Hemodynamic effects of direct biventricular compression studied in isovolumic and ejecting isolated canine hearts.

BACKGROUND: Biventricular direct cardiac compression (DCC) can potentially support the failing heart without the complications associated with a blood/device interface. The effect of uniform DCC on left and right ventricular performance was evaluated in 7 isolated canine heart preparations. METHODS AND RESULTS: A computer-controlled afterload system either constrained the isolated heart to contract isovolumically or simulated hemodynamic properties of physiological ejection. Biventricular DCC was provided by a chamber surrounding the heart that allowed adjustment of the compression pressure, onset time, and duration. Through a series of ventricular preloads, the effect of DCC on the end-systolic pressure-volume relationship (ESPVR) was evaluated under isovolumic and ejecting conditions. Under both conditions, DCC shifted the ESPVR of the left and right ventricles upward by an amount approximately equal to the compression pressure. The augmentation of end-systolic pressure for each initial preload tested, however, was less under ejecting conditions, because reductions in end-systolic and end-diastolic volumes occurred with ejection. Nevertheless, the net effect was to increase stroke volume. Measurement of M&f1;O2 demonstrated that at a given ventricular volume, M&f1;O2 did not change with DCC; however, peak ventricular pressure increased substantially, so that the effective pressure-volume area increased. CONCLUSIONS: Biventricular DCC can augment end-systolic pressure with no added costs of M&f1;O2. Under ejecting conditions, this augmentation of ventricular contracting ability manifests as increases in stroke volume. Thus, DCC represents a feasible alternative form of ventricular assist, and devices that support the heart in this manner should be further explored.

Malignant ventricular arrhythmias are well tolerated in patients receiving long-term left ventricular assist devices.

OBJECTIVES: We sought to quantitate the incidence of malignant ventricular arrhythmias and to identify subsequent hemodynamic changes and untoward events in patients who have received an implantable left ventricular circulatory assist device as an extended bridge to heart transplantation. BACKGROUND: Implantable long-term mechanical circulatory assist devices have been used clinically with increasing frequency and success for the past 4 years. Previous investigators have suggested that patients with malignant ventricular arrhythmias receiving a left ventricular assist device will require both left and right ventricular assistance to maintain vital organ perfusion. METHODS: We reviewed our 4-year experience with 21 patients who underwent implantation of a left ventricular assist device. Device flows and mean arterial pressure were used to assess systemic perfusion; central venous pressure provided a gauge of right ventricular function. Charts were screened for evidence of end-organ injury resulting from malignant ventricular arrhythmias. RESULTS: Malignant ventricular arrhythmias occurred in 4 patients (19%) before device placement and in 9 patients (43%) during device support. The latter nine patients formed the final study group; their arrhythmias occurred 0 to 186 days after device implantation and had a duration of 10 min to 12 days. The patients reported weakness or palpitation; however, none reported syncope or dyspnea. Mean arterial pressure and central venous pressure were insignificantly changed by the arrhythmias. Device flow decreased by 1.4 +/- 0.6 liters/min (p < 0.05) at the onset of the arrhythmias but returned to normal after cardioversion. No thromboembolic events or significant end-organ dysfunction occurred. CONCLUSION: Absence of right ventricular contraction during malignant ventricular arrhythmias is well tolerated in recipients of a left ventricular assist device. The diagnosis of malignant arrhythmia should be suspected if an unexplained decrease in left ventricular assist device flow occurs. Early electrical cardioversion is warranted to avoid both thrombus formation in the native heart and right ventricular myocardial injury from prolonged fibrillation.

Implantable left ventricular assist devices provide an excellent outpatient bridge to transplantation and recovery.

