Oren Bernstein, MD
Private Practice Anesthesiologist
Pacific Anesthesia, Inc.
The Queen’s Medical Center
The number of patients eligible for transcatheter aortic valve replacement is rapidly expanding, and TAVR centers are under growing production pressure with steadily increasing caseloads. As such, anesthesiologists can play an i mportant role in maximizing procedural efficiency while maintaining overall safety.
Transcatheter aortic valve replacement (TAVR) is a procedure nearing a tipping point in the treatment of aortic valve stenosis. Until recently, TAVR was commonly viewed as a palliative procedure reserved only for very elderly patients or the most extreme surgical risk cases, as the procedure was initially approved for only truly inoperable patients. Later, however, high-risk surgical patients also were granted access to the procedure. With the anticipated FDA approval of TAVR for intermediate-risk patients in the near future, the potential pool of TAVR recipients will expand dramatically, and by the end of the decade, even low-risk patients may be offered TAVR as first-line treatment.
As the indications for the procedure expand, TAVR referral centers will find themselves under increasing production pressure to accomplish their growing numbers of cases more efficiently. Whereas a “TAVR day” might previously have involved 1 or 2 cases, days of 3 or 4 procedures have become the new norm.
Anesthesiologists in high-volume centers can play an enormous supporting role in maximizing the efficiency and throughput of a day devoted to TAVR implantation by adopting a minimalist approach to anesthetic and periprocedural management, using monitored anesthesia care (MAC), and avoiding invasive lines and postprocedural admission to the ICU. This approach appears to be at least as safe as traditional management under general anesthesia—and possibly safer—with clear benefits including shorter procedural time and decreased overall cost.
Improving TAVR Outcomes
The most recent data presented at the 2016 meeting of the American College of Cardiology not only suggest noninferiority of TAVR compared with surgical AVR (SAVR) for high-risk patients, but actual superiority in valve hemodynamics and important clinical outcomes.1 In high-risk patients after 3 years, all-cause mortality or disabling stroke were demonstrated to be significantly lower in TAVR with the Medtronic CoreValve (37.3%) than with SAVR (46.7%). Valve gradients and effective orifice areas were also lower with TAVR than SAVR in this study. For intermediate-risk patients, data from the PARTNER 2 trial demonstrated TAVR to be noninferior to SAVR, with death or disabling stroke occurring in 19.3% of patients receiving TAVR and 21.1% of patients receiving SAVR in the intention-to-treat population.2 Notably, this trial involved the Sapien XT second-generation valve (Edwards Lifesciences), which has already been replaced by the improved Sapien 3 valve.
In a study comparing data between intermediate-risk patients receiving the Sapien 3 valve and matched surgical patients from the Partner 2 trial, TAVR showed superiority in all major clinical outcomes, at both 30 days and 1 year3(Table 1).
If these trends are confirmed in future studies, and are also proven to extend to low-risk patients in trials beginning soon, the majority of patients will probably opt for percutaneous valve implantation over the open surgical approach on cardiopulmonary bypass. This will be a major change in the treatment of aortic valve stenosis. In the future, it is likely that TAVR will become commonplace and SAVR will become relatively rare.
The accelerating expansion of TAVR has been driven by the rapid improvement in the delivery systems and valves themselves compared with the initial systems that came to market. The first-generation valves were implanted via enormous 22-24 Fr arterial sheaths, which carried a substantial risk for major vascular injury to the iliofemoral system, or required more invasive alternative surgical access techniques (ie, transapical or transaortic) that are associated with higher procedural morbidity and longer hospital length of stay (LOS). The valves themselves were prone to significant paravalvular leaks, even with post-implantation dilation via balloon.
With large delivery sheaths, surgical repair of vascular injuries was relatively common, and a transesophageal echocardiography (TEE) was routinely required to evaluate the degree of residual paravalvular leak (PVL) during valve implantation. Residual leaks graded moderate or greater are associated with worse outcomes after TAVR,2,4 and this rate was substantially higher than that in the surgical population for first- and second-generation devices.
In early-adopting TAVR centers, cardiac anesthesiologists formulated anesthetic plans for this new procedure with these technological imperfections in mind. General anesthesia for TAVR was quite obviously required, especially when considering the operative unfamiliarity that led to long procedure times and a significant risk for intraprocedural catastrophes.
