In 1897 Gustav Killian, a German laryngologist, performed the first airway examination in a human, with a rigid oesophagoscope, to remove a piece of bone lodged in a mainstem bronchus. The next milestone occurred with Shigeto Ikeda in 1964 and the development of the first flexible bronchoscope capable of examining the airways down to subsegmental bronchi. Flexible bronchoscopy (FB) has been used clinically since 1966 and has largely replaced rigid bronchoscopy as the technique of choice for airway examination and intervention. It can be of significant help in the diagnosis and management of pulmonary pathology in severely ill patients in the cardiothoracic intensive care unit (CICU). Intensivists, interventional pulmonologists and cardiothoracic surgeons are typical operators in this setting. Given its widespread use, it may be surprising that its role, efficacy and safety in critically ill patients has largely been limited in the published literature to small case series and expert opinions. A few meta-analyses have attempted to evaluate its role, particularly in the context of ventilator acquired pneumonia.
Patients in a CICU are typically selected for predominantly cardiac or respiratory support, with over 40% on mechanical ventilation. Patients are varied and include recipients of heart/lung transplants, immunocompromised established transplant recipients, those receiving anticoagulation or multiple antiplatelet agents, and those with chronic cardiac or respiratory disease. Polypharmacy is common, with the potential for multiple drug interactions adding to issues facing the bronchoscopist, which need to be carefully evaluated before deciding to proceed. Determining the appropriateness of the procedure, then optimising the patient and environment to ensure a safe and effective intervention, depends on a multidisciplinary approach with the physician/surgeon, intensivists, interventional pulmonologist and nurses all involved.
In this chapter, the role of bronchoscopy in the CICU, contraindications, preparation, procedural considerations and potential complications will be discussed.
One can generally group the indications for bronchoscopy into diagnostic and therapeutic reasons, though there is often overlap. Examples of this include haemoptysis, foreign body removal, central airway obstruction, tracheo-oesophageal fistula and removal of proximal sputum plugs. Studies suggest 65% to 79% of bronchoscopies performed in the ICU setting are conducted on patients being mechanically ventilated, and of these 47% to 75% have a therapeutic indication. In one review of bronchoscopies performed in critically ill patients on an ICU, 45% were performed to remove bronchial secretions, 35% for collecting samples from the lower respiratory tract, 7% for airway assessment, 2% for haemoptysis, 0.5% for assisting difficult tracheal intubation and 0.5% for removing foreign bodies.
Indications for bronchoscopy in a CICU are summarised in Table 5.1.
FB can be useful to facilitate difficult endotracheal intubations in patients with a Cormack–Lehane score of 3 or more, limited mobility of head and neck, cervical spine fractures, or where severe coagulopathy or excessive secretions might make intubation with a layngoscope inadvisable. Despite its advantages in these situations, it is typically only used in a small proportion (0.07–3.4%) of patients in the ICU setting. The oral route is preferred as a larger diameter endotracheal tube (ETT) can be passed. Adult bronchoscopes have a diameter of approximately 6 mm, able to pass through a size 7.5 ETT or larger – typical sizes used in adult patients. To prevent damage to the bronchoscope, a bite block is advised. As four minutes or more must be allowed to accomplish intubation, this technique is not recommended for apnoeic or near apnoeic patients.
Double Lumen Endotracheal Tubes
FB can be used to assist placement of a double lumen ETT used for differential ventilation or management of massive haemoptysis. A thin 4 mm diameter bronchoscope can pass through a 35F double lumen tube; the smallest size used in adults.
Changing Endotracheal Tubes
Occasionally an ETT may need to be changed due to cuff leakage or to facilitate a bronchoscopic intervention. If endotracheal reintubation is expected to be difficult, a flexible bronchoscope may be used to facilitate the exchange as an alternative to a bougie. The stomach contents are aspirated via a nasogastric tube, following which the new ETT is inserted over the bronchoscope and inserted into the posterior pharynx, whilst suctioning secretions. The existing ETT cuff is deflated and the bronchoscope/new ETT passed through the vocal cords alongside the old ETT. The old ETT is then withdrawn and the new tube adjusted accurately under direct vision prior to cuff inflation.
