteoporosis treated with romosozumab and those treated with denosumab. Over follow-up periods of 1 and 3 years, no significant differences were observed in the risk of MACE or overall CVAEs between the two treatment groups.

The cardiovascular safety of osteoporosis medications is clinically important because patients with osteoporosis typically include older, frail individuals with multiple comorbidities, particularly those with prior fractures, who inherently carry an elevated risk of cardiovascular disease [12,16–18]. Given the limited number of currently available osteoporosis medications and the substantial benefits of romosozumab in rapidly increasing bone mass and reducing fracture risk, further clarification of its cardiovascular safety profile is needed, particularly using data derived from real-world clinical practice [19].

Sclerostin, the molecular target of romosozumab, is expressed in areas of vascular calcification and within atherosclerotic plaques, which may represent a potential mechanistic link to cardiovascular outcomes [20]. However, the clinical relevance of this association remains uncertain, as animal studies and congenital sclerostin-deficient populations have not demonstrated clear evidence of increased lifetime cardiovascular risk [21,22]. A recent large-scale Japanese study reported no definitive increase in cardiovascular risk with romosozumab compared with bisphosphonates [23]. Comparative studies of romosozumab versus parathyroid hormone analogues have also reported either no difference in cardiovascular risk [10] or even a lower risk among patients treated with romosozumab [24,25]. Additionally, a recent meta-analysis integrating both randomized controlled trial and observational data found no significant overall increase in cardiovascular risk associated with romosozumab treatment [26]. These findings are consistent with our results. Post-marketing pharmacovigilance databases in Japan have suggested increased cardiovascular risks among romosozumab users; however, these findings have methodological limitations, including reliance on self-reported data and the potential inclusion of patients already at high cardiovascular risk due to the absence of cardiovascular-related contraindications or warnings at the time of reporting [27,28].

To achieve clinical equipoise, denosumab was selected as the active comparator when examining the association between romosozumab use and cardiovascular risk in this study. Denosumab is one of the most potent antiresorptive drugs and is widely used in patients at high or very high fracture risk [5,6]. Denosumab has consistently demonstrated a neutral cardiovascular safety profile, with no increased risk of MACE or overall CVAEs compared with placebo [12,29–31]. In this context, denosumab represents a clinically relevant comparator. In addition, by excluding patients with prior cardiovascular disease and applying large-scale PS matching that incorporated most known cardiovascular risk factors, we attempted to minimize baseline imbalances inherent to observational studies. Therefore, our findings suggest that romosozumab use may not be associated with increased cardiovascular risk compared with denosumab use. Nonetheless, further validation in additional studies is warranted.

Concerns regarding the cardiovascular safety of romosozumab primarily originate from the ARCH trial, which compared romosozumab with alendronate and included only postmenopausal women [2]. In contrast, the present study used denosumab as the active comparator and included both men and women, enabling sex-stratified analyses. In our meta-analysis, no significant differences in the risk of MACE or all CVAEs were observed between romosozumab and denosumab across sex or age subgroups. These findings suggest that the cardiovascular risk signal observed in the ARCH trial may have been context-dependent and influenced by the choice of comparator and the characteristics of the study population rather than reflecting a consistent treatment-related effect of romosozumab.

Although no statistically significant differences in cardiovascular risk were observed between romosozumab and denosumab, the absolute number of MACE events in the romosozumab group was relatively small, particularly during the first year of follow-up. Consequently, the study may have been underpowered to detect modest differences in cardiovascular risk between treatment groups. Therefore, the absence of statistical significance should not be interpreted as definitive evidence of the absence of risk. Although the findings are reassuring, further validation in larger studies with longer follow-up durations and a greater number of cardiovascular events will be necessary to more precisely characterize the cardiovascular safety profile of romosozumab.

This study contributes real-world evidence regarding the comparative cardiovascular risk among patients with osteoporosis initiating romosozumab or denosumab, an area in which data remain limited and inconsistent. Conducting the study within a common data model framework using ATLAS tools enabled independent validation through a distributed research network across three large tertiary-level institutions while ensuring transparency and reproducibility. However, this study has several limitations. Because this was an observational retrospective study, causality cannot be inferred. Additionally, the possibility of residual selection bias cannot be completely excluded despite efforts to minimize baseline differences between treatment groups through large-scale PS matching. BMD data were not available because the CDM framework used in this study had limited capacity to handle BMD data across multiple centers. Because the study relied on real-world clinical data, treatment decisions may have been influenced by clinicians avoiding romosozumab use in patients with a high cardiovascular risk profile. Therefore, the findings may be most applicable to patients with lower cardiovascular risk and without very recent cardiovascular events.

