Abstract

Background. Transcatheter aortic valve implantation (TAVI) is a treatment strategy for patients with severe aortic stenosis. Although general anaesthesia (TAVI-GA) and sedation (TAVI-S) have previously been described for TAVI, the difference in safety and efficacy of both methods has not been studied in a randomized trial.

Methods. The INSERT trial was a single centre, controlled parallel-group trial with balanced randomization. Sixty-six patients (68–94 yr) with acquired aortic stenosis undergoing transfemoral CoreValve™ were assigned to TAVI-GA or TAVI-S. Comparable operative risk was determined from risk-scores (EUROscore, STS-Score). Monitoring and anaesthetic drugs were standardized. Near-Infrared-Spectroscopy was used to monitor cerebral-oxymetry blinded. Primary outcome was the perioperative cumulative cerebral desaturation. As secondary outcomes, changes in neurocognitive function and respiratory and haemodynamic adverse events were evaluated.

Results. Of 66 included patients, 62 (TAVI-GA: n =31, TAVI-S: n =31) were finally analysed. Baseline characteristics were comparable. In 24 patients (39%) cerebral desaturation was observed. Cumulative cerebral desaturation was comparable (TAVI-GA:(median [IQR]) (0[0/1308] s%) vs. TAVI-S:(0[0/276] s%); P =0.505) between the groups. Neurocognitive function did not change within and between groups. Adverse events were more frequently observed in TAVI-S patients ( P <0.001). Bradypnoea ( n =16, 52%) and the need for airway manoeuvres ( n =11, 36%) or bag-mask-ventilation ( n =6, 19%) were the most common respiratory adverse events.

Conclusions. Cerebral desaturation occurred in both patient groups, but there was no significant difference between the two groups. Based on primary outcome, both methods were shown to be comparable. Neurocognitive outcome was similar. The higher incidence of adverse events in the sedation group suggests a potential advantage of general anaesthesia.

Clinical trial registration. NCT 01251328.

Editor's key points

  • Patients undergoing Transcatheter aortic valve implantation (TAVI) can be managed using sedation techniques or general anaesthesia.

  • This small study compared moderate to deep sedation using propofol and remifentanil with general anaesthesia.

  • There were no differences in cerebral oxygenation or neurocognitive outcome.

  • The incidence of adverse events, principally pain and respiratory depression, was much higher in the sedation group.

  • Moderate to deep sedation for TAVI has potential disadvantages compared with general anaesthesia.

Since 2002 transfemoral transcatheter aortic valve implantation (TAVI) 1 has continuously evolved with different approaches and techniques, 2 for treatment of severe aortic valve stenosis. A shift from general anaesthesia (TAVI-GA) towards sedation (TAVI-S) has been increasingly described. 3–6 Simplified monitoring, shortened procedure time, less haemodynamic instability, a decreased need for catecholamine therapy and shorter intensive care unit (ICU) stays were presented as benefits. 7,8 It is noteworthy that emergency conversion to general anaesthesia for TAVI-S has been reported in up to 17% of patients. 6,8

During procedural sedation, adverse events including hypoxemia have been reported in up to 21% of patients. 9 Cerebral oxymetry – as measured by Near-Infrared Spectroscopy (NIRS) – is a way of monitoring perfusion and oxygenation. 10,11 Induced cardiac arrest for balloon valvuloplasty and hypotension during valve release are two distinctive moments for cerebral desaturation to occur during TAVI, comparable with a decrease in cerebral saturation observed during brief cardiac arrest for ICD-testing. 12 In a different context perioperative cerebral desaturation has been linked to increased postoperative major organ morbidity and mortality. 13,14 Postoperative cognitive decline has been shown in desaturation of more than 3000 s%. 15 Neither the described benefits for TAVI-S, nor the equivalent safety of both procedures were examined in a randomized trial.

In this study we hypothesized that the cumulative perioperative cerebral desaturation is higher in patients undergoing TAVI-S because of a presumed lower PaO2 and SaO2 and as a result from a pilot study. As secondary endpoints we studied the difference in neurocognitive function, arterial blood-gas-analysis (ABG) and perioperative adverse events. In addition procedure-related time periods were evaluated.

Methods

The INSERT (Near INfrared Spectroscopy in aortic valvE ReplacemenT) trial was a single-centre, controlled, parallel-group study with balanced randomization (1:1). The trial was performed from February 2011 to April 2012 at the Institut für Anästhesiologie, Deutsches Herzzentrum München, a university-hospital specialized in cardiovascular diseases. Expert knowledge was obtained from approximately 600 previous TAVI patients. The study was approved by the medical ethics board of the Technische Universität München (Clinical trials: NCT01251328). Written informed consent was obtained from the patients before enrolment.

Patients

Inclusion criteria: patients with acquired severe aortic stenosis (age 18 yr and above), undergoing transfemoral CoreValve™ (Medtronic, Minneapolis, USA) implantation. The decision for transcatheter treatment was made by the cardiac surgeons. Patients had to be eligible to undergo the procedure using general anaesthesia or sedation and be able to undergo neurocognitive testing. Comparable operative risk was determined from comorbidities and risk-scores (EUROscore, STS-Score).

Patients with a predicted difficult airway, 16 severe pulmonary arterial hypertension (PAP syst >60 mm Hg) 17 or CPAP-therapy for obstructive sleep apnoea 18 were excluded from the study. Additional exclusion criteria were former/current alcohol or drug abuse, neurocognitive/neurodegenerative disease or psychiatric disorder, and use of antidepressants/sedatives.

