Author + information
- Received February 20, 2015
- Revision received June 8, 2015
- Accepted June 12, 2015
- Published online November 1, 2015.
- Carlo R. Bartoli, MD, PhD∗ (, )
- Jooeun Kang, BA,
- David J. Restle, BSE,
- David M. Zhang,
- Cameron Shabahang,
- Michael A. Acker, MD and
- Pavan Atluri, MD
- Division of Cardiovascular Surgery, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania
- ↵∗Reprint requests and correspondence:
Dr. Carlo R. Bartoli, Hospital of the University of Pennsylvania, Division of Cardiovascular Surgery, Stemmler Hall, Room 351, Philadelphia, Pennsylvania 19104.
Objectives The aim of this study was to investigate a potential therapy for left ventricular assist device (LVAD)–associated bleeding.
Background Nonsurgical bleeding is the most frequent complication of LVAD support. Recent evidence has demonstrated that supraphysiological shear stress from continuous-flow LVADs accelerates von Willebrand factor (vWF) metabolism by the action of a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 (ADAMTS-13) (the vWF protease). An acquired vWF deficiency causes bleeding. This suggests that ADAMTS-13 is a clinical target to reduce vWF degradation. We tested the hypothesis that inhibition of ADAMTS-13 with doxycycline, an inexpensive, clinically approved drug, reduces vWF degradation during shear stress.
Methods Whole blood was collected from human donors (n = 15), and purified, recombinant ADAMTS-13 protein was obtained. An enzyme-linked immunosorbent assay (ELISA) was used to quantify the dose relationship between doxycycline and ADAMTS-13 activity prior to shear stress (n = 10). To determine the effect of shear stress, plasma and recombinant ADAMTS-13 were exposed to LVAD-like supraphysiological shear stress (approximately 175 dyne/cm2). vWF multimers and degradation fragments were characterized with electrophoresis and immunoblotting (n = 10). Förster resonance energy transfer was used to quantify plasma ADAMTS-13 activity (n = 10). An ELISA was used to quantify vWF:collagen binding activity. Platelet aggregometry was performed with adenosine 5′-diphosphate, collagen, and ristocetin (vWF-platelet pathway) agonism (n = 10).
Results Doxycycline significantly decreased plasma ADAMTS-13 activity (p = 0.01) and the activity of recombinant human ADAMTS-13 protein by 21%. After plasma was exposed to shear stress, the same pattern of vWF degradation was observed as previously reported for LVAD patients, and vWF:collagen binding activity decreased significantly (p = 0.002). Doxycycline significantly decreased ADAMTS-13 activity (p = 0.04) and the activity of recombinant ADAMTS-13 by 18%, protected large vWF multimers from degradation, and significantly decreased the levels of the 5 smallest vWF fragments by 12 ± 2% (p < 0.05). As a result, vWF:collagen binding activity was significantly restored (p = 0.004). ADAMTS-13 inhibition with doxycycline did not hyperactivate platelets.
Conclusions Inhibition of ADAMTS-13 by doxycycline decreased vWF degradation and improved vWF function during supraphysiological shear stress without hyperactivating platelets. ADAMTS-13 is a clinical target to reduce vWF degradation, improve vWF function, and potentially reduce bleeding during LVAD support.
- heart failure
- left ventricular assist device (LVAD)
- mechanical circulatory support
- shear stress
- von Willebrand factor
Left ventricular assist devices (LVADs) improve clinical outcomes in patients with heart failure. However, abnormal shear stress from continuous-flow LVADs alters hemostasis. As a result, nonsurgical bleeding is the most frequent adverse event during LVAD support (1). Indeed, 30% to 75% of patients with an LVAD experience bleeding events (2–6). von Willebrand factor (vWF) plays a major role (7). Emerging evidence demonstrates that supraphysiological shear stress from LVADs accelerates vWF metabolism and depletes functional vWF (8). The effect is an acquired vWF deficiency that predisposes LVAD patients to bleeding events (9).
