Galectin‐3 and Hypertensive Heart Disease

The identification of new biomarkers of subclinical cardiovascular (CV) damage allowing for improvement in CV risk estimation represents a challenging task for CV medicine. In this issue of the Journal, Yao and colleagues1 describe the results of a study designed to assess the role of circulating galectin‐3 (Gal‐3) in predicting abnormal left ventricular (LV) geometries (ie, concentric LV hypertrophy, eccentric LV hypertrophy, and concentric LV remodeling) in a group of essential hypertensive patients.

Before addressing the strengths and limitations of the study, previous knowledge on this subject and related topics are summarized.

Experimental and human studies indicate that the changes of myocardial texture in hypertensive LV hypertrophy (LVH) are the result of two major pathological processes, namely hypertrophy of cardiomyocytes and interstitial fibrosis.2 Several lines of evidence show that the collagen fraction of the myocardium is higher in hypertensive patients with increased LV mass than in normotensive individuals.

The results of different methodologic approaches, including determination of circulating markers of collagen synthesis and degradation, of cardiac hormones and ultrasonographic procedures (ie, backscatter analysis of returning echoes and quantitative echo‐reflectivity analysis) have been validated against the collagen fraction of endomyocardial biopsies. Invasive and noninvasive studies have documented that increased amounts of fibrous tissue may be detected in the early phases of hypertension even in patients with LVH of mild degree.3 Cardiac fibrosis and myocardial damage may be reliably detected and monitored over time by serum biomarkers, as supported by an impressive number of clinical investigations in different settings.

The increased myocardial stress due to sustained hypertension causes myocardial fibrosis as a result of fibroblasts activation and collagen accumulation into extracellular matrix. The pathophysiology of this process also contributes to the complex interaction of factors regulating collagen synthesis, collagen degradation, and fibroblast activity. In the past several years, Gal‐3 has emerged as a powerful biomarker of myocardial inflammation and fibrosis.

This profibrotic and inflammatory substance is a member of endogenous P‐galactoside‐binding lectins, a family of widely distributed carbohydrate‐binding proteins that have been implicated in cell growth, adhesion, and migration. Gal‐3, secreted by activated macrophages in the extracellular space, interacts with fibroblast surface receptors to activate signaling pathways for a variety of cellular functions. Since the mid‐90s, an upregulation of this protein has been demonstrated in numerous pathological conditions such as neoplastic diseases, liver cirrhosis, lung fibrosis, chronic heart failure (HF), coronary artery disease, aortic stenosis, atrial fibrillation, and hypertension.4, 5

Among the manifestations of CV disease, most attention has been devoted to the association of Gal‐3 and HF because of the high prevalence of this condition associated with an elevated mortality risk in the general population worldwide. After the pioneering paper by Sharma and coworkers6 showing that a 4‐week continuous infusion of low‐dose Gal‐3 into the pericardial sac of healthy Sprague‐Dawley rats caused LV dysfunction with a threefold increase of collagen I over collagen III, a large amount of experimental and clinical information has been accumulated in this research area. Gal‐3 has been involved in the pathophysiology of HF because of its profibrotic and inflammatory effects. Increased circulating levels of Gal‐3 have been proposed to be a marker of active myocardial remodeling and to predict unfavorable outcomes. A number of recent studies have demonstrated that patients with HF have higher Gal‐3 levels than patients with preserved LV function. Elevated Gal‐3 levels have been shown to be directly related to the risk of hospitalization, CV mortality attributable to pump failure, or sudden death. The value of Gal‐3 in predicting CV disease and incident HF has been studied in general‐based population samples. In a subset of the Physicians' Health Study, Djoussé and colleagues7 documented that higher plasma levels of Gal‐3 were associated with a greater risk of incident HF after adjustment for several confounding factors. The highest quartile of log‐Gal‐3 was associated with a 57% higher risk of HF in a multivariable‐adjusted model compared with the lowest quartile. These findings are consistent with previous data from the Framingham Heart Study, which documented a 28% higher risk of HF per one standard deviation increase of log‐Gal‐3.8

Although the role of Gal‐3 in the pathogenesis of CV diseases has been consistently documented, the value of this biomarker in predicting CV disease and HF remains debated.9 Recently, data from a northern European population have demonstrated an association between Gal‐3 concentration and incident HF and all‐cause mortality,10 although the degree of association with incident HF and CV mortality was modest. Moreover, no significant improvement in risk prediction of CV outcomes was observed after the addition of Gal‐3 in Cox regression analyses including conventional risk factors and N‐terminal B‐type natriuretic peptide (NT‐pro‐BNP). In the Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training (HF‐ACTION),11 the authors prospectively assessed the relationship between NT‐pro BNP, Gal‐3, soluble T2, and death. Traditional risk factors along with NT‐pro‐BNP levels were strong predictors of death related to pump failure, and no significant incremental contributions were provided by soluble T2 and Gal‐3. Of note, the prediction of sudden cardiac death, although less robust, was somewhat improved by information provided by Gal‐3.

