sion, bronchospasm and laryngospasm, pneumothorax, and bleeding. Gas embolism has been reported using argon plasma coagulation [35].

Basic Diagnostic Procedures

Bronchoalveolar Lavage (BAL)

The bronchoalveolar lavage was frst introduced to clinical practice by Reynolds in 1974 [36].

Standardization of BAL was addressed in a report by the European Respiratory Society task force in 1999. The report considers in detail the four main problems that prevent accurate quantifcation of components in alveolar epithelial lining uid (ELF) using BAL:

\1.\ An unknown amount of dilution during lavage.

\2.\ Contamination of the ELF sample with material from the bronchi.

\3.\ Inadequate sampling due to incomplete mixing.

\4.\ Lung permeability varies, allowing loss of introduced lavage uid into the tissues and increased leakage of soluble components from the blood capillaries and tissues into the ELF [37].

To perform a BAL, the bronchoscope is wedged in the target bronchus while keeping the working channel in the lumen of the bronchus. A total of 4 aliquots (30–60 mL each) are instilled in the alveoli for a total of at least 100 mL and a maximum of 240 mL of sterile normal saline. Subsequently, the uid is suctioned into a trap with a pressure below 100 mmHg adjusted to avoid visible airway collapse.

In a healthy non-smoking subject, the BAL cellular composition is: macrophages (80–90%), lymphocytes (5–15%) with CD4/CD8 ratio of 1.5–1.8, neutrophils (1–3%), and eosinophils and mast cells <1% [38].

Cell counts on BAL can have non-specifc results in many conditions such as cryptogenic organizing pneumonia and usual interstitial pneumonia, making its utility, to some extent, contro-

versial [39, 40]. On the other hand, BAL plays an important role in diagnosing pulmonary infections, especially in immunocompromised hosts and mycobacterial infections [41, 42]. A higher yield can be achieved by adding TBLB [43].

It is important to mention that the presence of more than 5% squamous or epithelial cells represents contamination of the sample with bronchial secretions, rendering it a non-representative sample of alveolar cells [37].

Transbronchial Lung Biopsy (TBLB)

TBLB refers to sampling the lung parenchyma via exible biopsy forceps. Anderson and colleagues frst described this method and their results in the 1960s and 1970s [44, 45].

It is usually performed by frst wedging the bronchoscope in the bronchus. The forceps are then advanced in the closed position through the bronchoscope’s working channel under uoroscopic guidance, reaching the lung parenchyma where resistance is felt. The forceps are pulled back about 1 cm to open, re-advanced until the desired tissue is in contact with the forceps, and closed again to obtain a biopsy (Fig. 2.7).

The wedging position of the bronchoscope will facilitate further biopsies without the need to reposition the scope. It will also help isolate and tamponade any signifcant bleeding from the biopsy site.

The yield of TBLB increases with the number of biopsies taken. Descombes and colleagues showed that TBLB yield is increased from 38

Fig. 2.7  Biopsy forceps used with the exible bronchoscope

to 69% when 6 or more biopsies are performed [46]. The yield is also dependent on the pulmonary disease being investigated. The yield in usual interstitial pneumonitis (UIP) is only 30%, whereas higher yield of 70% or more is seen in pulmonary diseases with:

•\ A centrilobular distribution such as granulomatous lung diseases (hypersensitivity pneumonitis and sarcoidosis), eosinophilic pneumonia, and lymphangitic carcinomatosis [47, 48].

•\ Pulmonary infection in an immunocompromised host and mycobacterial infections [42, 43].

•\ Lung transplant patients with acute rejection or infection [49].

TBLB has a low complication rate with major bleeding (greater than 50 mL) averaging 1% and risk of pneumothorax between 1 and 4% [2, 50, 51].

Transbronchial Needle Aspiration (TBNA)

TBNA refers to sampling through the tracheal or bronchial wall. The mediastinal and hilar lymph nodes and lung and mediastinal masses can be sampled via this method. A thorough review of the patient’s chest computed tomography (CT) and knowledge of thoracic anatomy are essential before proceeding. A retractable hollow cytology needle (21 or 22 gauge) or histology needle (19 gauge) is used, with suction applied to the proximal end of the needle. Fluoroscopy should be used when performing TBNA of peripheral lung lesions.

Wang and colleagues in 1978 performed the frst successful TBNA of a paratracheal tumor via the exible bronchoscope. They later published their experience with TBNA for hilar and mediastinal adenopathy [52, 53].

Blind TBNA is the term used for standard TBNA without using EBUS guidance. The sensitivity of blind TBNA varies according to size, location, number of aspirates per lymph node’s

station, and the bronchoscopist experience. A sensitivity of 78% and specifcity of 99% has been reported for blind TBNA in patients with lung cancer [54, 55]. Baaklini and coworkers found a yield of 64% for pulmonary lesions located in the inner third of the lungs versus 35% for lesions located in the outer two-thirds of the lungs. They also showed a lower yield for smaller lesions (<2 cm) [56].

Blind TBNA also has a role in diagnosing peripheral lung masses and nodules, as Katis and colleagues frst showed in 1995 [57].

Despite its great diagnostic utility, both ACCP and UK surveys have shown low routine use of blind TBNA in malignant and non-malignant diseases with 11.8% and 2.3% respectively in the ACCP survey [16] and 10% use in the UK survey [2].

The overall complication rate for blind TBNA is quite low at 0.8% [55]. The most common is damage to the bronchoscope working channel [16].

The introduction of EBUS-TBNA for mediastinal and hilar adenopathy and central tumors has replaced blind TBNA to a great extent. More guidance tools to reach peripheral lung nodules, like radial probe EBUS and virtual bronchoscopy, have changed the way pulmonologists perform TBNA.

A brief review of advanced diagnostic bronchoscopy is to follow.

Bronchial Brushings

Bronchial brushings involve the introduction of a small-protected brush via the exible bronchoscope to the visible endoluminal lesion or peripheral pulmonary nodule with assistance viauoroscopy or guidance tools of bronchoscopy (radial EBUS or virtual bronchoscopy techniques). It is a useful tool to obtain both microbiological and cytological samples. Protected brushings have shown to increase the diagnostic yield in peripheral lung nodules [58].

A review of 30 studies published in 2003 assessed the performance characteristics of different modalities for suspected lung can-