Secondary exclusion criteria were abortion of procedure, perioperative emergency conversion from sedation to general anaesthesia and the need for cardiopulmonary resuscitation during the observation period, as the primary endpoint could not be determined in these conditions.

A sealed envelope containing the treatment assignment was allocated to the patients in the order of enrolment and opened at the patient's arrival in the operation room. The Institute of Medical Statistics and Epidemiology, Technische Universität München, provided the envelopes.

Treatment strategies

All patients received 3.75 mg of midazolam for oral premedication 30 min before arrival in the operating room. TAVI-GA and TAVI-S were performed using weight-adapted doses of propofol and remifentanil. TAVI-GA patients started with propofol 3 mg kg −1 h −1 and remifentanil 0.2 mcg kg −1 min −1 . A single dose of rocuronium (0.3 mg kg −1 ) was administered for tracheal intubation. In TAVI-S patients propofol 1 mg kg −1 h −1 and remifentanil 0.03 mcg kg −1 min −1 were used. In TAVI-S the cardiac surgeons infiltrated the groin with 10 ml of mepivacaine 1% at the beginning of the procedure. To provide optimal implantation conditions and to avoid patient movement especially during balloon valvuloplasty and valve implantation, moderate to deep sedation (according to the ASA 19 definition) was applied.

Depth of anaesthesia/sedation was monitored by Bispectral Analysis (BIS™, Covidien, Dublin, Ireland) throughout the whole course of anaesthesia and valve implantation. Values at distinctive time-points are presented in Table 2 . To achieve maximum oxygen supply, the inspiratory oxygen fraction ( FI O 2 ) was set to 1.0, and controlled ventilation was adjusted to an end-expiratory CO 2 of 4.6 kPa in TAVI-GA. TAVI-S patients received 8 Litre min −1 of oxygen for the same reason.

Apart from standard perioperative monitoring in accordance with ASA-guidelines, invasive arterial and central venous pressure blood pressure monitoring were applied. Perioperative transesophageal echocardiography was used in TAVI-GA patients, to monitor the result after implantation. Implantation was guided by fluoroscopy in both groups. A temporary 5-Fr pacemaker wire was inserted through the internal jugular vein for rapid ventricular pacing (RVP). Intraoperative hypotension (MAP<65 mm Hg) 20 was treated with a single i.v. norepinephrine bolus of five mcg. A continuous norepinephrine infusion was started in case five or more bolus administrations were needed. Haemodynamic parameters were documented at seven distinct perioperative time-points (at baseline, after induction, during and 60 s after RVP and valve implantation, and end of procedure). During the implantation of the self-expanding CoreValve™, two episodes of hypotension occured. The first one was during RVP-induced cardiac arrest for balloon valvuloplasty. The second episode was as a result of the temporary obstruction of the left ventricular outflow tract during valve implantation.

Two cardiac anaesthetists with experience in both procedures and perioperative care of TAVI patients (>250 patients each) were assigned to these implantations. One anaesthetist was engaged solely in study documentation, while the other anaesthetist conducted patient care. Cardiac surgeons were responsible for postoperative intensive care.

Primary endpoint

Procedure and anaesthesia-related impairment of circulation and gas-exchange may affect cerebral oxygenation 10 and initiate cerebral desaturation as demonstrated by our pilot-patients (see below). Therefore cerebral desaturation was chosen as a primary endpoint, assuming that is was greater in TAVI-S patients.

Perioperative NIRS measurement was conducted in a blinded ‘study’ mode. NIRS was monitored using INVOS 5100™ (Covidien, Dublin, Ireland), which has been previously described in detail. 11 Data were stored continuously in 6 s intervals, and analysed using the INVOS Monitoring-System-Analytics-Tool © (Covidien, Dublin, Ireland) after the last patient completed surgery. The room air rSO 2 baseline (rSO 2-RA baseline ) parameter was defined as the average rSO 2 value in a 30 s period, during the insertion of a radial artery catheter. Neither sedative drugs nor oxygen was administered before the measurement. An individual desaturation threshold (rSO 2-DESAT ) was determined for each patient. The threshold was defined as a decrease of 20% below the rSO 2-RA baseline or an absolute value below 50%, whichever was higher. 15 In case of interhemispheric rSO 2 differences (rSO 2-Δinterhem ), the lower value hemisphere was used for analysis. 21 Area-under-the curve (AUC) analysis was conducted to assess rSO 2 decreases during the entire case. 15 The AUC was calculated as  

AUC rSO2-DESAT=[rSO2-DESAT(%)rSO2current(%)]time (Sec)

The primary endpoint (AUC rSO 2 (total<DESAT) ) was defined as the calculated summation of all AUC rSO 2-DESAT.

Secondary endpoints

Neurocognitive outcome

All patients underwent neurocognitive testing the day before the intervention and on the seventh postoperative day or the day of hospital release (if earlier). The tests were provided by the Department of Psychiatry, Technische Universität München, and consisted of three parts. ‘DemTect’ is an age-adjusted screening test for cognitive dysfunction. The DemTect scale ranges from 0 to 18 points, with more than 13 points representing good age-based cognitive function. 22 ‘Zahlenverbindungstest’ (ZVT-1/2) examines the perceptual speed needed to connect 30 apparently arbitrary numbers. 23 ‘Regensburger-Wortflüssigkeitstest’ (RWT) 24 is an education-adjusted test to assess the formal lexical stream of speech. Patients had to name as many words with a given first letter (‘P’) in a period of 2 min. Speech fluency in the German language is mandatory for these tests. At the time of the preoperative testing, the randomized anaesthesia technique was unknown.

