During the past few decades, many lives have been saved by ventricular assist devices (VADs).1–5 Advances in VAD technology have led to the creation of nonpulsatile, continuous-flow (CF) rotary pumps, which are smaller, more durable, and easier to implant than older, pulsatile-flow devices.6 With CF-VADs, the survival rate is better than with pulsatile VADs.5 However, despite these improvements, many CF-VAD recipients develop gastrointestinal (GI) bleeding resulting from von Willebrand disease (vWD) acquired after device implantation.6–9 Demirozu et al.8 suggested that the impeller mechanism in CF-VADs causes von Willebrand factor (vWF) deformation, which reduces vWF multimer levels in the blood and thus leads to vWD.
In addition, CF-VAD implantation is expensive, accounting for approximately $141,000 per case, including the cost of the device (approximately $77,000).5 The high cost of these newer devices is problematic for heart failure patients in economically disadvantaged areas. Therefore, the availability of an easily implantable, cost-effective, and reliable VAD would make life-saving treatments more widely available to such patients.10
To provide a simple, economical alternative to the VADs currently available for use in Mexico, we are developing the Vitalmex Extracorporeal VAD-Pneumatic (EVAD-P; Vitalmex Internacional, SA de CV, Mexico City, Mexico). Compared with the commercially available paracorporeal centrifugal VAD systems, the EVAD-P device is approximately 50–60% cheaper, and the console is 30–40% less expensive. In addition to being economically affordable, the Vitalmex EVAD-P is pulsatile. Therefore, it should be associated with a lower incidence of bleeding complications than are CF-VADs, further reducing medical costs and postimplantation interventions.
In vitro studies of the EVAD-P were previously performed to study the functionality of the device under different physiological conditions, demonstrate that it produces adequate output levels, determine a more effective valve design, and test the device’s hemodynamic performance.11–13 In both in vitro experiments and computer-based simulations, a trileaflet valve was shown to work more efficiently than a mono-leaflet or bileaflet one, because of the trileaflet valve’s lower hydrostatic resistance.12 When hemodynamic performance was tested, the EVAD-P provided adequate flow rates and pressures to maintain physiological parameters within their normal ranges.13
The initial in vivo evaluation of the device (then known as the Innovamedica VAD) was reported in 2012 by Tuzun et al.,10 who validated the pump’s short-term (30 ± 5 day) effectiveness in sheep. The results showed that the device is easy to implant, provides adequate hemodynamic support, and can safely provide 30 days of support.10 The current study confirms and extends the previous findings, showing that the EVAD-P can provide up to 90 ± 5 days of mechanical circulatory support (MCS).
Materials and Methods
Sixteen healthy Suffolk crossbreed sheep, each weighing 55–75 kg, were enrolled in this study. All animals were handled in accordance with the requirements of the Laboratory Animal Welfare Act (Public Law [PL] 89–544) and its 1970 (PL 91–579), 1976 (PL 94–279), and 1985 (PL 99–198) amendments. Our Institutional Animal Care and Use Committee approved all protocols used.
Two sheep were excluded from the study: one in which cardiac fibrillation caused death before device implantation, and one in which surgical complications unrelated to the device occurred during implantation. Each of the remaining 14 sheep was assigned to an experimental group. In group I (n = 7), the EVAD-P was evaluated for a goal of 30 days (±5 days) to confirm the feasibility of this MCS duration; in group II (n = 2), the device was evaluated for an intended study duration of 90 ± 5 days; and in group III (n = 5), a modified version of the device was also studied for 90 ± 5 days.
