Author + information
- Juan J. Badimon, PhD∗ ( and )
- Carlos G. Santos-Gallego, MD
- Atherothrombosis Research Unit, Mount Sinai Heart, Icahn School of Medicine at Mount Sinai, New York, New York
- ↵∗Address for correspondence:
Dr. Juan J. Badimon, Atherothrombosis Research Unit, Mount Sinai Heart, Icahn School of Medicine at Mount Sinai, One Gustave L Levy Place, 1428 Madison Avenue, Atran Building, 6th Floor, Room 6.30, Box 1030, New York, New York 10029.
- arterial pulsatility
- blood flow
- extracorporeal membrane oxygenation
- mechanical circulatory support
- von Willebrand factor
The use of mechanical circulatory support (MCS), either as a bridge to transplantation or as destination therapy, improves outcomes in patients with end-stage heart failure (HF) refractory to medical therapy (1). First-generation pulsatile flow (PF) MCS has been gradually replaced by second- or third-generation continuous flow (CF) MCS due to improvements in survival and quality of life (2). The most frequent adverse event associated with CF-MCS is bleeding, affecting up to 50% of patients (3,4). The supraphysiological shear-stress forces generated by these devices induce an acquired von Willebrand factor (VWF) disease/defect, which is responsible for bleeding complications (5). Of note, the fact that bleeding is 10-fold higher in CF- than PF-MCS (6) and the inverse relationship between pulsatility index in MCS and bleeding (7) strongly suggest some causal relationship between pulsatility and bleeding. In this issue of the Journal, Vincent et al. (8) shed light on this relationship by demonstrating that pulsatility triggers the endothelial release of VWF and suggesting that preservation of pulsatility during CF-MCS would mitigate the acquired VWF defect associated with assisting devices and reduce bleeding.
VWF is synthesized and stored in the endothelial Weibel-Palade bodies (WPBs) and platelet α-granules. VWF is composed of several structural domains (A to D); the VWF subunits dimerize in the endoplasmic reticulum through intermolecular disulfide bridges (9,10). The acidic pH in the Golgi stimulate the formation of high molecular weight (HMW) VWF multimers through disulfide bridges between D3 domains. The growing VWF multimer organizes into a right-handed helix, thereby folding itself into a tubular conformation for storage in the WPBs (9,10). When the WPBs fuse with the endothelial membrane, the VWF multimers lose their tubulated conformation, unfurling into long strings (9,10). After secretion, these ultra-large VWF multimers are proteolyzed by A disintegrin and metalloprotease with thrombospondin type I repeats-13 (ADAMTS13) into smaller multimers that circulate in plasma. VWF circulates in plasma in a coiled, globular, inactive form and does not interact with platelets. Damage to the endothelial surface exposes subendothelial collagen, allowing HMW to bind collagen at the A3 domain (9,10). These bound HMW multimers uncoil in adhesive strings, exposing the binding site for GPIbα in the A1 domain and allowing the binding of platelet GPIbα. The formation of bridges with neighboring platelets facilitates the formation of stable thrombus (9,10).
VWF is unique among hemostatic factors due to its highly multimeric structure, which allows it to function as an endogenous sensor of hemodynamic forces. Under conditions of low shear stress (below the usual physiological range), VWF self-associates into compacted HMW multimers that promote platelet adhesion. Physiological high shear stress (100 to 5,000 s−1) converts this coiled VWF multimer into an elongated shape, thus unfolding the A2 domain; this conformational change in VWF, which regulates its proteolysis, exposes specific residues to which the metalloprotease ADAMTS13 binds and cleaves. As a result, the shorter multimers become less hemostatically competent than the intact large HMW multimers, thus impeding platelet aggregation and clot formation (9–11). Pathological conditions associated with supraphysiological shear stress (shear rate >10,000 s−1) and turbulent blood flow can cause excessive degradation of HMW multimers by ADAMTS13, leading to acquired VWF defect and major bleeding (9,10).
Therefore, increased shear stress will cause acquired VWF defect (reduced HMW monomers and impaired VWF collagen-binding functionality). This form of acquired VWF deficiency explains the increased hemorrhage in patients with severe aortic stenosis (12) (Heyde syndrome) and the resolution of bleeding upon aortic valve replacement. Other conditions with increased shear stress will also cause VWF defect, such as obstructive hypertrophic cardiomyopathy (13) and, specifically to our case, MCS (5,14). Importantly, the response of destruction of HMW multimers is highly dynamic; loss of HMW multimers occurs rapidly after the onset of supraphysiological shear stress, and the multimer distribution normalizes within minutes after restoration of normal blood flow. As a corollary, assessment of HMW multimers can be used to monitor paravalvular leak during the TAVR procedure: no leak is present if HMW normalizes after TAVR (negative predictive value of 98%), whereas HMW remains reduced if paravalvular leak is present (15).