Arterial blood gas analysis (ABG)

ABG ( PaO2,PaCO2,SaO2, haemoglobin, pH, and lactate) was performed at baseline, after induction of anaesthesia, 60 s after RVP, and 120 s after extubation.

Adverse events and procedure-related time periods

Anaesthesia and procedure-related adverse events were predefined and documented at the time of occurrence. ‘Difficult airway’ was defined according to ASA guidelines as a clinical condition in which a trained anaesthetist experienced difficulty with face-mask ventilation of the upper-airway, difficulty with tracheal intubation, or both. 16 ‘Difficult central venous catheter’ was defined as a clinical condition in which a trained anaesthetist needed more than 2 attempts of puncturing the internal jugular vein, despite the use of ultrasound, difficulty in advancing the guidewire, or both. ‘Defibrillation’ was used in a clinical condition with tachyarrhythmia and hypotension. ‘Jaw-thrust-manoeuvre’ was used in case of upper airway obstruction because of loss of pharyngeal muscle tone. A ‘nasopharyngeal airway’ was inserted in situations of a permanent need of jaw-thrust-manoeuvre. ‘Bag-valve mask ventilation’ was used in patients with bradypnoea/apnoea and/or hypoxia. ‘Bradypnoea’ was defined as a respiratory rate <8 min −1 . ‘Hypoxia’ was defined as a SpO2<90% over 10 s. 9 ‘Unrest/pain’ was defined as any undesired targeted/untargeted movement from the patient or expression of pain throughout the whole anaesthetic and interventional procedure.

Statistical analysis

Categorical data are presented as absolute and relative frequencies. Group comparisons were performed using χ 2 tests or Fisher's exact test depending on the expected cell counts of corresponding contingency tables. The distribution of continuous data is given by the median and interquartile range (IQR=25–75th percentile). Mann-Whitney U -tests and Hodges-Lehmann 95% confidence intervals were used to assess differences between groups. All statistical tests were performed using a two-sided 5% significance level. Derived from a pilot study, expected mean AUC rSO 2 (total<DESAT) of 2000 s% (TAVI-S) and 1000 s% (TAVI-GA) with standard deviations of 1500 and 500 have been used for a conservative sample size calculation. A group size of 28 individuals was found to result in a 90% power (β=10%) for a two-sided two-sample t-test to detect such a difference of α=5% level of significance (N-Query Advisor 7.0). The application of Pitmans ARE (Asymptotic Relative Efficiency)=0.864 lead to an adjusted sample size of 28/0.864=32.4 which is approximately 33 subjects per treatment arm. SPSS Vers.21 (IBM, Germany) and SAS Enterprise Guide 4.3 (SAS Institute, Cary, USA) were used for analysis.

Results

In the study period 100 patients were undergoing transfemoral CoreValve™ implantation, 66 patients (33 TAVI-GA, 33 TAVI-S) were enrolled. Details of patient enrolment are shown in Fig. 1 . Two TAVI-GA patients were excluded because of the need for perioperative cardio-pulmonary resuscitation (CPR). Both patients required CPR after valve implantation. The first one as a result of prolonged hypotension. At the end of the procedure the patient needed a second episode of resuscitation and drainage of pericardial effusion. After four days of ICU care the patient was discharged on POD 8. The other patient was resuscitated as a result of prolonged ventricular fibrillation. At the end of operation the patient was transferred awake to the ICU for a three-day stay and discharged on POD 9. One TAVI-S patient was excluded because of CPR and emergency conversion to general anaesthesia, as a result of signs of heart failure and respiratory distress at the beginning of sedation. After recompensation transfemoral CoreValve™ was successfully implanted. The patient was extubated at the end of procedure and discharged home POD 13.

Fig 1

CONSORT Flow Diagram.

Fig 1

CONSORT Flow Diagram.

One additional TAVI-S patient was excluded because of difficulties with femoral vascular access and abortion of the procedure. In each group, 31 allocated patients were analysed.

As shown in Table 1 , the patients' preoperative baseline characteristics were comparable. Median age of TAVI-S patients was 4 yr higher leading to a higher logistic EUROscore, while other risk scores (STS) were comparable.

Table 1

Preoperative patient characteristics. Data are given as median and 25–75th percentile for continuous variables. Categorical variables are given as percentage. BMI, Body Mass Index; NYHA, New York Heart Association classification; ASA, American Society of Anesthesiologists classification; STS, Society of Thoracic Surgeons

 TAVI-GA TAVI-S 
n =31  n =31  
Female patients n ( %)  18 (58) 13 (42) 
Age (yr) 80 (75/84) 84 (79/86) 
BMI (kg/m 2 )  25 (23/30) 27 (24/29) 
Aortic Valve Area (cm 2 )  0.70 (0.56/0.83) 0.63 (0.60/0.80) 
Left-Ventricular Ejection Fraction (%) 55 (43/60) 57 (40/62) 
Pulmonary Artery Pressure (mm Hg) 43 (35/50) 47 (42/60) 
Mean pressure gradient aortic valve (mm Hg) 54 (37/69) 46 (40/56) 
Maximum pressure gradient aortic valve (mm Hg) 64 (45/88) 65 (45/82) 
NYHA (III/IV) n (%)  24 (77) 22 (71) 
EUROScore log (%)* 9.72 (7.01/18.26) 11.66 (10.32/21.43) 
STS Mortality Score (%) 4.3 (2.8/10.2) 5.0 (3.6/7.4) 
STS Morbidity Score (%) 23.3 (16.9/32.4) 25.4 (19.6/29.7) 
 TAVI-GA TAVI-S 
n =31  n =31  
Female patients n ( %)  18 (58) 13 (42) 
Age (yr) 80 (75/84) 84 (79/86) 
BMI (kg/m 2 )  25 (23/30) 27 (24/29) 
Aortic Valve Area (cm 2 )  0.70 (0.56/0.83) 0.63 (0.60/0.80) 
Left-Ventricular Ejection Fraction (%) 55 (43/60) 57 (40/62) 
Pulmonary Artery Pressure (mm Hg) 43 (35/50) 47 (42/60) 
Mean pressure gradient aortic valve (mm Hg) 54 (37/69) 46 (40/56) 
Maximum pressure gradient aortic valve (mm Hg) 64 (45/88) 65 (45/82) 
NYHA (III/IV) n (%)  24 (77) 22 (71) 
EUROScore log (%)* 9.72 (7.01/18.26) 11.66 (10.32/21.43) 
STS Mortality Score (%) 4.3 (2.8/10.2) 5.0 (3.6/7.4) 
STS Morbidity Score (%) 23.3 (16.9/32.4) 25.4 (19.6/29.7) 