OBJECTIVES: Our recent experience with outpatient left ventricular assist device (LVAD) support is presented to demonstrate the possibilities and limitations of long-term outpatient mechanical circulatory assistance. BACKGROUND: The experience with inpatient LVAD support as a bridge to transplantation has proved the efficacy of such therapy in improving circulatory hemodynamic status, restoring normal end-organ function and facilitating patient rehabilitation. With miniaturization of the power supplies and controllers, such mechanical circulatory support can now be accomplished in an outpatient setting. METHODS: Between March 1993 and February 1997, 32 patients (26 male, 6 female, mean [+/-SEM] age 49 +/- 15 years) underwent implantation of the ThermoCardiosystems (TCI) Heartmate vented electric (VE) LVAD. The VE LVAD is powered by batteries worn on shoulder holsters and is operated by a belt-mounted system controller, allowing unrestricted patient ambulation and hospital discharge. RESULTS: Mean duration of support was 122 +/- 26 days (range 3 to 605), with a survival rate to transplantation or explantation of 78%. Nineteen patients were discharged from the hospital on mean postoperative day 41 +/- 4 (range 17 to 68), for an outpatient support time of 108 +/- 30 days (range 2 to 466). Four patients underwent early transplantation and could not participate in the discharge program, and three patients currently await discharge. The complication rate was not statistically different from that encountered in our previous 52 patients with a pneumatic LVAD. CONCLUSIONS: Outpatient LVAD support is safe and provides improved quality of life for patients awaiting transplantation. Wearable and totally implantable LVADs should be studied as permanent treatment options for patients who are not candidates for heart transplantation.

Use of explanted human hearts as a model for the study of cardiac pathophysiologic conditions.

The purpose of this paper is to describe an isolated heart model that uses human hearts to study cardiomyopathy. Techniques of organ preparation and perfusion are described that result in successful restoration of explanted human hearts to a beating condition. Native hearts of transplant recipients were placed on an isolated perfusion circuit immediately after explant. After appropriate organ and circuit preparation these hearts were restored to a functional state. Studies were done to assess the stability over time and response to inotropic stimulation. Six of seven hearts were returned to a functional condition. Left ventricular pressure generation ranged from 56 to 118 mm Hg (mean 84.8 +/- 34.11) at physiologic loading conditions. Hearts remained functional from 67 to 271 minutes (mean 152 +/- 71.7) and retained up to 70% of functional capacity after 120 minutes. Hearts performed in isovolumic and ejecting modes. Hearts had a marked response to inotrope administration not previously described. We conclude that human hearts can be reproducibly restored to a functional condition after explant from transplant recipients and can be maintained in a beating state with reasonably stable pressure generation for an extended period of time, which makes this a useful model with which to study cardiac pathophysiologic conditions. These hearts demonstrate an appropriate response to inotropes not previously observed, most likely because of improved myocardial preservation and stringent control of perfusate chemical make-up.

Miniature axial flow pump for ventricular assistance in children and small adults.

We investigated the efficacy of the Jarvik 2000 intraventricular assist device (Jarvik Research, Inc., New York, N.Y.) in an ovine model. The device is an axial flow pump measuring 1.8 cm in diameter by 5 cm long, has a displacement volume of 12 ml, and can deliver flow from 2 to 7 L/min. Seven devices were implanted through a left thoracotomy into the left ventricle with an outflow graft to the descending aorta. Animals were treated with warfarin sodium and aspirin to maintain prothrombin times approximately 1.5 times control. Animals were followed up for 3 to 123 days. Two animals died of operative complications at days 3 and 5. One device failed at 58 days because of thrombus formation at the inflow side of the impeller. The remaining four animals were killed at days 19, 42, 42, and 123, respectively, because of broken electric power cables. Hematocrit values rose significantly higher than preoperative levels (22.8% +/- 3.8% to 30.5% +/- 3.4%); premortem elevations of values higher than baseline values of plasma free hemoglobin (10.4 +/- 7.8 mg/dl to 17.1 +/- 7.4 mg/dl) and lactate dehydrogenase (391.5 +/- 113.7 units/L to 771.2 +/- 370.8 units/L) were statistically insignificant. Serum creatinine and bilirubin levels were normal. No end-organ dysfunction arising from long-term support was evident clinically or at postmortem examination, nor was there any evidence of embolism or damage to intracardiac structures. We found the Jarvik 2000 intraventricular assist device to be easily implantable, safe, nonhemolytic, and able to provide physiologic flow with power requirements under 10 watts.

Systolic ventricular interaction in normal and diseased explanted human hearts.