To date, available literature addressing the relationship between circulating Gal‐3, brachial BP, and subclinical cardiac damage in general population samples and in hypertensive cohorts is scanty.

López and coworkers12 investigated the association of Gal‐3 with the degree of synthesis and deposition of fibrillar collagen types I and III in endomyocardial biopsies from hypertensive patients with HF. The authors found that an excess of Gal‐3 was present both in the myocardium and plasma of these patients. They were unable, however, to find an association between circulating or myocardial Gal‐3 and molecular or histological markers of myocardial fibrosis.

In a recent study carried out in 966 randomly selected participants from a community sample, Gal‐3 concentrations were not found to be independently associated with office or, more importantly, ambulatory BP.13 This result is in contrast to prior study findings showing a relationship between Gal‐3 and either the hypertensive condition or office BP values. Seferovic and colleagues14 compared the circulating levels of Gal‐3 in type 2 diabetic patients without renal disease, microvascular diabetic complications, or insulin therapy and in treated essential hypertensive patients and found similar levels of this biomarker in the two groups.

In the study by Yao and colleagues,1 a total of 107 uncomplicated essential hypertensive patients without comorbidities (mean age 58±11 years) and 108 age‐ and sex‐matched controls (58±4 years) underwent clinic BP measurement, NT‐pro‐BNP and Gal‐3 determinations, and standard echocardiography. The major findings of this report are summarized as follows: (1) a progressive increase in circulating Gal‐3 concentrations occurred across normotensive controls (5.6 ng/mL), hypertensive patients with normal LV mass/geometry (7.3 ng/mL), and hypertensive patients with LV remodeling (12.5 ng/mL); (2) in multivariate analysis, only Gal‐3 and body mass index (BMI) were independently correlated to LV remodeling; (3) NT‐pro‐BNP did not differ in hypertensive patients with (30 pg/mL) and without (30 pg/mL) LV remodeling; and (4) receiver operating characteristic curve analyses showed that Gal‐3 performed better than NT‐pro‐BNP in detecting LV remodeling as documented by a greater area under the curve (0.698 [P<.001] vs 0.530 [P=.347], respectively).

Strengths and limitations of this study merit to be briefly discussed. First, the association of Gal‐3 with LV remodeling was independent from potential confounders such as age, sex, and BMI. Moreover, LV mass, a continuous measure of subclinical cardiac damage, was the unique variable independently correlated with Gal‐3. This was not the case for systolic BP, relative wall thickness, body size, and age. On the whole, these results provide the first reliable evidence of a relationship between circulating Gal‐3 and subclinical LV structural changes in hypertensive patients with preserved LV systolic function and no history of HF. From a clinical perspective, the value of Gal‐3 determination in identifying hypertensive patients with a high likelihood of LVH and LV concentric remodeling remains hypothetical and needs further investigation. It should be noted that the Gal‐3 cutoff of 11.4 ng/mL, which yields the best performance in diagnosing LV remodeling, was markedly lower than the 90th percentile (17.6 ng/mL) of the reference interval distribution of Gal‐3 previously reported among healthy asymptomatic patients.15

In the study by Yao and colleagues,1 the majority of patients with (82%) and without (73%) LV remodeling were taking antihypertensive drugs that interfered with the renin‐angiotensin‐aldosterone system and approximately half of them were taking calcium channel blockers. The impact of antihypertensive drugs on tissue and circulating levels of Gal‐3 is largely unknown and may have influenced the relationship between this marker with clinical and echocardiographic variables. In the setting of hypertension, preserved LV systolic function, as in the patients included in the analysis, is often associated with impaired diastolic function. This condition is strongly dependent on the extent myocardial fibrosis; unfortunately, no indexes of diastolic function were assessed and correlated with Gal‐3. Surprisingly, left atrial diameter, an indirect marker of LV diastolic performance, did not significantly differ in hypertensive patients with and without LV remodeling. Finally, in the present study, a single Gal‐3 concentration was measured; however, some authors have recently suggested that repeat measurements of Gal‐3 may improve its prognostic value in HF patients. Whether this is also the case in different clinical conditions is a subject for future investigation.16

Conclusions

The interesting findings of Yao and colleagues1 represent a valuable point for future research on the role of Gal‐3, an emerging biomarker of fibrosis and inflammation, in the diagnostic approach to hypertensive patients.

Disclosure

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.