Primary endpoint

There was no significant difference in AUC rSO 2 (total<DESAT) (Fig. 2 C) between TAVI-GA (median [IQR]) (0 [0/1308] s%) and TAVI-S (0 [0/276] s%, P =0.505).

Fig 2

Perioperative evolution rSO 2 . ( a and b ) Green lines show the individual perioperative rSO 2 values. The blue line shows the Median and the 25–75th percentile. ( c ) Cummulative Area under the curve (AUC) for individual patients in both groups below the desaturation (DESAT) threshold. The pink line indicates 3000 s% ( d ) Perioperative mean arterial pressure. Boxplots show Median, 25–75th percentile, Minimum, Maximum and outliers within 5–95th percentile. TAVI-GA – general anaesthesia, TAVI-S sedation, RVP – Rapid Ventricular Pacing for balloon valvuloplasty.

Fig 2

Perioperative evolution rSO 2 . ( a and b ) Green lines show the individual perioperative rSO 2 values. The blue line shows the Median and the 25–75th percentile. ( c ) Cummulative Area under the curve (AUC) for individual patients in both groups below the desaturation (DESAT) threshold. The pink line indicates 3000 s% ( d ) Perioperative mean arterial pressure. Boxplots show Median, 25–75th percentile, Minimum, Maximum and outliers within 5–95th percentile. TAVI-GA – general anaesthesia, TAVI-S sedation, RVP – Rapid Ventricular Pacing for balloon valvuloplasty.

Sensitivity analysis

There was no difference in rSO 2 values between the two groups (Table 2 ). All patients showed a decrease in rSO 2 during RVP and valve implantation (Fig. 2 A and B). The baseline values of rSO 2 before these interventions were comparable (Table 2 ). Twenty-four patients (39%) experienced perioperative decrease in rSO 2 below the desaturation threshold (TAVI-GA: n =13, TAVI-S: n =11; P =0.602). Maximal calculated desaturation was 27 174 s% in TAVI-GA and 14 502 s% in TAVI-S patients. Eight of these patients (TAVI-GA: n =5, TAVI-S: n =3; P =0.707) experienced a cumulative perioperative desaturation of more than 3000 s% (Fig. 2 C).

Table 2

Perioperative course of rSO 2 and procedure related values. Data are given as median and 25–75th percentile for continuous variables. Categorical variables are given as percentage. rSO 2 , Cerebral Saturation; ICU, Intensive Care Unit; BIS, Bispectral index; RVP, Rapid ventricular pacing. *Indicates a P <0.05

