Author + information
- Received September 6, 2018
- Revision received February 1, 2019
- Accepted February 4, 2019
- Published online May 6, 2019.
- Rabea Asleh, MD, PhD, MHAa,
- Maurice Enriquez-Sarano, MDa,
- Allan S. Jaffe, MDa,
- Sheila M. Manemann, MPHb,
- Susan A. Weston, MSb,
- Ruoxiang Jiang, BSb and
- Véronique L. Roger, MD, MPHa,b,∗ ()
- aDepartment of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota
- bDepartment of Health Sciences Research, Mayo Clinic, Rochester, Minnesota
- ↵∗Address for correspondence:
Dr. Véronique L. Roger, Department of Cardiovascular Diseases and Department of Health Sciences Research, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905.
Background Galectin-3 (Gal-3) is implicated in cardiac fibrosis, but its association with adverse outcomes after myocardial infarction (MI) is unknown.
Objectives The purpose of this study was to examine the prognostic value of Gal-3 in a community cohort of incident MI.
Methods A population-based incidence MI cohort was prospectively assembled in Olmsted County, Minnesota, between 2002 and 2012. Gal-3 levels were measured at the time of MI. Patients were followed for heart failure (HF) and death.
Results A total of 1,342 patients were enrolled (mean age 67.1 years; 61.3% male; 78.8% non–ST-segment elevation MI). Patients with elevated Gal-3 were older and had more comorbidities. Over a mean follow-up of 5.4 years, 484 patients (36.1%) died and 368 (27.4%) developed HF. After adjustment for age, sex, comorbidities, and troponin, patients with Gal-3 values in tertiles 2 and 3 had a 1.3-fold (95% confidence interval [CI]: 0.9-fold to 1.7-fold) and a 2.4-fold (95% CI: 1.8-fold to 3.2-fold) increased risk of death, respectively (ptrend < 0.001) compared with patients with Gal-3 values in tertile 1. Patients with Gal-3 values in tertiles 2 and 3 had a higher risk of HF with hazard ratios of 1.4 (95% CI: 1.0 to 2.0) and 2.3 (95% CI: 1.6 to 3.2), respectively (ptrend < 0.001). With further adjustment for soluble suppression of tumorigenicity-2, elevated Gal-3 remained associated with increased risk of death and HF. The increased risk of HF did not differ by HF type and was independent of the occurrence of recurrent MI.
Conclusions Gal-3 is an independent predictor of mortality and HF post-MI. These findings suggest a role for measuring Gal-3 levels for risk stratification post-MI.
Over the last 2 decades, population-based studies have demonstrated major changes in the epidemiology of myocardial infarction (MI). Patients with MI now present at older ages and without ST-segment elevation, and they more frequently survive the acute phase (1–3). Even after rapid restoration of coronary blood flow by acute intervention, the risk of adverse outcomes including heart failure (HF) and death remains substantial, although its presentation is also evolving (4). Indeed, when HF occurs, ejection fraction (EF) is more often preserved, which may reflect evolving mechanisms of the left ventricular (LV) remodeling process leading to progressive dysfunction (5). Thus, it is important, given the changing epidemiology of MI, to evaluate risk stratification in the current population of patients with MI and to investigate whether novel biomarkers, and in particular those implicated in cardiac fibrosis and remodeling, add prognostic information over clinical information.
We recently reported on risk stratification approaches among a prospective contemporary cohort of patients with MI in the community. We reported that the performance of scores historically recommended for MI risk stratification was disappointing in contemporary community cohorts (6), underscoring the need to revisit these approaches. We examined the contribution of a novel biomarker of fibrosis, soluble suppression of tumorigenicity 2 (sST2), and found that sST2 was frequently elevated in MI and that higher values were associated with a large excess risk of death and HF independently of troponin (7). These data underscored the importance of studying risk stratification in contemporary cohorts of optimal clinical relevance and the prognostic value of markers of fibrosis, such as sST2. Galectin-3 (Gal-3), a β-galactoside–binding lectin mainly secreted by activated macrophages, is also reflective of fibrosis and cardiac remodeling in response to myocardial injury (8–10). Although biologically plausible, the hypothesis of an association between Gal-3 and post-MI outcomes remains unproven, as small reports in convenience samples of Gal-3 in acute coronary syndrome (ACS) yielded conflicting results (11–13). Further, a comparison of Gal-3 and sST2 using rigorous risk prediction statistical methods is lacking after MI. To address this gap in knowledge, we used robust and complementary methods including calibration, discrimination, and reclassification analyses to investigate the association between Gal-3 and outcomes after MI, and its incremental value over troponin and sST2. We studied a large geographically defined contemporary cohort of prospectively enrolled incident (first ever) MI to avoid contamination of the results by pre-existing myocardial injury and scarring.
