Ventricular-assist Device Therapy in Children

Yuki K, Sharma R, DiNardo J
Best Practice & Research Clinical Anaesthesiology 2012; 26: 247-264.

Reviewers: Rachel A. Koll, MD and Lori A. Aronson, MD
Cincinnati Children’s Hospital Medical Center

As studies are starting to demonstrate that ventricular assist devices (VAD) offer better survival to transplant than extracorporeal membrane oxygenation (ECMO), having a comprehensive knowledge of these devices is invaluable.1-3 Additionally, translating adult devices to pediatric patients is never easy. As such, the National Heart, Lung and Blood Institution awarded contracts to promote development of mechanical circulatory support (MCS) specifically designed for infants and children.4 Clinical trials through the Pumps for Kids, Infants and Neonates (PUMPKIN) program will hopefully be underway soon. This comprehensive review article of mechanical circulatory support (MCS) primarily focuses on VADs for children with circulatory failure and outlines the various types of devices as well as anesthetic management related to VAD placement.

History and Indications
Yuki et al begin by discussing the history of pediatric VADs. In adults with refractory heart failure, VADs have become part of the standard of care as temporary or long-term support. This role is still evolving for pediatric patients. The earliest use of VAD in pediatric patient was in 1967, followed by ECMO in the 1970s. These centrifugal pumps evolved into paracorporeal pulsatile devices used in adults. When translated to children, size discrepancies resulted in increased complications related to thromboembolic and hemorrhagic neurologic events. To date, we are still seeing these issues with our current VADs.5

The current indication for MCS depends on size and type of support needed (cardiopulmonary versus cardiac) and duration of support. Short term VAD support, involving days to weeks, is often seen in myocarditis and post-cardiotomy syndrome as a bridge to recovery or bridge to decision. Long term support often includes bridge to transplant. ECMO may be preferred in the face of intracardiac shunts, respiratory failure and pulmonary hypertension where cardiopulmonary support is required.

Classification of Devices
The article continues with a very thorough classification of MCS devices. Devices can be in series or parallel, pulsatile or nonpulsatile, paracorporeal or intracorporeal, triggering (manual, asynchronous, synchronous and fill to empty). Briefly, a device may be used in parallel (inflow via atrium and outflow to corresponding great vessel) or in series (inflow via ventricle and outflow to corresponding great vessel) with the native ventricular system. The series arrangement is beneficial in that it provides better filling and higher device output; however, the ventricular apex may become akinetic or prone to thrombus after device removal, which must be considered for patients who are bridged to recovery.

Pulsatile devices are either pneumatic or electromechanical, while nonpulsatile devices are centrifugal or axial. The pump housing may be paracorporeal (outside the body) or implantable (intracorporeal). Intracorporeal devices are size-limited typically to >1.5 m2. Triggering modes serve different purposes. Manual triggering is useful for de-airing and delivering a single ejection. Asynchronous mode offers a fixed rate and variable stroke volume which may require increased preload at faster rate. Fill to empty is essentially a volume mode that is preload dependent to eject and therefore has a variable rate. If a VAD is in series and triggers by fill to empty, it is synchronous if filling is during diastole and counter-synchronous if filling is during systole. Finally, synchronous triggering is linked to the R wave to allow ejection and may be used to reduce afterload.

Short-term Support Devices
Short-term devices include ECMO and paracorporeal centrifugal pumps. ECMO is the most common MCS in children, is utilized for cardiopulmonary support and allows for rapid deployment so it can be quickly engaged during CPR. Disadvantages include complexity of the system, personnel requirements, aggressive anticoagulation, blood prime and transfusion exposure, intubation, sedation, more limited duration (1-3 weeks) and possibly left atrial decompression issues. Intermediate to long-term support is provided by pulsatile and non-pulsatile VADs.