The use of general anesthesia mandated preinduction placement of an arterial catheter and postinduction placement of a dedicated central venous catheter for vasopressors and volume administration, usually in the right internal jugular vein. Some centers floated pulmonary artery catheters to monitor hemodynamics, and some routinely left patients intubated after the procedure as they would after SAVR.
A Dramatic Evolution
Since those first years of TAVR, the procedure has dramatically evolved and improved. The arterial sheaths have been reduced to a more reasonable 14 to 16 Fr—similar to those used for balloon valvuloplasty (BAV). The next-generation systems will feature even smaller 12-Fr sheaths. Correspondingly, the risk for vascular complications has declined, as has the requirement for nonfemoral approaches with higher morbidity (ie, transapical and transaortic).
Moreover, surgical cutdown is almost never necessary for femoral access using the modern systems, because a truly percutaneous approach is feasible with the much smaller sheaths. The valves themselves now feature “skirts” that significantly reduce the incidence and severity of paravalvular leaks, such that the rate of moderate or greater aortic regurgitation is now half that of the first-generation devices: 3.7% at 30 days with the Sapien 3 versus 7% reported in the PARTNER A trial with the original Sapien valve,2,5 which has addressed a major Achilles heel of the first iterations of the procedure. Consequently, aortography along with transthoracic echocardiography (TTE) is usually sufficient to determine that the residual PVL is acceptable, and TEE is no longer absolutely indicated. In addition, TAVR operators have become experienced and facile, reducing procedural time and complications.
Another major change in TAVR has been the risk for stroke. For reasons not entirely known, first-generation TAVR procedures carried a stroke risk of about 5%.5 This was clearly more than the stroke risk from SAVR, and was seen as a major disadvantage of the transcatheter approach. Fortunately, the risk for stroke conferred by third-generation TAVR implantation seems much lower than found previously, and may ultimately prove even lower than the stroke risk from SAVR.3
Given the modernization and improvement of transfemoral (TF) TAVR, anesthesiologists can also adapt their anesthetic care to reflect the new realities of the procedure. Doing so brings benefits to the patient, the hospital performing TAVR, and the health system. TF TAVR can now be readily accomplished under MAC for most patients; the previously routine placement of invasive monitors for the procedure can be minimized; and Foley catheters need not be placed for what is now a short procedure with minimal fluid administration.
In hospitals with a reliable step-down unit, the ICU can be avoided entirely when invasive lines are not placed. Without compromising safety, anesthesiologists can thereby reduce procedural time, decrease recovery time, and help reduce costs—all factors that will maximize the efficiency of their facility’s program. By doing less to our patients, we indeed can do more for them and the health care system.
Consider that for many decades, BAV has been performed by interventional cardiologists in the catheterization lab without the support of anesthesiologists. Via 12- or 14-Fr sheaths, BAV involves percutaneous femoral arterial access, rapid ventricular pacing to minimize cardiac output, and the deployment of a balloon in the aortic root.
Although important differences obviously exist, in many ways TAVR is a glorified BAV procedure, one in which the valve is rendered completely functional after balloon deflation. The critical portions are very similar. Nurse-administered sedation has been used for these patients for many years without anesthesiology involvement, with a good track record of safety.
MAC Versus GETA
When considering adopting the MAC approach, anesthesiologists will frequently balk at the idea of giving up the situational control that general endotracheal anesthesia (GETA) affords. Indeed, in the event of a vascular catastrophe, GETA allows total focus on resuscitation. We are inherently averse to the idea of needing to manage the airway while simultaneously needing to tend to vasoactive administration of medication and fluid management.
These concerns were completely justified in those early days of TAVR, when adverse events were far more likely to occur. The first iterations of TAVR involved a very real possibility of surgical intervention, including emergent cardiopulmonary bypass. While such risk will never be entirely eliminated, experienced centers that have attained procedural proficiency using third-generation delivery systems have a much lower incidence of complications requiring surgical intervention.
Although proper preparation is mandatory and rescue equipment should be available in the event of an unforeseen catastrophe, in the author’s experience, managing the airway during these rare events does not hinder the experienced anesthesiologist’s ability to perform a proper resuscitation. Therefore, the low risk for a major intraprocedural complication should not be the factor dictating the anesthetic technique.