In summary, in this multicenter observational study, we did not observe evidence of an increased cardiovascular risk associated with romosozumab use compared with denosumab. Although these findings are reassuring, further validation in studies with larger sample sizes, longer follow-up, and a greater number of cardiovascular events is warranted.

Supplementary Material

Article information

CONFLICTS OF INTEREST

Seng Chan You reports grants from Daiichi Sankyo. Seng Chan You is the chief executive officer of PHI Digital Healthcare. Other authors have no potential conflicts of interest to disclose.

AUTHOR CONTRIBUTIONS

Conception or design: S.H.A., K.J.K., S.C.Y., J.S.O., N.H. Acquisition, analysis, or interpretation of data: S.H.A., D.B., K. J.K., J.C., H.Y., C.Y.J., S.C.Y., J.S.O., N.H. Drafting the work or revising: S.H.A., D.B., K.J.K., S.C.Y., J.S.O., N.H. Final approval of the manuscript: S.H.A., D.B., K.J.K., J.C., H.Y., C.Y.J., S.C.Y., J.S.O., N.H.

Fig. 1

Study flow for the analysis of major adverse cardiovascular events (MACE). YUHS, Yonsei University Health System; AMC, Asan Medical Center; KUMC, Korea University Medical Center.

Fig. 2

Incidence and risk of (A) major adverse cardiovascular event (MACE) and (B) all cardiovascular adverse events (CVAEs) in romosozumab-versus denosumab-treated patients. Forest plots showing hazard ratios with 95% confidence intervals (CIs) for the incidence of (A) MACE and (B) all CVAEs in patients treated with romosozumab versus denosumab, based on 1- and 3-year intention-to-treat analyses at each institution. Incidence rates (per 1,000 person-years [PYs]) and the number of events are presented in the accompanying table. Hazard ratios were estimated using Cox proportional hazards regression models after large-scale propensity score (PS) matching. Meta-analyses were conducted using a random-effects model. The I² statistic and corresponding P values indicate the degree of heterogeneity across institutions; I²=0.0% in all comparisons suggests no observed heterogeneity. When cohort balance before and after PS matching was assessed, all matching-variable absolute standardized mean differences were <0.25 in Yonsei University Health System (YUHS) and Asan Medical Center (AMC), whereas several exceeded 0.25 in Korea University Medical Center. Accordingly, YUHS and AMC were included in the meta-analysis.

Fig. 3

Kaplan–Meier curves for the cumulative incidence of (A) major adverse cardiovascular event (MACE) and (B) all cardiovascular adverse events (CVAEs) by treatment group. Shaded areas in the survival curves indicate 95% confidence intervals. YUHS, Yonsei University Health System; AMC, Asan Medical Center.

Fig. 4

Subgroup analyses of the risk of (A) major adverse cardiovascular event (MACE) and (B) all cardiovascular adverse events (CVAEs) in romosozumab-versus denosumab-treated patients according to sex and age. Forest plots showing hazard ratios with 95% confidence intervals (CIs) for the incidence of MACE and all CVAEs in patients treated with romosozumab versus denosumab, based on 1- and 3-year intention-to-treat analyses. Incidence rates (per 1,000 person-years [PYs]) and the number of events are presented in the accompanying table. Hazard ratios were estimated using Cox proportional hazards regression models after large-scale propensity score (PS) matching. When cohort balance before and after PS matching was assessed, all matching-variable absolute standardized mean differences were <0.25 in Yonsei University Health System (YUHS) and Asan Medical Center (AMC), whereas several exceeded 0.25 in Korea University Medical Center. Accordingly, YUHS and AMC were included in the meta-analysis. YUHS had no events and was therefore not included in the meta-analytic estimates for the MACE outcome in men.