Centrifugal pumps are also used short-term, and they are extracorporeal and non-pulsatile VADs used in parallel. They are very preload and afterload dependent, and like ECMO require full anticoagulation as well as an intubated, sedated patient. Examples include the BioMedicus (Medtronic BioMedicus, Minneapolis,MN, USA), CentriMag (Levitronix, Zurich, Switzerland), RotaFlow (Jostra, Hirrlingen, Germany), the Capiox (Terumo, Ann Arbor, MI, USA) and the Tandem Heart.

Intermediate to Long-term Support
For intermediate to long-term support, there are currently many options for paracorporeal and intracorporeal devices. Refer to Table 1 (page 252) or the manufacturer for details of specific devices. Some manufacturers are Thoratec Laboratories, Pleasanton, CA, Abiomed, Danver, MA, Berlin Heart AG, Berlin, Germany, and Jarvik Heart, New York, NY. In general, VADs allow for extubation, ambulation, activities of daily living, and enteral nutrition. Less aggressive anticoagulation is acceptable. While the rate of adverse events related to VAD use, namely thrombosis, bleeding, neurologic complications, and infection in the pediatric population is similar to adult data, they are not inconsequential. Optimal anticoagulation strategies have yet to be determined, but many practitioners follow the Edmonton antithrombotic protocol, which consists of aspirin, dypyridamole and enoxaparin for infants, and aspirin, dypyridamole, and warfarin (INR 2.7-3.5) for children over 12 months of age (Table 5 page 260). Several devices are presented in detail, but note there is wide variation between centers regarding device selection and anticoagulation strategies.

The review provides two very helpful tables. Table 2 (page 253) compares the advantages and disadvantages of the pulsatile vs. axial devices. Some advantages for axial devices are simpler design, less noise, lower infection rates and more mobility, but they are very preload and afterload sensitive. Table 3 (page 254) provides an algorithm for troubleshooting pulsatile VADs. Inadequate filling involves assessing preload, intrathoracic pressure changes with ventilation, and tamponade, and suggests possibly adjusting diastolic pressure (vacuum) and systolic ejection time. Inadequate ejection may be due to increased afterload with consideration for adjusting the systolic drive pressure or systolic ejection time.

Anesthetic Management for VAD Placement
A particularly helpful inclusion addresses anesthetic management for placement of a VAD as well as after device placement. Not surprisingly, patients presenting for VAD placement will have low cardiac output. After a thorough preanesthetic evaluation, be prepared to deal with end-organ dysfunction, especially coagulopathy related to hepatic congestion and renal impairment. Evaluate for arterial and venous access sights, and recognize that blood product exposure will be necessary, and perhaps massive. Any present pacemaker or defibrillator will need to be reprogrammed.

Induction of anesthesia and pre-bypass management requires recognition that the patient has limited contractile reserve, delayed circulation time of drugs and may be quite sensitive to changes in preload and afterload from medications, stimulation as well as initiation of positive pressure ventilation. Patients who require VAD placement will have limited ability to respond to acute increases in afterload, but will lack the ability to compensate for myocardial depression and vasodilation. During positive pressure ventilation, maintain an appropriate minute volume at the lowest possible mean airway pressure. Avoid circumstances or maneuvers that might increase pulmonary vascular resistance, as this will have the effect of inadequate preload delivery to the left ventricle and a severe drop in cardiac output. In addition to standard ASA monitors, induction will likely require arterial and central monitoring. After induction, transesophageal echocardiography (TEE) is indicated, but may compress the heart or respiratory structures such as the left atrium, left mainstem bronchus, or an aberrant subclavian artery. It is imperative that any atrial or ventricular level shunt be identified for closure before VAD placement. If not, left atrial pressure can be reduced more than right atrial pressure, leading to a right-to-left shunt and hypoxemia or paradoxical embolus. Use lactate trends and mixed venous saturation to guide therapy.