In patients with a bleeding diathesis, management of anticoagulation poses clinical challenges. Therapeutic options include cessation of antiplatelet therapy and reduction of INR, which may reduce bleeding but place LVAD patients at risk for thrombosis. Importantly, targeted therapies to reduce LVAD-associated vWF degradation and bleeding have not been described.
vWF, a large multimeric glycoprotein, circulates in the plasma and plays an initial and critical role in primary hemostasis (10). Absence of large vWF multimers causes bleeding. vWF metabolism is closely regulated by the rheological conditions of the blood (11). Shear stress triggers vWF activation and degradation by the vWF-specific protease, a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 (ADAMTS-13) (12). Recently, we described the mechanism of pathological vWF degradation during LVAD support (8). In short, supraphysiological shear stress from the LVAD significantly increases vWF metabolism by ADAMTS-13. In this process, ADAMTS-13 cleaves large, biologically active vWF multimers into small vWF fragments, which have reduced hemostatic function.
These findings suggest that ADAMTS-13 is a potential clinical target to prevent bleeding in LVAD patients. Specifically, inhibition of ADAMTS-13 may reduce vWF degradation during LVAD support. Importantly, a number of agents are available that inhibit ADAMTS-13, including doxycycline (13), which is an inexpensive and clinically approved tetracycline antibiotic. We tested the hypothesis that inhibition of ADAMTS-13 by doxycycline reduces vWF degradation during LVAD-like supraphysiological shear stress.
Blood sample collection
Fresh whole blood was obtained via venipuncture from consenting volunteer human donors (n = 15) by a physician per University of Pennsylvania guidelines. Blood samples were anticoagulated in buffered sodium heparin tubes (BD Biosciences, Franklin Lakes, New Jersey) for vWF immunoblotting and ADAMTS-13 activity assays and in sodium citrate blood tubes (BD Biosciences) for platelet aggregation assays. Plasma was used for ex vivo assays, and platelet-rich plasma (PRP) was used for platelet aggregation experiments.
Ex vivo effect of doxycycline on plasma ADAMTS-13 activity without shear stress
The effect of doxycycline on ex vivo plasma ADAMTS-13 activity without shear stress was determined with a commercial enzyme-linked immunosorbent assay (ELISA) kit (Technoclone, Vienna, Austria). A dose-response experiment was performed with doxycycline (0.6, 2, 5, and 10 mg/dl) in human plasma samples of equal total volume (n = 10). Glutathione-S-transferase (GST)–conjugated recombinant vWF-73, the vWF fragment that contains the cleavage site for ADAMTS-13 proteolysis, was incubated with anti-GST–coated wells. Plasma samples were added. Native ADAMTS-13 from each plasma sample cleaved the GST-conjugated vWF-73 and exposed the N-10 cleavage site (14). A horseradish peroxidase (HRP)–conjugated monoclonal antibody to the N-10 cleavage site was co-incubated in solution and bound the newly cleaved vWF site. Tetramethylbenzidine was added to elicit a colorimetric reaction, which was quenched with a stopping solution. A standard curve was constructed from reference plasma samples with known ADAMTS-13 activity. ADAMTS-13 activity in each experimental sample was determined by interpolation of the colorimetric intensity values from the standard curve.
In vitro effect of doxycycline on purified ADAMTS-13 protein activity without shear stress
To further investigate the mechanism, the effect of doxycycline on purified ADAMTS-13 protein activity without shear stress was determined in vitro. Doxycycline 5 mg/dl and recombinant full-length human ADAMTS-13 protein (0.5 μg/ml , R&D Systems, Minneapolis, Minnesota) were suspended in a buffered solution (EBM-2, Lonza, Basel, Switzerland) without blood cells or plasma proteins. The ELISA was performed as previously described.
