Classification based on echocardiography can identify the risk of long-term heart failure (HF) in asymptomatic patients, according to research results published in JACC: Cardiovascular Imaging.

Using data from the STANISLAS cohort (ClinicalTrials.gov Identifier: NCT01391442), the researchers sought to verify the external validity of echocardiographic phenotyping by quantifying the phenotype with long-term incident HF and risk of cardiovascular death.

The STANISLAS cohort is a monocentric, family, longitudinal cohort of people living in the Nancy region in France, established between 1993 and 1995.


Continue reading

Patients underwent echocardiographic examinations measuring left atrial volume (LA), diastolic function, mitral inflow pattern, E / A ratio, mean e ‘and mean E / e’, and systolic deformity of the body. left ventricle (LV). Other markers including central and peripheral arterial pressure and the index of increase and inflammatory mediators associated with diastolic dysfunction in ICFpEF were also assessed.

Pooled analysis was performed based on the echocardiographic data to identify echocardiographic patterns via the K-means R program. In total, 30 different quality measurements for different numbers of clusters were used to find the optimal number of clusters.

A total of 827 participants (mean age, 60 ± 5 years; 48.2% male) were included in the study. An additional 1,394 participants from the Malmö Preventive Project Cohort, a population-based longitudinal cohort of residents of Malmö, Sweden, who underwent echocardiography between 2002 and 2006 and had no history of HF were also was included (mean age, 67 ± 6 years; 70% Men).

Cluster analysis found 3 groups with different echocardiographic phenotypes. The largest group (n = 334), labeled “near normal” had the highest e ‘and E / A ratio and the highest LV systolic absolute pressure. The next largest group consisted of 323 patients and was labeled as the “diastolic changes” phenotype, had lower e ‘and higher E / e’ ratios. The final group (n = 170) – called the “diastolic changes with structural remodeling” phenotype – had the highest LV mass and volumes, the highest LA volume, and the lowest LV absolute systolic pressure. e ‘was lower and the E’e ratio was higher in this group.

The assessment of the 2016 DD class with conventional echo variables was very low for the 3 phenotypes; about half of people with diastolic changes with a structural remodeling phenotype had additional echo biomarkers.

Similar patterns of echocardiographic phenotypes have been noted between K-means clustering and hierarchical clustering. In addition, the LCM approach showed consistent phenotypes with K-means clustering.

The diastolic changes phenotype predominantly included females, while the diastolic changes with the structural remodeling phenotype predominantly included males. Both phenotypes were associated with older age, higher BMI, and more cardiovascular risk factors compared to the generally normal phenotype; no significant difference in these clinical features was observed.

Those in the mostly normal phenotype had more favorable levels of central and peripheral pressure, as well as vascular stiffness, compared to diastolic changes, and diastolic changes were structural remodeling phenotypes.

Among the 32 circulating biomarkers, 14 were significantly different for the 3 phenotypes. These biomarkers were associated with different pathophysiological domains, such as inflammation and extracellular matrix remodeling. The diastolic phenotype exhibited the highest levels of circulating biomarkers associated with inflammation, while biomarkers typically associated with remodeling – GDF15, PIIINP, ST-2, troponin-1, and CNP – increased from predominantly normal to diastolic phenotypes and diastolic with structural remodeling.

A decision tree identified e ‘, LVEDVi and LVMi as the most relevant variables in classifying participants in an echocardiographic profile, with good overall accuracy at 79%. The addition of clinical variables such as age, sex, BMI, hypertension, diabetes, dyslipidemia, coronary heart disease, and smoking did not change the decision algorithm.

The Malmö Preventive Project cohort was used to externally validate the phenotypes of the e’VM algorithm. The predominantly normal phenotype (n = 440) had the most favorable values ​​in terms of clinical, laboratory and echocardiographic profiles. The diastolic phenotype (n = 512) had the lowest proportion of males and the diastolic phenotype with structural remodeling had the highest levels of NT-proBNP and the highest LV mass and LA area indices.

Over a median follow-up period of 10.3 years (range 9.8 to 11.1), 10.1% of participants in the Malmö cohort met the primary endpoint of the study. Compared to the predominantly normal phenotype, diastolic and diastolic phenotypes with structural remodeling were significantly associated with increased rates of the primary endpoint (gross risk ratio [HR], 2.47; 95% CI, 1.38-4.41 and crude HR, 4.67; 95% CI, 2.67-8.14).

After adjustment for the ARIC HF and NT-proBNP risk score, these phenotypes remained significantly associated with the primary endpoint (adjusted RR, 0.187 and 3.02; 95% CI, 1.04-3.37 and 1, 71-5.34).

Echocardiographic phenotypes significantly improved prognostic performance in addition to the ARIC HF risk score, and discriminating values ​​were “consistently observed in addition to traditionally defined abnormalities in cardiac structure or function”.

Limitations of the study include observational design, low number of participants with incident HF, and young cohort age, among other limitations.

“These echocardiographic phenotypes shed new light on our understanding of asymptomatic cardiac dysfunction, and our findings may have important clinical implications in the design of strategies for preventing HF,” the researchers concluded. “Another prospective multicenter study is needed to assess the applicability of the e’VM algorithm.”

Disclosure: Several study authors declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of author disclosures.

Reference

Masatake K, Huttin O, Magnusson M, et al; on behalf of the STANISLAS Study. Echocardiographic phenotypes derived from machine learning predict the incidence of heart failure in asymptomatic individuals. JACC Cardiovascular Imaging. Published online September 15, 2021. doi: 10.1016 / j.jcmg.2021.07.004