Key PointsQuestion
Is fibroblast growth factor 23 associated with recurrent cardiovascular events in patients stabilized after an acute coronary syndrome?
Findings
Among 4947 patients with recent acute coronary syndrome enrolled in the SOLID-TIMI 52 trial, fibroblast growth factor 23 was independently and significantly associated with an increased risk of cardiovascular death or heart failure hospitalization and its individual components after adjustment for baseline clinical characteristics and established biomarkers (high-sensitivity troponin I, brain-type natriuretic peptide, and high-sensitivity C-reactive protein). Although this association was robust in men, C-terminal fibroblast growth factor 23 did not appear to be useful for risk stratification in women.
Meaning
In patients stabilized after acute coronary syndrome, elevated fibroblast growth factor 23 concentrations may be associated with recurrent major cardiovascular events independent of established clinical risk factors and cardiorenal biomarkers; however, this association may be attenuated in women.
Importance
Elevated fibroblast growth factor 23 (FGF-23) concentrations are associated with myocardial fibrosis and renin-angiotensin system upregulation, potentially providing prognostic information distinct from standard cardiovascular (CV) biomarkers.
Objective
To evaluate the association of FGF-23 with recurrent CV events in patients after an acute coronary syndrome (ACS).
Design, Setting, and Participants
C-terminal FGF-23 was measured in plasma samples using an established enzyme-linked immunosorbent assay system for 4947 patients within 30 days of ACS (median, 14 days) and with 1 additional CV risk factor in the Stabilization of Plaques Using Darapladib-Thrombolysis in Myocardial Infarction 52 (SOLID-TIMI 52) trial of the lipoprotein-associated phospholipase A2 inhibitor darapladib vs placebo performed from December 1, 2009, to April 24, 2014 (median follow-up, 2.5 years). Analyses were adjusted for clinical risk factors, renal function, and established cardiorenal biomarkers. This secondary analysis was performed from September 25, 2014, to October 1, 2017.
Exposure
The FGF-23 concentration at baseline.
Main Outcomes and Measures
The primary end point for this post hoc analysis was the composite of CV death or hospitalization for heart failure.
Results
In this study, baseline FGF-23 concentrations were available for 4947 patients (median age, 64.0 years; interquartile range, 59.0-71.0 years; 1276 [25.8%] female). Patients with higher FGF-23 concentrations were older and more likely female, with a greater proportion of hypertension, diabetes, and previous myocardial infarction. After multivariable adjustment for baseline clinical characteristics and established biomarkers (high-sensitivity troponin I, brain-type natriuretic peptide, and high-sensitivity C-reactive protein), FGF-23 concentration in the top quartile was independently associated with an increased risk of CV death or heart failure hospitalization (adjusted hazard ratio [HR], 2.35; 95% CI, 1.82-3.02; P < .001) and its individual components. Elevated FGF-23 concentration was also associated with an increased risk of all-cause mortality (adjusted HR, 2.27; 95% CI, 1.73-2.97; P < .001) and CV death, myocardial infarction, or stroke (adjusted HR, 1.42; 95% CI, 1.17-1.71; P < .001). When analyses were stratified by patient sex, the association between FGF-23 and CV risk, including CV death or heart failure, appeared to be attenuated in women (adjusted HR, 1.11; 95% CI, 0.70-1.76; P = .67) compared with men (HR, 3.11; 95% CI, 2.29-4.22; P < .001; P < .001 for interaction).
Conclusions and Relevance
In patients stabilized after ACS, elevated FGF-23 concentrations may be associated with recurrent major CV events and all-cause mortality, providing information independent of established clinical risk factors and cardiorenal biomarkers. A potential sex difference in these findings deserves further study.