The EVAD-P system used in groups I and II has been previously described by Tuzun et al.10 Both sheep in group II experienced an unexpected device shutdown, which indirectly resulted in the demise of one animal; this led us to modify the EVAD-P system before testing began in group III. We modified several aspects of the inflow cannula (Figure 1): the tip of the cannula was changed from a 90° angle to a 45° angle to allow more ergonomic eventual placement in humans; the aluminum insert at the tip of the cannula was replaced with a stainless steel insert to reduce the risk of thrombosis; and the eyes on the suction tip were enlarged to ensure that the main opening and the eyes remained unobstructed at high flow rates. In addition, we reduced the mean beat rate from 90 to 56 bpm and increased the mean SV from approximately 45 to 58 ml; these changes allowed the pump to fill and empty completely with each beat and resulted in an optimal flow rate. Additional changes included modifying the console and revising the controller software. This involved increasing the maintenance period, adding an electrostatic discharge protection circuit to reduce the risk of system freezes, and improving the useful lifetime of the spring actuator. Figure 2 shows the current console and portable system cart.
All animals were premedicated and subjected to general anesthesia. The surgical procedure was identical to that previously reported by Tuzun et al.10
Postoperative Care and Data Collection
After surgery, each sheep received a heparin drip (25,000 units/250 ml D5W) as necessary to maintain the activated clotting time above baseline. When the international normalized ratio (INR) reached an acceptable level, the heparin drip was replaced by warfarin administration (2.5–20 mg intravenous [IV]/oral, once/day), which was continued in an effort to keep the INR between 2.0 and 3.0. In addition, beginning on postoperative day (POD) 3, group III sheep received antiplatelet therapy in the form of dipyridamole (50 mg oral, twice/day) and aspirin (325 mg oral, once/day).
Throughout the study, the heart rate, pump beat rate, and flow rate were recorded at hourly intervals. Blood specimens—for assessing chemistry, hematology, and coagulation profiles and plasma-free hemoglobin (PFH) levels—were obtained at least weekly. Urine output and body temperature were monitored routinely.
Macroscopic and Microscopic Explant Analyses
All 13 sheep whose lives were electively terminated were fully heparinized before being euthanized. The extracorporeal (pump and cannula) and intracorporeal (cannula) portions of the EVAD-P were removed, rinsed with sterile saline, and photographed. The pump circuit was removed and inspected for evidence of fibrin formation and gross thrombus. All pump and cannula components were placed in formalin for histologic/microscopic analysis. Each cannulation site was examined and graded with respect to endothelial or other related injury. The 14 animals in which device implantation was successful had a complete necropsy, including examination and weighing of major end organs. Tissue specimens from organs of interest were embedded in paraffin wax blocks for slide preparation. Selected microscopic histopathology studies were performed, providing extensive microscopic evaluation of major organ systems.
Descriptive Statistical Analyses
When calculating the results, we included only data from the 11 sheep that survived for ≥10 days. Hemodynamic data are presented as the mean value ± standard deviation (SD) for the daily average of the hourly recordings obtained from each animal. Hematology and biochemistry results are presented as the mean ± SD of the values obtained from each animal at various predetermined time points throughout the study.
Group I Animals (n = 7)
Two of the seven sheep in group I underwent early termination of the study (Table 1). One animal was euthanized because of a celiac thrombus, and the other was euthanized as a result of intractable sepsis (final blood cultures confirmed Enterobacter cloacae infection) not related to the device. In the five surviving sheep, a fixed beat rate of 90 bpm was maintained, with a SV of 45.2 ± 2.4 ml (blood flow, 4.1 ± 0.3 L/min). None of the surviving animals had anorexia, infection, or neurologic deficits.
Group II Animals (n = 2)
As shown in Table 1, neither of the two sheep in group II survived to study completion. The first sheep died on POD 28 of a chronic infection that led to acute septic shock, resulting in metabolic acidosis, ventricular fibrillation, and ultimately a cardiac arrest while the sheep was sedated with general propofol anesthesia and receiving mechanical ventilation. Postmortem examination revealed that the acute septic shock was a result of multiple intravascular bacterial thrombi on the EVAD-P, which may have been caused by a console shutdown that occurred 3 days before the animal’s death. The second sheep was electively euthanized on POD 49. On POD 33, the console had malfunctioned and been exchanged. Despite IV administration of heparin, the flow began to decrease within 24 hours after the console exchange. Although the sheep was in good health, it was eventually euthanized because an occlusive thrombus in the inflow cannula substantially reduced the flow.
