Tracheal extubation under deep sevoflurane anesthesia: A novel strategy for weaning difficulties in intensive care.

Various criteria for weaning patients from ventilators in intensive care have been widely published. These criteria are increasingly incorporated into guidelines, protocols, and more recently, care pathways. We present a case where a patient's lungs were ventilated for 4 days with an infective exacerbation of chronic obstructive pulmonary disease (COPD). We successfully weaned off mechanical ventilation and rapidly extubated the patient's trachea utilizing deep sevoflurane anesthesia. Published weaning indices suggest that this would have been an inappropriate course of action at the time. However, our patient clearly benefited and avoided the need for tracheostomy and prolonged ventilation.

Introduction

Treatment of infection, failure to wean, tracheostomy and then a slow wean from ventilation over days or weeks. This is the usual course in intensive care for a patient intubated and ventilated due to an infective exacerbation of severe chronic obstructive pulmonary disease (COPD).[1] With patient’s consent, we present such a case who we successfully weaned off mechanical ventilation, and his trachea extubated early in his illness despite troublesome and severe bronchospasm. Despite falling outside of the usual weaning parameters set out in guidelines, we accomplished this utilizing deep sevoflurane anesthesia and avoided the need for tracheostomy and prolonged ventilation.

Case Report

A 59-year-old 60 kg male presented with an acute severe exacerbation of chronic obstructive pulmonary disease (COPD). His COPD had been diagnosed 15 years previously and had clearly been troublesome, as he was well known to the respiratory physicians. Features of his presentation suggested an infective exacerbation - core temperature of 38.5°C, white cell count (WCC) of 22 × 109 L-1, and C-reactive protein (CRP) 134 mgL-1 Apart from controlled hypertension and hypercholesterolemia, there was no other relevant past medical history. Recent pulmonary function tests indicated severe COPD with a forced expiratory volume in 1 second (FEV1) of 0.69 liters (19% of predicted), forced vital capacity (FVC) 2.95 liters (66% of predicted), FEV1/FVC 0.23 (30% of predicted), and peak expiratory flow rate (PEFR) less than 50% of predicted. These results were typical for the patient. His regular medications consisted of simvastatin 40 mg, ramipril 2.5 mg, clopidogrel 75 mg, aspirin 75 mg, and prednisolone 30 mg, all once-a-day, along with salbutamol (100 mcg metered dose inhaler) (MDI) and beclomethasone (200 mcg MDI) one puff each, 6 and 12 hourly, respectively.

Eight hours following admission, he was intubated and ventilated in the intensive care unit (ICU). Initial arterial blood gases (ABGs) showed severe respiratory acidosis (pH 7.02, pCO2 18 kPa, pO2 12.4 kPa, HCO3- 26 mmolL-1, BE 1.7). From the start, it was very difficult to ventilate him, but we were able to achieve adequate tidal volumes (VT)(= 0.5 L) following muscle relaxation with vecuronium bromide 10 mg intravenous (IV). Inspiratory pressures (Pinsp) were 32 cm H2O on pressure controlled (PCV) mode. Arterial and central venous catheterization was performed for hemodynamic management. He was commenced on cefotaxime 1 gm IV and clarithromycin 0.5 gm IV (both 12 hourly), salbutamol 5 mg and ipratropium bromide 500 mcg 6 hourly (both nebulized), hydrocortisone sodium succinate 50 mg IV 6 hourly and low molecular weight heparin (LMWH). Aminophylline was commenced at 41.6 mg/hour following a bolus dose of 300 mg IV. He required minimal vasopressor support (nor-adrenaline 0.025 mcg/kg/min) to support his blood pressure.