Patients with suspected upper airway obstruction are ideal candidates for FB directed extubation. Examples include bilateral vocal cord paralysis, tracheomalacia or tracheal stenosis due to causes such as multiple intubation attempts, prolonged intubation or airway injury. The bronchoscope is inserted into the ETT just beyond the tip, the cuff deflated and the bronchoscope/ETT slowly removed en bloc. During withdrawal, if there is endoscopic evidence of significant subglottic or glottic obstruction, the ETT can be safely reinserted under direct vision.
Pneumonia is the most common infection in the CICU. The overall incidence of VAP ranges from 9% to 25% in the general ICU population, with rates up to 70% in those with acute respiratory distress syndrome (ARDS). The risk increases over the first 10 days and can affect two thirds of patients who have been ventilated for more than 30 days. Mortality ranges from 35% to 90%. Bronchoscopically directed lavage, protected brushing and occasionally deep lung biopsies (transbronchial lung biopsies) are often used to determine the cause of infection. Qualitative or quantitative (number of colony forming units, or number of intracellular organisms) techniques can be used to analyse the samples obtained. This results in a wide range of reported sensitivities of bronchoscopic techniques (51–100%) with the overall impression that bronchial lavage and brushing are safe techniques for microbiological diagnosis in ventilated patients. However, well-conducted systematic reviews looking at bronchoscopic versus ‘non-invasive’ techniques (tracheal suctioning, blind catheter brushing) in reducing mortality and ICU stay in clinically diagnosed VAP patients have shown no statistically significant differences in mortality, duration on mechanical ventilation, length of ICU stay or antibiotic change. The British Thoracic Society therefore recommends directed non-invasive diagnostic strategies in preference to bronchoscopy in ventilated patients suspected to have VAP.
Patients with solid organ transplants receiving immunosuppressive medications in the CICU are at higher risk of developing opportunistic fungal infections, including Pneumocystis jirovecii, invasive aspergillosis, candidiasis, cryptococcus, bacterial infections and viral infections such as cytomegalovirus. In areas of high prevalence in the UK, or in patients from endemic areas worldwide (Sub-Saharan Africa, Far East, Southeast Asia, parts of Eastern Europe and South America), tuberculosis must also be excluded.
Bronchoalveolar lavage (BAL) with 120–200 ml of instilled 0.9% saline can be performed via the bronchoscope into an affected lobe. If diffuse changes are present, bilateral BAL (preferentially of the upper lobes) appears to provide the highest sensitivity. BAL is reported to have sensitivity for P. jirovecii of 90–98% and is considered the gold standard. Diagnostic sensitivity for pulmonary tuberculosis can be increased from up to 30% based on microscopy alone to 86% with rapid PCR techniques, and is preferable to transbronchial lung biopsies in the majority of ICU patients due to the potential risk of pneumothorax or bleeding. For patients with suspected invasive aspergillosis, BAL galactomannan testing has shown superiority to fungal staining and culture, with sensitivity of 94% and specificity of 79% and should be considered if available.
FB has been used in ventilated patients with lobar or whole lung atelectasis who have failed to respond to treatments such as physiotherapy, nebulised saline, mucolytics or repositioning (including prone ventilation). In several small case series it has been shown to be effective in immediate reversal of lobar atelectasis. Physiotherapy and conventional non-invasive measures are still recommended first line treatment with bronchoscopy reserved for the following situations:
Life threatening whole or near-whole lung collapse;
Lobar atelectasis due to proximal sputum plug with a lack of visible air bronchograms on radiology;
Failure to respond to chest physiotherapy and other measures;
Where physiotherapy or repositioning is not feasible (e.g. thoracic trauma, spinal fractures)
Patients with neuromuscular disorders and impaired cough;
Cystic fibrosis patients (copious inspissated secretions);
Lung transplant patients with tenacious plug composed of necrotic tissue and mucus.