 TAVI-GA TAVI-S P -value  
n =31  n =31  
Procedure related timeline 
Induction time (min) 45 (40/50) 43 (35/50) 0.276 
Duration of procedure (min) 72 (62/85) 75 (62/90) 0.667 
Skin closure to extubation or equivalent alertness (min) 1 (−2/4) −1 (−4/3) 0.235 
Arrival of patient to handover to ICU (min) 170 (155/185) 170 (153/185) 0.983 
Duration RVP (s) 22 (17/28) 21 (18/25) 0.621 
Duration Valve implantation (s) 230 (170/339) 300 (179/376) 0.371 
rSO 2 values  
rSO 2 at room air baseline (%)  60 (53/66) 57 (54/67) 0.916 
rSO 2 interhemispheric difference (%)  0 (−3/3) 0 (−5/3) 0.789 
rSO2 with oxygen after induction (%) 65 (58/70) 68 (63/78) 0.111 
rSO 2 before RVP (%)  68 (60/71) 70 (64/79) 0.107 
rSO 2 minimal value during RVP (%)  58 (51/64) 60 (56/67) 0.323 
rSO 2 before valve implantation (%)  68 (58/72) 69 (64/78) 0.161 
rSO 2 minimal value during valve impl. (%)  60 (51/66) 62 (54/71) 0.430 
rSO 2 120 S after extubation (%)  73 (62/78) 72 (66/76) 0.921 
Drug usage 
Remifentanil total (µg/kg) 25.3 (18.3/34.7) 4.5 (3.7/5.2) <0.001* 
Propofol total (mg/kg) 8.3 (6.1/9.9) 2.6 (2.1/3.6) <0.001* 
Norepinephrine total (µg/kg) 1.95 (1.07/4.78) 1.67 (0.62/3.13) 0.495 
Depth of Anaesthesia 
BIS pre-induction 97 (92/98) 95 (89/97) 0.155 
BIS at entering OR 40 (37/46) 83 (73/86) <0.001 
BIS before RVP 40 (36/43) 72 (58/81) <0.001 
BIS during RVP (minimal value) 39 (34/40) 60 (51/74) <0.001 
BIS during valve release (minimal value) 39 (31/46) 56 (50/75) <0.001 
BIS during skin closure 86 (75/95) 89 (86/92) 0.486 
Postoperative 
Stay on ICU (days) 2 (1/4) 4 (1/6) 0.019* 
Stay on Ward (days) 4 (3/6) 5 (2/6) 0.603 
Total stay in hospital (days) 7 (6/8) 8 (7/11) 0.050 
In hospital mortality n (%)  1 (3) 1.000 
Adverse events 
Unrest/Pain n (%)  19 (61) <0.001* 
Bradypnoea n (%)  n.a. 16 (52) n.a. 
SpO 2 <90% n (%)  5 (16) 0.053 
Jaw-Thrust/nasopharyngeal airway n (%)  n.a. 11 (36) n.a. 
Bag-Mask Ventilation n (%)  n.a. 6 (19) n.a. 
Difficult tracheal intubation n (%)  3 (10) n.a. n.a. 
Bronchospasm n (%)  1 (3) 1.000 
Difficult central venous catheter n (%)  1 (3) 3 (10) 0.612 
Defibrillation n (%)  1 (3) 1.000 
Perioperative Stroke n (%)  1 (3) 1.000 
Cumulative no. of patients with adverse events n (%)  4 (13) 29 (94) <0.001* 
Renal parameters 
Creatinine at time of admission (mg/dl) 0.91 (0.78/1.03) 0.94 (0.74/1.34) 0.205 
Creatinine at time of discharge (mg/dl) 0.92 (0.81/1.10) 0.86 (0.67/1.32) 0.899 
Need for dialysis n (%)  2 (6) 0.492 
Procedural outcome n =29  n =27   
Mean pressure gradient aortic valve at time of discharge (mm Hg) 11 (7/12) 12 (9/15) 0.079 
Aortic Valve Area at time of discharge (cm 2 )  1.55 (1.29/1.79) 1.50 (1.35/1.80) 0.975 
Aortic Regurgitation at discharge n =29  n =29  0.315 
Grade 0 (%) 13 (45) 7 (26)  
Grade I (%) 14 (48) 16 (59)  
Grade II (%) 2 (7) 4 (15)  
 TAVI-GA TAVI-S P -value  
n =31  n =31  
Procedure related timeline 
Induction time (min) 45 (40/50) 43 (35/50) 0.276 
Duration of procedure (min) 72 (62/85) 75 (62/90) 0.667 
Skin closure to extubation or equivalent alertness (min) 1 (−2/4) −1 (−4/3) 0.235 
Arrival of patient to handover to ICU (min) 170 (155/185) 170 (153/185) 0.983 
Duration RVP (s) 22 (17/28) 21 (18/25) 0.621 
Duration Valve implantation (s) 230 (170/339) 300 (179/376) 0.371 
rSO 2 values  
rSO 2 at room air baseline (%)  60 (53/66) 57 (54/67) 0.916 
rSO 2 interhemispheric difference (%)  0 (−3/3) 0 (−5/3) 0.789 
rSO2 with oxygen after induction (%) 65 (58/70) 68 (63/78) 0.111 
rSO 2 before RVP (%)  68 (60/71) 70 (64/79) 0.107 
rSO 2 minimal value during RVP (%)  58 (51/64) 60 (56/67) 0.323 
rSO 2 before valve implantation (%)  68 (58/72) 69 (64/78) 0.161 
rSO 2 minimal value during valve impl. (%)  60 (51/66) 62 (54/71) 0.430 
rSO 2 120 S after extubation (%)  73 (62/78) 72 (66/76) 0.921 
Drug usage 
Remifentanil total (µg/kg) 25.3 (18.3/34.7) 4.5 (3.7/5.2) <0.001* 
Propofol total (mg/kg) 8.3 (6.1/9.9) 2.6 (2.1/3.6) <0.001* 
Norepinephrine total (µg/kg) 1.95 (1.07/4.78) 1.67 (0.62/3.13) 0.495 
Depth of Anaesthesia 
BIS pre-induction 97 (92/98) 95 (89/97) 0.155 
BIS at entering OR 40 (37/46) 83 (73/86) <0.001 
BIS before RVP 40 (36/43) 72 (58/81) <0.001 
BIS during RVP (minimal value) 39 (34/40) 60 (51/74) <0.001 
BIS during valve release (minimal value) 39 (31/46) 56 (50/75) <0.001 
BIS during skin closure 86 (75/95) 89 (86/92) 0.486 
Postoperative 
Stay on ICU (days) 2 (1/4) 4 (1/6) 0.019* 
Stay on Ward (days) 4 (3/6) 5 (2/6) 0.603 
Total stay in hospital (days) 7 (6/8) 8 (7/11) 0.050 
In hospital mortality n (%)  1 (3) 1.000 
Adverse events 
Unrest/Pain n (%)  19 (61) <0.001* 
Bradypnoea n (%)  n.a. 16 (52) n.a. 
SpO 2 <90% n (%)  5 (16) 0.053 
Jaw-Thrust/nasopharyngeal airway n (%)  n.a. 11 (36) n.a. 
Bag-Mask Ventilation n (%)  n.a. 6 (19) n.a. 
Difficult tracheal intubation n (%)  3 (10) n.a. n.a. 
Bronchospasm n (%)  1 (3) 1.000 
Difficult central venous catheter n (%)  1 (3) 3 (10) 0.612 
Defibrillation n (%)  1 (3) 1.000 
Perioperative Stroke n (%)  1 (3) 1.000 
Cumulative no. of patients with adverse events n (%)  4 (13) 29 (94) <0.001* 
Renal parameters 
Creatinine at time of admission (mg/dl) 0.91 (0.78/1.03) 0.94 (0.74/1.34) 0.205 
Creatinine at time of discharge (mg/dl) 0.92 (0.81/1.10) 0.86 (0.67/1.32) 0.899 
Need for dialysis n (%)  2 (6) 0.492 
Procedural outcome n =29  n =27   
Mean pressure gradient aortic valve at time of discharge (mm Hg) 11 (7/12) 12 (9/15) 0.079 
Aortic Valve Area at time of discharge (cm 2 )  1.55 (1.29/1.79) 1.50 (1.35/1.80) 0.975 
Aortic Regurgitation at discharge n =29  n =29  0.315 
Grade 0 (%) 13 (45) 7 (26)  
Grade I (%) 14 (48) 16 (59)  
Grade II (%) 2 (7) 4 (15)  