Group III Animals (n = 5)
In group III, three sheep were euthanized because of morbidities not related to the EVAD-P: a right tibia/fibula accidental dislocation, an intractable coagulopathy that occurred within the initial recovery period after implantation, and an intractable GI hemorrhage caused by an abomasal ulcer (Table 1). The other two sheep survived for 91 and 93 days, respectively, at a fixed beat rate of 56 bpm, with an SV of 58.0 ± 2.3 ml and a flow of 3.5 ± 0.2 L/min (Figure 3). Neither surviving animal had anorexia or neurologic deficits.
Hemodynamic Performance (All Animals)
In the immediate postoperative period, the heart rate, arterial pressure, and arterial blood gas measurements indicated adequate cardiac function with left ventricular support. In the 11 sheep that survived for ≥10 days, the pump flow was 3.9 ± 0.3 L/min; the SV was 45.6 ± 1.6 ml (flow, 4.0 ± 0.3 L/min) for group I, 41.4 ± 2.9 ml (flow, 3.6 ± 0.1 L/min) for group II, and 58.0 ± 2.3 ml (flow, 3.5 ± 0.2 L/min) for group III.
In three group I cases and two group II cases, the console had to be replaced because of unexpected system shutdown related to a malfunctioning console. During these events, alarms notified the caretakers in a timely fashion. They manually pumped the EVAD-P while the malfunctioning console was removed. The consoles were exchanged without incident. Whereas there was no effect on the group I animals, the system shutdowns may have contributed to the early deaths of both animals in group II. Because of the subsequent modifications made to the console, similar shutdowns were not observed in group III.
Hematology and Biochemistry Summary
All hematologic results were provided by Equine Laboratories (Houston, TX). Postoperative serum levels of blood urea nitrogen (BUN) (15.6 ± 6.7 mg/dl) and creatinine (1.2 ± 0.6 mg/dl) indicated normal renal function (Figure 4). As seen in Figure 5, the average postoperative concentration of PFH (10.3 ± 7.6 mg/dl) was greater than the average baseline value (4.0 ± 1.9 mg/dl). The INR was maintained at 2.3 ± 0.5 throughout the study. Postoperatively, fibrinogen levels (381.3 ± 183.9 mg/dl) and hematocrit (23.9 ± 4.5%) were below the normal ovine range. Even though the postoperative hemoglobin level (10.4 ± 1.6 g/dl) was greater than the baseline level (9.1 ± 1.7 g/dl), the results remained within the normal range. Temporary increases in the levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatine kinase (CK), and lactate dehydrogenase (LDH) were observed during the week after surgery (Figure 6). These values gradually returned to their usual ranges within the following week, indicating normal hepatic function. All other hematologic and biochemical values were within their customary ranges throughout the study.
Macroscopic and Microscopic Histopathologic Findings in Animals Surviving for ≥10 Days
Device and cannulas
Multiple acute thrombi were found on the cannulas used in the two group II sheep. In two of the seven surviving group I animals and both of the surviving group III animals, small thrombi were present on the pump bladder valves (Figure 7), but the valves continued to function normally.
In the seven sheep that tolerated the device to term, all major organs were unremarkable. In four of these animals, postmortem examination of the explanted organs revealed small infarcted areas in the kidneys, as well as thromboembolic incidents, but these were not clinically significant. Normal hematologic findings indicated no adverse effects on end-organ function. The tissue changes observed during necropsy were not excessive, being within the expected degree of change for this type of device implantation.
After confirming the feasibility of using the EVAD-P system for MCS in healthy sheep for 30 days, we extended the study duration to 90 ± 5 days to evaluate the device’s potential for providing longer MCS support. The results showed that the EVAD-P was safe and provided typical clinical ranges of hemodynamic support in a nondiseased sheep model for up to 93 days.
