Anaesthetic management for endobronchial valve insertion: lessons learned from a single centre retrospective series and a literature review

This is a new procedure and not frequently performed. It is not surprising that each anaesthetist employs their own favourite technique (and even this may change with the experience of previous cases). Broadly the techniques used in our institution fall into two groups; general anaesthesia with a tracheally intubated patient, and ‘monitored anaesthesia care’ with a sedated unintubated patient.

Dexmedetomidine using a bolus followed by infusion, was used in eight cases of MAC along with midazolam and fentanyl. In nine cases, glycopyrrolate was administered prior to commencing dexmedetomidine and six had airway topicalisation. Total intravenous anaesthesia (TIVA) with propofol was used in all GA via ETT cases. In 8 out of 10 of the GA via ETT cases, combinations of propofol, remifentanil and rocuronium were used. The details of the cases analysed presented in Fig. 4. Sugammadex was used as the reversal agent in 9 of the 10 GA cases. While the GA cases were positioned supine, MAC cases were positioned 15–20 degrees head up.

The technical aspects of this procedure from a pulmonology perspective have been extensively reviewed recently by two groups [1, 2]. The patients are selected according to generally agreed spirometry criteria, namely hyperinflation with a residual volume > 175% of predicted, and forced expiratory volume < 50% of predicted [1]. The efficacy of this technique depends on the absence of collateral ventilation between the lobe with implanted valves and the ipsilateral lobe [1]. Precise measurement of the amount of collateral ventilation is performed using endobronchial flow catheters. The Chartis Diagnostic System (Pulmonx, Redwood City, California, USA) is one such system [1]. This is an endobronchial flow catheter and associated computer software (Fig. 5). It is generally recommended that both the Chartis measurement and EBV implantation be performed as a single-stage procedure whenever possible, to minimize the risk of bronchoscopy induced COPD exacerbations [1]. The Chartis system comprises a balloon catheter attached to a sensor remotely housed in a console. The catheter is passed through the working channel of the bronchoscope; the main bronchus of the target lobe is blocked by inflating the balloon; and collateral ventilation is assessed by measuring the expiratory flow over time (Fig. 6). If no collateral flow is observed, a flexible delivery catheter is then used to deploy the EBV into the targeted bronchial lumen. An inbuilt measurement gauge in the catheter helps in choosing the appropriate size of the valve for the target lumen [7]. In our review, all the procedures were done by the same interventional pulmonologist, while different consultant anaesthetists participated in the anaesthetic management. An arterial line was inserted in all patients and serial blood gas analyses were performed.

Mild hypotension was observed in 5 patients. The hypotension was treated with a single bolus of metaraminol in 4 cases and a bolus of ephedrine in 1 case. Intravenous infusions of inotropes or vasopressors were not required. One case of profound hypotension - thought likely to be anaphylaxis - occurred in the GA via ETT group, and the procedure was abandoned. There was one case of inadequate topicalisation. All patients were discharged to the respiratory high dependency unit.

Postoperative complications were encountered in seven patients. One episode of rapid AF occurred on extubation. There were exacerbations of chronic obstructive pulmonary disease in three instances, and two cases of haemoptysis (one being described as massive) and one valve migration (Table 2). There were no pneumothoraces encountered and the mean hospital length of stay was 5 days. As an aid to comprehending the distribution of these complications, they are also included in Fig. 4.

A TIVA technique may be preferable compared to a volatile based anaesthetic regime. Leakage of volatile agents is a concern due to repeated airway manipulation. Additionally, propofol may have a favourable effect of maintaining hypoxic pulmonary vasoconstriction, an effect beneficial in this scenario [14]. Remifentanil may be a good narcotic choice considering its potency and rapid onset and offset, allowing immediate return of respiratory function post procedure. Moreover, by continuing a low dose infusion during extubation, cough and hypertensive responses could be lessened. Sugammadex may be preferable to prevent residual neuromuscular blockade in the presence of compromised baseline respiratory function.