Ex vivo model of supraphysiological shear stress to induce LVAD-associated vWF degradation
To investigate mechanisms of pathological vWF degradation with an LVAD as well as therapeutic interventions, we developed and validated a tightly controlled ex vivo model to reproduce vWF degradation from an LVAD (16). This model produced an identical profile of vWF degradation fragments as that observed in patients with a continuous-flow LVAD (8,16). Briefly, plasma samples (n = 10) placed in cryogenic tubes (Thermo Fisher Scientific, Rockford, Illinois) were exposed to a continuous cyclone spin in a laboratory vortex mixer for 4 h (Vortex-Genie 2, Scientific Industries, Bohemia, New York). At 2,400 rpm, the vortex mixer generated a shear stress of 175 dyne/cm2, which is approximately an order of magnitude greater than physiological values (17). In comparison, continuous-flow LVADs generate shear stress 1 to 2 orders of magnitude greater than physiological values (18). Paired baseline samples were gently rocked for the duration of the experiment. Samples were frozen and stored at −80°C for electrophoresis and immunoblotting.
Agarose gel electrophoresis and immunoblotting for high-molecular-weight vWF multimers
The relative sizes of high-molecular-weight plasma vWF multimers were analyzed by standard immunoblotting techniques, as previously described (19,20). Plasma samples were each combined 1:20 with loading buffer (1 mM EDTA, 10 mM Tris-HCl, 8 M urea, 2% sodium dodecyl sulfate [SDS], bromophenol blue, pH 8.0). Samples were heated in a 60°C water bath for 20 min. Samples were loaded onto 1.8% agarose-SDS gels (8.3 cm × 8 cm × 1.5 mm, 1% SDS, 0.375 M Tris-HCl, Life Technologies, Grand Island, New York). Electrophoresis was performed at 60 V for 1.75 h at 4°C in 1× Tris-acetate SDS running buffer (Life Technologies) in an XCell SureLock mini-cell electrophoresis system (Life Technologies).
Proteins were transferred for 8.5 min with the iBlot dry transfer device onto iBlot polyvinylidene difluoride gel transfer stacks (Life Technologies). Blots were blocked for 1 h in milk blocking buffer (5% dried milk powder in Tris-buffered saline [TBS], 0.05% Tween-20, Thermo Fisher Scientific). Blots were incubated with a rabbit antihuman vWF primary antibody (1:500, Dako, Carpinteria, California) overnight at 4°C in TBS-Tween/5% milk. Blots were washed in TBS-Tween and incubated with goat antirabbit immunoglobulin G HRP-conjugated secondary antibody (1:3,000, Cell Signaling, Danvers, Massachusetts) for 2 h at room temperature in TBS-Tween/5% milk. Blots were developed with Luminata Forte Western blot HRP substrate (Millipore, Billerica, Massachusetts) and imaged with an ImageQuant LAS 4000 (GE Healthcare, Piscataway, New Jersey).
Polyacrylamide gel electrophoresis and immunoblotting for vWF degradation fragments
The relative sizes of plasma vWF degradation fragments were analyzed by standard immunoblotting techniques, as previously described (20). Plasma samples were combined 1:40 in NuPAGE LDS sample buffer (Life Technologies). Samples were incubated at 70°C for 10 min and loaded onto NuPAGE 3% to 8% Tris-acetate gels (Life Technologies). Pre-stained Precision Plus Protein Kaleidoscope standard protein ladder (Bio-Rad, Hercules, California) and HighMark pre-stained standard protein ladder (Life Technologies) were loaded. Electrophoresis was performed at 150 V for 1.25 h at 4°C with 1× Tris-acetate SDS running buffer (Life Technologies). Proteins were transferred, blocked, probed for vWF, and imaged as described earlier.
A protein loading control was performed by immunoblotting for human plasma albumin. Blots were washed and blocked for 1 h at room temperature in TBS-Tween/5% milk. Blots were incubated overnight in primary goat antihuman albumin HRP-conjugated antibody (1:5,000, 4°C, Abcam, Cambridge, Massachusetts) in TBS-Tween/5% milk. Blots were washed, developed, and imaged as described earlier.
vWF degradation fragment immunoblotting analysis
Relative levels of vWF degradation fragments were quantified. ImageJ (National Institutes of Health, Bethesda, Maryland) was used to measure the chemiluminescent intensity of 11 vWF fragment bands that were identified.