A major advantage of the “minimalist approach” to TAVR is in overall hospital cost and potential hospital LOS. A 2014 analysis of the minimalist bundle versus a “standard” more invasive approach demonstrated statistically significantly shorter LOS (3 vs 5 days) and decreased cost ($45,485 vs $55,377).6 Procedural success and freedom from significant paravalvular leaks were similar in both cohorts, meaning intraprocedural TEE imaging was not essential in excluding higher-grade leaks.
The choice of MAC over GETA (Table 2) is therefore associated with substantial economic benefit when deployed as part of a minimalist bundle, although it should be noted that the data do not yet definitively demonstrate a safety benefit for either approach. Recent data do suggest a 30-day morbidity and mortality benefit from moderate sedation compared with GETA,7 but it cannot yet be concluded that anesthetic technique alone was responsible for the effect. Either anesthetic technique can be administered safely. While the induction of general anesthesia and institution of positive pressure ventilation is associated with a potential for significant physiologic insult, anesthesiologists performing TAVR have significant experience in anesthetizing patients with aortic valve stenosis for SAVR, and are well equipped to treat any hemodynamic compromise.
Still, patients with severe aortic stenosis undergoing induction of general anesthesia must have an arterial line placed beforehand, as avoidance of cardiovascular collapse requires beat-to-beat blood pressure monitoring and immediate treatment of downward trends. In contrast, slowly titrated sedatives used during the MAC approach do not carry the same hemodynamic consequences. In one of the first steps of the TAVR procedure, a small arterial sheath is placed in the contralateral groin from the main introducer sheath, and this line can readily be transduced for hemodynamic monitoring during the critical portions of the procedure.
The anesthesiologist, therefore, does not need to place a radial arterial line. While arterial line placement in experienced hands is usually rapid, some patients have such arteriopathy that adequate arterial access is difficult and time-consuming. In academic centers with trainees performing lines, the procedure can require even more time. Under MAC, assuming the patient’s initial hemodynamics are adequate and reliable cuff pressures are available, a dedicated arterial line is not necessary to monitor those few minutes between the onset of sedation and the placement of the femoral arterial line.
Under general anesthesia, even patients with preserved left ventricular function often require vasopressors to maintain adequate perfusion pressures. Accordingly, an internal jugular central venous catheter is often placed to allow such administration, which sometimes must be continued postprocedurally for a short time. The MAC approach is associated with much less hypotension, and vasopressor infusions are needed far less frequently. Assuming adequate peripheral venous access is present, the anesthesiologist can ask for access to the side arm of the femoral venous sheath, which is routinely placed to allow transvenous pacing. A sterile length of large-bore tubing is handed off to the anesthesiologist, who now has adequate central access if catastrophic complications occur or if vasopressors are required. If the patient is stable at the end of the procedure, this line is removed and the patient goes to recovery with only peripheral IV catheters in place.
By avoiding arterial and central venous catheterization, significant time is saved and the associated risks from the procedures are avoided. The TAVR day is thereby made more efficient, so that more cases can be performed in a single day. This optimizes the output of the TAVR program and is economically beneficial to the hospital. In one study, room time was decreased from an average of 218 minutes with the “standard” approach to 150 minutes, a savings of nearly half an hour per case.7 This time savings is not trivial on days when 3 or 4 valves are implanted in succession.
Pulmonary artery catheterization is generally not required for patients undergoing third-generation TAVR, regardless of anesthetic technique. Unlike SAVR, there are no significant fluid shifts. Moreover, most TAVR patients can entirely bypass the ICU, going from the procedural suite to a PACU and then directly to a step-down unit. Avoiding pulmonary artery catheterization and other invasive lines facilitates such a fast-track approach. Simply stated, pulmonary artery catheterization is unlikely to change periprocedural management and is invasive and expensive. Its risks and costs outweigh any benefit in TF TAVR. For SAVR patients, being sent to the ICU is standard. For modern TAVR, this approach is unnecessary; it confers no demonstrable benefit and comes at substantial cost.
