Distinct molecular portraits of human failing hearts identified by dedicated cDNA microarrays

This study aimed to investigate whether a molecular profiling approach should be pursued for the classification of heart failure patients.

Applying a subtraction strategy we created a cDNA library consisting of cardiac- and heart failure-relevant clones that were used to construct dedicated cDNA microarrays. We measured relative expression levels of the corresponding genes in left ventricle tissue from 17 patients (15 failing hearts and 2 nonfailing hearts). Significance analysis of microarrays was used to select 159 genes that distinguished between all patients. Two-way hierarchical clustering of the 17 patients and the 159 selected genes led to the identification of three major subgroups of patients, each with a specific molecular portrait. The two nonfailing hearts clustered closely together. Interestingly, our classification of patients based on their molecular portraits did not correspond to an identified etiological classification. Remarkably, patients with the highest medical urgency status (United Network for Organ Sharing, Status 1A) clustered together.

With this pilot feasibility study we demonstrated a novel classification of end-stage heart failure patients, which encourages further development of this approach in prospective studies on heart failure patients at earlier stages of the disease.