What are the important anesthetic considerations for patients supported by LVADs?

What are the important anesthetic considerations for patients supported by LVADs?

Preoperative Considerations

The preoperative clinical status of LVAD-supported patients depends on multiple issues, such as the amount of end-organ damage sustained during low-output states prior to ventricular assist device (VAD) implantation, post-implantation complications, and underlying surgical prob-lems. Many LVAD recipients are ambulatory and otherwise uncompromised. Others experience varying degrees of renal, hepatic, pulmonary, and/or central nervous system insufficiency. Preoperative evaluation of neurologic dys-function and other major organ system problems is essen-tial. Any further deterioration in the perioperative period may preclude full recovery or disqualify a patient from later heart transplantation.

One of the most serious complications of extracorpo-real circulation is thromboembolism, and LVADs are no exception. The Heartmate’s blood chamber is designed with an antithrombogenic surface and requires no formal anticoagulation; however, the Novacor’s polyurethane-lined blood chamber mandates anticoagulation. Initially heparin is used and then long-term warfarin therapy is commenced. International normalized ratios (INRs) are maintained at 2.5–3.5 times normal. In elective situa-tions, Novacor patients may discontinue warfarin therapy preoperatively and convert to carefully monitored heparin infusions. In most cases, heparin infusions should not be discontinued preoperatively. The majority of surgical procedures (except for neurosurgical cases) can proceed safely in the presence of anticoagulation; however, scrupu-lous attention to hemostasis is required intraoperatively. Fresh frozen plasma or cryoprecipitate may be infused to decrease the level of anticoagulation toward the lower limit of manufacturer’s recommendations. Frequent partial thromboplastin time (PTT) measurements are important to balance the dual potential complications of hemorrhage and thromboembolism. Anesthesiologists must determine (perhaps in consultation with the sur-geon) a safe anticoagulation regimen for the perioperative period.

Adherence to strict aseptic technique is mandatory for all invasive procedures and prophylactic perioperative antibiotics are routinely employed. Infection of an LVAD is a catastrophic complication. They are very large foreign bodies that cannot be sterilized.

As with all critical life-support equipment in the oper-ating room, an LVAD must be connected to a reliable power supply. Its battery life is limited.

Preoperative considerations and practices regarding pacemakers and AICDs are much the same in LVAD-supported patients as in other patients. The pre-set pace-maker mode is ascertained and it is interrogated for proper functioning. Usually, atrioventricular sequential pacing will be in use (DDD or DOO mode), because this frequently preserves RV output (and therefore LV filling) in these patients. Magnets should be available in case of pacemaker malfunction. Modern pacemakers will usually convert to an asynchronous mode (e.g., AOO, VOO, or DOO) when a magnet is applied, and should revert to their prior programming when the magnet is removed. In any event, pacemaker-dependent patients should have their device interrogated postoperatively to assure proper func-tioning. Many, but not all LVAD patients will have an AICD. Unipolar electrocautery emits a high-frequency signal that could potentially be interpreted as ventricular fibrilla-tion, resulting in unnecessary defibrillatory discharges. Consequently, AICDs are usually deactivated in the imme-diate preoperative setting, assuming that a defibrillator is immediately available. Where possible, bipolar electro-cautery should be preferentially used. In an emergency situation, a magnet may be used to deactivate the AICD. Most AICDs (Medtronics, St. Jude, Biotronik) will remain deactivated as long as the magnet remains in place; however, a Guidant AICD is permanently deactivated by application of a magnet. Removal and subsequent reapplication of the magnet is required to reactivate a Guidant AICD.

Cardiovascular collapse in LVAD patients is treated with standard advanced cardiac life support (ACLS) protocols. However, one should never perform chest compressions on a ventricular assist device (VAD)-supported patient. Dislodgment of intracardiac cannulae will result in imme-diate exsanguination and certain death. The Novacor is very well shielded, and will not be affected by defibrillation or electrocautery. Unfortunately, the Heartmate may be reset to a fixed-rate mode by electrocautery and potentially damaged by external defibrillation. For Heartmate patients, preoperative consultation with the physician managing the LVAD is advisable to discuss the implications of electrocautery and external defibrillation.

Intraoperative considerations

Novacor’s requisite anticoagulation contraindicates major conduction anesthesia, which relegates most patients to general anesthesia. In some cases, sedation with local skin infiltration, or a regional intravenous technique (Bier block) is appropriate. Heartmate patients are not anticoagulated and remain potential candidates for major conduction anes-thesia. Unfortunately, no specific recommendations exist regarding this issue at the time of writing.

The pump’s pre-peritoneal location places the LVAD-supported patient at increased risk for aspiration pneumoni-tis. Consequently, “full stomach” precautions (e.g., gastric acid prophylaxis and rapid sequence induction with cricoid pressure) should be considered. Extubation criteria of the LVAD-supported patient are the same as in any other patient.

The anesthetic drugs used should be appropriate for the planned operation, and should take into account any alter-ations of physiology resulting from insufficiency of, or prior injury to, major organ systems. For example, it may be disadvantageous to use pancuronium or vecuronium in the patient with renal insufficiency or biliary obstruction, respectively. Succinylcholine may be contraindicated in patients with recent cerebrovascular accidents. VADs do not specifically contraindicate any particular anesthetic agents, but the anesthetic plan should consider the poten-tially dysfunctional unassisted RV. Consequently, particular attention should be paid to optimizing RV preload, after-load, and inotropic support as required.

Long-term, implanted LVADs are typically set to auto-matically eject as soon as the blood chamber is full. The faster the device fills, the faster it pumps and the higher the pump output. Hypovolemia results in slow pump filling, decreased LVAD output, and hypotension. Consequently, the goal of fluid management is to maintain normal or slightly elevated intravascular volume. Markedly increased systemic vascular resistance (SVR) impairs forward flow, resulting in incomplete pump emptying, which leads to stagnation of blood in the pump and increased risk of thrombosis. Therefore, maintenance of normal or slightly low systemic vascular resistance is desirable. Management must be individualized. Inotropes, vasodilators, and vaso-pressors are administered to achieve optimal hemodynam-ics. LVADs generally function well as long as there is sufficient intravascular volume to fill the pump.

Depth of anesthesia is judged, in part, by hemodynamic parameters. However, LVAD-supported patients may not manifest increased pulse rates, a classic sign of light anesthe-sia. As previously discussed, LVADs eject as soon as the blood chamber fills, and it is this rate of ejection which constitutes the LVAD-supported patient’s pulse rate. Therefore, the pulse rate is rarely the same as the ECG-derived heart rate. For this reason, intraoperative tachycardia, as measured by the pulse rate, is reflective only of the speed of LVAD filling, and not of light anesthesia. However, while relative hyper-tension is reflective of relative volume overload and higher pump outputs, it could also reflect heightened adrenergic activity with increased SVR (Table 11.4). Nevertheless, lack of an acutely increased blood pressure with surgical stimula-tion is not always a reliable indicator of adequate depth of anesthesia in an LVAD-supported patient.