Recognizing the evolution of TAVR, a motivated high-volume program will therefore adopt a minimally invasive approach and conduct MAC anesthesia by default. In the author’s TAVR institution, patients are screened for the following contraindications to MAC: significant obstructive sleep apnea (unless the patient agrees to very light or no sedation), an anticipated truly difficult airway, potential need for surgical vascular access or repair, a hostile aortic root, and an approach other than transfemoral.
Patients with severe cardiac disease, including depressed ejection fraction, pulmonary hypertension, and right ventricular dysfunction are all candidates for MAC, and indeed eschewing the induction of general anesthesia in these patients avoids the deleterious physiologic insult that may otherwise occur. TAVR patients enter the hospital in a state of physiologic homeostasis that is often precarious and easily disturbed, but presently stable. The MAC approach tends to keep this tenuous balance intact, whereas general anesthesia with positive pressure ventilation tends to require vasoactive agents to restore it.
Regardless of whether MAC or GETA is chosen for a particular patient, the goals are the same: Assure patient safety and comfort, and allow for early mobilization and physical therapy. Anesthetic agents chosen should reliably allow a patient to be awake, oriented, and interactive at the end of the procedure. Certainly this is easier after MAC, but GETA can be conducted in a fashion that achieves these goals too.
Regardless of the anesthetic approach, both benzodiazepines and opiates will optimally be avoided in elderly patients naive to these medication classes. Ketamine is an attractive sedative in many anesthetic arenas, including cardiac surgery, but the author cautions against its routine use in the elderly population undergoing TF TAVR, which is a brief procedure not associated with significant discomfort afterward. Postprocedural disorientation may result with even small doses of ketamine, although its successful use in TAVR sedation protocols has been reported.
Choice of Sedatives
If MAC is deemed appropriate, there are a variety of sedatives to choose from. Two of the most attractive agents that allow for the goal of rapid recovery are dexmedetomidine and propofol. These agents have very different properties, but most importantly, both will allow for an interactive patient at the end of the procedure. No one drug will be ideal for all patients. Some patients may strongly prefer deeper sedation; others may opt for staying awake and interactive. Either medication can be used safely, depending on the patient’s particular physiologic and psychological milieu. It should also be noted that TAVR can and has been accomplished without sedation at all—words of comfort and liberal local anesthesia will usually suffice if sedation of any kind is deemed to be unsafe.
Dexmedetomidine is an agent that will allow for true “conscious sedation,” in which the patient is awake and aware but calm and comfortable. It does have a modest analgesic effect, such that the patient is tolerant of stimuli that would otherwise be unpleasant. Many patients actually prefer to remain cognizant of the procedure and their surroundings but do desire analgesia and anxiolysis, and dexmedetomidine is ideal for this cohort. It is not a respiratory depressant, which is useful in patients with minimal cardiopulmonary reserve.
There are disadvantages to dexmedetomidine worth considering. Especially when loaded as a standard 10-minute bolus, it can produce significant bradycardia. A transvenous pacemaker is routinely placed during TAVR, mitigating this concern. Another issue is its long half-life. Although the patient will not be oversedated postprocedurally and will remain lucid, some sense of lingering drug effect may be felt for an hour or 2 afterward.
The biggest downside to dexmedetomidine is that some patients will simply find the sedative effect inadequate, and may begin to move in a way that can threaten the safe conduct of the procedure. Dexmedetomidine monotherapy may best be avoided in patients with arthritic joints and backs, for whom lying flat for an extended time can be very uncomfortable.
All anesthesiologists are familiar with the use of propofol for procedural sedation. It has many of the properties of an ideal sedative, although with a few obvious drawbacks that require caution and diligence when it is chosen. Propofol can be used for light, moderate, or deep sedation that will be reliably satisfactory to both the patient and proceduralist. Patients anxious about the procedure will usually be amnestic of any noxious stimuli. The brisk neurologic recovery after propofol sedation is unmatched by any other sedative. While not analgesic, proper local anesthetic infiltration of the puncture sites is all that is required in this regard for TF TAVR. The author’s usual use of propofol for TAVR involves deeper sedation during local infiltration and sheath placement, then ideally lightening the sedation as tolerated, so that the patient is meaningfully interactive during BAV and valve deployment.