OBJECTIVE: The purpose of this study was to quantify the magnitude of interaction between the right and left ventricles in conditions of heart failure. METHODS: Human hearts were taken from transplant recipients diagnosed with diluted cardiomyopathy at the time of transplantation and were restored to beating condition with use of an isolated perfusion circuit. Left ventricular-right ventricular interaction was determined by ramping volume in the left ventricle while holding right ventricular volume constant. Right ventricular pressure gain was plotted against left ventricular pressure and the slope of the linear regression determined the left ventricular-right ventricular interaction. A similar procedure was used to determine right ventricular-left ventricular interaction. Two normal hearts were obtained from transplant donors not suitable for cardiac donation to serve as control hearts. RESULTS: Mean left ventricular-right ventricular interaction was 0.22 in the hearts with dilated cardiomyopathy compared with 0.06 in the control hearts. Mean right ventricular-left ventricular interaction was 0.14 in the hearts with dilated cardiomyopathy compared with 0.09 in the control hearts. A marked increase in left ventricular-right ventricular interaction was noted in the hearts with dilated cardiomyopathy compared with control hearts. Although observed values of right ventricular-left ventricular interaction also correspond to previously published results, no significant increase was observed in the dilated cardiomyopathy condition. CONCLUSIONS: These studies confirm previously published values for systolic ventricular interaction obtained with animal models and demonstrate a marked increase in the dependence of the right ventricle on left ventricular function to maintain systolic pressure generation during conditions of dilated cardiomyopathy.

Reevaluating the significance of pulmonary hypertension before cardiac transplantation: determination of optimal thresholds and quantification of the effect of reversibility on perioperative mortality.

OBJECTIVES: Right-sided circulatory failure resulting from severe preoperative pulmonary hypertension is a source of mortality early after cardiac transplantation. We undertook the present study (1) to analyze the association of elevated pulmonary hemodynamic indices with 30-day mortality, (2) to define threshold ranges associated with an increase in 30-day mortality, and (3) to evaluate the effect of vasodilator reversibility on 30-day mortality. METHODS: Preoperative hemodynamic profiles were evaluated in 476 patients who ultimately underwent cardiac transplantation. From these data, receiver-operating characteristic curves and stratum-specific likelihood ratios were generated to compare the efficacy of each hemodynamic index. A subset of patients with elevated hemodynamic profiles at baseline additionally underwent graded sodium nitroprusside infusion. RESULTS: Analysis of receiver-operating characteristic curves demonstrated no statistically significant difference among the indices in their ability to predict 30-day mortality. Analysis of stratum-specific likelihood ratios demonstrated three risk strata that correlated with significant differences in 30-day mortality, with patients in the high-risk stratum having a 3.2 to 4.4 increase in odds of 30-day mortality when compared with patients in the low-risk stratum. Nitroprusside data demonstrated that although 30-day mortality was better in patients with reversible pulmonary hypertension than in those with fixed pulmonary hypertension, it was not comparable with that of patients without pulmonary hypertension at baseline. CONCLUSIONS: Candidates for cardiac transplantation may be categorized into three risk strata on the basis of their preoperative pulmonary hemodynamic profile; the association of this profile with 30-day mortality is not linear. Reversibility with nitroprusside appears to confer some improvement in the risk of 30-day mortality, but it may not eliminate the risk entirely.

Delayed reversal of impaired metabolic vasodilation in patients with end-stage heart failure during long-term circulatory support with a left ventricular assist device.

BACKGROUND: Whether increased cardiac output during chronic circulatory support with a left ventricular assist device (LVAD) is associated with improved metabolic vasodilation in the peripheral circulation of patients with congestive heart failure is unknown. METHODS: Forearm blood flow, determined by venous occlusion plethysmography, mean arterial pressure, and cardiac output were measured at rest and after 5 minutes of arterial occlusion (a maximal metabolic vasodilatory stimulus) in 14 patients with severe heart failure before LVAD implantation, and in the early (<4 weeks) and late (8 to 12 weeks) postoperative recovery phases after LVAD implantation. Nine normal subjects served as controls. Vascular conductance was calculated as the ratio of forearm blood flow and mean arterial pressure. RESULTS: Mean arterial pressure and cardiac output increased to normal values in the early and late recovery phases after LVAD implantation. Resting forearm blood flow and vascular conductance were similar to normal subjects in the early and late recovery phases after LVAD implantation. Peak forearm blood flow and vascular conductance were significantly less than control subjects in the early preoperative recovery phase (p < 0.05) but were similar to control subjects in the late postoperative recovery phase after LVAD implantation. CONCLUSIONS: In spite of early normalization of cardiac output, mean arterial pressure, and resting forearm blood flow during chronic circulatory support with the LVAD, peak forearm blood flow, and peak vascular conductance did not increase to values similar to those observed in normal subjects until the late postoperative recovery period. The delayed effect of the LVAD on metabolic vasodilation may be related to flow-dependent changes in the peripheral vasculature of patients with heart failure.