Fibroblast growth factor 23 (FGF-23) is a protein hormone that was first described for its role in mineral metabolism.1 Produced most abundantly by osteocytes, FGF-23 prevents reabsorption of phosphate in the renal proximal convoluted tubule and inhibits conversion of 25-hydroxyvitamin D to its active form.1,2 These effects limit hyperphosphatemia in renal failure, a disease associated with elevated concentrations of FGF-23.1,2 Prior work has suggested a possible causal link between FGF-23 and adverse cardiovascular (CV) remodeling, in particular to the development of left ventricular hypertrophy,3-5 alterations in myocyte calcium handling,6 renin-angiotensin system upregulation,7 and promotion of vascular calcification.8
Given these pleotropic effects on CV structure and function, FGF-23 might be expected to provide prognostic value independent of established biomarkers. Previous observations have found elevated FGF-23 concentrations to be associated with adverse CV outcomes in patients without established CV disease9-15 and in patients with stable coronary artery disease and preserved left ventricular ejection fraction.16
The prognostic utility of FGF-23 in patients stabilized after a recent acute coronary syndrome (ACS) remains unknown, however. We therefore examined the association of FGF-23 with recurrent CV events in patients after ACS in the Stabilization of Plaques Using Darapladib-Thrombolysis in Myocardial Infarction 52 (SOLID-TIMI 52) trial.17
Study Design and Participants
The study design, baseline patient characteristics, and primary results of the SOLID-TIMI 52 trial (NCT01000727)18 have been previously described.17,19 In brief, SOLID-TIMI 52 was a multinational, double-blind, placebo-controlled trial performed from December 1, 2009, to April 24, 2014, that randomized 13 026 patients within 30 days of ACS (median, 14 days) to the lipoprotein-associated phospholipase A2 inhibitor darapladib or placebo, with a median follow-up of 2.5 years. In addition to the recent ACS, patients were required to have 1 additional CV risk factor for inclusion in the trial. Major exclusion criteria relevant to the current analysis included coronary artery bypass graft surgery completed or planned for the index ACS, liver disease, estimated glomerular filtration rate (eGFR) less than 30 mL/min/1.73 m2 by the Modification of Diet in Renal Dysfunction formula, New York Heart Association class III or IV heart failure, poorly controlled hypertension, and life expectancy less than 2 years because of a non-CV condition. Written informed consent was provided by all participants, and the study protocol was approved by all relevant institutional review boards (see the full list of all participating institutions in SOLID-TIMI 52 in eAppendix 1 in Supplement 1 in the report by O’Donoghue et al17). The SOLID-TIMI 52 trial and this post hoc secondary analysis (performed from September 25, 2014, to October 1, 2017) comply with the Declaration of Helsinki.20 All data were deidentified.
On the basis of prior work16 showing the prognostic value of an elevated FGF-23 concentration for CV death or heart failure hospitalization but not atherothrombotic events in patients with stable ischemic heart disease, the primary prespecified outcome of interest for the current analysis was the composite of CV death or heart failure hospitalization. Other end points captured in the SOLID-TIMI 52 trial were also examined, including the primary trial end point of major coronary events (coronary heart disease death, myocardial infarction [MI], or urgent revascularization for myocardial ischemia) and the key secondary end point of major adverse CV events (MACE), defined as CV death, MI, or stroke. All deaths, heart failure hospitalizations, cardiac ischemic events, and cerebrovascular events were adjudicated by an independent and blinded clinical events committee according to prespecified criteria.17,19 Atrial fibrillation was captured as a term in the safety database but was not adjudicated.
Measurement of Biomarkers
The FGF-23 concentrations were analyzed at baseline in 4947 patients randomly selected from the overall trial cohort. Patients included in the biomarker subgroup had baseline characteristics similar to the overall trial population (eTable 1 in the Supplement). The FGF-23 concentration was measured in duplicate using a C-terminal human enzyme-linked immunosorbent assay (Immutopics)21 with a median coefficient of variation of 3.86% (eTable 2 in the Supplement). Mean (SD) normal values for this assay in adults with preserved renal function were 55 (50) reference units (RU)/mL.21 All assays for FGF-23 and other biomarkers, including high-sensitivity troponin I (hsTnI) (Architect i2000SR, Abbott Laboratories), high-sensitivity C-reactive protein (hsCRP) (cobas 6000, Roche Diagnostics), brain-type natriuretic peptide (BNP) (Architect i2000SR, Abbott Laboratories), and cystatin-C (Randox), were performed in the TIMI Clinical Trials Laboratory (Boston, Massachusetts) by laboratory personnel masked to treatment allocation and clinical outcome.
All patients with FGF-23 concentrations were included in this exploratory analysis and categorized by quartile of baseline FGF-23 concentration. Baseline characteristics by FGF-23 quartile were described with median and interquartile range (IQR) for continuous variables and number and percentage for categorical variables. Comparisons of baseline characteristics across FGF-23 quartiles were made with analysis of variance (linear trend) for continuous variables and the Cochran-Armitage trend test for categorical variables.
Cumulative rates of all outcomes and their components were calculated for each quartile using the Kaplan-Meier method reported at 3 years and tested for significance with the trend test. The association between FGF-23 concentration and trial end points was examined using Cox proportional hazards regression models to produce hazard ratios (HRs) and 95% CIs for elevation in FGF-23 concentrations, which were analyzed as categorical and continuous variables (as restricted cubic splines). Clinical risk predictors were included in the risk adjustment model (age, sex, body mass index [BMI], current smoker, race/ethnicity [white vs nonwhite], geographic region, hypertension, hyperlipidemia, diabetes, MI before qualifying event, index diagnosis, days from qualifying event, baseline low-density lipoprotein cholesterol concentration, and randomized treatment arm), in addition to eGFR (Modification of Diet in Renal Dysfunction formula), cystatin-C, hsTnI, hsCRP, and BNP measurements. On the basis of inspection of the distribution of continuous variables, the following variables were log-transformed: BMI, low-density lipoprotein cholesterol, cystatin-C, hsTnI, hsCRP, BNP, and FGF-23. Additional restricted cubic splines were used for age and biomarker data (hsTnI, hsCRP, and BNP) in which nonlinear associations were still significant after transformation. Cystatin-C and eGFR were highly correlated (Spearman correlation coefficient r = −0.68) with each other; thus, only eGFR was retained in the final model.