Study design and setting
This prospective study was conducted in Olmsted County, Minnesota (2017 population of approximately 154,930), using the resources of the Rochester Epidemiology Project (14,15), which is a medical records linkage system allowing virtually complete capture of health care use and outcomes of nearly all persons living in the county. Complete recording of clinical events in Olmsted County is possible because of its relative isolation from other urban centers, thus enabling the delivery of most health care to local residents by few health care facilities, including Mayo Clinic, Olmsted Medical Center, and their affiliated hospitals. All medical diagnoses are maintained through an electronic index, and patients can be identified through their inpatient and outpatient contacts across the local medical providers (15). The demographic and ethnic characteristics of Olmsted County residents are representative of the state of Minnesota and the Midwest region of the United States, with similar broad disease trends and age-specific and sex-specific mortality rates to that observed in national data (15). This study was approved by the Mayo Clinic and Olmsted Medical Center Institutional Review Boards.
Identification and validation of the MI cohort
Olmsted County residents admitted to Mayo Clinic hospitals from November 1, 2002, through December 31, 2012, with a cardiac troponin T level of 0.03 ng/ml or higher were identified within 12 h of the blood draw. Written consent was obtained from all participants or next of kin, and samples were stored for future use. MI cases were validated using standard epidemiological algorithms integrating cardiac pain, electrocardiogram (ECG) changes, and elevated troponin T as previously described (1,16) and following current guidelines (17); only incident (first-ever) MI cases were included in this study. A significant change (increase or decrease) in troponin levels was considered when the difference between successive troponin measurements was 0.03 ng/ml or higher to ensure that this is greater than the level of imprecision of the assay at all concentrations (17). All cases were reviewed to ensure there were no alternative causes resulting in biomarker elevation.
Cardiac troponin T was measured as part of standard care using a sandwich electrochemiluminescence immunoassay on the Elecsys 2010 (Roche Diagnostics, Indianapolis, Indiana) in the laboratories of the Department of Medicine and Pathology at Mayo Clinic.
sST2 was measured from stored plasma samples using a high-sensitivity sandwich monoclonal immunoassay (Presage ST2 assay, Critical Diagnostics, San Diego, California). All samples were collected via peripheral vein into EDTA-containing tubes, centrifuged immediately, and stored at −70°C prior to analysis at core laboratories. The antibodies used in the Presage assay were generated from a recombinant protein based upon the human cDNA clone for the complete soluble sequence (18). This platform offers improved accuracy in quantifying sST2 levels, particularly at lower concentrations. This specific assay has high sensitivity; the reliability of running the Presage sST2 assay on EDTA plasma samples stored at −70°C (as per biomarker core laboratory) has been established in numerous studies (19–21). Calibration and standardization of this assay were performed according to the manufacturer’s protocol. Previous reports document the intra-assay and interassay coefficients of variation as <2.5% and <4.0%, respectively (18).
Gal-3 concentrations were measured from the stored plasma samples by enzyme-linked immunosorbent assay (BG Medicine, Waltham, Massachusetts) on an automated plate reader with high sensitivity and reliability measures. According to the manufacturer`s protocol, the measurement values of Gal-3 using this assay can range between 1.4 and 94.8 ng/ml, and the overall intra-assay and interassay coefficients of variation have been reported as approximately 3.4% and 8.5%, respectively.
Cardiovascular disease risk factors, comorbid conditions, MI severity indicators, and use of angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and beta-blockers at hospital dismissal were collected from the medical records by trained nurse abstractors. Body mass index (kg/m2) was calculated using the current weight and earliest adult height. Cigarette use was classified as current versus no smoker. Clinical definitions were used for hypertension, diabetes mellitus, hyperlipidemia, and prior HF. Comorbidity was assessed by the Charlson comorbidity index (22), which consists of 17 conditions weighted according to the degree to which they predict death. The presence of ST-segment elevation MI was determined using the Minnesota Code Modular ECG Analysis System (23). Ejection fraction was recorded based on the closest EF measurement within 30 days after the MI diagnosis date as previously described (24).