The LVAD should function well upon insertion providing that there is adequate intravascular volume and right ventricular function to fill the pump. TEE can assess volume status and may identify the cause of inadequate preload such as improper positioning or obstruction of the inflow cannula or impaired RV function. While the RV function is expected to decrease due to leftward shift of the septum and loss of the LV contribution to RV systole, this is offset by increasing RV preload and reducing RV afterload. RV dysfunction, manifesting as high CVP, RV distension, tricuspid regurgitation and/or hypokinesis can be treated with pulmonary vasodilators (milrinone, nitric oxide) and inotropes. Surgical interventions from tricuspid annuloplasty to ECMO or BiVAD placement may be necessary in the face of severe RV dysfunction.

BiVAD insertion in the pediatric patient is not without risk, as pulmonary endothelial injury, pulmonary edema, and hemorrhage may occur due to the large RVAD stroke volumes that overwhelm the pulmonary vasculature. This is especially true for patients with elevated pulmonary artery pressures.

As previously mentioned, the majority of VAD complications are due to bleeding or thrombosis. Acutely, patients tend towards bleeding due to suppressed platelet and coagulation function after the device is placed. Antifibrinolytic medications may be beneficial as well as intraoperative thromboelastogram. However, a hypercoagulable phase later ensues. There are large transfusion requirements during VAD placement, and recombinant factor VIIa has been used with varying degrees of success. At this time, factor VIIa is not FDA approved in this setting and cannot be recommended by the literature, but use is at the anesthesiologist’s discression in the face of potentially fatal hemorrhage.

Finally, a description of anesthetic management of a patient with VAD presenting for other surgery is discussed. Ketamine seems to have the lowest incidence of hypotension on induction, especially in fixed rate, variable stroke volume devices like the Berlin Heart Excor. As such, most patients respond to fluid boluses of 10 ml/kg. If this fails, the addition of alpha-adrenergic receptor agonist or vasopressin is suggested. If hypotension is refractory to treatment, consider RV failure and increased pulmonary arterial pressures as the cause; instigate maneuvers to decrease PVR, administer PDE inhibitors and nitric oxide. Table 6 (page 261) compares a pulsatile VAD (Berlin Heart EXCOR) to an axial VAD (HeartMate II) regarding hemodynamic stability management. While device inspection and assessment is different, possible interventions are quite similar. For both devices, hypotension should be treated as discussed above. Hypertension is similarly treated for both types of VADs with increased sedation, analgesia, and titration of vasodilators. Additionally, avoid increasing systolic drive pressure in the pulsatile VAD.

Conclusion
In summary, this article provides a wealth of information regarding the complicated and ever-changing topic of pediatric VAD management. It is thorough and detailed, providing both a helpful introduction to the various devices as well as useful, clinically relevant information regarding the management of anesthesia for placement of the device as well as treating patients with an existing device. While clinical trials need to better assess these devices, hopefully the PUMPKIN program will offer more devices with fewer complications in the near future.

References

Jeewa A, Manlhiot C, McCrindle BW, et al A. Outcomes with ventricular assist device versus extracorporeal membrane oxygenation as a bridge to pediatric heart transplantation. International Center for Artificial Organs and Transplantation and Wiley Periodicals 2010 Dec;34(12):1087-91.
Sharma MS, Forbess JM, Guleserian KJ. Ventricular assist devicesupport in children with heart failure: The Children’s Medical Center of Dallas Experience. Artificial Organs 2012, 36 (7): 635-648.
Morales DL, Almond CS, Jaquiss RD, et al. Bridging children of all sizes to cardiac transplantation: the intial multicenter North American Experience with the Berlin Hear EXCOR ventricular assist device. J Heart Lung Transplant Jan;30(1):1-8.
Baldwin JT, Borovetz HS, Duncan BW et al. The national heart, lung and blood institute pediatric circulatory support program. Circulation 2006 Jan 3; 113(1):147-155.
Fraser CD, Jaquiss RD, Rosenthal DN et al. Prospective trial of a pediatric assist device. N Engl J Med 2012, 367(6):532-541.