We also explored the FGF-23 cut point for CV death or heart failure by the Youden index in the context of receiver operating curve analysis because descriptive data showed evidence of a threshold effect.22 We observed that an FGF-23 concentration of 92.8 RU/mL was an optimal cutoff value that maximized discrimination between those who had a clinical event (CV death or heart failure hospitalization) and those who did not, and this value was close to the third quartile of FGF-23 distribution (93.5 RU/mL). On the basis of an apparent threshold in risk between the third and fourth quartiles, categorical FGF-23 data were further analyzed as a binary variable (quartile 4 vs quartiles 1-3). There was no interaction by treatment arm, and all analyses were therefore performed in the full biomarker subgroup (n = 4947).
Metrics of discrimination (C-statistic, categoryless net reclassification improvement, and integrated discrimination improvement) were calculated for the adjusted multivariable models with and without FGF-23. To evaluate the incremental value of FGF-23 for prediction over an adjusted multivariable model based on clinical risk factors alone, P values were also calculated based on the likelihood ratio test.23 A test for FGF-23 concentration interaction by sex was performed by including an interaction term in the adjusted Cox proportional hazards regression model.
All analyses were performed by the TIMI Study Group with a commercial statistical software package (SAS, version 9.4; SAS Institute Inc). Assessment of all trial outcomes was performed on an intention-to-treat basis. A 2-sided, unpaired P<.05 was considered to be significant for all tests.
Baseline FGF-23 concentrations were available for 4947 patients (median age, 64.0 years; IQR, 59.0-71.0 years; 1276 [25.8%] female). The median FGF-23 concentration was 63.1 RU/mL (IQR, 46.0-93.5 RU/mL). Patients with higher FGF-23 concentrations tended to be older (67 years [IQR, 60-74 years] vs 63 years [IQR, 58-69 years]), more likely female (499 [40.4%] vs 183 [14.8%]), more likely to present with a non–ST-elevation MI (636 [51.5%] vs 467 [37.8%]), and more likely to have a history of hypertension (987 [79.9%] vs 853 [69.0%]), diabetes (463 [37.5%] vs 398 [32.2%]), and prior MI (478 [38.7%] vs 346 [28.0%]) (Table 1). The FGF-23 concentrations were moderately correlated with eGFR (r = −0.29) and cystatin-C (r = 0.39) and weakly with hsTnI (r = 0.05), hsCRP (r = 0.14), and BNP (r = 0.20) (eTable 3 in the Supplement).
Unadjusted Risk of CV Outcomes
During a median follow-up period of 2.5 years, 205 CV deaths and 183 hospitalizations for heart failure occurred. There were 648 MACE, including 404 fatal or nonfatal MIs and 118 fatal or nonfatal strokes.
When FGF-23 was modeled as a continuous variable, there was a significant 75% higher risk of CV death or heart failure hospitalization for each SD increase in log-transformed FGF-23 (HR, 1.75; 95% CI, 1.61-1.91; P < .001). A similar higher risk was observed for each of the individual components of the composite end point and MACE.
When the FGF-23 concentrations were categorized by quartile, patients with higher baseline FGF-23 concentrations had a significant increase in the 3-year rate of CV death or heart failure hospitalization with evidence of an apparent threshold effect between quartiles 3 and 4 (4.4% in quartile 1, 4.7% in quartile 2, 4.9% in quartile 3, and 17.5% in quartile 4; P < .001) (eTable 4 and eFigures 1 and 2 in the Supplement). Similar observations were seen for the rates of all-cause mortality, with a marked inflection in risk for patients with FGF-23 concentrations between the third and fourth quartiles (5.0% in quartile 1, 5.2% in quartile 2, 4.3% in quartile 3, and 14.6% in quartile 4; P < .001) (eTable 4 in the Supplement). As such, subsequent analyses modeled FGF-23 as a dichotomous variable by applying a threshold of quartile 4 vs quartiles 1 through 3. The observed risk of CV death or heart failure and MACE for patients with FGF-23 concentrations in the top quartile emerged early and persisted over time (Figure 1 and eFigure 3 in the Supplement).
Adjusted Risk of CV Outcomes
After multivariable adjustment for the clinical characteristics and the renal and CV biomarkers previously described, patients with FGF-23 concentrations in quartile 4 had more than a 2-fold higher risk of CV death or heart failure hospitalization than patients with FGF-23 concentrations in quartiles 1 through 3 (adjusted HR, 2.35; 95% CI, 1.82-3.02; P < .001), including increased risks of CV death and heart failure hospitalization individually (Figure 2). The FGF-23 concentrations in quartile 4 were also associated with an increased risk of all-cause mortality, CV death, MI or stroke, fatal or nonfatal stroke, and atrial fibrillation (Figure 2). The association with MI was directionally consistent but of lesser magnitude.