References

• 1. Cosman F, Crittenden DB, Adachi JD, Binkley N, Czerwinski E, Ferrari S, et al. Romosozumab treatment in postmenopausal women with osteoporosis. N Engl J Med 2016;375:1532–43.ArticlePubMedPMC
• 2. Saag KG, Petersen J, Brandi ML, Karaplis AC, Lorentzon M, Thomas T, et al. Romosozumab or alendronate for fracture prevention in women with osteoporosis. N Engl J Med 2017;377:1417–27.ArticlePubMed
• 3. Lewiecki EM, Blicharski T, Goemaere S, Lippuner K, Meisner PD, Miller PD, et al. A phase III randomized placebo-controlled trial to evaluate efficacy and safety of romosozumab in men with osteoporosis. J Clin Endocrinol Metab 2018;103:3183–93.ArticlePubMed
• 4. Langdahl BL, Libanati C, Crittenden DB, Bolognese MA, Brown JP, Daizadeh NS, et al. Romosozumab (sclerostin monoclonal antibody) versus teriparatide in postmenopausal women with osteoporosis transitioning from oral bisphosphonate therapy: a randomised, open-label, phase 3 trial. Lancet 2017;390:1585–94.ArticlePubMed
• 5. Shoback D, Rosen CJ, Black DM, Cheung AM, Murad MH, Eastell R, et al. Pharmacological management of osteoporosis in postmenopausal women: an Endocrine Society guideline update. J Clin Endocrinol Metab 2020;105:dgaa048..ArticlePubMedPMCPDF
• 6. Camacho PM, Petak SM, Binkley N, Diab DL, Eldeiry LS, Farooki A, 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.ArticlePubMedPMC
• 7. Kim KJ, Kim KM, Lee YK, Kim J, Jang H, Kim J, et al. Twenty-year trends in osteoporosis treatment and post-fracture care in South Korea: a nationwide study. J Bone Metab 2025;32:57–66.ArticlePubMedPMCPDF
• 8. Cummings SR, McCulloch C. Explanations for the difference in rates of cardiovascular events in a trial of alendronate and romosozumab. Osteoporos Int 2020;31:1019–21.ArticlePubMedPDF
• 9. Farr JN. Cardiovascular safety and romosozumab: the plot thickens. J Clin Endocrinol Metab; 2024. 110:p. e190–1.ArticlePubMedPMCPDF
• 10. Masuda S, Fukasawa T, Matsuda S, Yoshida S, Kawakami K. Comparative effectiveness and cardiovascular safety of romosozumab versus teriparatide in patients with osteoporosis: a population-based cohort study. Osteoporos Int 2024;35:2165–74.ArticlePubMedPDF
• 11. Tsur A, Cahn A, Levy L, Pollack R. Cardiovascular outcomes of romosozumab treatment: real-world data analysis. JBMR Plus; 2025. 9:p. ziae146..ArticlePubMedPMCPDF
• 12. Rodriguez AJ, Abrahamsen B. Cardiovascular safety of antifracture medications in patients with osteoporosis: a narrative review of evidence from randomized studies. JBMR Plus 2021;5:e10522.PubMedPMC
• 13. Kim KJ, Boo D, Choi J, Yoon H, Jung CY, Ahn SH, et al. Comprehensive evaluation of treatment patterns in post-menopausal patients with osteoporosis without fractures: insights from tertiary care institutions and nationwide OMOPCDM data. Endocrinol Metab (Seoul) 2025;40:737–47.ArticlePubMedPMCPDF
• 14. Walker AM, Patrick AR, Lauer MS, Hornbrook MC, Marin MG, Platt R, et al. A tool for assessing the feasibility of comparative effectiveness research. Comp Eff Res 2013;3:11–20.Article
• 15. Stuart EA, Lee BK, Leacy FP. Prognostic score-based balance measures can be a useful diagnostic for propensity score methods in comparative effectiveness research. J Clin Epidemiol 2013;66(8 Suppl):S84–90.e1.ArticlePubMedPMC
• 16. Tanko LB, Christiansen C, Cox DA, Geiger MJ, McNabb MA, Cummings SR, et al. Relationship between osteoporosis and cardiovascular disease in postmenopausal women. J Bone Miner Res 2005;20:1912–20.ArticlePubMedPDF
• 17. Szulc P, Samelson EJ, Kiel DP, Delmas PD. Increased bone resorption is associated with increased risk of cardiovascular events in men: the MINOS study. J Bone Miner Res 2009;24:2023–31.ArticlePubMedPMCPDF