Ex vivo effect of doxycycline on plasma ADAMTS-13 activity after exposure to supraphysiological shear stress
The effect of doxycycline on ex vivo plasma ADAMTS-13 activity after exposure to supraphysiological shear stress was determined with a commercial Förster resonance energy transfer (FRET) kit (American Diagnostica Inc., Stamford, Connecticut), according to manufacturer instructions. Doxycycline 5 mg/dl was added to plasma samples (n = 10). Samples were exposed to continuous supraphysiological shear stress in the vortex mixer as described earlier. Afterward, Alexa fluorochrome–conjugated recombinant vWF protein was added to plasma samples. At 37°C, proteolytic cleavage of the vWF-Alexa substrate uncoupled Alexa fluorochromes and resulted in an increase in fluorescence. The increase in fluorescence over time (maximum velocity) was measured with a spectrofluorometer. A standard curve was constructed with known plasma concentrations of ADAMTS-13. ADAMTS-13 activity was determined by interpolation of the maximum velocity values from the standard curve.
In vitro effect of doxycycline on purified ADAMTS-13 protein activity after exposure to supraphysiological shear stress
To further investigate the mechanism, the effect of doxycycline on purified ADAMTS-13 protein activity after exposure to shear stress was determined in vitro. Doxycycline 5 mg/dl and recombinant full-length human ADAMTS-13 protein (0.5 μg/ml ) were suspended in EBM-2 without blood cells or plasma proteins. FRET was performed as previously described.
Effect of doxycycline on vWF:collagen binding activity
The effect of doxycycline on ex vivo plasma vWF:collagen binding activity without and with shear stress was determined with a commercial kit (Technozym vWF:CBA ELISA, Technoclone, Vienna Austria), according to manufacturer instructions. Doxycycline (20 mg/dl) was added to plasma samples (n = 10). Samples were exposed to continuous supraphysiological shear stress in the vortex mixer as described earlier. Afterward, samples were incubated with wells coated with human collagen III. An HRP-conjugated polyclonal anti-vWF antibody was co-incubated in solution and bound the vWF–collagen III complex. Tetramethylbenzidine was added to elicit a colorimetric reaction that was quenched with a stopping solution. A standard curve was constructed from reference plasma samples with known vWF concentrations. vWF:collagen binding activity in each experimental sample was determined by interpolation of the colorimetric intensity values from the standard curve.
Platelet aggregometry was performed to determine if ADAMTS-13 inhibition activated platelets. Fresh whole blood was anticoagulated in buffered sodium citrate blood tubes (BD Biosciences). Blood was transferred to plastic cuvettes and centrifuged at 210 × g for 12 min at 22°C. The supernatant PRP fraction was aspirated into a fresh tube. Samples were exposed to continuous supraphysiological shear stress (approximately 175 dyne/cm2, 4 h) in the vortex mixer as described earlier.
Platelet aggregation was measured with a 2-channel platelet aggregometer (Chrono-Log Corporation, Haverton, Pennsylvania) (21). PRP samples were stimulated to aggregate separately with ADP (10 μM), ristocetin (4 mg/ml), and collagen (10 μg/ml). Aggregation for each agonist was measured at a fixed interval of 10 min. Percent aggregation was measured as absorbance.
GraphPad version 5.00 (Prism, GraphPad Software, Inc., La Jolla, California) was used to perform statistical analyses and plot data. A repeated-measures analysis of variance with a Tukey post-hoc test of means was performed to compare ADAMTS-13 activity with increasing doxycycline concentrations prior to shear stress. Wilcoxon matched-pairs signed-rank tests were used to compare plasma ADAMTS-13 activity with and without doxycycline, vWF degradation fragment values, and vWF:collagen binding activity between baseline versus shear stress and shear stress versus shear stress plus doxycycline. A p < 0.05 (95% CI) was considered statistically significant. Presented p values are raw and did not account for multiple comparisons. All data were presented as mean ± standard error.
