You can respond to this article at http://www.anaesthesiacorrespondence.com
Arrhythmogenic risk of pulmonary artery catheterisation in patients with severe aortic stenosis undergoing transcatheter aortic valve implantation
Article first published online: 5 NOV 2012
Anaesthesia © 2012 The Association of Anaesthetists of Great Britain and Ireland
Volume 68, Issue 1, pages 46–51, January 2013
How to Cite
Bergmann, L., Großwendt, T., Kahlert, P., Konorza, T., Wendt, D., Thielmann, M., Heusch, G., Peters, J. and Kottenberg, E. (2013), Arrhythmogenic risk of pulmonary artery catheterisation in patients with severe aortic stenosis undergoing transcatheter aortic valve implantation. Anaesthesia, 68: 46–51. doi: 10.1111/anae.12069
- Issue published online: 11 DEC 2012
- Article first published online: 5 NOV 2012
- Manuscript Accepted: 2 OCT 2012
Many clinicians consider severe aortic stenosis to be a contraindication to pulmonary artery catheterisation, except during open heart surgery with cardiopulmonary bypass. This is due to the perceived high risk of arrhythmia, although the true incidence of ventricular tachycardia and fibrillation remains unclear. We conducted a retrospective study to estimate the incidence of severe arrhythmias during pulmonary artery catheterisation in 380 patients with severe aortic stenosis scheduled for transcatheter aortic valve implantation. Ventricular fibrillation was seen in only one patient (0.26%), and this was successfully terminated by external defibrillation. No episodes of ventricular tachycardia were recorded and there were also no arrhythmias during removal of the catheter. We have therefore concluded that pulmonary artery catheterisation in patients with severe aortic stenosis is not associated with a high incidence of ventricular fibrillation or tachycardia, allowing pulmonary artery pressure monitoring to be performed relatively safely in such patients.
Following its introduction by Swan and Ganz in 1970, pulmonary artery (PA) catheterisation has become a mainstay of invasive haemodynamic monitoring . However, observational and randomised controlled trials have failed to prove significant outcome benefits associated with PA catheterisation [2, 3]. This has not deterred many clinicians from using it to diagnose and manage pulmonary hypertension, especially in critically ill patients [4-6]. One of the most commonly reported complications during insertion and removal of the PA catheter is arrhythmia – indeed, the incidence of ventricular arrhythmias was as high as 67.6% in a large observational study of 6245 patients undergoing both cardiac and non-cardiac surgical procedures . In addition, the prevalence of ventricular arrhythmias is higher in patients suffering from aortic stenosis than in control subjects [8, 9]. It is also worth noting that patients with ventricular fibrillation (VF) and aortic stenosis are considered particularly difficult to resuscitate . For these reasons, PA catheterisation is not recommended in patients with severe aortic stenosis in situations other than open heart surgery with cardiopulmonary bypass [11, 12], despite the lack of publications reporting the actual incidence of severe arrhythmogenic complications induced by PA catheterisation in such patients.
Our hypothesis was that the incidence of severe arrhythmias during PA catheterisation is lower than has been considered before, and therefore we decided to study a relatively large cohort of high-risk patients with severe aortic stenosis scheduled for transcatheter aortic valve implantation (TAVI).
Following approval of the local ethics committee and with written informed consent provided for data collection and publication, we retrospectively analysed all patients who underwent TAVI at our institution, whether they received general anaesthesia or sedation and monitored care, between November 2005 and July 2011. Patients were considered candidates for TAVI if they had severe aortic stenosis and cardiac symptoms, and if conventional surgery to replace the aortic valve was associated with high operative risk. Aortic stenosis was considered to be severe if aortic valve area, as determined by transthoracic and transoesophageal echocardiography, was < 1.0 cm2 or, if indexed to body surface area, < 0.6 cm2.m−2 .