All participants were followed for HF and death through their complete inpatient and outpatient community medical records from the index MI to date of last follow-up or December 31, 2014, whichever occurred first. HF was identified using International Classification of Diseases-9th Revision (ICD-9) code 428 and validated using the Framingham criteria (25) following a previously reported approach, with minimal missing data and excellent interobserver agreement (26). HF type was determined by echocardiographic measurements of left ventricular EF using the closest measurement to HF diagnosis (with a predefined maximum period of 90 days). HF was classified as heart failure with reduced ejection fraction (HFrEF) when EF was <50% and as heart failure with preserved ejection fraction (HFpEF) when EF was ≥50%. Death (occurrence and date) was ascertained from the medical records and death certificates obtained from the county and state by the Rochester Epidemiology Project. Deaths were classified as cardiovascular (ICD-9 codes 390-459, ICD-10 codes I00-I99) based on American Heart Association classifications (27).
Continuous variables were expressed as mean ± SD or median (25th, 75th percentile), as appropriate; categorical variables were expressed as percentages. Baseline clinical characteristics were compared across Gal-3 tertiles using linear regression and Mantel-Haenszel chi-square tests, as appropriate.
Differences in all-cause death after MI were assessed using the Kaplan-Meier method according to Gal-3 tertiles and compared with the log-rank test. The cumulative incidence of HF following MI was estimated according to Gal-3 tertiles, treating death as a competing risk. Cox proportional hazards regression models were constructed to examine the association between Gal-3 and outcomes (death, cardiovascular death, HF, HFpEF, and HFrEF) after MI. Splines indicated a linear relationship between Gal-3 and all outcomes. Results are reported for each 10-U increase in Gal-3 and also according to Gal-3 tertiles using the lowest tertile as the referent to aid in interpretation. Models were adjusted for age and sex with further adjustment for the Charlson comorbidity index and maximum troponin T measurement. HF prior to the MI was also adjusted for in the models predicting HF, HFpEF, and HFrEF. Analyses were repeated for patients who survived at least 30 days after the incident MI. The proportional hazards assumption was tested using the scaled Schoenfeld residuals and found to be valid.
Several measures were used to assess the incremental prognostic value of Gal-3, modeled as a log-transformed continuous variable, with and without the inclusion of sST2, also log-transformed, in the reference models, which included age, sex, Charlson comorbidity index, and maximum troponin T. Calibration was assessed using the likelihood ratio test, Akaike information criterion (AIC), and Bayesian information criterion (BIC). Additionally, a group-based measure of calibration was assessed with a novel method using a model-based framework that provides a natural extension to survival data (28). Discrimination was assessed using the C-statistic. Reclassification was assessed within the Cox model framework using the continuous net reclassification improvement (NRI) (29,30), in addition to its 2 components, event and nonevent NRI, with a 5-year cutoff to determine event status. Data analyses were performed using SAS software, version 9.4 (SAS institute Inc, Cary, North Carolina) and R version 3.4 (R Foundation for Statistical Computing, Vienna, Austria).
Between November 1, 2002, and December 31, 2012, 1,342 patients with incident MI were enrolled in the study and had stored plasma samples available for Gal-3 and sST2 measurement. The mean age of the cohort was 67.1 ± 14.9 years, 61.3% were male, and 78.8% presented with non–ST-segment elevation MI. The median (25th, 75th percentile) time from MI event to blood sample collection for Gal-3 and sST2 was 2 days (1, 3 days). The median (25th, 75th percentile) Gal-3 level was 18.1 ng/ml (13.3, 25.5 ng/ml); 457 (34%) had elevated Gal-3 compared with the published reference value of 22.1 ng/ml (31). The median (25th, 75th percentile) sST2 level was 47.0 ng/ml (32.1, 99.3 ng/ml); 676 (50%) had elevated sST2, compared with published normal values from the Framingham Heart Study (32).
Patients were divided into 3 groups according to Gal-3 concentrations: tertile 1 (<15.1 ng/ml), tertile 2 (15.1 to 22.4 ng/ml), and tertile 3 (>22.4 ng/ml). The range of Gal-3 values in each tertile were 4.46 to 15.00 ng/ml in tertile 1, 15.02 to 22.40 ng/ml in tertile 2, and 22.404 to 115.0 in tertile 3. Patients with higher Gal-3 levels were older; were more likely to be female; had more diabetes and hypertension; had more comorbidities than patients with lower Gal-3 levels; were more likely to present with Q waves on the ECG, anterior MI, and Killip class >1; were less likely to present with ST-segment elevation (Table 1); had lower EF; and were less likely to be current smokers, have familial coronary disease, and be prescribed beta-blockers at discharge.