Doxycycline reduced plasma and purified ADAMTS-13 activity without shear stress
A dose-response experiment demonstrated that without shear stress, doxycycline (0.6 to 10 mg/dl) significantly (analysis of variance p = 0.01) decreased ex vivo plasma ADAMTS-13 activity as determined by ELISA (0.6 mg/dl, p = 0.13; 2 mg/dl, p = 0.004; 5 mg/dl, p = 0.04; 10 mg/dl, p = 0.02). In a separate experiment, doxycycline (5 mg/dl) decreased the in vitro activity of purified recombinant ADAMTS-13 by 21%.
Supraphysiological shear stress induced LVAD-like vWF degradation
Human plasma circulated in an ex vivo model of LVAD blood trauma (16) caused extensive degradation of large vWF multimers (Figures 1A and 2A) and significantly (p < 0.01) increased 11 of 11 vWF degradation fragments (Figure 2). The same profile of vWF multimers and degradation fragments was observed as previously reported in LVAD patients (8,16).
Doxycycline reduced plasma and purified ADAMTS-13 activity with supraphysiological shear stress
After exposure to shear stress, doxycycline (5 mg/dl) significantly (p = 0.04) decreased ex vivo plasma ADAMTS-13 activity from 728 ± 27 ng/ml to 652 ± 52 ng/ml as determined by Förster resonance energy transfer (Figure 1B). In a separate experiment, doxycycline (5 mg/dl) decreased the in vitro activity of purified recombinant ADAMTS-13 by 18%.
Doxycycline protected large vWF multimers from degradation and reduced vWF degradation fragments
Reduced ADAMTS-13 activity with doxycycline protected large vWF multimers from degradation (Figures 1A and 2A). In parallel, doxycycline (5 mg/dl) decreased 10 of 11 vWF degradation fragments (Figure 2). Importantly, production of the 5 smallest vWF fragments (<460 kDa) (Figure 2B, bands A to E), which represent the smallest (least functional) vWF peptide chains, decreased significantly by 12 ± 2% (p < 0.05 for each fragment). Furthermore, ADAMTS-13 proteolysis of a single vWF monomer or the terminal monomer on a vWF multimer generates the 2 smallest vWF fragments, which are 140 and 176 kDa (22,23). These 2 fragments decreased the most, which indicated that doxycycline effectively reduced the direct end-products of ADAMTS-13 enzymatic cleavage.
Doxycycline improved vWF:Collagen binding after supraphysiological shear stress
Shear stress significantly decreased vWF:collagen binding from 1.14 ± 0.08 U/ml to 0.85 ± 0.07 U/ml (p = 0.002). Doxycycline partially reversed this effect (Figure 3A). During exposure to shear stress, doxycycline improved vWF:collagen binding from 0.85 ± 0.07 U/ml to 0.96 ± 0.09 U/ml (p = 0.004).
Doxycycline did not hyperactivate platelets
Shear stress increased platelet aggregation through the ADP pathway from 111 ± 43 %/min to 349 ± 50 %/min (p = 0.004) and modestly decreased platelet aggregation through the ristocetin-vWF pathway from 795 ± 14 %/min to 704 ± 46 %/min (p = 0.04). Doxycycline did not significantly alter platelet function through the ADP or ristocetin-vWF pathways (Figure 3B). However, during exposure to shear stress, doxycycline significantly increased platelet aggregation through the collagen pathway (p = 0.04), which was not significantly different than baseline function.