The anaesthesia recording system (Siemens, Infinity Delta XL, Erlangen, Germany) provided analysis, storage and trend printing of the following parameters: heart rate and rhythm; ST-segment changes; arterial, central venous and PA pressures; cardiac output; and calculated pulmonary and systemic vascular resistance. The departmental database and data from the hospital information system were used to document and follow-up patients to hospital discharge or death.
After oral premedication with midazolam 0.05–0.1 mg.kg−1 both the evening before surgery and in the morning before induction of anaesthesia, a five-lead electrocardiography system was applied to monitor heart rate, rhythm and ST segments (leads II and V5). A pulse oximetry probe was attached, and a peripheral venous cannula was placed. For continuous measurement of arterial pressure and blood sampling, a catheter was inserted into the left radial artery under local anaesthesia. After placement of external defibrillator pads and using ultrasound guidance, a triple-lumen central venous catheter and an 8.5-French introducer were inserted via the right internal jugular vein, and position was confirmed by intravascular electrocardiography (B. Braun Melsungen AG, Alphacard®, Melsungen, Germany). The patient was then tilted to the left with his/her head up at an angle of 15–20°, following which a 7.5-French balloon-tipped flow-directed PA catheter (Becton Dickinson, Tense Belgium, Belgium) was advanced through the introducer. The PA catheter was advanced to 15 cm, and the balloon was then inflated with 1.5 ml air. The catheter was then advanced through the right ventricle into the PA and subsequently wedged while directly observing the pressure waveform. The PA catheter was then used for continuous monitoring and measurement of PA and central venous pressures, intermittent cardiac output using thermodilution and measurement of blood temperature [14, 15]. All anaesthetists had prior education and experience in PA catheterisation.
Sedation was performed using a continuous intravenous infusion of remifentanil (0.01–2 µg.kg−1.min−1) and titrated injections of midazolam (0.01–0.02 mg.kg−1) as required, combined with local anaesthesia (up to 400 mg lidocaine) of the groin. General anaesthesia was induced with fentanyl (2 µg.kg−1) and etomidate (0.3 mg.kg−1), followed by rocuronium (0.6 mg.kg−1). Following tracheal intubation, patients' lungs were mechanically ventilated with oxygen in air and anaesthesia was maintained with inhaled isoflurane in end-tidal concentrations of 0.4–0.6%, with additional intravenous fentanyl boluses of 2 µg.kg−1.
Before the procedure, all patients received oral aspirin (100 mg) and clopidogrel (300 mg), and intravenous cefazolin (2 g). During the intervention, heparin (100–200 IU.kg−1) was administered to elevate the activated clotting time to more than 250 s. Further anaesthesia management and technical aspects of the TAVI procedure have been described in detail before [16-18]. After the procedure, all patients were transferred to the intensive care unit.
When mean PA pressure exceeded 30 mmHg and was associated with haemodynamic deterioration, nebulised iloprost (20 µg), intravenous milrinone (50 µg.kg−1 over 10 min as a loading dose, followed by a continuous infusion of 0.4 µg.kg−1.min−1) or nitroglycerine (2–5 mg.h−1) were administered, at the discretion of the responsible anaesthetist.
New severe ventricular arrhythmias at any point during insertion of the PA catheter until its placement in the final wedge position were recorded and diagnosed as either ventricular tachycardia (VT, defined as more than five ventricular beats lasting more than 30 s with a heart rate greater than 100 beats per minute) or VF. All statistical analyses were performed using SPSS (version 17.0; SPSS, Chicago, IL, USA).
During the four-and-a-half year study period, 380 patients underwent TAVI and their data and outcomes were reviewed. The procedure was carried out under sedation and monitored care in 203 (53%) patients, and the remainder received general anaesthesia. There was a high incidence of comorbidities in this cohort of relatively elderly patients (Table 1). Of note, moderate or severe pulmonary hypertension (mean PA pressure > 35 mmHg) was documented in 135 (39%) patients; left ventricular function was severely impaired in 36 (9%) patients (ejection fraction < 25%); and 220 (58%) patients had moderate or severe mitral regurgitation.