Inclusion of baseline left ventricular ejection fraction (available in 4163 patients) in the multivariable model yielded similar findings (CV death or heart failure: adjusted HR, 2.39; 95% CI, 1.81-3.15; P < .001 for quartile 4 vs quartiles 1-3).
Furthermore, FGF-23 provided an incremental value for risk stratification when analyses were stratified by baseline eGFR (≤60 or >60 mL/min/1.73 m2) (eFigure 4 in the Supplement). In the described risk model including clinical variables, established biomarkers, eGFR, and FGF-23, the adjusted HR for CV death or heart failure hospitalization was 1.51 (95% CI, 1.18-1.93) for low eGFR (≤60 mL/min/1.73 m2) and 2.44 (95% CI, 1.92-3.09) for elevated FGF-23 concentration (quartile 4). When dividing patients into 4 groups with high eGFR (>60 mL/min/1.73 m2) and low FGF-23 concentration (quartiles 1-3) as the reference, the adjusted HR of CV death or heart failure hospitalization was 1.56 (95% CI, 1.07-2.27) for patients with low eGFR and low FGF-23 concentration, 2.61 (95% CI, 1.95-3.50) for patients with high eGFR and high FGF-23 concentration, and 2.88 (95% CI, 2.10-3.95) for patients with the highest-risk combination of low eGFR and high FGF-23 concentration (Figure 3A and Table 2). A similar association was seen with triple stratification by FGF-23, hsTnI, and BNP, with an adjusted HR of 15.40 (95% CI, 8.87-26.73) for patients with elevated concentrations of each biomarker compared with patients with low FGF-23, hsTnI, and BNP concentrations (Figure 3B and Table 2).
Discrimination and Reclassification
Compared with the base model of clinical variables, inclusion of FGF-23 (restricted cubic spline) significantly improved the model for risk of CV death or heart failure hospitalization (likelihood ratio test P < .001), with an improvement in the C statistic from 0.72 (95% CI, 0.69-0.75) to 0.76 (95% CI, 0.73-0.79). The FGF-23 (restricted cubic spline) was also associated with better net reclassification improvement (0.53; 95% CI, 0.42-0.64), with positive event net reclassification improvement (0.19; 95% CI, 0.09-0.30) and nonevent net reclassification improvement (0.34; 95% CI, 0.31-0.37) and the integrated discrimination improvement indicator (0.05; 95% CI, 0.03-0.06). Additional discrimination and reclassification values are provided in eTable 5 in the Supplement.
Women had significantly higher baseline median FGF-23 concentrations than men (77.7 RU/mL [IQR, 55.5-125.5 RU/mL] vs 59.1 RU/mL [IQR, 44.0-84.7 RU/mL]; P < .001). After multivariable adjustment, the HR of CV death or heart failure hospitalization for each 1-SD increase in log-transformed FGF-23 concentration was 1.01 (95% CI, 0.82-1.25; P = .93) in women compared with 1.58 (95% CI, 1.38-1.80; P < .001) in men (P<.001 for interaction). Similar findings were observed for results stratified by sex when modeled by quartile of FGF-23 concentration, with an attenuated association in women (quartile 4 vs quartiles 1-3: HR, 1.11; 95% CI, 0.70-1.76; P = .67) compared with men (HR, 3.11; 95% CI, 2.29-4.22; P < .001; P<.001 for interaction). The association appeared similarly attenuated in women for additional CV outcomes of interest (eTable 6 in the Supplement).
In a large, contemporary cohort of patients with recent ACS, we found that an elevated FGF-23 concentration was associated with the risk of adverse CV outcomes, including CV death or heart failure hospitalization. Furthermore, FGF-23, a protein reflective of physiologic axes separate from those affecting traditional CV biomarkers, provided prognostic information that was incremental to established clinical predictors and multiple biomarkers, including eGFR, BNP, hsCRP, and hsTnI. To our knowledge, this is the first time this association has been explored in patients stabilized after recent ACS. An apparent attenuation of this association in women additionally merits further investigation.
The median FGF-23 concentration of 63.1 RU/mL (IQR, 46.0-93.5 RU/mL) observed here is similar to that seen in patients with stable ischemic heart disease (50.6 RU/mL; IQR, 38.7-69.9 RU/mL)16 but higher than concentrations in the general population9,24 and lower than those in patients with chronic kidney disease.3,4
The observed correlation between FGF-23 concentration and subsequent CV risk after ACS is intriguing with respect to the protein’s postulated direct and indirect CV effects. In non-ACS populations, an elevated FGF-23 concentration has been associated with mortality and heart failure but with mixed associations with atherothrombotic outcomes.9-11,16,25-27 In the current analysis, we observed an association between FGF-23 and the risk of CV death, heart failure hospitalization, and all-cause mortality among patients after ACS. However, we also found an independent association between FGF-23 concentration and the risk of MACE with directional consistency across all components.