Monitored anaesthesia care with sedation along with airway topicalisation are guided by Institutional practices. Adequate airway topicalisation may help in preventing cough and movement. Mucosal atomizing devices such as the DeVilbiss (DeVilbiss health care, Devilbiss International), can generate local anaesthetic droplets between 6 and 12 μm, which are delivered to the distal tracheobronchial tree [15]. Excessive suctioning to clear secretions can result in airway oedema, mucosal bleeding and inflammation, and affect the valve sizing judgements [1]. Apart from reducing the secretions, an antisialagogue might also help in reducing the dilution of the delivered local anaesthetic into the tracheobronchial tree during airway topicalisation. However, antisialagogue benefits should be weighed against the risk of their chronotropic effects [1]. Achieving optimal sedation is always challenging, and never more so than during the Chartis Procedure. The patients’ should display adequate tidal volume to validate the detection of a collateral-free target lobe. This should be balanced against minimizing coughing and/or secretion production [1]. Avoiding respiratory failure and CO₂ retention is a major concern during sedation techniques. Deep sedation should be attempted only if appropriate rescue resources are readily available [1]. Titrated doses of a short acting benzodiazepine and a narcotic may be an attractive option, as the effects can be readily reversed. Dexmedetomidine may be a promising selection in view of its sympatholytic and respiratory stability effects. The sympatholytic effect needs careful monitoring in these patients. For MAC technique, co–induction using reversible agents which reduce the requirements for dexmedetomidine, and thereby avoiding large drops in arterial pressure seemed popular with our anaesthetists.

An increasing trend towards increasing GA utilization was noted in our review. It became evident that GA via ETT enabled quicker Chartis measurements. The absence of cough and movement made the technique easier. A GA technique also meant that any prolonged procedure could be easily tolerated. An increase in the anaesthetic times of about 24 min were noted for the MAC group compared with the GA via ETT group (61 vs 37 min). Reassuring the patients, and gaining their cooperation takes time; likewise, obtaining effective topical anaesthesia of the airway further prolongs the anaesthetic time required. Currently, GA via ETT is the preferred choice at our institution. A supraglottic airway together with neuromuscular blockade may also offer similar bronchoscopic conditions without the need for intubation. However, the bronchoscope may be inserted and removed on more than one occasion, and trauma to the vocal cords and surrounding structures from repeated passage of the bronchoscope; along with the risk of displacement of a supraglottic device, means that an ETT may prove to be the preferred choice. Making such decisions must involve the pulmonologist, and close cooperation between the proceduralist and the anaesthetist is absolutely essential.

A wide range of complications can occur during the procedure, such as respiratory and also haemodynamic instability. None of the patients own medications (such as beta blockers and calcium channel blockers) were omitted preoperatively and there is a possibility of these agents worsening dexmedetomidine induced hypotension. This does not seem to have been a significant problem in our patients. The liberal use of glycopyrrolate prior to commencing dexmedetomidine may have averted significant bradycardia. The hypotension observed in the GA via ETT cases was in line with the expected drop after any general anaesthetic. It responded to boluses of vasopressors namely ephedrine and metaraminol, and no infusions (either vasopressors or inotropes) were required.

Pneumothorax is a major complication of EBV procedures, attributed to the sudden expansion of the lobe adjacent to the target lobe after the collapse of the treated lobe, and patients showing significant volume reduction on post procedure Chest X-ray may be at an increased risk [1]. The incidence varies between 15 and 25% [16], and the median onset is 2 days post procedure [16, 17]. It can also occur within hours after valve insertion and anytime during the procedure [2]. Two studies have looked into the factors responsible for pneumothorax occurrence after EBV insertion [17, 18]. One found that a strategy of post procedure bed rest and cough reduction helped decrease the incidence [18].

Presently, it is unclear whether the type of anaesthetic technique has any role in precipitating pneumothorax. Although, there is a theoretical possibility of increased pneumothorax with positive pressure ventilation in these high risk patients especially with bullous lung disease, the risk may be reduced if a specific individually optimised ventilatory approach is adhered to. If High Frequency Jet Ventilation is available this may be an attractive option, however, we do not have this device in our institution. If a small pneumothorax does occur, − in the absence of tension - careful monitoring should suffice. A chest drain is indicated for larger pneumothoraces. Infections, COPD exacerbations, pneumonia and temporary haemoptysis are other possible postoperative complications [2].

The small numbers, the retrospective nature and missing data are some of the limitations of this review.