We report the first data on ADAMTS-13 inhibition with a clinically approved drug as a potential therapeutic strategy to reduce pathological vWF degradation during LVAD support. Specifically, inhibition of ADAMTS-13 by doxycycline: 1) decreased plasma and recombinant ADAMTS-13 activity without shear stress; 2) decreased plasma and recombinant ADAMTS-13 activity after exposure to supraphysiological shear stress; 3) decreased shear stress–induced degradation of large vWF multimers; 4) decreased shear stress–induced vWF degradation fragments by more than 10%; 5) increased vWF:collagen binding; 6) normalized platelet aggregation through the collagen pathway; and 7) did not hyperactivate platelets. Overall, inhibition of ADAMTS-13 by doxycycline protected vWF from degradation during supraphysiological shear stress. We speculate that doxycycline and/or ADAMTS-13 inhibition may serve a therapeutic role to decrease vWF degradation, improve vWF activity, and reduce bleeding during LVAD support. It is unlikely that this strategy will precipitate hypercoagulability and thrombosis, given our finding that doxycycline did not increase platelet aggregation above baseline.
These findings warrant further clinical investigation. Clinical studies may provide valuable data toward the development of therapy for LVAD patients with nonsurgical bleeding. If doxycycline preserves vWF activity in LVAD patients, clinical studies in which doxycycline is given during bleeding episodes and/or as maintenance therapy in patients with recurrent bleeding events may be warranted. This approach may provide an urgently needed therapeutic option for the management of bleeding during LVAD support.
Supraphysiological shear stress causes pathological vWF degradation and an acquired vWF deficiency
vWF, a 20,000-kDa multimeric glycoprotein, is assembled from 225-kDa monomers. At sites of vascular injury, turbulent blood flow and altered shear stress activate vWF multimers. As one of the initial steps of primary hemostasis, the vWF protein unravels from a globular to a linear quaternary conformation. This change in 3-dimensional structure exposes vWF binding sites for subendothelial collagen and platelets, which propagate clot formation (10). It is well established that a plasma metalloprotease, ADAMTS-13, proteolytically cleaves vWF continuously in the blood (24). During this process, large vWF multimers are degraded into small peptide fragments (22,23).
Shear stress is a powerful trigger for vWF metabolism by ADAMTS-13 (11,12). An important clinical relationship between abnormal shear stress and degradation of vWF is well described (10). In patients with critical aortic stenosis, increased shear stress through the stenotic aortic annulus reduces large vWF multimers (25). The acquired vWF deficiency produces a bleeding diathesis and gastrointestinal bleeding, or Heyde's syndrome.
Similarly, LVADs also significantly increase shear stress (18). Current continuous-flow LVADs generate shear stress 1 to 2 orders of magnitude greater than physiological values (18). As a result, during each circulatory cycle, the blood of an LVAD patient is exposed to supraphysiological shear stress. It had been speculated that the acquired von Willebrand syndrome in LVAD patients was the result of shear-induced degradation of vWF (2,3,5–7,19,26,27). Recently, we confirmed the mechanism of accelerated vWF degradation during LVAD support (8). Supraphysiological shear stress from the LVAD significantly increases vWF metabolism by ADAMTS-13. In this process, ADAMTS-13 cleaves large, biologically active vWF multimers into small vWF fragments, which exhibit reduced hemostatic function.
ADAMTS-13 inhibition as a clinical strategy to reduce LVAD-associated vWF degradation
Our current understanding of the role of ADAMTS-13 in LVAD-associated vWF degradation suggests that ADAMTS-13 inhibition may be an effective clinical strategy to reduce vWF degradation and bleeding. Multiple agents, including hemoglobin (28), interleukin-6 (29), and monoclonal antibodies (30) inhibit ADAMTS-13. Tetracycline antibiotics also inhibit matrix metalloproteinases (31), which include the ADAMTS family of proteases. Importantly, doxycycline is a known inhibitor of ADAMTS-13 (13).