Sedation was maintained with propofol and alfentanil, with the requirement for each amounting to 3.3 mg/kg/hour (20 ml/hr of 1 % solution) and 5.5 mcg/kg/min (20 ml/hour of 1 mg/ml solution), respectively. It was not possible to satisfactorily ventilate the patient with lower levels of sedation with the endotracheal tube in situ. Even slight reduction in the level of sedation provoked episodes of severe bronchospasm with desaturation (SpO2 dropping to 60-70%) despite administration of 100% O2. In spite of liberal use of bronchodilators, it took an average 45 minutes to 1 hour to settle him following such episodes. Stimuli such as endotracheal suctioning and rolling of the patient triggered troublesome episodes of bronchospasm and desaturation, associated with ABG evidence of severe respiratory acidosis (pCO2 up to 12.5 kPa). All these events prompted high levels of sedation, the addition of terbutaline (0.5 mg S/C, 6 hourly) to his treatment regimen and an increase of salbutamol frequency to 4 hourly.

By day 4, his infective markers had improved significantly (white cell count normalized, C-reactive protein reduced to normal, and the patient was afebrile). However, there was no improvement in his brittle airway reactivity. When stable, the patient had VT of 0.6 liters at Pinsp of 26 cm H2O, PEEP of 8 cm H2O (on PCV mode) and SpO2 of 99% with a FiO2 of 0.35. (pH 7.43, PO2 12.5 kPa, PCO2 5.4 kPa, BE + 2).

When heavily sedated, the patient arguably met the criteria for consideration for extubation (according to published recommendations),[2] but certainly not when the sedation was reduced. Notably, he would not be able to pass the spontaneous breathing test (SBT), his compliance fell to 20 ml/cm H2O, and his PaO2/FiO2 ratio fell to as low as 40 (PaO2 in mmHg).

There was much discussion regarding the next move in his treatment strategy. Although no guidelines or protocols would have supported it, experience and instinct lead us to consider extubation under deep inhalational anesthesia with sevoflurane. The alternative would have been a tracheostomy and the slow wean over days or weeks typical of many COPD patients. This would risk all of the well-recognized ventilator-associated complications.

An anesthetic machine (Dräger Cato edition) with a sevoflurane vaporizer (Dräger Vapor 2000) was brought to ICU. Sedation was turned off, and the patient gently manually ventilated with 100% O2 and sevoflurane using a Bain’s circuit, achieving end-tidal sevoflurane concentrations of 3-4 kPa. His blood pressure (BP) was supported by a small increase in nor-adrenaline from 0.025 to 0.05 mcg/kg/min. 15 minutes following cessation of sedation, 400 mcg of naloxone was administered IV and the patient started to breathe spontaneously. At this stage, his BP was 140/70 mm Hg, heart rate (HR) 70/min, RR 9/min, and SpO2 was 98%. 15 minutes later, the trachea was extubated following gentle endotracheal suctioning in the sitting position with a nasopharyngeal airway in situ. He was oxygenated with an anesthetic mask attached to Water’s circuit for the next 30 minutes with a FiO2 of 1.0 at 10 l/min until he was responsive to verbal commands. 45 minutes following extubation, his BP and HR were 140/84 mm Hg and 70/min, respectively. Initial ABGs were encouraging in view of his previous condition: pH 7.3, pCO2 8.9 kPa, pO2 12 kPa, HCO3-26 mmolL-1, BE 4.5, FiO2 0.6.

2 hours later, he had an episode of respiratory depression, which was treated with a further bolus of 400 mcg of naloxone IV. 2-3 minutes later, he was fully conscious with a RR of 18/min and was able to cough. Naloxone was then administered as an infusion overnight (500 mcg/hour), with no further episodes of respiratory depression. He subsequently made an uneventful recovery and was discharged from the hospital 3 days later. Hemophilus influenzae was identified as the infective cause for the exacerbation of his chest disease.