• 18. Park SY, Lee J, Lee Y, Park JW, Lee YK. Bisphosphonate use and reduced all-cause mortality in Korean women older than 50 years: a nationwide sample cohort study. J Bone Metab 2025;32:323–30.ArticlePubMedPMCPDF
• 19. Ramchand SK, Seeman E. Advances and unmet needs in the therapeutics of bone fragility. Front Endocrinol (Lausanne) 2018;9:505.ArticlePubMedPMC
• 20. Langdahl BL, Hofbauer LC, Forfar JC. Cardiovascular safety and sclerostin inhibition. J Clin Endocrinol Metab 2021;106:1845–53.ArticlePubMedPDF
• 21. Turk JR, Deaton AM, Yin J, Stolina M, Felx M, Boyd G, et al. Nonclinical cardiovascular safety evaluation of romosozumab, an inhibitor of sclerostin for the treatment of osteoporosis in postmenopausal women at high risk of fracture. Regul Toxicol Pharmacol 2020;115:104697.ArticlePubMed
• 22. van Lierop AH, Appelman-Dijkstra NM, Papapoulos SE. Sclerostin deficiency in humans. Bone 2017;96:51–62.ArticlePubMed
• 23. Tominaga R, Ikenoue T, Ishii R, Niihata K, Aita T, Okuda T, et al. Comparative cardiovascular safety of romosozumab versus bisphosphonates in Japanese patients with osteoporosis: a new-user, active comparator design with instrumental variable analyses. J Bone Miner Res 2025;40:906.PubMed
• 24. Stokar J, Szalat A. Cardiovascular safety of romosozumab vs PTH analogues for osteoporosis treatment: a propensity-score-matched cohort study. J Clin Endocrinol Metab; 2025. 110:p. e861–7.ArticlePubMedPDF
• 25. Hartz MC, Johannessen FB, Harslof T, Langdahl BL. The effectiveness and safety of romosozumab and teriparatide in postmenopausal women with osteoporosis. J Clin Endocrinol Metab; 2025. 110:p. e1640–52.ArticlePubMedPDF
• 26. Wong RMY, Wong PY, Liu C, Wong HY, Fong MK, Zhang N, et al. Treatment effects, adverse outcomes and cardiovascular safety of romosozumab - existing worldwide data: a systematic review and meta-analysis. J Orthop Translat 2024;48:107–22.ArticlePubMedPMC
• 27. Kotake K, Mitsuboshi S, Omori Y, Kawakami Y, Kawakami Y. Evaluation of risk of cardiac or cerebrovascular events in romosozumab users focusing on comorbidities: analysis of the Japanese Adverse Drug Event Report Database. J Pharm Technol 2023;39:23–8.ArticlePubMedPMCPDF
• 28. Vestergaard Kvist A, Faruque J, Vallejo-Yague E, Weiler S, Winter EM, Burden AM, et al. Cardiovascular safety profile of romosozumab: a pharmacovigilance analysis of the US Food and Drug Administration Adverse Event Reporting System (FAERS). J Clin Med 2021;10:1660.ArticlePubMedPMC
• 29. Seeto AH, Abrahamsen B, Ebeling PR, Rodriguez AJ. Cardiovascular safety of denosumab across multiple indications: a systematic review and meta-analysis of randomized trials. J Bone Miner Res 2021;36:24–40.ArticlePubMedPDF
• 30. Spangler L, Nielson CM, Brookhart MA, Hernandez RK, Stad RK, Lin TC, et al. Cardiovascular safety in postmenopausal women and men with osteoporosis treated with denosumab and zoledronic acid: a post-authorization safety study. JBMR Plus 2023;7:e10793.ArticlePubMedPMC
• 31. Bliuc D, Tran T, Chen W, Alarkawi D, Alajlouni DA, Blyth F, et al. Antiresorptive medication use is not associated with acute cardiovascular risk: an observational study. J Clin Endocrinol Metab; 2023. 108:p. e110–9.ArticlePubMedPDF

Figure & Data

References

Citations

Citations to this article as recorded by  

• Romosozumab and Cardiovascular Safety: Bridging the Gap between Clinical Trials and Real-World Evidence
  Beom-Jun Kim
  Endocrinology and Metabolism.2026; 41(3): 390.     CrossRef
• Reframing Osteoblast Homeostasis as a Systemic Endocrine–Metabolic Node: Mechanisms, Inter‐Organ Crosstalk, and Translational Strategies
  Yang Li, Zheng Chen, Guangchen Li, Shiwei Ji, Ruofang Yin, Shizhuo Wang
  Comprehensive Physiology.2026;[Epub]     CrossRef