Bronchoscopy in conjunction with CT can be useful to identify the endobronchial source of haemoptysis in intubated patients with persistent or excessive bleeding from the ETT. It is also typically employed (usually through an ETT) for diagnosis and control of massive haemoptysis, defined as 400 ml in 24 hours, or 200 ml in any one event. Through the bronchoscope, iced saline, fibrin precursors or topical adrenaline 1/10,000 can be instilled to attempt haemostasis in a bleeding segment. Alternatively, the bronchoscope can be used to facilitate the passage of specific endobronchial blockers (e.g. Cohen flexitip® or Arndt®) and a Fogarty or Swan-Ganz catheter to occlude the bleeding lobe or segment. Finally, an ETT can be directed bronchoscopically into the normal lung to isolate the bleeding side and avoid spillover of blood, or help direct a dual lumen ETT to achieve the same effect. Where facilities and expertise exist, rigid bronchoscopy may be preferable in cases of massive haemoptysis as it provides a secure airway, large volume suction capability and easy access to the airways with endobronchial blocking devices. All these measures attempt to secure the airway, ensure adequate oxygenation, prevent soiling of the normal lung and in some cases allow definitive airway intervention – or buy time for a surgical or interventional radiology solution.
Tracheobronchial injuries affect up to 2.8% of severe blunt chest trauma and accidental deaths. Sternal or upper rib fractures can indicate significant blunt force injury and likelihood of internal problems so bronchoscopic examination is mandatory. Physical and radiological signs include hypoxaemia, haemoptysis, pneumothorax, surgical emphysema, haemothorax, pneumomediastinum, flail chest and the so-called ‘falling lung sign’ on chest radiograph (pneumothorax with atelectatic lung collapsing away from the mediastinum) which is pathognomonic of rupture of a mainstem bronchus.
The incidence of BPF after pulmonary resection varies between 4.5–20% after pneumonectomy and 0.5% after lobectomy. Risk factors for this serious complication include right-sided pneumonectomy, a long bronchial stump, residual cancer at the bronchial margin, devascularisation of the bronchial stump, prolonged ventilation and reintubation after resection. Many patients may not be fit enough to have repeat surgery where intercostal drainage and sometimes pleurodesis methods fail. The extent of the air leak may prevent lung reinflation and delay ambulation, even with Heimlich valves or portable suction devices (e.g. Thopaz®, Medela Switzerland). Bronchoscopy can be used in these patients to identify the predominant lobe/segment causing the air leak (by employing an endobronchial occluding balloon), followed by insertion of one-way valves preventing airflow into that segment. Examples of valves used for this purpose include Emphasys® (Pulmonx, Redwood, USA), Spiration IBV® (Olympus Medical, Japan) or Watanabe spigots® (Novotech, France). These devices require considerable expertise to deploy accurately and are usually inserted by interventional pulmonologists.
In the non-intubated adult patient, a standard 5.7 mm diameter bronchoscope occupies only 10% of the cross-sectional area of the trachea. Therefore, in spontaneously breathing patients endotracheal pressures generated are similar to those in patients without bronchoscopy.
In an intubated, ventilated patient, the effect is quite different. The obstructive effect of the bronchoscope is added to that of the ETT, with the potential to cause quite dramatic changes in respiratory mechanics, gas exchange and haemodynamics. Indeed, a 5.7 mm bronchoscope occupies 40% of the cross-sectional area of a 9.0 mm inner diameter ETT, 51% of a 8.0 mm ETT and 66% of a 7.0 mm ETT.
Complicating the effect of the bronchoscope on the patient’s physiology are patient-specific factors including the underlying diagnosis (for example, atelectasis, ARDS, central airway obstruction, obesity, hypotension), the effects of sedation and the procedure being undertaken (such as suctioning to clear airway debris).
All these factors need consideration prior to undertaking FB, and in intubated patients particular attention needs to be paid to ensuring an adequate sized airway and choosing an appropriately sized bronchoscope for the task (a thinner scope may be safer but will have less suctioning capacity and may not allow passage of certain instruments such as a cryoprobe).
Table 5.2 summarises the main physiological effects of bronchoscopy in intubated patients.
Flexible bronchoscopes typically have working channels 2.0–3.2 mm in diameter, allowing varying suction capacity or ability to pass specialised instruments to perform interventions.
Additionally, linear endobronchial ultrasound scopes available today (with approximate external diameter 6.9 mm) can be used via larger ETTs to facilitate transbronchial needle aspiration of mediastinal lymph nodes or masses.
Advanced bronchoscopic interventions are generally the remit of the interventional pulmonologist or thoracic surgeon.
Procedures likely to be encountered in the CICU are detailed in Table 5.3.