Discussion

We are unaware of sevoflurane being used in this category of patients in intensive care to facilitate tracheal extubation. Our patient’s lungs were mechanically ventilated due to an infective exacerbation of type II respiratory failure. He required heavy sedation for effective ventilation. By day 4, although his infective markers had improved significantly, reduction of sedation reliably lead to episodes of severe bronchospasm accompanied by rapid desaturation and respiratory acidosis. This happened despite ‘maximal’ bronchodilator therapy. At this stage, we faced the all too familiar ‘slow wean’ usually involving the concomitant use of percutaneous tracheostomy. However, we were particularly concerned about the associated risks of long-term ventilation in this patient due to him always requiring particularly heavy sedation. Long-term mechanical ventilation (MV) is associated with a vast spectrum of complications including pulmonary and cardio-vascular complications, ventilator-associated pneumonia, ileus, malnutrition, neuro-muscular dysfunction, decubitus ulcers, and psychological dysfunction.[3]

Even after a tracheostomy, patients may require a variable period of mechanical ventilation during the slow weaning process. In a 3 - center clinical trial of early (within 48 hours) versus delayed (day 14-16) tracheostomy in 120 medical ICU patients, Rumbak et al., found the duration of MV to be 7.6 ± 4 days in the early tracheostomy group as compared to 17.4 ± 5.3 days in the late tracheostomy group.[1] Exposure of our patient to this slow weaning process and the associated risks of tracheostomy (hemorrhage, misplacement, pneumothorax, subcutaneous emphysema, and tracheal or esophageal injury)[4] struck us as questionable in view of the fact that except for his airway irritability, the infective pathology had been successfully treated.

Volatile anesthetics have long been proposed for the treatment of bronchospasm precipitated by endotracheal intubation in humans.[5] Numerous case reports describe the use of halothane for management of status asthmaticus. However, its use is associated with possible adverse effects. Halothane sensitizes the myocardium to potential arrhythmias in patients with elevated catecholamine levels. This particularly holds true for patients being treated with methylxanthines and beta-agonists. Halothane has also been implicated in the pathogenesis of malignant hyperthermia and is potentially hepatotoxic. Isoflurane and enflurane have also been used in ventilated patients to treat bronchospasm, though their use in awake patients is limited by their airway irritant and respiratory depressant properties.[6]

Sevoflurane (fluoromethyl-2, 2, 2-trifluoro-1-ethyl ether) was first synthesized in 1968 and used clinically in 1971. It is non-irritant to the airway and protects against bronchospasm in animals. In addition, it does not sensitize the myocardium to catecholamines.[6] Sevoflurane was shown to act both at the proximal airway level and the lung periphery, thereby reducing resistive and visco-elastic pressures and static elastance in a model of chronic allergic asthma.[7] Furthermore, sevoflurane has a faster onset of bronchodilator action in comparison to isoflurane, specifically in patients with COPD.[5] Its low blood-gas partition coefficient is an added advantage, permitting rapid induction and emergence from anesthesia.[8]

Dexmedetomidine could have been an option in our case, given the favorable properties of anxiolysis, sedation, and analgesia without causing respiratory depression. Although there are numerous reports of its benefits in awake intubation and emergence from anesthesia, there are no data on its role in the asthmatic patient.[9] Moreover, it is associated with more adverse hemodynamic effects in comparison to agents in conventional practice.[10]

Our patient did not meet either the oxygenation or ventilatory parameters required for extubation, as he had episodes of severe bronchospasm on reduction of sedation.[2] However, and importantly, his infective pathology had been successfully treated as depicted by overall improvement in his clinical status and laboratory data. Instinct, experience, and intuition were the main factors guiding our decision to proceed to tracheal extubation, and our patient clearly benefited from the chosen strategy.

In view of the relative rarity of severe intractable bronchospasm in the intensive care setting along with the many potentially confounding factors, a properly conducted randomized controlled trial proving the efficacy of this technique would be impossible. Regrettably, case reports such as this remain in the lower echelons of the evidence pyramid. However, a degree of support for a similar treatment strategy elsewhere would now be supported by the reporting of this successful outcome.