Secondary endpoints

Neurocognitive outcome

All three neurocognitive tests revealed no significant pre- to postoperative difference neither within nor between the two patient groups (Fig. 3 ). Memory function (DemTect) was comparable between TAVI-GA (0 [−2/2] points) and TAVI-S (0 [−2/2] points, P =0.817). Both parts of the ZVT (ZVT 1: TAVI-GA: 3 [−4/14] s, TAVI-S: −2 [−7/8] s, P =0.307; ZVT 2: TAVI-GA: 2 [−5/8] s, TAVI-S: 3 [−4/18] s, P =0.362) revealed no difference in pre- to postoperative testing. Executive function (RWT) was also similar in both groups (TAVI-GA: −1 [−3/3] s, TAVI-S: 0 [−3/2] words, P =0.815).

Fig 3

Pre- to postoperative difference in neurocognitive testing. P -values between TAVI-patients in general anaesthesia (TAVI-GA, blue) and sedation (TAVI-S, green). Neurocognitive tests: DemTect, RWT – Regensburger Wortflüssigkeitstest, ZVT – Zahlenverbindungstest part 1 and 2.

Fig 3

Pre- to postoperative difference in neurocognitive testing. P -values between TAVI-patients in general anaesthesia (TAVI-GA, blue) and sedation (TAVI-S, green). Neurocognitive tests: DemTect, RWT – Regensburger Wortflüssigkeitstest, ZVT – Zahlenverbindungstest part 1 and 2.

Arterial blood gas analysis

As shown in Fig. 4 , baseline ABG measurements were similar in both groups. In the TAVI-S group PaCO2 increased significantly after the start of sedation (TAVI-GA: (5.2 [4.6/5.5] kPa; TAVI-S: 6.7 [6.0/7.5] kPa, P <0.001) and continued to stay higher perioperatively. Simultaneously, respiratory acidosis was observed (pH TAVI-GA: (7.45 [7.41/7.48]; pH TAVI-S: 7.35 [7.31/7.42], P <0.001). PaO2 was significantly higher in TAVI-GA patients during this period (TAVI-GA: (48 [39.3/53.3] kPa; TAVI-S: 22.8 [16.3/27.2] kPa, P <0.001). At the end of the procedure, ABG values were similar in both groups. Haemoglobin values remained comparable during the procedure (Fig. 4 ).

Fig 4

Blood gas analysis of arterial blood samples. TAVI-GA – general anaesthesia (blue), TAVI-S – Sedation (green). Boxplots show Median, 25–75th percentile, Minimum, Maximum and outliers within 5–95th percentile.

Fig 4

Blood gas analysis of arterial blood samples. TAVI-GA – general anaesthesia (blue), TAVI-S – Sedation (green). Boxplots show Median, 25–75th percentile, Minimum, Maximum and outliers within 5–95th percentile.

Adverse events and procedure-related time periods

Several perioperative adverse events occurred during the study, as shown in Table 2 . In 29 of 31 (94%) TAVI-S and in 4 of 31 TAVI-GA patients (13%) ( P <0.001) adverse events were observed. One TAVI-S suffered a perioperative severe stroke (confirmed by neuroimaging) without signs of cerebral desaturation. This patient was transferred to an external ICU and died on POD 19. A second TAVI-S patient was re-admitted to ICU on POD 8 and died on POD 22 from septic multi-organ failure.

Induction time, procedure time, and time from arrival of the patient until handover to the ICU were comparable as shown in Table 2 . The anaesthetic drug dosage was higher in TAVI-GA patients. As to be expected, perioperative BIS values were lower in TAVI-GA (Table 2 ) during the operation. There was no statistically significant difference in catecholamine use. The ICU stay was statistically significantly shorter in TAVI-GA patients (TAVI-GA: 2 [1/4] days; TAVI-S: 4 [1/6] days; P =0.019)).

Procedural outcome and postoperative kidney injury

Procedural outcome defined as aortic regurgitation, mean pressure gradient and aortic valve area at discharge were comparable in TAVI-GA and TAVI-S patients (Table 2 ). Two TAVI-S patients required temporary postoperative dialysis. Pre- and postoperative creatinine values did not differ between TAVI-GA and TAVI-S patients (Table 2 ).