|Age; years||80.4 (7.5)|
|Weight; kg||72.9 (12.9)|
|Body mass index; kg.m−2||26.5 (4.3)|
|Aortic valve area; cm2||0.6 (0.2)|
|Coronary artery disease||223 (57%)|
|Previous cardiac surgery||62 (16%)|
|Previous cardiac stent implantation||106 (28%)|
|Peripheral vascular disease||83 (22%)|
|Atrial fibrillation||85 (22%)|
|Serum creatinine; µmol.l−1||133.5 (70.7)|
|Chronic pulmonary disease||154 (41%)|
|Prior cerebral ischaemic event||24 (6%)|
|Mitral regurgitation, moderate or severe||220 (58%)|
|Severely impaired left ventricular function||36 (9%)|
|Pulmonary hypertension, moderate or severe||135 (39%)|
In 226 (59.5%) patients, the PA catheter was inserted when the patient was awake. The main access route was the right jugular vein (362, 95.3%), 14 were inserted through the left jugular vein (3.7%) and four using the right subclavian vein (1.0%). Severe arrhythmias occurred in only one out of 380 patients (0.3%) during PA catheter placement. In this patient, VF occurred while the catheter was advanced through the right ventricle, after previous uneventful placement of the central venous catheter, and the VF was immediately terminated by defibrillation. Of note, the electrocardiograph from this patient showed incomplete left bundle branch block; however, pre-operative serum electrolyte concentrations and acid-base balance were within the normal range. No further attempt at PA catheterisation was made in this patient.
The PA catheters were removed 24 h after the TAVI procedure on the intensive care unit, and no severe arrhythmia was noted during PA catheter removal. No other complications including PA rupture, haemorrhage, balloon failure or knotting of the catheter were observed in our patient cohort.
With regard to management of pulmonary hypertension during TAVI, four (1.0%) patients received nebulised iloprost, seven (1.8%) patients milrinone and 68 (17.9%) patients nitroglycerine; the remainder received no specific therapy.
To the best of our knowledge, this study is the first to assess formally the incidence of severe arrhythmias during PA catheterisation in patients with severe aortic stenosis. Our data show that the risk of developing VT or VF during PA catheterisation is very low, despite the fact that the majority of our patients were elderly and could be described as relatively sick because of their comorbidities and cardiac disease. We therefore suggest that PA catheterisation can be performed relatively safely, even in patients with severe aortic stenosis. Patients scheduled for TAVI tend to have a high incidence of pulmonary hypertension. As TAVI requires rapid pacing, cardiac function is inevitably compromised for a short period, and recovery may be prolonged in some patients, leading to hypoperfusion of major organs. In our opinion, PA catheterisation is essential for the diagnosis and management of acute life-threatening impairment of right heart function during and immediately after TAVI, and thereby, indirectly, left heart function as well.
On the other hand, several reports show that premature ventricular contractions are frequent in patients with severe aortic valve disease, ranging from 34% to 65% [9, 19, 20]. Death among symptomatic patients is relatively common (8–34%) and if sudden, is likely to be due to severe arrhythmia [9, 21, 22]. Conversely, in patients with asymptomatic aortic stenosis, sudden death occurs only occasionally (0–5%) [23, 24]. The frequency and complexity of ventricular arrhythmias are closely related to myocardial function, and therefore more common in patients with higher left ventricular systolic stress and decreased systolic function. The rate of severe arrhythmia in our study of high-risk patients was much lower than we had anticipated. Earlier observations published between 1964 and 2007 showed significantly higher complication rates in a heterogeneous patient population (Table 2) [7, 25-27, 29]. Moreover, patients with underlying valve disease from PA catheterisation, due to the potential arrhythmogenic risk , were not studied in one recent prospective study. Such reports in the literature have almost certainly led to the common perception that PA catheterisation in such patients is associated with an unacceptably high risk, except immediately before open heart surgery with cardiopulmonary bypass [11, 12].