Transient normalization of systolic and diastolic function after support with a left ventricular assist device in a patient with dilated cardiomyopathy.

A 19-year-old man who had fulminant heart failure caused by an idiopathic dilated cardiomyopathy was supported with a left ventricular assist device for 183 days as a bridge to heart transplantation. At the time of intended transplantation it was noted that the patient's heart had returned to normal size, had a normal ejection fraction, and was able to maintain normal pressures and flows. In view of the apparent recovery of cardiac properties, the left ventricular assist device was explanted and the transplantation was not performed. However, the heart dilated, ejection fraction worsened, and the patient died of heart failure exacerbated acutely by a systemic viral illness. Although such recovery of systolic function is uncommon, as use of the left ventricular assist devices becomes more widespread other physicians might encounter similar findings and, in this regard, they might find our experience useful as they contemplate their treatment options.

Comparison of rejection in the atrioventricular node and bundles with the working myocardium in transplanted hearts.

At present the monitoring of heart transplant patients for rejection is done by endomyocardial biopsy. It has been proposed that ECG assessments of conduction delay may provide a noninvasive means of detecting rejection. To determine whether rejection in the atrioventricular node and conduction bundles reflects rejection in the working myocardium, we studied 21 transplanted hearts from 21 cardiac allograft recipients. Twenty of the hearts were obtained at autopsy, and one was obtained from a patient who underwent a second transplant procedure. The atrioventricular conduction tissues, the ventricular and atrial free walls, and the interventricular and interatrial septa from these hearts were examined for histologic evidence of rejection. Of the 21 hearts, 17 showed histologic changes of acute allograft rejection. Of the 17, rejection involved the conducting system and the myocardium equally in 11. In 6 of the 17 hearts, rejection involved the working myocardium more severely than it did the conducting system. Of interest, nonrejection pathologic changes were also noted in the conducting systems of several hearts. Severe accelerated arteriosclerosis was found in the artery to the atrioventricular node in one case, and lesser degrees of accelerated arteriosclerosis in this artery were found in two additional cases. In one case, lymphoid infiltrates, consistent with posttransplant lymphoproliferative disorder, were noted in the vicinity of the atrioventricular node, and several hearts demonstrated operative trauma or ischemic changes that appeared to involve the conducting system. These results suggest that although there may be a morphologic basis for using electrophysiologic changes in the conducting system to monitor heart allograft recipients for rejection, caution should be exercised in interpreting these changes.

Duration of left ventricular assist device support affects transplant survival.

BACKGROUND: Left ventricular assist devices have become an important tool in the successful treatment of heart failure as bridges to transplantation. The optimal duration of device support before heart transplantation is debated. We report the effect of left ventricular device support duration on survival after heart transplantation. METHODS: All patients bridged to heart transplantation with the ThermoCardiosystems Heartmate 1,000 IP left ventricular assist device between January 1, 1986, and October 15, 1994, were included in our study. Parameters studied included duration of support, measures of end-organ function, and complications while supported with the device. Patients supported < 30 days were compared with patients supported > 30 days before undergoing transplantation. RESULTS: Patients supported for < 30 days had a threefold increased perioperative mortality compared with patients supported > 30 days (p = 0.031). Laboratory values of end-organ function were similar before left ventricular device insertion in both groups, although at the time of transplantation patients supported < 30 days had a significantly elevated bilirubin level compared with patients supported > 30 days (p < 0.001). Patients supported > 30 days had significantly more infections than the < 30 days group (p = 0.0345). CONCLUSIONS: Patients supported for > 30 days with left ventricular assist devices have improved post-transplant perioperative survival because of normalization of end-organ function and improved physiologic status secondary to aggressive physical rehabilitation. Patients should be supported for > 30 days in combination with physical rehabilitation, to improve early survival after heart transplantation.