Doxycycline is an inexpensive, clinically approved bacteriostatic tetracycline antibiotic. Oral doxycycline is safe and well tolerated for long-term therapy (months to years). For example, physicians frequently prescribe doxycycline for the prolonged management of inflammatory acnes (32). Therefore, it is conceptually appealing to consider that doxycycline administration in LVAD patients would be safe and cost effective and may reduce proteolytic degradation of vWF by ADAMTS-13. As an additional benefit, doxycycline kills aerobic skin flora such as Staphylococci (including methicillin-resistant Staphylococcus aureus), which cause 47% of LVAD-related infections (33).
Of note, the antimicrobial dose in humans of 2 to 4 mg/kg of oral doxycycline produces plasma levels of 0.3 to 0.5 mg/dl (34). To demonstrate a strong inhibitory effect, we performed experiments with a doxycycline concentration of 5 to 20 mg/dl, which is above the antimicrobial therapeutic concentration. The median lethal dose for oral doxycycline in rats is more than 2,000 mg/kg (35). Therefore, 5 to 20 mg/dl is likely well within the safety margin for doxycycline. Nonetheless, further investigation is needed to evaluate whether supratherapeutic doses of doxycycline may be tolerated by LVAD patients.
Balancing bleeding versus thrombosis during LVAD support
LVAD thrombosis is a potentially devastating complication of LVAD support (36). The underlying cause(s) of LVAD thrombosis in patients with therapeutic anticoagulation and antiplatelet therapy remain unclear (37). Multiple mechanisms predispose to clot formation. LVADs activate platelets (38–40) and the coagulation cascade (40–42). Hemorheological derangements such as abnormal shear stress (43), turbulence and stasis (44), diminished pulsitility (45), heat from the device (46), activation of leukocytes (47), endothelial cells (42), and the complement system (40), and generation of microparticles (48) may also promote clotting.
As a result, LVADs produce a hypercoagulable state. However, bleeding is the most frequent adverse event during LVAD support (1). This creates a challenging clinical dilemma. In patients predisposed to thrombosis, therapies to prevent bleeding (that promote clotting) may not be appropriate. It is reasonable to consider that systemic ADAMTS-13 inhibition may pose risks. Indeed, a congenital or acquired absence of ADAMTS-13 function produces thrombotic thrombocytopenia purpura (TTP), a clotting diathesis in which unmetabolized vWF accumulates, activates platelets, and triggers microthrombosis (49). As such, ADAMTS-13 inhibition should be performed with caution.
It is unknown how much ADAMTS-13 inhibition may be safely tolerated in LVAD patients. For example, 83 ± 8% inhibition of ADAMTS-13 activity with an antihuman ADAMTS-13 monoclonal antibody was recently reported in an LVAD mock circulatory loop (30). The investigators suggested that this approach may reinstate normal vWF levels and prevent bleeding. However, it is possible that such widespread inhibition of ADAMTS-13 may inadvertently provoke a hypercoagulable, TTP-like state and thrombosis. Conservative inhibition of ADAMTS-13 may be a safer strategy in LVAD patients. We demonstrated modest ADAMTS-13 inhibition of approximately 10%, which did not provoke platelet aggregation and is unlikely to precipitate thrombosis. Further investigation of the safety threshold for ADAMTS-13 inhibition is warranted.
Notably, we recently demonstrated that supraphysiological shear stress from the LVAD itself causes mechanical demolition of vWF multimers. During this process, vWF monomers are physically ripped apart by shear stress (8). This nonproteolytic mechanism of vWF degradation is independent of ADAMTS-13. Importantly, doxycycline therapy likely would not affect this parallel mechanism of vWF degradation. This point may be favorable for therapy to protect against TTP. In the setting of ADAMTS-13 inhibition, nonproteolytic vWF degradation from LVAD shear stress itself may protect against thrombogenesis from accumulation of large vWF multimers.