Pathophysiologic and Clinical Considerations
The renin-angiotensin system is a central mediator of the CV effects of FGF-23. Klotho, the FGF-23 coreceptor in the kidney and vasculature, is a renin-angiotensin system inhibitor,28-30 and FGF-23 suppresses renal expression of angiotensin-converting enzyme 2,7 a negative renin-angiotensin system regulator.31 Elevation of FGF-23 concentration is associated with left ventricular hypertrophy independent of blood pressure and renal function,3,4,32 whereas FGF receptor blockade appears to reverse murine left ventricular hypertrophy and myocardial fibrosis.33
It has previously been demonstrated that patients with stable coronary artery disease in the Prevention of Events with Angiotensin Converting Enzyme Inhibition (PEACE) trial derived enhanced benefit from the angiotensin-converting enzyme inhibitor trandolapril if they had an elevated baseline concentration of FGF-23.16 In the Evaluation of Cinacalcet Hydrochloride Therapy to Lower Cardiovascular Events (EVOLVE) trial, significant reductions in CV death and heart failure were reported among patients undergoing hemodialysis whose FGF-23 concentrations decreased by at least 30% with cinacalcet.34 Together, FGF-23 elevation and Klotho deficiency support renin-angiotensin system activation29,30 and its known deleterious downstream effects, including cardiac remodeling and heart failure.
The association between FGF-23 and ischemic events is less clear. In the arterial system, FGF-23 prevents vascular calcification via Klotho,8,28,35 and FGF-23 knockout mice have developed rapid vascular calcification and died by 12 weeks of age.36,37 However, a study9 in patients did not find consistent associations between FGF-23 and vascular events. Our results in a post-ACS population show an independent association between FGF-23 and the risk of subsequent stroke and a weaker association with MI. The Multi-Ethnic Study of Atherosclerosis (MESA) found an elevated FGF-23 concentration to be associated with incident coronary heart disease events in the general population,32 whereas other nontrial cohorts of lower-risk patients did not find statistically significant associations between increased stroke and/or MI.9 In a cohort of patients enrolled 6 to 12 months after ACS, elevated FGF-23 concentration was associated with higher incidence of a composite outcome, including ischemic events, heart failure, and all-cause mortality.38
The statistically significant interaction with patient sex was unexpected. Higher concentrations of FGF-23 in women compared with men have been reported in observational cohorts irrespective of FGF-23 assay used,11,16 although differential prognostic value of this biomarker in women has not previously been shown. Estrogen is known to influence concentrations of FGF-23–modifying compounds, including parathyroid hormone, phosphate, and vitamin D,24 and may contribute to an observed sex difference in the association between serum phosphorus concentration and CV outcomes.39
Because the current study enrolled patients after ACS, most women were likely to have been postmenopausal. Menopause results in urine phosphorus retention, which in turn is believed to increase FGF-23 concentration.40 In fact, postmenopausal women not receiving exogenous estrogen have higher FGF-23 concentrations than postmenopausal women receiving estrogen therapy or men.40 Conceivably, low FGF-23 concentrations in this context may reflect intrinsically low FGF-23 concentrations or low FGF-23 concentrations related to exogenous estrogen.
One cannot exclude that the observed difference by patient sex could have occurred by chance. Of note, the prognostic utility of FGF-23 previously appeared to be broadly comparable in men and women in the Atherosclerosis Risk in Communities (ARIC) study and MESA.11,32 However, the current study measured the C-terminal fragment of FGF-23 as opposed to the intact peptide that was previously measured in the older studies. As such, it remains unclear whether the observed findings may reflect differences between men and women in the posttranslational processing of FGF-23. Further study will be useful in that regard.
Although the present analysis benefits from a well-characterized trial cohort, a large sample size and number of events, and adjudicated outcomes, it has several limitations. First, the analysis is observational and does not allow for causal inference. Second, left ventricular ejection fraction was only assessed in 84% of patients; however, the findings remained consistent after inclusion of left ventricular ejection fraction in the adjustment model. Finally, any clinical cut points for FGF-23 will require prospective validation in a separate cohort in addition to validation of the observed interaction by patient sex.
In a large, contemporary cohort of patients with recent ACS, we found elevated FGF-23 concentrations to be associated with a higher risk of adverse CV outcomes independent of established risk predictors and cardiorenal markers. A potential sex difference in these findings deserves further study.
Accepted for Publication: February 21, 2018.
Corresponding Author: Michelle L. O’Donoghue, MD, MPH, Thrombolysis in Myocardial Infarction (TIMI) Study Group, Cardiovascular Division, Brigham and Women’s Hospital, Harvard Medical School, 60 Fenwood Rd, Ste 7022, Boston, MA 02115 (modonoghue@partners.org).
