Left Ventricular Assist Device (LVAD) Implantation Transiently Increases Plasma TGF-β1 in Heart Failure Patients: Potential Impact of Lvad on Platelet Activation

Although survival and quality of life have improved in patients with advanced heart failure (HF) after implantation of left ventricular assist devices (LVADs), they still pose risks of hemocompatibility-related complications, including thrombosis and bleeding. Development of biomarkers predictive of these LVAD-associated complications could guide decision making for both clinicians and patients. Recently, we showed higher plasma TGF-β1 levels within one-week after implantation with a miniaturized mechanical-bearing axial-flow pump HeartMate II (HM-II), and reasoned that platelet activation by the rotor may have caused the release of TGF-β1 in plasma in HF patients (Mancini et al. Transl. Res. 2018; 192:15-29). Recent clinical trials with the newest LVAD, the Heartmate 3 (HM-3), which uses a fully magnetically-levitated pump, showed superior clinical outcomes, including significantly reduced incidences of pump thrombosis and stroke (Mehra et al. N Engl J Med. 2019; 380:1618-1627). In this study, we evaluated release of TGF-β1 in plasma following implantation of HM-II and HM-3 LVADs compared to either coronary artery bypass graft (CABG) surgery or extracorporeal membrane oxygenation (ECMO) therapy. We measured serial total TGF-β1 levels in 38 Stage-D HF patients (11 received HM-II and 27 received HM-3). As a control, we collected blood samples from 10 patients undergoing CABG surgery, and 10 patients receiving ECMO therapy following acute onset cardio-pulmonary failure. Blood samples were collected before and 4-8 hours after procedures, and thereafter daily for up to one week. Plasma was prepared by centrifuging blood at 12,000 rpm for 5 min at 4°C within 10 min of blood drawing, which reduces in vitro release of TGF-β1 from platelets and thus allows accurate measurement of plasma TGF-β1. Total TGF-β1 levels were measured after acidification and neutralization of samples using DUO-ELISA kit (R&D Systems). Baseline total plasma TGF-β1 levels were higher in HF patients before LVAD implantation than in healthy controls [4.7 ± 1.9 ng/mL in HF patients (n= 38); 3.3 ± 0.8 ng/mL in healthy controls (n= 6); p=0.006)]. Total TGF-β1 levels surged transiently to 14.6 ± 6.1 ng/mL within 4-8 hours after LVAD implantation [(p<0.0001 compared to patients 4-12 hours after CABG surgery (3.6 ± 1.4 ng/mL) or ECMO therapy (4.9 ± 1.3 ng/mL)]. Interestingly, however, we found that the transient surge of TGF-β1 in HM-3 recipients was significantly lower than in HM-II recipients (Figure-1; p=0.04). TGF-β1 levels then gradually decreased and reached near basal levels 2-3 days after LVAD implantation, but remained significantly elevated in plasma of HM-II recipients until day 5 (p=0.049). TGF-β1 levels remained unchanged in both CABG and ECMO patients at all time points (Figure 1). We conclude that LVAD implantation causes a transient surge in total plasma TGF-β1 within a few hours after the procedure, presumably due to platelet activation by LVAD, not the surgery itself, as CABG or circulating blood through ECMO did not cause the surge. The observation that a reduced initial surge and lower levels of TGF-β1 in HM-3 vs. HM-II recipients needs further investigation to determine whether these differences are due to LVAD-specific factors (different rotors causing variable shear effects) or to confounding differences in implantation procedures, such as, by-pass time, cardiac tissue injury, number of platelet transfusions, blood suction with catheters etc. or other unknown factors. Our data suggest that serial TGF-β1 measurements after LVAD implantation may serve as a surrogate biomarker for platelet activation in association with hemocompatibility-related adverse events (Uriel et al. Circ. 2017; 135:2003-2012). Disclosures No relevant conflicts of interest to declare.