Nonsurgical bleeding is the most common complication of continuous-flow LVAD therapy. vWF plays a major role. Cleavage of vWF by ADAMTS-13 is the major mechanism of vWF degradation. Limited treatment options exist. Our data with human blood and recombinant ADAMTS-13 protease demonstrated that ADAMTS-13 inhibition protected vWF from degradation and improved vWF binding activity. This was achieved with doxycycline, an inexpensive, clinically available drug. These findings warrant further clinical investigation. ADAMTS-13 inhibition should be considered as a potential therapeutic strategy during bleeding episodes and/or as maintenance therapy in LVAD patients with recurrent bleeding events.
COMPETENCY IN MEDICAL KNOWLEDGE: Nonsurgical bleeding is the most frequent complication of LVAD support. Supraphysiological shear stress from the LVAD accelerates vWF degradation by ADAMTS-13 (the vWF protease), which causes an acquired vWF deficiency and a bleeding diathesis. Inhibition of ADAMTS-13 with doxycycline reduced vWF metabolism and improved vWF function during supraphysiological shear stress.
TRANSLATIONAL OUTLOOK: ADAMTS-13 is a clinical target to reduce pathological vWF degradation during LVAD support. ADAMTS-13 inhibition should be investigated clinically as a potential therapeutic strategy during bleeding episodes and/or as maintenance therapy in LVAD patients with recurrent bleeding events.
This project was performed with grant support from the University of Pennsylvania Clinical Translational Research Center. The authors acknowledge and thank George Hung, BS, and Alen Trubelja, BS, for assistance. Robert Dowling, MD, and Sujith Dassanayaka, MS, provided valuable feedback during the project and drafting of the manuscript.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13
- enzyme-linked immunosorbent assay
- Förster resonance energy transfer
- horseradish peroxidase
- left ventricular assist device
- sodium dodecyl sulfate
- Tris-buffered saline
- thrombotic thrombocytopenia purpura
- von Willebrand factor
- Received February 20, 2015.
- Revision received June 8, 2015.
- Accepted June 12, 2015.
- American College of Cardiology Foundation
- Meyer A.L.,
- Malehsa D.,
- Bara C.,
- et al.
- Uriel N.,
- Pak S.W.,
- Jorde U.P.,
- et al.
- Klovaite J.,
- Gustafsson F.,
- Mortensen S.A.,
- Sander K.,
- Nielsen L.B.
- Geisen U.,
- Heilmann C.,
- Beyersdorf F.,
- et al.
- Zhang X.,
- Halvorsen K.,
- Zhang C.Z.,
- Wong W.P.,
- Springer T.A.
- Tsai H.M.,
- Sussman I.I.,
- Nagel R.L.
- Tsai H.M.,
- Sussman I.I.,
- Ginsburg D.,
- Lankhof H.,
- Sixma J.J.,
- Nagel R.L.
- Kokame K.,
- Matsumoto M.,
- Fujimura Y.,
- Miyata T.
- ↵ACTIFLUOR ADAMTS-13 activity assay. Stamford, CT: American Diagnostica.
- Restle D.J.,
- Zhang D.M.,
- Hung G.,
- et al.
- Turner N.A.,
- Nolasco L.,
- Ruggeri Z.M.,
- Moake J.L.
- Crawley J.T.,
- de Groot R.,
- Xiang Y.,
- Luken B.M.,
- Lane D.A.
- Heilmann C.,
- Geisen U.,
- Beyersdorf F.,
- et al.
- Studt J.D.,
- Kremer Hovinga J.A.,
- Antoine G.,
- et al.
- Bernardo A.,
- Ball C.,
- Nolasco L.,
- Moake J.F.,
- Dong J.F.
- Gordon R.J.,
- Weinberg A.D.,
- Pagani F.D.,
- et al.
- Holmes N.E.,
- Charles G.P.
- ↵(2008) Material safety data sheet: doxycycline hyclate. Version 4.0 (Pfizer Inc. New York, NY).
- Wever-Pinzon O.,
- Selzman C.H.,
- Drakos S.G.,
- et al.
- Ledford I.D.,
- Labedi M.,
- Kfoury A.G.,
- et al.
- Diehl P.,
- Aleker M.,
- Helbing T.,
- et al.