Gal-3 and death
During a mean follow-up of 5.4 ± 3.5 years, 484 (36.1%) patients died. Elevated Gal-3 was associated with increased mortality, with 5-year estimates of 10.2%, 24.4%, and 51.9% for patients with values in tertiles 1, 2, and 3, respectively (p < 0.001) (Central Illustration, panel A).
Elevated Gal-3 was strongly associated with death in the unadjusted model, with the association remaining strong after adjustment for key clinical characteristics including age, sex, Charlson index, and maximum troponin T measurement (Table 2). From the adjusted model, the association followed a dose-response pattern with a 30% increased risk of death for each 10-U increase in Gal-3 (hazard ratio [HR]: 1.30; 95% confidence interval [CI]: 1.24 to 1.36) or, when analyzing tertiles of Gal-3, a 26% increased risk of death for patients with Gal-3 values in tertile 2 (HR: 1.26; 95% CI: 0.93 to 1.70) and more than a 2-fold increased risk for patients with Gal-3 values in tertile 3 (HR: 2.35; 95% CI: 1.76 to 3.15) compared with patients with Gal-3 values in tertile 1 (ptrend < 0.001). In sensitivity analyses, we examined the effect of adjusting for a number of additional variables. First, the Charlson comorbidity index was replaced with the individual comorbidities that comprise the Charlson index. This did not materially change the association between Gal-3 level and death (HR: 1.34; 95% CI: 1.27 to 1.41 for each 10-U increase in Gal-3; p < 0.001). Second, further adjustment for discharge medications (including beta-blockers and angiotensin-converting enzyme inhibitors/angiotensin receptor blockers) and EF within 30 days post-MI each resulted in a slight attenuation of the association between Gal-3 level and death (discharge medications: HR: 1.23; 95% CI: 1.16 to 1.30; p < 0.001; and EF: HR: 1.24; 95% CI: 1.16 to 1.32; p < 0.001 for each 10-U increase in Gal-3). Finally, additional adjustment for time from MI symptom onset to Gal-3 measurement, estimated glomerular filtration rate, current smoking status, hypertension, and hyperlipidemia did not change the associations (data not shown). When restricted to individuals who survived at least 30 days after the MI, the associations from the adjusted model were similar (HR: 1.25; 95% CI: 1.18 to 1.32; p < 0.001 for a 10-U increase in Gal-3; or HR: 1.25; 95% CI: 0.92 to 1.70 and HR: 2.20; 95% CI: 1.62 to 2.99 for patients with Gal-3 values in tertiles 2 and 3, respectively; ptrend < 0.001).
Similarly, for cardiovascular death (n = 188), higher Gal-3 tertile was associated with an increased risk of cardiovascular death that followed a dose-response pattern (Table 2) and remained independent of key clinical characteristics known to predict risk after MI.
Gal-3 and HF
During the same follow-up, 368 (27.4%) patients experienced an HF event (53 in Gal-3 tertile 1, 115 in Gal-3 tertile 2, and 200 in Gal-3 tertile 3). Higher Gal-3 tertile was significantly associated with a higher cumulative incidence of HF (Central Illustration, panel B) treating death as a competing risk. The cumulative incidence of HF at 5 years of follow-up was 11.6%, 22.9%, and 43.8% for patients with Gal-3 values in tertiles 1, 2, and 3, respectively (p < 0.001). Elevated Gal-3 levels were associated with a significant increase in HF risk (Table 2). After adjustment for age, sex, Charlson index, maximum troponin T, and HF prior to the MI, there was a 15% increased risk of HF for each 10-U increase in Gal-3 (HR: 1.15; 95% CI: 1.09 to 1.22). When analyzing tertiles of Gal-3, a dose-response pattern was evident with a 40% increased risk of HF for patients with Gal-3 values in tertile 2 (HR: 1.40; 95% CI: 1.00 to 1.96) and more than a 2-fold increased risk for patients with Gal-3 values in tertile 3 (HR: 2.25; 95% CI: 1.61 to 3.15; ptrend < 0.001). Further adjustment for: 1) recurrent MI as a time-dependent covariate; 2) the individual comorbidities that comprise the Charlson index; and 3) time from MI symptom onset to Gal-3 measurements, estimated glomerular filtration rate, current smoking status, hypertension, hyperlipidemia, EF post-MI, and discharge medications did not materially change the associations (data not shown). When restricted to individuals who survived at least 30 days after the MI, the associations from the fully adjusted model were similar (HR: 1.16; 95% CI: 1.09 to 1.24; p < 0.001 for a 10-U increase in Gal-3; HR: 1.43; 95% CI: 1.02 to 2.00 and HR: 2.18; 95% CI: 1.54 to 3.07 for patients with Gal-3 values in tertiles 2 and 3, respectively; ptrend < 0.001). Stratified by HF type, the association between Gal-3 levels and HF risk did not differ appreciably between HFrEF and HFpEF and remained significant after adjustment (Table 2).