Published Online: April 18, 2018. doi:10.1001/jamacardio.2018.0653
Author Contributions: Drs Bergmark and Udell contributed equally to this work. Dr O'Donoghue had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Bergmark, Udell, Morrow, Cannon, Steen, Braunwald, Sabatine, O'Donoghue.
Acquisition, analysis, or interpretation of data: Bergmark, Udell, Morrow, Cannon, Steen, Jarolim, Budaj, Hamm, Guo, Im, Kuder, Sabatine, O'Donoghue.
Drafting of the manuscript: Bergmark, O'Donoghue.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Guo, Im, Kuder.
Obtained funding: Morrow, Cannon, Braunwald.
Administrative, technical, or material support: Bergmark, Cannon.
Study supervision: Cannon, Hamm, Braunwald, Sabatine, O'Donoghue.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Bergmark reported consulting for Janssen Pharmaceuticals and Daiichi Sankyo. Dr Udell reported consulting for Amgen, Boehringer Ingelheim, Janssen, Merck & Co, Novartis, and Sanofi Pasteur; receiving honoraria from Novartis (registry steering committee), Sanofi Pasteur, and Servier; and receiving grant support from Novartis. Dr Udell reported receiving funding from a Heart and Stroke Foundation of Canada National New Investigator/Ontario Clinician Scientist Award, Ontario Ministry of Research and Innovation Early Researcher Award, the Women’s College Research Institute, and the Department of Medicine, Women’s College Hospital. Dr Morrow reported receiving grants to the TIMI Study Group from Abbott Laboratories, Amgen, AstraZeneca, Daiichi Sankyo, Eisai, GlaxoSmithKline, Merck, Novartis, and Roche Diagnostics and consultant fees from Abbott Laboratories, AstraZeneca, diaDexus, GlaxoSmithKline, Merck & Co, Peloton, Roche Diagnostics, and Verseon. Dr Cannon reported receiving research grants from Amgen, Arisaph, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi Sankyo, Janssen, Merck, and Takeda and consulting fees from Alnylam, Amgen, Arisaph, AstraZeneca, Boehringer-Ingelheim, Bristol-Myers Squibb, GlaxoSmithKline, Kowa, Lipimedix, Merck & Co, Pfizer, Regeneron, Sanofi, and Takeda. Dr Jarolim reported receiving research support through his institution from Abbott Laboratories, AstraZeneca, Daiichi Sankyo, GlaxoSmithKline, Janssen Scientific Affairs, Merck & Co, Roche Diagnostics, Takeda Global Research and Development Center, and Waters Technologies Corporation. Dr Budaj reported receiving grants and personal fees from GlaxoSmithKline during the conduct of the study as well as grants and personal fees from AstraZeneca, Sanofi, Bristol-Myers Squibb/Pfizer, Novartis, and GlaxoSmithKline and grants from Boehringer Ingelheim and Eisai outside the submitted work. Dr Hamm reported receiving personal fees from AstraZeneca, Boehringer Ingelheim, Daiichi Sankyo, Novartis, Pfizer, Sanofi, GlaxoSmithKline, Eli Lilly and Corporation, Berlin Chemie AG, Heart.org, Bayer Healthcare, Gilead, The Medicines Company, Merck Sharp & Dohme, Abbott Vascular, BRAHMS GmbH, Siemens Healthcare, Boston Scientific, Roche Diagnostics, Correvio, Medtronic, Edwards Lifesciences, and Symetis SA. Dr Braunwald reported receiving research grant support from Eli Lilly and Company, AstraZeneca, Novartis, Merck, Daiichi Sankyo, and GlaxoSmithKline; uncompensated consultancies and lectures with Merck & Co and Novartis; and consultancies with The Medicines Company and Theravance. Dr Sabatine reported receiving research grants to his institution from Abbott Laboratories, Amgen, AstraZeneca, Critical Diagnostics, Daiichi Sankyo, Eisai, Gilead, GlaxoSmithKline, Intarcia, Merck & Co, Roche Diagnostics, Sanofi, Takeda, Novartis, Poxel, Janssen Research Development, and MedImmune and has been a consultant for Alnylam, AstraZeneca, CVS Caremark, Merck & Co, and Ionis. Dr O’Donoghue reported receiving research grant support from GlaxoSmithKline, Eisai, Merck & Co, Janssen, Amgen, The Medicines Company, and AstraZeneca. No other disclosures were reported.
Funding/Support: Research reported in this publication was supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under contract HHSN268201000033C. Dr Bergmark was supported by grant T32HL007604 (Training Grant in Cardiovascular Research) from the National Institutes of Health. Stabilization of Plaques Using Darapladib-Thrombolysis in Myocardial Infarction was sponsored by GlaxoSmithKline.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Marc S. Sabatine, MD, MPH, is Deputy Editor of JAMA Cardiology, but he was not involved in any of the decisions regarding review of the manuscript or its acceptance.