Desaturated venous-to-arterial shunting reduces right-sided heart failure after cardiopulmonary bypass.

BACKGROUND: Right-sided circulatory failure is a major cause of morbidity in heart transplant and ventricular assist device recipients. METHODS: Several systems for managing right-sided circulatory failure with use of a right-to-left shunt without need for an oxygenator or systemic heparinization were designed and used clinically. RESULTS: A right-to-left shunt was successfully used to treat severe right-sided circulatory failure in both a transplant and a left ventricular assist device recipient. CONCLUSIONS: If constructed between the femoral vein and artery, such a shunt could (1) be easily inserted and removed, (2) selectively infuse the lower extremities with desaturated blood while maintaining cerebral and cardiac perfusion with saturated blood, (3) selectively reduce the risk of paradoxical emboli to the head and heart, and (4) provide a known and adjustable degree of shunting depending on the condition of the patient, a major advantage of this system compared with creation of an atrial septal defect.

Use of a pulsatile right ventricular assist device and continuous arteriovenous hemodialysis in a 57-year-old man with a pulsatile left ventricular assist device.

BACKGROUND: Despite advances in the perioperative treatment of both heart transplant and left ventricular assist device recipients, right-sided circulatory failure refractory to medical management remains a major source of morbidity in the immediate postoperative period. In addition, hypervolemia is a frequent complication encountered in the treatment of these patients because of their large fluid intake requirements and relative potential for kidney failure. METHODS: Previous reports have documented the use of continuous-flow devices to support the failing right-sided circulation of patients after both left ventricular assist device insertion and orthotopic heart transplantation. However, such continuous-flow devices may carry the attendant risks of hemolysis and bleeding and may further require 24-hour monitoring by trained personnel. We report the temporary-use pulsatile Abiomed BVS 5000 right ventricular assist device and continuous arteriovenous hemodialysis in the recipient of a pulsatile TCI HeartMate 1000 IP left ventricular assist device both after left ventricular assist device implantation and after orthotopic heart transplantation. RESULTS: The patient was well at 13 months follow-up. CONCLUSIONS: The use of right ventricular assist devices and continuous arteriovenous hemodialysis in both transplant and left ventricular assist device recipients undoubtedly will remain important as the popularity of these two therapeutic modalities continues to grow.

Early progressive mobilization of patients with left ventricular assist devices is safe and optimizes recovery before heart transplantation.

BACKGROUND: As early perioperative survival with left ventricular assist device support improves, additional emphasis must be placed on patient rehabilitation. Especially as mean left ventricular assist device support times exceed 100 days, it is essential for health professionals to mobilize patients to prevent the incidence of the deleterious effects of bedrest. The timing of optimal functional performance and the safety of rehabilitation has not been described. METHODS: We retrospectively studied patients surviving left ventricular assist device implantation (34 of 41 patients; 27 men, 7 women; age 51 +/- 11 years). Physical therapy consisted of progressive mobilization leading to treadmill exercise or cycling. All patients exercised in the "pump on full" mode with flows >or= 3.0 L/min. RESULTS: Twenty of thirty-four patients initiated ambulation at 7 to 10 days with independent ambulation by 14 days in 55% of the group. Treadmill exercise was tolerated by 82% of the patients, begun at postoperative day 21 by eighteen patients. The greatest improvement in exercise performance was seen by 6 to 8 weeks (20 to 30 minutes at 3.17 +/- 0.79 metabolic equivalents). Maximal functional capacity achieved was influenced by medical complications. A total of 1878 treatment sessions lasting 1390 hours was performed. Only four minor incidents occurred representing 2.9 incidents/1000 patient hours; all involved a transient decrease in pump flow. None of these events resulted in an increase in morbidity or mortality. CONCLUSIONS: Progressive mobilization in patients with left ventricular assist device is safe. Patients return to independence in activities of daily living and tolerate prolonged workloads of up to 5 metabolic equivalents. There is rapid improvement in functional capacity until 6 weeks after operation. Delay in transplantation until this time may optimize postoperative recovery.