Incremental prognostic value of Gal-3
In the final stage of the analyses, log-transformed sST2 was added to the reference model to evaluate the incremental value of Gal-3 over sST2. With sST2 in the model, it is important to note that higher values of Gal-3 remained associated with higher risk of death, although the HRs were attenuated (HR: 1.21; 95% CI: 1.16 to 1.27; p < 0.001 for a 10-U increase in Gal3; HR: 1.09; 95% CI: 0.80 to 1.47 and HR: 1.64; 95% CI: 1.22 to 2.22 for patients with Gal-3 values in tertiles 2 and 3, respectively; ptrend < 0.001). When examining HF as the outcome, the association between higher Gal-3 and increased risk of HF remained when sST2 was added to the reference model (HR: 1.08; 95% CI: 1.02 to 1.15; p = 0.009 for a 10-U increase in Gal3; HR: 1.24; 95% CI: 0.89 to 1.74 and HR: 1.72; 95% CI: 1.22 to 2.43 for patients with Gal-3 values in tertiles 2 and 3, respectively; ptrend = 0.001).
For both outcomes (death and HF), the likelihood ratio test indicated improved model fit when Gal-3 and sST2, both log-transformed, were each added to the reference model and when Gal-3 was added to the model that included sST2 (Table 3). The AIC and BIC were lower in the models that included Gal-3 and sST2 compared with the reference model, and were lower in the model that included Gal-3 and sST2 compared with the model with only sST2 included. The group-based calibration showed that the models that predicted death were well-calibrated when the biomarkers were included. When predicting HF, the group-based calibration indicated that the only model that was well-calibrated was the model that included Gal-3. Regarding discrimination, the individual additions of Gal-3 and sST2 to the reference model slightly increased the C-statistic for the prediction of death and HF, as did the addition of Gal-3 to the model for death with sST2 included. In examining reclassification, of note is that adding Gal-3 to the model with sST2 showed improvement in classification as measured by the continuous NRI, especially for death. The nonevent NRI is the larger component of the continuous NRI, which might suggest that the addition of Gal-3 and sST2 to the model helps decrease the predicted risk for those who do not experience the event more so than to increase the predicted risk for those who do experience the event.
Herein, we report on the value of Gal-3 for post-MI risk stratification in a large, nonselected community cohort. Our finding that, in a nonselected geographically defined cohort, Gal-3 is elevated in 34% of the patients presenting with an incident MI was unexpected. The large increase in the risk of death and HF with increasing levels of Gal-3 was striking, as it was independent of indicators of MI severity, comorbidity, and sST2, which as we have previously shown is also associated with death and HF after MI. The dose-response pattern uncovered by the tertile analyses supports causality. Because these data represent the comprehensive experience of a community, they are of optimal clinical relevance and draw attention to new approaches for contemporary risk stratification after MI in clinical practice.
Importantly, troponin T levels were not associated with Gal-3 levels, and the association between Gal-3 and outcomes was unchanged by the inclusion of troponin T in the models. This is important as troponin is a major prognostic indicator that serves as a marker of MI severity, reflecting the amount of myocardial necrosis. Gal-3 elevation reflects cardiac inflammation and fibrosis and thus different mechanistic pathways than troponin. It was thus logical to hypothesize that it would be independently associated with post-MI outcomes. Our data provide robust evidence that this is the case by demonstrating a strong association between Gal-3 and outcomes that is independent of key clinical factors and troponin, and follows a dose-response pattern that supports causality. For HF, the association with Gal-3 was similar regardless of the type of HF (HFpEF or HFrEF). These data represent the comprehensive experience of a large contemporary MI population, thus supporting a potential use of Gal-3 measurement in clinical practice as a novel biomarker for post-MI risk stratification.