2.Ix
JH, Shlipak
MG, Wassel
CL, Whooley
MA. Fibroblast growth factor-23 and early decrements in kidney function: the Heart and Soul Study.
Nephrol Dial Transplant. 2010;25(3):993-997.
PubMedGoogle ScholarCrossref 3.Gutiérrez
OM, Januzzi
JL, Isakova
T,
et al. Fibroblast growth factor 23 and left ventricular hypertrophy in chronic kidney disease.
Circulation. 2009;119(19):2545-2552.
PubMedGoogle ScholarCrossref 4.Negishi
K, Kobayashi
M, Ochiai
I,
et al. Association between fibroblast growth factor 23 and left ventricular hypertrophy in maintenance hemodialysis patients: comparison with B-type natriuretic peptide and cardiac troponin T.
Circ J. 2010;74(12):2734-2740.
PubMedGoogle ScholarCrossref 6.Touchberry
CD, Green
TM, Tchikrizov
V,
et al. FGF23 is a novel regulator of intracellular calcium and cardiac contractility in addition to cardiac hypertrophy.
Am J Physiol Endocrinol Metab. 2013;304(8):E863-E873.
PubMedGoogle ScholarCrossref 7.Dai
B, David
V, Martin
A,
et al. A comparative transcriptome analysis identifying FGF23 regulated genes in the kidney of a mouse CKD model.
PLoS One. 2012;7(9):e44161.
PubMedGoogle ScholarCrossref 8.Lim
K, Lu
T-S, Molostvov
G,
et al. Vascular Klotho deficiency potentiates the development of human artery calcification and mediates resistance to fibroblast growth factor 23.
Circulation. 2012;125(18):2243-2255.
PubMedGoogle ScholarCrossref 9.Jiang
M, Gong
D, Fan
Y. Elevated fibroblast growth factor-23 and risk of cardiovascular disease or mortality in the general population: a meta-analysis.
Int J Cardiol. 2016;222:342-345.
PubMedGoogle ScholarCrossref 10.Ix
JH, Katz
R, Kestenbaum
BR,
et al. Fibroblast growth factor-23 and death, heart failure, and cardiovascular events in community-living individuals: CHS (Cardiovascular Health Study).
J Am Coll Cardiol. 2012;60(3):200-207.
PubMedGoogle ScholarCrossref 11.Lutsey
PL, Alonso
A, Selvin
E,
et al. Fibroblast growth factor-23 and incident coronary heart disease, heart failure, and cardiovascular mortality: the atherosclerosis risk in communities study.
J Am Heart Assoc. 2014;3(3):e000936, 000931-000911.
Google ScholarCrossref 12.Gutiérrez
OM, Mannstadt
M, Isakova
T,
et al. Fibroblast growth factor 23 and mortality among patients undergoing hemodialysis.
N Engl J Med. 2008;359(6):584-592.
PubMedGoogle ScholarCrossref 13.Scialla
JJ, Xie
H, Rahman
M,
et al; Chronic Renal Insufficiency Cohort (CRIC) Study Investigators. Fibroblast growth factor-23 and cardiovascular events in CKD.
J Am Soc Nephrol. 2014;25(2):349-360.
PubMedGoogle ScholarCrossref 14.Tuñón
J, Fernández-Fernández
B, Carda
R,
et al. Circulating fibroblast growth factor-23 plasma levels predict adverse cardiovascular outcomes in patients with diabetes mellitus with coronary artery disease.
Diabetes Metab Res Rev. 2016;32(7):685-693.
PubMedGoogle ScholarCrossref 15.Seiler
S, Cremers
B, Rebling
NM,
et al. The phosphatonin fibroblast growth factor 23 links calcium-phosphate metabolism with left-ventricular dysfunction and atrial fibrillation.
Eur Heart J. 2011;32(21):2688-2696.
PubMedGoogle ScholarCrossref 16.Udell
JA, Morrow
DA, Jarolim
P,
et al. Fibroblast growth factor-23, cardiovascular prognosis, and benefit of angiotensin-converting enzyme inhibition in stable ischemic heart disease.
J Am Coll Cardiol. 2014;63(22):2421-2428.
PubMedGoogle ScholarCrossref 17.O’Donoghue
ML, Braunwald
E, White
HD,
et al; SOLID-TIMI 52 Investigators. Effect of darapladib on major coronary events after an acute coronary syndrome: the SOLID-TIMI 52 randomized clinical trial [published correction appears in
JAMA. 2014;312(14):1473].
JAMA. 2014;312(10):1006-1015.
PubMedGoogle ScholarCrossref 19.O’Donoghue
ML, Braunwald
E, White
HD,
et al. Study design and rationale for the Stabilization of pLaques usIng Darapladib-Thrombolysis in Myocardial Infarction (SOLID-TIMI 52) trial in patients after an acute coronary syndrome.