Complications of endomyocardial biopsy in heart transplant patients.

Despite the increasing use of alternative techniques, endomyocardial biopsy remains the primary method for diagnosing cardiac allograft rejection. Improved patient longevity and an increasing number of transplantation centers are resulting in an increase in the number of biopsies performed. Although endomyocardial biopsy is usually considered a safe procedure, no large studies of the risks of endomyocardial biopsy specifically in the transplant population have been undertaken. To determine the risk of endomyocardial biopsy in these patients, we reviewed 2454 endomyocardial biopsies performed from January 1983 to December 1990 in 133 cardiac allograft patients at our institution. At the time of each endomyocardial biopsy, a worksheet was completed detailing the patient's interval history, the site of vascular introduction, the number of attempts, the number of specimens, and any complications encountered. A total of 74 (3.0%) complications occurred. Fifty-six (2.3%) complications were associated with catheter insertion, including carotid puncture (1.8%), vasovagal reaction (0.1%), and prolonged bleeding (0.4%). Complications during biopsy included arrhythmias (0.25%) and conduction abnormalities (0.2%). In addition, we observed five episodes (three patients) of allergic reaction to a reusable bioptome and one case of pacemaker dislodgement. All complications were without significant long-term sequelae. In contrast to the cardiomyopathy population, no ventricular perforations or deaths occurred. Thus although endomyocardial biopsy has some risk, it continues to be a safe and effective way of monitoring rejection.

The influence of infection on survival and successful transplantation in patients with left ventricular assist devices.

BACKGROUND: Mechanical cardiac assistance has recently emerged as a tenable option in the treatment of end-stage heart failure. In spite of recent technical improvements that have reduced the incidence of life-threatening complications, the reported frequency of infections in these patients has remained high. METHODS: Over a 5-year period, 60 patients underwent insertion of a left ventricular assist device (LVAD) at our institution. Detailed medical records were kept prospectively for all patients, and a variety of endpoints were analyzed, including the incidence, nature, and sequelae of infections before and after LVAD implantation and after transplantation. RESULTS: Twenty-nine of 60 patients (48%) undergoing LVAD insertion subsequently had development of infections. The most frequent sites of infection were blood, LVAD drivelines, and central venous catheters, representing 61% of all infections. At the time of LVAD implantation, 13 of 60 patients (22%) had culture-proven infections. In spite of an increased incidence of subsequent infection (77% vs 40%), there were no differences in rates of mortality (31% vs 26%), LVAD endocarditis, (23% vs 11%) and eventual transplantation (62% vs 57%) between these patients and those without periimplantation infections. Although the overall mortality rate was not influenced by infections during LVAD support (28% vs 26%), the development of LVAD endocarditis was associated with a high mortality rate. Finally, although patients with infections during LVAD support had significantly longer median support times than those who remained infection free (101 vs 49 days, respectively), there was no difference in the rate of successful transplantation (59% vs 58%) or in the rate of infection after transplantation (35% vs 28%). CONCLUSIONS: Infections are common in patients undergoing LVAD support, but they do not adversely affect survival, the rate of successful transplantation, or the incidence of posttransplantation infection. Periimplantation infections may increase the risk of subsequent infections, but they also do not influence survival or transplantability. Patients with development of LVAD endocarditis are at increased risk for morbidity and death and require early and aggressive therapy, potentially including device explantation.

Technique for removal of left ventricular assist devices.

As the use of long-term implantable left ventricular assist devices increases, more surgeons will be required to remove the devices for transplantation or in the event of infection. We outline several maneuvers that facilitate device removal and several pitfalls that complicate the procedure. Important considerations include preventive measures at the time of device implantation and prevention of air embolism.

Low thromboembolic risk without anticoagulation using advanced-design left ventricular assist devices.