Experimental data suggest that Gal-3 plays a role in cardiac fibrosis, which is important for the occurrence of adverse cardiac remodeling that precedes clinical HF (9,10). Gal-3 is highly expressed by macrophages in animal hypertrophied heart models and induces cardiac fibroblast proliferation and deposition of type I collagen, leading to fibrosis and pathological remodeling (10,33). In human studies, Gal-3 has been shown to be a predictor of EF and infarct size at 4 months after MI (34). With cardiac magnetic resonance, elevated Gal-3 levels were associated with adverse post-MI LV remodeling at 6 months (35), and in patients with ST-segment elevation MI, Gal-3 was an independent predictor of LV remodeling at 1 and 6 months (36).
Although previously suggested, the prognostic role of Gal-3 in ACS was not convincingly established. In a case-control study (37), Gal-3 levels were higher among MI cases, and Gal-3 was associated with 30-day mortality and HF after MI. Lisowska et al. (12) reported that Gal-3 was associated with more severe coronary disease in patients with ischemic heart disease and was an independent predictor of death after MI. In a post hoc analysis of the CLARITY-TIMI 28 trial, an association between increasing Gal-3 levels and cardiovascular death or HF at 30 days has been reported (38), but was attenuated after adjusting for other prognostic factors. These studies all have notable limitations, including small sample size, heterogeneous design and adjustment approaches, and patient selection, which all compromise external validity and hinder inference. As most studies only reported short-term outcomes, the long-term prognostic value of Gal-3 was still undefined. The present study substantively augments previous reports by convincingly demonstrating that Gal-3 is associated with a large excess risk of adverse outcomes after MI, including death and HF, independently of major confounders, in a large prospective community cohort of patients with first MI followed for an extensive period.
Other biomarkers recommended for post-MI risk stratification
Troponin is the biomarker of choice for post-MI risk stratification and, as underscored in the current AHA/ACC guidelines, the use of newer biomarkers should be considered if they provide additional prognostic information over and above that provided by troponin. For example, the AHA/ACC guidelines mention that measuring “especially B-type natriuretic peptide (BNP), may be reasonable to provide additional prognostic information” (Class of Recommendation IIb; Level of Evidence: B). This recommendation is formulated only for non–ST-segment elevation MI, and no additional biomarker is mentioned for ST-segment elevation MI.
We recently reported on the value of sST2 to predict adverse outcomes after acute MI independently of troponin (7). Herein, Gal-3 provided incremental prognostic information over clinical indicators and sST2. Both Gal-3 and sST2 are considered markers of cardiac fibrosis. We recognized that the appropriate “multimarker” approach to predict outcomes after MI remains to be fully defined as also is the case for the evolving field of contemporary post-MI risk stratification (6). In this context, our findings contribute to addressing this question by convincingly demonstrating the incremental prognostic value of Gal-3 over that of clinical prognostic indicators and other biomarkers.
Study limitations and strengths
Some limitations to our study should be acknowledged. As in any observational study, we cannot exclude residual confounding, but this is likely to have minimal impact as we adjusted for numerous relevant covariates known to affect the risk of adverse clinical events. We could not directly compare our data with natriuretic peptides, which were not measured. However, natriuretic peptides reflect different mechanistic pathways, and reports have indicated the independent association of Gal-3 and sST2 with outcomes after MI and in heart failure (39). In a case-control study of patients after ACS, Grandin et al. (40) have shown that patients with elevated Gal-3 and BNP levels were at the highest odds of developing HF, suggesting a potential incremental value of Gal-3 for assessment of HF risk after ACS. Moreover, there was a very weak pattern of correlation between Gal-3 and BNP, which argues against confounding by BNP and further supports that the expression of these 2 proteins is mediated by different biological pathways. Last, our findings will need replication in racial and ethnic groups underrepresented in the cohort.
Our study has several notable strengths. We used the comprehensive data resources of the Rochester Epidemiology Project to examine the impact of Gal-3 on clinical outcomes after MI. We report on a large and contemporary community cohort of patients with a first MI validated by standardized criteria (1), prospectively enrolled with measurement of 2 novel markers of cardiac fibrosis: Gal-3 and sST2. We report for the first time on Gal-3 as an independent prognostic marker after MI in a community cohort of unselected patients, which is of optimal clinical relevance. The focus on incident (first ever) MIs is important as the findings cannot by design reflect prior clinical ischemic damage. Finally, the use of rigorous risk prediction statistical methods substantially strengthens the inference that can be drawn from our results.