In many cases, heart transplantation is the most effective treatment for end-stage heart disease; however, because of the long waiting period for a donor organ and the limited number of donors, many patients are left without treatment options.14 In the past several decades, mechanical circulatory devices have become widely used for patients with end-stage heart failure. Despite the success of CF-VADs in providing safe and long-term MCS, the current high cost of this technology is problematic, especially for patients in economically disadvantaged countries. The Vitalmex EVAD-P is intended to offer the simplicity and functionality of previous pneumatic pulsatile models while incorporating features designed to reduce costs.10 This device is extracorporeal and made of clear polycarbonate (Figure 8), allowing caretakers to routinely inspect the system for thrombus development and, using a sterile technique, to replace either the shell or the bladder without interrupting MCS for more than a short time. Because the device consists of reusable elements that can be replaced without surgical intervention, it should be marketable and economical in poorer countries where current alternative devices would be unaffordable.
The results of our study are encouraging, showing that the EVAD-P can provide safe, pulsatile MCS for up to 93 days. The BUN and creatinine levels observed in our sheep suggest that their renal function was not adversely affected. Although there was a slight increase in the BUN levels after 30 days (Figure 4), this increase did not affect renal function and was probably due to mild dehydration, because the animals were regulating their own fluid status at that point. The increase in the levels of ALT, AST, CK, and LDH observed after the implantation procedure is commonly seen after other thoracic operations, because liver function is affected by anesthetic drugs, intraoperative tissue damage, surgical duration, and blood transfusion (if required).15 Generally, these variables return to their normal ranges during the first postoperative week, as seen in the current study. Whereas the mean postoperative PFH levels in our study were slightly higher than at baseline, they were considered within the normal range, indicating negligible hemolysis (Figure 5). The number of surviving sheep started to decrease on POD 10 and, by 30 days, only two survivors were left (Table 1). As a result, the SDs shown in Figures 4–6 provided little predictability of the population variance and, therefore, prompted our recommendation for more 90 day studies. The main purpose of these graphs is to show that the levels were within acceptable ranges and were typical of similar VAD studies. The hemoglobin levels and hematocrit were maintained close to baseline with proper fluid replacement therapy. Although our necropsy findings included thrombus formation and renal infarcts, more than 30 years of experience with VAD testing in ruminants has shown us that renal infarcts occur in almost all of these animals, regardless of the type of device tested. Thus, the presence of these kidney infarcts did not elicit concern about the observed renal thromboembolic events.
Thrombus formation can be a serious complication associated with VADs. Because the VAD surfaces are in contact with blood, there is a high chance that thrombi will develop and obstruct normal blood flow.16 To reduce thrombogenicity, anticoagulants are routinely given after device implantation. Despite such therapy in our study, thrombus formation was associated with the early deaths of one animal each in groups I and II. These deaths prompted us to modify the EVAD-P system in an effort to minimize this complication. By changing the shape of the inflow cannula, the beat rate from 90 to 56 bpm, and the SV from approximately 45 to 58 ml, we believe that we decreased the risk of thrombogenesis. The larger SV and the decreased beat rate allowed the device to fill and empty completely with each beat, providing better washing and minimizing thrombus formation. In addition, the anticoagulation regimen was revised to include antiplatelet therapy. The effectiveness of these modifications was confirmed by the successful results seen in group III.
Whenever anticoagulation regimens are needed, meticulous observation is required to prevent thromboembolic and bleeding events.17 Compared with nonpulsatile VADs, pulsatile VADs involve significantly fewer GI bleeding complications.6–9 Because pulsatile-flow devices mimic the natural cardiac physiology, there is a reduced risk of excessive GI bleeding.7 Crow et al.7 found that the rate of GI bleeding events was nearly 10 times higher in patients with nonpulsatile pumps than in patients with pulsatile devices. Recent studies have linked these bleeding complications to acquired vWD, a disease that affects the blood’s ability to clot. Uriel et al.6 found that 44.3% of patients receiving nonpulsatile VAD support had major postoperative GI bleeding. All of those patients had absent or reduced high-molecular-weight vWF levels, pointing to a diagnosis of acquired vWD.6 However, Meyer et al.9 found that after the nonpulsatile VAD was removed, the vWF levels returned to normal, suggesting that excess bleeding is not due to anticoagulation therapy but, rather, is associated with implantation of a nonpulsatile device.