BACKGROUND: A major limitation of cardiac assist devices has been the high incidence of thromboembolic events and their requirement for systemic anticoagulation. The Thermo Cardiosystems HeartMate 1000 IP left ventricular assist device (LVAD) employs a design that may reduce thromboembolic risk and obviate the need for systemic anticoagulation. METHODS: Two hundred twenty-three patients with nonreversible heart failure were supported with the HeartMate LVAD as a bridge to heart transplantation. All patients were monitored prospectively for thromboembolic events. Anticoagulation regimens and occurrence of subclinical thromboembolic events, including those seen by transcranial Doppler examinations in selected patients, were also recorded. RESULTS: Total time of LVAD support use was 531.2 patient-months. Twenty-three patients (10%) received warfarin postoperatively for 42.4 patient-months (8.2% of total support time). Six patients (2.7%) had thromboembolic events, representing 0.011 events per patient-month of device use. CONCLUSIONS: The thromboembolic complication rate associated with this LVAD is acceptably low despite the minimal anticoagulation employed in this series, allowing consideration of long-term device use for the treatment of heart failure.

Experience with right ventricular assist devices for perioperative right-sided circulatory failure.

BACKGROUND: Right-sided circulatory failure remains a significant source of morbidity and mortality for both cardiac transplant and left ventricular assist device recipients. METHODS: We reviewed our experience with 11 patients who required a right ventricular assist device (RVAD) after either orthotopic heart transplantation or left ventricular assist device implantation. Variables analyzed included total time of RVAD support, hemodynamic and hematologic parameters, and parameters of end-organ perfusion. These were assessed at five time points: (1) at least 2 weeks before RVAD implantation, (2) intraoperatively just before RVAD insertion, (3) while on RVAD support, and, for those who survived, (4) just before RVAD explantation, and (5) off RVAD support. Survival was assessed as the ability to be weaned successfully from RVAD support. Urine output and serum transaminase levels were recorded throughout the period of RVAD support. RESULTS: Five patients received an ABIOMED 5000 BVS RVAD, and 6 received a Bio-Medicus centrifugal pump. Nine patients in the study underwent orthotopic heart transplantation and had development of right-sided circulatory failure from 0 to 96 hours after donor organ insertion, and 2 patients underwent left ventricular assist device implantation 12 to 48 hours before RVAD support. The mean time of RVAD support for survivors was 133.6 +/- 33.6 hours (range, 107 to 190 hours). Six patients were successfully separated from RVAD support, and 5 patients died while on RVAD support. Causes of death included sepsis (2), biventricular failure (2), and coagulopathy (1). Continuous arteriovenous hemodialysis was employed in 3 of 6 survivors and 1 of 5 nonsurvivors. CONCLUSIONS: Right ventricular assist devices work most effectively if implanted early enough to avoid significant, potentially irreversible end-organ injury. We liberally employ continuous arteriovenous hemodialysis, minimize the use of heparin immediately postoperatively, keep patients sedated, and continue RVAD support until the patient displays signs of hemodynamic and end-organ recovery as heralded by (1) a decrease in central venous pressure and, more importantly, a decrease in pulmonary artery diastolic pressure, (2) an increase in urine output, and (3) a decrease in serum transaminase levels.

Characterization of hematopoietic cells arising on the textured surface of left ventricular assist devices.

BACKGROUND: Textured biomaterial surfaces in implantable left ventricular assist devices induce development of a nonthrombotic neointimal surface and allow elimination of anticoagulation therapy in device recipients. Characterization of the hematopoietic cells formed within the neointimal surfaces of these devices will contribute to our understanding of this unique neointima. METHODS: The blood-contacting surface of seven ThermoCardiosystems left ventricular assist devices was removed, washed with phosphate-buffered saline solution, and digested with 0.1% collagenase for 15 to 20 minutes. The hematopoietic cells released from the explants were isolated and analyzed by flow cytometry and immuno-histochemical staining. RESULTS: More than 80% +/- 6% of hematopoietic cells isolated in this fashion are of myelomonocytic origin and express CD14, CD15, and CD33 surface molecules. Four percent of cells express the CD34 surface marker, which suggests that the neointima is colonized by pluripotent hematopoietic stem cells. Continuous culture of these hematopoietic cells in the presence of the cytokines interleukin-3, c-kit ligand, granulocyte colony-stimulating factor resulted in tenfold expansion by day 7 and 25-fold expansion by day 14. CONCLUSIONS: Pluripotent hematopoietic cells with a high proliferative capacity colonize textured surfaces of left ventricular assist devices and may contribute to the development of a biologically nonthrombogenic neointima.