Am Heart J. 2011;162(4):613-619.e1.
PubMedGoogle ScholarCrossref 21.Jonsson
KB, Zahradnik
R, Larsson
T,
et al. Fibroblast growth factor 23 in oncogenic osteomalacia and X-linked hypophosphatemia.
N Engl J Med. 2003;348(17):1656-1663.
PubMedGoogle ScholarCrossref 23.Pepe
MS, Kerr
KF, Longton
G, Wang
Z. Testing for improvement in prediction model performance.
Stat Med. 2013;32(9):1467-1482.
PubMedGoogle ScholarCrossref 24.Ärnlöv
J, Carlsson
AC, Sundström
J,
et al. Higher fibroblast growth factor-23 increases the risk of all-cause and cardiovascular mortality in the community.
Kidney Int. 2013;83(1):160-166.
PubMedGoogle ScholarCrossref 25.Haring
R, Enserro
D, Xanthakis
V,
et al. Plasma fibroblast growth factor 23: clinical correlates and association with cardiovascular disease and mortality in the Framingham Heart Study.
J Am Heart Assoc. 2016;5(7):e003486, 003481-003488.
Google ScholarCrossref 26.Koller
L, Kleber
ME, Brandenburg
VM,
et al. Fibroblast growth factor 23 is an independent and specific predictor of mortality in patients with heart failure and reduced ejection fraction.
Circ Heart Fail. 2015;8(6):1059-1067.
PubMedGoogle Scholar 27.Wohlfahrt
P, Melenovsky
V, Kotrc
M,
et al. Association of fibroblast growth factor-23 levels and angiotensin-converting enzyme inhibition in chronic systolic heart failure.
JACC Heart Fail. 2015;3(10):829-839.
PubMedGoogle ScholarCrossref 29.Mitani
H, Ishizaka
N, Aizawa
T,
et al. In vivo klotho gene transfer ameliorates angiotensin II–induced renal damage.
Hypertension.2002;39(4):838-843.
Google ScholarCrossref 30.Yu
L, Meng
W, Ding
J, Cheng
M. Klotho inhibits angiotensin II-induced cardiomyocyte hypertrophy through suppression of the AT1R/beta catenin pathway.
Biochem Biophys Res Commun. 2016;473(2):455-461.
PubMedGoogle ScholarCrossref 32.Kestenbaum
B, Sachs
MC, Hoofnagle
AN,
et al. Fibroblast growth factor-23 and cardiovascular disease in the general population: the Multi-Ethnic Study of Atherosclerosis.
Circ Heart Fail. 2014;7(3):409-417.
PubMedGoogle ScholarCrossref 33.Di Marco
GS, Reuter
S, Kentrup
D,
et al. Treatment of established left ventricular hypertrophy with fibroblast growth factor receptor blockade in an animal model of CKD.
Nephrol Dial Transplant. 2014;29(11):2028-2035.
PubMedGoogle ScholarCrossref 34.Moe
SM, Chertow
GM, Parfrey
PS,
et al. Cinacalcet, FGF23 and cardiovascular disease in hemodialysis: the EVOLVE trial.
Circulation. 2015;132:27-39.
PubMedGoogle ScholarCrossref 35.Hu
MC, Shi
M, Zhang
J,
et al. Klotho deficiency causes vascular calcification in chronic kidney disease.
J Am Soc Nephrol. 2011;22(1):124-136.
PubMedGoogle ScholarCrossref 36.Shimada
T, Kakitani
M, Yamazaki
Y,
et al. Targeted ablation of Fgf23 demonstrates an essential physiological role of FGF23 in phosphate and vitamin D metabolism.
J Clin Invest. 2004;113(4):561-568.
PubMedGoogle ScholarCrossref 37.Grabner
A, Faul
C. The role of fibroblast growth factor 23 and Klotho in uremic cardiomyopathy.
Curr Opin Nephrol Hypertens. 2016;25(4):314-324.
PubMedGoogle ScholarCrossref 38.Tuñón
J, Cristóbal
C, Tarín
N,
et al. Coexistence of low vitamin D and high fibroblast growth factor-23 plasma levels predicts an adverse outcome in patients with coronary artery disease.
PLoS One. 2014;9(4):e95402.
PubMedGoogle ScholarCrossref 39.Yoo
KD, Kang
S, Choi
Y,
et al. Sex, age, and the association of serum phosphorus with all-cause mortality in adults with normal kidney function.
Am J Kidney Dis. 2016;67(1):79-88.
PubMedGoogle ScholarCrossref 40.Ix
JH, Chonchol
M, Laughlin
GA, Shlipak
MG, Whooley
MA. Relation of sex and estrogen therapy to serum fibroblast growth factor 23, serum phosphorus, and urine phosphorus: the Heart and Soul Study.
Am J Kidney Dis. 2011;58(5):737-745.
PubMedGoogle ScholarCrossref