In this community cohort of patients with incident MI, elevated Gal-3 levels at the time of the MI were strongly associated with mortality and HF over a long period of follow-up independently of key clinical prognostic indicators and of troponin and sST2. The large excess risk of HF associated with Gal-3 does not differ by HF type, supporting its role in predicting HFrEF and HFpEF similarly. Our findings suggest a role for measuring Gal-3 levels for risk stratification post-MI.
COMPETENCY IN MEDICAL KNOWLEDGE: Elevated plasma levels of Gal-3 at the time of MI are a strong predictor of long-term mortality and HF, independent of clinical prognostic indicators and other biomarkers, including troponin and sST2.
TRANSLATIONAL OUTLOOK: Additional studies should better define the incremental value of Gal-3 measurements compared with other biomarkers for risk stratification among survivors of acute MI.
The authors thank Ellen Koepsell, RN, and Deborah Strain for their study support.
This work was supported by a grant from the National Heart, Lung, and Blood Institute (R01-HL120957) and the Rochester Epidemiology Project (R01-AG034676). The funding sources played no role in the design, conduct, or reporting of this study. Dr. Jaffe has served as a consultant for Beckman, Coulter, Siemens, Abbott, ET Healthcare, Quidel, Roche, Alere, NeurogenomeX, Sphingotec, Single, and Novartis. Dr. Roger has received grants from the National Heart, Lung, and Blood Institute (RO1 HL 120959) and Patient Centered Outcomes Research Institute (PCORI) CDRN-1501-26638-1 during the conduct of the study; and has served as a consultant for Sanofi. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Listen to this manuscript's audio summary by Editor-in-Chief Dr. Valentin Fuster on JACC.org.
- Abbreviations and Acronyms
- acute coronary syndrome
- B-type natriuretic peptide
- heart failure
- heart failure with preserved ejection fraction
- heart failure with reduced ejection fraction
- International Classification of Diseases
- left ventricle
- soluble suppression of tumorigenicity 2
- Received September 6, 2018.
- Revision received February 1, 2019.
- Accepted February 4, 2019.
- 2019 American College of Cardiology Foundation
- Roger V.L.,
- Weston S.A.,
- Gerber Y.,
- et al.
- Benjamin E.J.,
- Blaha M.J.,
- Chiuve S.E.,
- et al.
- Cohn J.,
- Ferrari R.,
- Sharpe N.
- Gerber Y.,
- Weston S.A.,
- Enriquez-Sarano M.,
- et al.
- Jenkins W.S.,
- Roger V.L.,
- Jaffe A.S.,
- et al.
- Yu L.,
- Ruifrok W.P.,
- Meissner M.,
- et al.
- Sharma U.C.,
- Pokharel S.,
- van Brakel T.J.,
- et al.
- Weir R.A.,
- Petrie C.J.,
- Murphy C.A.,
- et al.
- Lisowska A.,
- Knapp M.,
- Tycinska A.,
- et al.
- O'Donoghue M.L.,
- Morrow D.A.,
- Cannon C.P.,
- et al.
- Alpert J.S.,
- Thygesen K.,
- Antman E.,
- Bassand J.P.
- Dunlay S.M.,
- Roger V.L.,
- Weston S.A.,
- Jiang R.,
- Redfield M.M.
- Roger V.L.,
- Go A.S.,
- Lloyd-Jones D.M.,
- et al.
- ↵Mayo Clinic Mayo Medical Laboratories. Test ID: GAL3 Galectin-3, Serum. Available at:. https://www.mayomedicallaboratories.com/test-catalog/Clinical+and+Interpretive/86202. 1932019.
- Coglianese E.E.,
- Larson M.G.,
- Vasan R.S.,
- et al.
- Perea R.J.,
- Morales-Ruiz M.,
- Ortiz-Perez J.T.,
- et al.
- Di Tano G.,
- Caretta G.,
- De Maria R.,
- et al.
- Calvier L.,
- Martinez-Martinez E.,
- Miana M.,
- et al.
- Grandin E.W.,
- Jarolim P.,
- Murphy S.A.,
- et al.