Discussion

The results of this study show that cerebral desaturation occurred both during general anaesthesia and sedation without a statistically significant and clinical difference in primary outcome (Table 2 , Fig. 2 ). Sedation was associated with more adverse events, such as respiratory events. Longer periods of cerebral desaturations are associated with adverse outcome. In this study episodes of perioperative cerebral desaturation were monitored especially during balloon valvuloplasty and valve implantation, but were too short to produce desaturation to a larger extend. Our data are comparable with those recently published in a cohort of patients undergoing cardiac surgery 21 and baseline values described in patients undergoing transapical TAVI. 25 Although NIRS is a localized trend monitoring and only parts of the frontal lobe are monitored for signs of hypoperfusion and/or desaturation, it is the best available method to monitor the common endpoint of cerebral perfusion and oxygenation. In addition the chosen desaturation threshold complies with the current consensus. 15 Although extracranial contamination has been described, with INVOS™ we used the technique that has been used in most trials. 14,15,21

Three neurocognitive tests examining different types of cognitive function did not reveal any significant difference between pre- and postoperative testing. These data are in accordance with results recently published, 26 showing that more than 90% of sedated TAVI patients had no signs of cognitive deficit. From our point of view, neither general anaesthesia nor sedation impairs postoperative neurocognitive function in this specific group of patients (Fig. 3 ).

Perioperative PaCO2 was significantly higher in TAVI-S patients with consecutive respiratory acidosis. In moderate to deep sedation PaCO2 is often difficult to control as most sedatives and opioids affect respiration to a certain degree. 25 Respiratory depression, hypercarbia, and acidosis may increase pulmonary artery pressure (PAP) and may lead to right ventricular failure. 27 Pulmonary hypertension (PH) has been described in TAVI patients in up to 50%. 28 As PAP is easier to control, we consider general anaesthesia as favourable in patients with pre-existing PH.

Shorter procedure times for TAVI-S have been described 7,29 but these studies were not randomized. In addition different ways of haemodynamic monitoring, different techniques for femoral access and different periods of the learning curve were compared. Our study was performed with an experienced team, and the same haemodynamic monitoring was used, showing no time-benefit for sedation. On the contrary, ICU stay was significantly shorter in TAVI-GA patients. The reason for this difference remains unclear.

Several adverse events were observed in our study. A high rate of respiratory events and the need for bag-valve mask ventilation in 19% of TAVI-S patients, occurred even in the presence of experienced anaesthetists (Table 2 ). This number is in agreement with a previously described rate of unplanned perioperative intubations of 17%. 6 Unrest and pain may not be considered severe adverse events but are disturbing for the surgeon. Movement of the legs may also cause injury to the femoral vessels. Despite moderate to deep sedation and infiltration of local anaesthesia, unrest and pain were common during sedation. These adverse events in TAVI-S patients expose the potential risks of moderate to deep sedation for this procedure. There was no significant impact of anaesthesia to the implantation result. The tendency towards more aortic regurgitation in TAVI-S is in accordance with the recent published FRANCE-2 registry. 30 With the introduction of novel devices and the reduction of the diameter of the delivery system to 14 French, unrest and pain may be less problematic. As the experience with these new devices is rapidly growing, moderate or even minimal sedation might be sufficient and decrease the rate of respiratory events. 16 Further studies are needed to prove this concept.

Limitations

This study was powered for the primary endpoint of cerebral desaturation and not for the secondary outcome parameters of neurocognitive outcome and adverse events. This might be a limitation for these results. As we did not measure PAP, the effect of hypercapnia on the right ventricular function and PAP in our TAVI-S patients remains unclear. Although the study was blinded for most parts, the anaesthesia technique could not be blinded for obvious reasons.

Conclusion

In summary, TAVI in sedation did not offer an advantage with respect to cerebral desaturation and neurocognitive outcome compared with general anaesthesia. Moreover sedation was associated with more adverse events, such as respiratory events, and ICU longer stays. Therefore these data suggest an advantage of general anaesthesia vs moderate to deep sedation.

Authors' contributions

Study design/planning: N.P.M., A.H., A.K., B.B.

Study conduct: N.P.M.

Data analysis: N.P.M., A.H.

Writing paper: N.P.M., A.H.

Revising paper: all authors

Declaration of interest

None declared.

Funding

Funding was restricted to institutional and departmental resources.

Acknowledgements

The authors would like to thank our study nurse Doris Kienmoser, MTA for her help in conducting this study.

References

1
Cribier
A
Eltchaninoff
H
Bash
A
et al
Percutaneous transcatheter implantation of an aortic valve prosthesis for calcific aortic stenosis: first human case description
.
Circulation
 
2002
;
106
:
3006
8
2
Rodes-Cabau
J
.
Transcatheter aortic valve implantation: current and future approaches
.
Nat Rev Cardiol
 
2012
;
9
:
15
29
3
Rex
S
.
Anesthesia for transcatheter aortic valve implantation: an update
.
Curr Opin Anaesthesiol
 
2013
;
26
:
456
66
4
Bufton
KA
Augoustides
JG
Cobey
FC
.
Anesthesia for transfemoral aortic valve replacement in North America and Europe
.
J Cardiothorac Vasc Anesth
 
2013
;
27
:
46
9
5
Guarracino
F
Cabrini
L
Baldassarri
R
et al
Noninvasive ventilation for awake percutaneous aortic valve implantation in high-risk respiratory patients: a case series
.
J Cardiothorac Vasc Anesth
 
2011
;
25
:
1109
12
6
Bergmann
L
Kahlert
P
Eggebrecht
H
Frey
U
Peters
J
Kottenberg
E
.
Transfemoral aortic valve implantation under sedation and monitored anaesthetic care--a feasibility study
.
Anaesthesia
 
2011
;
66
:
977
82
7
Dehedin
B
Guinot
PG
Ibrahim
H
et al
Anesthesia and perioperative management of patients who undergo transfemoral transcatheter aortic valve implantation: an observational study of general versus local/regional anesthesia in 125 consecutive patients
.
J Cardiothorac Vasc Anesth
 
