Galectin-3 as a Risk Predictor of Mortality in Survivors of Out-of-Hospital Cardiac Arrest

Out-of-hospital cardiac arrest (OHCA) is defined as non-traumatic, unexpected circulatory arrest occurring within one hour of the symptom onset in an apparently healthy subject¹. With an increased rate of successful resuscitation from non-traumatic OHCA, there is increasing emphasis on early risk-stratification and management of patients with return-of-spontaneous-circulation². There are insufficient data on biomarkers that can help risk-stratify patients and guide therapy. Here, we report the novel role of galectin-3 among other cardiac risk markers to predict both short- and long-term all-cause mortality in patients successfully resuscitated from OHCA.

In this prospective multicenter study conducted from March-2016 to December-2018, we enrolled 189 patients (142 OHCA survivors, 18 controls with structural heart disease without OHCA, and 29 healthy controls) from four tertiary care centers at Buffalo-Niagara metropolitan area. We collected clinical covariates including age/gender, prior cardiac events, LV ejection fraction (EF), renal dysfunction, and use of diuretics or other guideline-directed heart failure (HF) medications. The events surrounding OHCA including initial ECG rhythm, QTc, and time-to-resuscitation were studied on the EMS dataset. Serum samples were obtained at a mean of 255-minutes from cardiac arrest for biomarker analysis. Research protocols were approved by the institutional review board (IRB) at the University of Buffalo, which waived the informed consent. The data supporting the findings of this study are available from the corresponding author upon reasonable request.

After clinical data retrieval by chart-review and follow-up of events with standardized telephone interview with the patients or family members, we performed Univariate cox-proportional-hazard analysis to study the association of primary outcome of long-term all-cause mortality with clinical, laboratory and imaging covariates. Additionally, serum biomarkers including galectin-3, galectin-3-binding protein (G3BP), troponin-I and BNP were measured and included in the univariate model for comparison.

Overall, there were 72 survivors and 70 non-survivors on an average of 5±7-month follow-up. The highest mortality (46%) was noted within the first 4-weeks, with a modest 3% mortality after discharge. Non-survivors had higher prevalence of diabetes, renal disease, anemia, QT-prolongation, diuretic use, unwitnessed arrest, longer cardio-pulmonary resuscitation (CPR), and non-shockable rhythm. Non-survivors also had elevated galectin-3 (ng/ml, survivors:27.7±21.7; non-survivors:48.7±27.7, p<0.0001), and BNP (pg/ml, survivors:318.6±559.1; non-survivors:1166.8±1601.1, p<0.0001), without differences in the peak troponin-I, CK, and G3BP levels.

The significant predictors from the univariate model were then entered into multivariate model. We generated receiver operating characteristic (ROC) curves for each significant variable on the multivariate cox-proportional-hazard analysis. For biomarkers with a significant area under-the-curve (AUC), optimal cut-points were chosen for predicting 5-month all-cause mortality. Kaplan-Meier curves were then constructed for 30-day and long-term survival with the population stratified based on these cut-points.

We found that galectin-3, BNP, CPR-duration and diuretic use were each independently associated with mortality. Combined, they significantly increased the risk of death. Maximized-Youden-index obtained by ROC-curves demonstrated that cardiac arrest survivors with galectin-3 levels > 26.6ng/ml had higher risk of death (HR:3.5,95% CI:1.94–6.13, p<0.0001). Additionally, individuals on diuretics at the time of OHCA (HR:2.44,95% CI:1.5–3.96, p=0.04), BNP levels >205pg/ml (HR:3.16,95% CI:1.61–6.18, p=0.0008) and CPR-duration >7 minutes (HR:3.23, 95% CI:1.65–6.33, p=0.0006) also had increased risk of death. When elevated galectin-3 level of >26.6ng/ml was present with any of the other risk factors (e.g., on diuretics at the time of OHCA, BNP >205pg/ml or CPR>7 minutes), the risk of death was very high (HR:20.37,95% CI:2.81–147.48, p<0.002) (Figure 1).

Figure 1.

Combinatorial analysis of all-cause mortality using Kaplan-Meier survival plot in resuscitated out-of-hospital sudden cardiac arrest patients (OHCA). Survival plots compared as (a) no risk factors, (b) presence of one or more risk factors without elevated galectin-3, and (c) elevated galectin-3 (>26.6 ng/ml) with presence of one or more of the other co-existing risk factors. Risk factors are significant variables derived from multivariate analysis including, i) being on diuretics at the time of OHCA, ii) resuscitation time of >7 minutes, and iii) BNP levels of >205 pg/ml. Proportional risk distribution analysis showed, 9% without any risk factors, 3% with elevated galectin-3 only, 13% with elevated galectin-3 and 1 risk factor, 44% with elevated galectin-3 and ≥1 risk factor(s), and 15% with >1 risk factors without galectin-3 elevation.

Although diuretic use was associated with higher mortality, risk-distribution analysis showed no differences in the proportion of patients with severely reduced (EF<35%) vs., preserved cardiac function. The frequency of prior HF-admissions was similar in survivors and non-survivors. Nearly half of the patients presented with VT/VF, with others presenting as asystole, pulseless-electrical-activity or an indeterminate-rhythm. Two-thirds of the patients had ischemic etiology. In a subgroup analysis, non-survivors with prior ischemic-heart-disease with OHCA had significantly elevated galectin-3 (ng/ml, survivors:28.5±20.8; non-survivors:47.1±26.7, p=0.001). However, in acute MI, serum galectin-3 did not differ between survivors and non-survivors, but trended higher in non-survivors.

Currently, limited data are available to risk-stratify the survivors of OHCA for subsequent clinical outcomes, including the follow-up beyond hospital discharge. Our analyses show that galectin-3 is associated with poor-survival in resuscitated patients after OHCA. Galectin-3 enhances the predictive utility when combined with other clinical covariates. Since our first pre-clinical study published in 2004 identifying increased galectin-3 expression in cardiac fibrosis³, and the first clinical study published in 2006 showing increased circulating galectin-3 in HF⁴, several other groups have reported an association between galectin-3 expression and cardiac remodeling, HF, and death. There were no data examining the association between circulating galectin-3 and post-resuscitation outcomes. Unlike other biomarkers, galectin-3 is expressed at the early-stage of cardiac fibrosis, suggesting its involvement in the pathogenesis. Galectin-3 inhibitors are currently being tested⁵. Our study emphasizes the importance of further studying this unique macrophage-secreted protein en route to the development of arrhythmic mortality and supports the therapeutic potential of inhibiting this molecule.

Non-Standard Abbreviations and Acronyms

OHCA

Out-of-hospital cardiac arrest

G3BP

galectin-3-binding protein

AUC

area under-the-curve