The advantages of the EVAD-P—cost-effectiveness, pulsatility, system simplicity, reduced risk of bleeding complications, and ability to visually monitor for thrombus formation—support the continued evaluation of this device. The EVAD-P has the potential to be a bridge to transplantation or to provide destination therapy for heart failure patients in economically disadvantaged countries, who would otherwise have no MCS options. Our study provides initial evidence that the EVAD-P can offer hemodynamic support without causing adverse effects on end-organ function, indicating that the device is functionally robust in the long term. In animal studies of VAD systems, adverse events unrelated to the device sometimes occur. Because of such adverse events, only two of our group III animals successfully reached the study end-point. Although the results from these two animals are promising, additional studies are recommended to further confirm the long-term safety of this device.
1. Frazier OH, Benedict CR, Radovancevic B, et al.Improved left ventricular function after chronic left ventricular unloading.Ann Thorac Surg199662675681discussion 681
2. Goldstein DJ, Oz MC, Rose EAImplantable left ventricular assist devices.N Engl J Med199833915221533
3. Nakatani S, McCarthy PM, Kottke-Marchant K, et al.Left ventricular echocardiographic and histologic changes: Impact of chronic unloading by an implantable ventricular assist device.J Am Coll Cardiol199627894901
4. Poirier VLLVADs – a new era in patient care.J Cardiovasc Manag2000112634
5. Neyt M, Van den Bruel A, Smit Y, et al.Cost-effectiveness of continuous-flow left ventricular assist devices.Int J Technol Assess Health Care201329254260
6. Uriel N, Pak SW, Jorde UP, et al.Acquired von Willebrand syndrome after continuous-flow mechanical device support contributes to a high prevalence of bleeding during long-term support and at the time of transplantation.J Am Coll Cardiol20105612071213
7. Crow S, John R, Boyle A, et al.Gastrointestinal bleeding rates in recipients of nonpulsatile and pulsatile left ventricular assist devices.J Thorac Cardiovasc Surg2009137208215
8. Demirozu ZT, Radovancevic R, Hochman LF, et al.Arteriovenous malformation and gastrointestinal bleeding in patients with the HeartMate II left ventricular assist device.J Heart Lung Transplant201130849853
9. Meyer AL, Malehsa D, Bara C, et al.Acquired von Willebrand syndrome in patients with an axial flow left ventricular assist device.Circ Heart Fail20103675681
10. Tuzun E, Winkler JA, Contreras AL, Sacristan E, Cohn WEIn vivo
performance evaluation of the Innovamedica pneumatic ventricular assist device.ASAIO J20125898102
11. Sacristán E, Corona F, Suárez B, et al.Development of a universal second generation pneumatic ventricular assist device and drive unit.Engineering in Medicine and Biology Society, 2003. Proceedings of the 25th Annual International Conference of the IEEE20031427430
12. Escobedo C, Tovar F, Suarez B, Hernández A, Corona F, Sacristán EExperimental and computer-based performance analysis of two elastomer VAD valve designs20057Engineering in Medicine and Biology Society, 2005. Proceedings of the 27th Annual International Conference of the IEEE76207623
13. Corona F, Sacristán E, Barragan R, et al.Hemodynamic performance in-vivo of a new ventricular assist device2005in 27th Annual IEEE-EMBS Conference (Vol. 25, No. 4)
14. DeRose JJ Jr, Umana JP, Argenziano M, et al.Implantable left ventricular assist devices provide an excellent outpatient bridge to transplantation
and recovery.J Am Coll Cardiol19973017731777
15. Hojjati M, Ghofrani M, Vala’i NA quasi-experimental study on the effect of upper gastrointestinal surgery on liver function tests.Hepatogastroenterology19984517021705
16. John R, Kamdar F, Liao K, et al.Low thromboembolic risk for patients with the Heartmate II left ventricular assist device.J Thorac Cardiovasc Surg200813613181323
17. Kurien S, Hughes KAAnticoagulation and bleeding in patients with ventricular assist devices: Walking the tightrope.AACN Adv Crit Care2012239198