Recent reports of myocardial recovery after mechanical unloading with left ventricular assist devices (LVADs) have challenged the prevailing notion that end-stage heart failure (HF) is irreversible. To improve our understanding of this phenomenon, we comprehensively analysed the structural, functional, and energetic changes in failing human cardiomyocytes after LVAD implantation.

Based on a prospectively registered protocol (PROSPERO-CRD42022380214), 30 eligible studies were identified from 940 records with a pooled population of 648 patients predominantly with non-ischaemic cardiomyopathy.

LVAD led to a substantial regression in myocyte size similar to that of donor hearts (standardised mean difference, −1.29; p<0.001). The meta-regression analysis revealed that HF duration was a significant modifier on the changes in myocyte size. There were some suggestions of fibrosis reversal (−5.17%; p=0.009); however, this was insignificant after sensitivity analysis. Developed force did not improve in cardiac trabeculae (n=5 studies); however, non-physiological isometric contractions were tested. At the myocyte level (n=4 studies), contractile kinetics improved where the time-to-peak force reduced by 41.7%–50.7% and time to 50% relaxation fell by 47.4%–62.1% (p<0.05). Qualitatively, LVAD enhanced substrate utilisation and mitochondrial function (n=6 studies). Most studies were at a high risk of bias.

The regression of maladaptive hypertrophy, partial fibrosis reversal, and normalisation in metabolic pathways after LVAD may be a testament to the heart’s remarkable plasticity, even in the advanced stages of HF. However, inconsistencies exist in force-generating capabilities. Using more physiological force-length work-loop assays, addressing the high risks of bias and clinical heterogeneity are crucial to better understand the phenomenon of reverse remodelling.