2011
;
25
:
1036
43
8
Dvir
D
Jhaveri
R
Pichard
AD
.
The minimalist approach for transcatheter aortic valve replacement in high-risk patients
.
JACC Cardiovasc Interv
 
2012
;
5
:
468
9
9
Weaver
CS
Hauter
WE
Brizendine
EJ
Cordell
WH
.
Emergency department procedural sedation with propofol: is it safe?
J Emerg Med
 
2007
;
33
:
355
61
10
Murkin
JM
.
Cerebral oximetry: monitoring the brain as the index organ
.
Anesthesiology
 
2011
;
114
:
12
3
11
Ghosh
A
Elwell
C
Smith
M
.
Review article: cerebral near-infrared spectroscopy in adults: a work in progress
.
Anesth Analg
 
2012
;
115
:
1373
83
12
de Vries
JW
Visser
GH
Bakker
PF
.
Neuromonitoring in defibrillation threshold testing. A comparison between EEG, near-infrared spectroscopy and jugular bulb oximetry
.
J Clin Monit
 
1997
;
13
:
303
7
13
Casati
A
Fanelli
G
Pietropaoli
P
et al
Continuous monitoring of cerebral oxygen saturation in elderly patients undergoing major abdominal surgery minimizes brain exposure to potential hypoxia
.
Anesth Analg
 
2005
;
101
:
740
7
,
table of contents
14
Murkin
JM
Adams
SJ
Novick
RJ
et al
Monitoring brain oxygen saturation during coronary bypass surgery: a randomized, prospective study
.
Anesth Analg
 
2007
;
104
:
51
8
15
Slater
JP
Guarino
T
Stack
J
et al
Cerebral oxygen desaturation predicts cognitive decline and longer hospital stay after cardiac surgery
.
Ann Thorac Surg
 
2009
;
87
:
36
44
;
discussion -5
16
American Society of Anaesthesiologists
.
Practice guidelines for management of the difficult airway: an updated report by the American Society of Anesthesiologists Task Force on Management of the Difficult Airway
.
Anesthesiology
 
2003
;
98
:
1269
77
17
Roques
F
Nashef
SA
Michel
P
et al
Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19030 patients
.
Eur J Cardiothorac Surg
 
1999
;
15
:
816
22
;
discussion 22-3
18
Gross
JB
Bachenberg
KL
Benumof
JL
et al
Practice guidelines for the perioperative management of patients with obstructive sleep apnea: a report by the American Society of Anesthesiologists Task Force on Perioperative Management of patients with obstructive sleep apnea
.
Anesthesiology
 
2006
;
104
:
1081
93
;
quiz 117-8
19
American Society of Anaesthesiologists
.
Continuum of depth of sedation: Definition of general anesthesia and lebals of sedation/analgesia. Standards, Guidelines, Statements and Other Documents
.
20
Rivers
E
Nguyen
B
Havstad
S
et al
Early goal-directed therapy in the treatment of severe sepsis and septic shock
.
N Engl J Med
 
2001
;
345
:
1368
77
21
Heringlake
M
Garbers
C
Kabler
JH
et al
Preoperative cerebral oxygen saturation and clinical outcomes in cardiac surgery
.
Anesthesiology
 
2011
;
114
:
58
69
22
Kalbe
E
Kessler
J
Calabrese
P
et al
DemTect: a new, sensitive cognitive screening test to support the diagnosis of mild cognitive impairment and early dementia
.
Int J Geriatr Psychiatry
 
2004
;
19
:
136
43
23
Oswald
W
Roth
E
.
Der Zahlenverbindungs-Test (ZVT)
  .
Göttingen, Toronto, Zürich
:
Hogrefe
,
1987
24
Aschenbrenner
S
Tucha
O
Lange
K
.
RWT - Regensburger Wortflüssigkeits-Test Göttingen
  .
Berlin, Toronto, Seattle
:
Hogrefe
,
2000
25
Paarmann
H
Heringlake
M
Heinze
H
et al
Non-invasive cerebral oxygenation reflects mixed venous oxygen saturation during the varying haemodynamic conditions in patients undergoing transapical transcatheter aortic valve implantation
.
Interact Cardiovasc Thorac Surg
 
2012
;
14
:
268
72
26
Ghanem
A
Kocurek
J
Sinning
JM
et al
Cognitive Trajectory After Transcatheter Aortic Valve Implantation
.
Circ Cardiovasc Interv
 
2013
;
6
:
615
24
27
Minai
OA
Yared
JP
Kaw
R
Subramaniam
K
Hill
NS
.
Perioperative risk and management in patients with pulmonary hypertension
.
Chest
 
2013
;
144
:
329
40
28
Ben-Dor
I
Goldstein
SA
Pichard
AD
et al
Clinical profile, prognostic implication, and response to treatment of pulmonary hypertension in patients with severe aortic stenosis
.
Am J Cardiol
 
2011
;
107
:
1046
51
29
Ben-Dor
I
Looser
PM
Maluenda
G
et al
Transcatheter aortic valve replacement under monitored anesthesia care versus general anesthesia with intubation
.
Cardiovasc Revasc Med
 
2012
;
13
:
207
10
30
Oguri
A
Yamamoto
M
Mouillet
G
Gilard
M
Laskar
M
Eltchaninoff
H
et al
Clinical Outcomes and Safety of Transfemoral Aortic Valve Implantation Under General Versus Local Anesthesia: Subanalysis of the French Aortic National CoreValve and Edwards 2 Registry
.
Circ Cardiovasc Interv
 
2014
;
7
:
602
10

Author notes

Comments

0 Comments