The biomarker plasma galectin-3 in advanced heart failure and survival with mechanical circulatory support devices.

BACKGROUND: During screening of heart transplantation (HTx) candidates supported by ventricular assist devices (VADs) for plasma biomarkers we found that galectin-3 (Gal-3) was increased pre-operatively in patients who later died during VAD support. Therefore, we analyzed the predictive value of plasma Gal-3 in the context of other potential clinical risk factors for death on device (DOD) in a cohort of 175 VAD patients. METHODS: We analyzed numerous clinical factors and plasma Gal-3 levels of 175 VAD patients before device implantation. Eighty VAD patients were successfully bridged to HTx (BTT, 45.7%), 80 (45.7%) died on VAD, 2 recovered on device (BTR, 1.1%) and 13 (7.4%) were still on device. Uni- and multivariate analyses were performed to assess the importance of Gal-3 with respect to other clinical factors. Myocardial gene expression of Gal-3 was investigated in apex samples by RT-PCR (n = 30) and Western blotting (n = 45). RESULTS: Plasma Gal-3 levels were higher in VAD patients than in controls (16.6 ± 9.3 vs 9.5 ± 3.9 ng/ml, p < 0.0001). Cox regression showed several clinical factors and type of VAD as independent outcome predictors, but Gal-3 was not among them. Using the regression equation we grouped patients according to their factor constellation for prediction of survival on VAD. We propose a calculation method for VAD survival prediction. Gal-3 mRNA and protein were detectable in failing myocardium, but did not correlate with its plasma concentration. CONCLUSIONS: Galectin-3 levels are associated with severe heart failure but do not provide sufficient discrimination for prediction of outcomes after VAD implantation. Importantly, we were unable to confirm myocardial tissue as a primary source for the observed plasma elevations of Gal-3.

Safety of Long-Term Mechanical Support With Berlin Heart EXCOR in Pediatric Patients.

Background. The Berlin Heart EXCOR Pediatrics is utilized at our center as a bridge to transplantation or bridge to recovery. This retrospective study reviews our results regarding the safety of long-term support and outcome. Methods. Between January 2008 and December 2010, 12 patients (6 females and 6 males) underwent implantation of a ventricular assist device. The median weight was 14.2 kg (range 4.2-51.6 kg) and the median age was 4.12 years (range 0.25-11.83 years). All patients were on inotropes, five patients required mechanical ventilation and three patients experienced cardiopulmonary resuscitation. Results. Eight patients received a left ventricular assist device and four patients received a biventricular assist device. Of the 12 patients, 8 were bridge to heart transplantation, in 2 patients explantation was possible, and 1 patient died on support. The median support time for these 11 patients was 151 days (range 4-488 days), with 2124 days of cardiac support. One patient is on support. Survival rate was 91.6%. Seven patients had a blood pump change once. Four patients had local signs of infection. There was no mediastinitis and thromboembolism. One patient had intracerebral hemorrhage. There was no death after heart transplantation or after explantation of the device. Conclusions. The Berlin Heart EXCOR is effective in bridging children of almost all ages and sizes to cardiac transplantation or myocardial recovery. Our experience proved that long-term support is possible with a low rate of adverse events.

Global gene expression analysis in nonfailing and failing myocardium pre- and postpulsatile and nonpulsatile ventricular assist device support.

Mechanical unloading by ventricular assist devices (VAD) leads to significant gene expression changes often summarized as reverse remodeling. However, little is known on individual transcriptome changes during VAD support and its relationship to nonfailing hearts (NF). In addition no data are available for the transcriptome regulation during nonpulsatile VAD support. Therefore we analyzed the gene expression patterns of 30 paired samples from VAD-supported (including 8 nonpulsatile VADs) and 8 nonfailing control hearts (NF) using the first total human genome array available. Transmural myocardial samples were collected for RNA isolation. RNA was isolated by commercial methods and processed according to chip-manufacturer recommendations. cRNA were hybridized on Affymetrix HG-U133 Plus 2.0 arrays, providing coverage of the whole human genome Array. Data were analyzed using Microarray Analysis Suite 5.0 (Affymetrix) and clustered by Expressionist software (Genedata). We found 352 transcripts were differentially regulated between samples from VAD implantation and NF, whereas 510 were significantly regulated between VAD transplantation and NF (paired t-test P < 0.001, fold change >or=1.6). Remarkably, only a minor fraction of 111 transcripts was regulated in heart failure (HF) and during VAD support. Unsupervised hierarchical clustering of paired VAD and NF samples revealed separation of HF and NF samples; however, individual differentiation of VAD implantation and VAD transplantation was not accomplished. Clustering of pulsatile and nonpulsatile VAD did not lead to robust separation of gene expression patterns. During VAD support myocardial gene expression changes do not indicate reversal of the HF phenotype but reveal a distinct HF-related pattern. Transcriptome analysis of pulsatile and nonpulsatile VAD-supported hearts did not provide evidence for a pump mode-specific transcriptome pattern.