Therefore, the authors hypothesized that the CF-MCS–induced VWF defect could be modulated by the endothelial response to the level of pulsatility (8). The authors, using an in vitro endothelial-free mock circulatory loop to avoid the release of new VWF from endothelial cells, demonstrated the cleavage of the large VWF multimers when exposed to high shear stress. They confirmed the previous reports by this same group of VWF defect under CF-MCS (reduction in both HMW monomer and VWF collagen-binding activity) (5).
The main novelty of the study emanates from the in vivo experiments showing a dose-response relationship between pulsatility and VWF in a porcine model. Both HMW multimers and VWF collagen-binding activity were boosted with increased pulsatility levels (but similar shear stress), with a strong direct correlation (r = 0.73) between pulse pressure and HMW multimers. The authors used parallel groups to substantiate HMW multimer loss with continuous flow and their recovery with pulsatile flow. Finally, the authors confirm this observation in the clinical scenario—with just 1 single patient—demonstrating the disappearance of large HMW multimers under CF-MCS, while a return to PF-MCS caused a rapid restoration of functional VWF.
The acute release of VWF was demonstrated by the significant increase in VWF antigen observed in the swine cross-over model as soon as the endothelium was exposed to high pulsatility. Because WPB store other proteins such as Factor VIII coagulant activity (FVIIIc) and angiopoietin-2, the concomitant increase of VWF antigen, FVIIIc, and angiopoietin-2 upon high pulsatility is strongly suggestive of an acute release of new VWF multimers from the WPB. In fact, the previous report that pulsatile stretch triggers a rapid exocytosis of endothelial WPB and a rapid release (within minutes) of VWF (through vascular endothelial growth factor receptor-2 signaling pathways) support the current findings (8,16). The finding of significantly increased levels of VWF antigen, FVIIIc, and angiopoietin-2 (3 proteins stores inside the WPB) with the recovery of normal pulsatility strongly suggest that the restoration of HMW-multimer profile is due to an acute release of VWF stored within endothelial WPB. Finally, the authors confirm in a single human patient that CF-MCS lowered HMW multimers while return to PF-MCS caused a rapid restoration of functional VWF.
Some questions remain unanswered and warrant further investigation. The most obvious limitation of the present work is that the authors have demonstrated that pulsatility improves a surrogate endpoint (an increase in HMW multimers) but not a clinically relevant endpoint (e.g., reduction in bleeding events). Nonetheless, the authors have to be congratulated for this mechanistic study, which paves the way to a clinical trial actually studying clinical events (bleeding episodes, morbidity, and mortality). Additionally, the fact that the studies in 3 different experimental settings point in the same direction of pulsatility enhancing VWF strengthens the results. We also take into account 2 significant differences between the translational and clinical findings. First, the fact that their animal findings are confirmed in 1 single patient is a clear limitation; thus, these preliminary results need to be corroborated in a larger sample size. The second and probably more important difference is that the translational studies performed in normal animals involved a healthy and fully functional endothelium offering no obvious limitation to endothelial VWF production and release in response to rheological changes. On the other hand, whether the less-than-healthy, and probably dysfunctional, endothelium of an end-stage HF patient is able to respond to pulsatility and produce large quantities of VWF in a maintained fashion remains to be ascertained. Finally, the authors have demonstrated the physiological process (i.e., pulsatility) but not the specific molecular pathways that drive VWF production/release; the discovery of this molecular mechanisms could lead to the development of pharmacological agents to target bleeding (both CF-MCS–related and perhaps even unrelated hemorrhagic episodes).
In summary, the authors should be commended first for the mechanistic study (it is the pulsatility that increases HMW multimers in the circulation) and for their scientific creativity to evaluate the relationship between pulsatility and shear stress. The methods are solid and robust, with studies at different levels (in vitro, both parallel-group and cross-over swine studies, and human study). Finally, the study has a clear clinical application (pulsatility may be utilized to reduce the loss of HMW multimers and thereby attenuate the risk of MCS-related bleeding). But, utmost caution should be taken when speculating that the development of CF-MCS generating pulsatility could mitigate this acquired VWF, reduce bleeding, and improve clinical outcomes. The potential importance of these preliminary observations deserve to be further investigated in larger sample sizes incorporating clinically relevant endpoints.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Both authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2018 American College of Cardiology Foundation
- Uriel N.,
- Pak S.W.,
- Jorde U.P.,
- et al.
- Wever-Pinzon O.,
- Selzman C.H.,
- Drakos S.G.,
- et al.
- Vincent F.,
- Rauch A.,
- Loobuyck V.,
- et al.
- Springer T.A.
- Leebeek F.W.,
- Eikenboom J.C.
- Badimon L.,
- Badimon J.J.,
- Turitto V.,
- Fuster V.
- Le Tourneau T.,
- Susen S.,
- Caron C.,
- et al.
- Meyer A.L.,
- Malehsa D.,
- Budde U.,
- et al.
- Van Belle E.,
- Rauch A.,
- Vincent F.,
- et al.