|Wennevold et al. ||135||2 (1.5%)||2 (1.5%)|
|Sprung et al. ||44||8 (18.2%)||0|
|Sprung et al. ||119||5 (4.2%)||2 (1.7%)|
|Shah et al. ||6245||< 1%||< 1%|
|Patel et al. ||142||24 (16.9%)||0|
|Gwak et al. ||100||2 (2.0%)||2 (2.0%)|
Independently of aortic stenosis, other studies have reported that PA catheterisation is still associated with severe ventricular arrhythmias, with an incidence between 3% and 37% [7, 25-28]. In 1964, Wennevold et al. described right heart catheterisation of 135 patients under direct visualisation with fluoroscopy, and reported the incidence of VT and VF to be 1.5% each . In 1981, Sprung et al. reported VT in 18.2% of 44 critically ill intensive care patients undergoing PA catheterisation. They also found that the development of VT was strongly correlated with the duration of the catheterisation procedure  . Two years later, Sprung et al. prospectively studied 119 critically ill patients in a medical intensive care unit and observed sustained VT in five patients (4.2%) and VF in two (1.7%) patients, requiring defibrillation . In a large study by Shah et al., 6245 patients were included over a period of 5.5 years, of whom 4990 were undergoing cardiac surgery and 1255 undergoing non-cardiac surgery, and who required PA catheterisation for extended haemodynamic monitoring. Persistent premature ventricular contractions occurred in 193 (3.1%) patients during PA catheterisation; however, VT or VF requiring defibrillation was seen in less than 1% of patients .
Clinical experience has a direct influence on the time required for PA catheterisation [28, 30]. In our study, the procedure was performed by either experienced consultant anaesthetists or postgraduate year-four anaesthesia residents, directly supervised by the consultant. As a result, avoidance of a long transit time of the PA catheter through the cardiac chambers may have contributed to the low incidence of arrhythmias we have observed. Due to the fact that this was a retrospective study, we were not able to state the precise time we needed to place the PA catheter in its final position. Further prospective studies are needed to confirm this statement. In our institution, PA catheterisation is also routinely used in cardiac surgery and during liver transplantation, and this could have contributed to the low complication rate, because the responsible anaesthetist was very familiar with the catheterisation technique.
Other risk factors for developing severe ventricular arrhythmias during PA catheterisation include predisposition to ventricular arrhythmia, previous liver transplantation, complicated myocardial infarction, sepsis and acidosis (pH < 7.2) [24, 26, 27]. Except for aortic stenosis, our patient cohort did not match any of these previously described risk factors. However, we are aware that our study suffers from some limitations. First, it was an observational, retrospective study, with inherent weaknesses in design. Also, we did not address the question of whether results obtained from PA catheterisation impacted on clinical outcomes during and after TAVI procedures. Finally, our study assessed arrhythmias solely during PA catheter insertion and removal, and did not include a control group of patients not undergoing PA catheterisation, or indeed healthy subjects undergoing PA catheterisation. In addition, despite the low incidence of arrhythmias demonstrated in this study, we strongly believe that placing defibrillator pads before the procedure is essential in these high-risk patients for immediate conversion of new onset atrial fibrillation or VF. Further studies are required to investigate whether PA catheter-guided therapy during and after TAVI affects outcome.
In conclusion, PA catheterisation in high-risk patients with severe aortic stenosis is associated with a low incidence of severe ventricular arrhythmias. We therefore recommend that PA catheterisation should not be withheld in such patients on safety grounds alone, especially considering their utility for assisting in the titrated management of pulmonary hypertension and right heart failure.
No external funding or competing interests declared.
- 11Cardiovascular Monitoring. In: Miller RD, ed. Miller's Anesthesia, 6th edn. Philadelphia, USA: Elsevier Churchill Livingston, 2005: Chapter 32. 1266–362., .
- 12Aortic stenosis and non-cardiac surgery. Continuing Education in Anaesthesia, Critical Care & Pain 2011; 5: 1–4., .
- 19Determinants of the incidence and severity of ventricular arrhythmias in aortic valve disease. American Journal of Cardiology 1983; 51: 1103–9., , , , , .