Despite propofol’s many attractive properties, vigilance and caution are mandatory during its use for TAVR. Even though carefully titrated infusions do not carry the same hemodynamic consequences as bolus doses, close attention must still be paid to systemic blood pressure. The drug can depress respiration and can also lead to airway obstruction.
Very importantly, airway secretions must be diligently managed, and any upper airway obstruction must be promptly and properly relieved. A patient cannot suddenly cough from airway secretions during critical portions of TAVR. Oropharyngeal or nasopharyngeal adjuncts can be used to ensure airway patency if necessary, although it is usually highly prudent to decrease sedation depth if airway obstruction is encountered. Alternatively, the anesthesiologist can induce true general anesthesia and insert a laryngeal mask airway if the sedation strategy is unsatisfactory to either the patient or operator.
If general anesthesia as the primary plan is indicated for the procedure, it should be conducted in a manner that will allow extubation on the procedural table and should avoid lingering drug effects. To that end, the author advises that opiates can be avoided entirely in naive elderly patients, even if an endotracheal tube is placed.
Esmolol and topical lidocaine will reliably blunt the sympathetic response to intubation if the patient’s hemodynamics are sufficiently robust that this is a concern. Rocuronium and vecuronium should usually be avoided in elderly patients for the short TAVR procedure given their decreased drug clearance, unless the new reversal agent sugammadex is used. Otherwise, cisatracurium is probably the best nondepolarizing neuromuscular blocking drug available for TAVR due to its predictable offset, and small doses generally suffice. Residual weakness must be diligently avoided.
The author prefers desflurane over sevoflurane for its marginally quicker offset and faster neurologic recovery, but either agent can be used per the anesthesiologist’s preference and experience. Extubation in the catheterization laboratory and avoidance of benzodiazepines and opiates enable the patient to achieve the goals of early mobilization and rehabilitation.
Health systems face substantial economic pressure to reduce LOS, equipment costs, and other drivers of overall expenditure. Consequently, facilities have sought ways to decrease recovery times, discharge patients earlier, and maximize procedural efficiency.
Modern TAVR with a minimally invasive approach limits physiologic trespass in the patient, reduces cost, reduces procedure time, and may reduce hospital LOS (Table 3). Many patients can be safely discharged on postprocedure day 1 or 2, and most can entirely avoid the ICU, going instead to a properly monitored step-down unit. Invasive lines with their inherent costs and risks can be avoided. Bladder catheterization is not routinely required. Patient satisfaction with this approach is high.
Given the rapid improvement in TAVR technology, it is incumbent on us as physician anesthesiologists to adapt our approach to the procedure as it exists now, rather than carry forward our biases about the difficulties encountered at the procedure’s inception. As stated in the excellent fictional work The House of God, “The delivery of good medical care is to do as much nothing as possible.”8We anesthesiologists would be wise to heed these words and perform our TAVR anesthetic care accordingly.
- Deeb GM, Reardon MJ, Chetcuti S, et al. Three-year outcomes in high-risk patients who underwent surgical or transcatheter aortic valve replacement.J Am Coll Cardiol. 2016;67(22):2565-2574.
- Leon M, Smith CR, Mack MJ, et al. Transcatheter or surgical aortic-valve replacement in intermediate-risk patients.N Engl J Med. 2016;374(17):1609-1620.
- Thourani VH, Kodali S, Makkar RR, et al. Transcatheter aortic valve replacement versus surgical valve replacement in intermediate-risk patients: a propensity score analysis.Lancet. 2016;387(10034):2218-2225.
- Kodali SK, Williams MR, Smith CR, et al. Two-year outcomes after transcatheter or surgical aortic-valve replacement.N Engl J Med. 2012;366(18):1686-1695.
- Leon MB, Smith CR, Mack M, et al. Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery.N Engl J Med. 2010;363(17):1597-1607.
- Giri J. Moderate vs. general anesthesia for transcatheter aortic valve replacement: an STS/ACC transcatheter valve therapy registry analysis. Presented at: Society for Cardiac Angiography and Interventions; May 6, 2016;Orlando,
- Babaliaros V, Devireddy C, Lerakis S, et al. Comparison of transfemoral transcatheter aortic valve replacement performed in the catheterization laboratory (minimalist approach) versus hybrid operating room (standard approach): outcomes and cost analysis.JACC Cardiovasc Interv. 2014;7(8):898-904.