Chapter 8

Biopsy techniques

Samuel V. Kemp

Dept of Respiratory Medicine, Royal Brompton Hospital, London, UK.

Correspondence: Samuel V. Kemp, Dept of Respiratory Medicine, Royal Brompton Hospital, Fulham Road, London, SW3 6NP, UK. E-mail: s.kemp@rbht.nhs.uk

The term “biopsy” is derived from the Greek words bios (life) and opsis (sight), and was first used in 1879 by Ernest Besnier [1], a French Dermatologist also credited with the term “lupus pernio” [2]. It is defined as “An examination of tissue removed from a living body to discover the presence, cause, or extent of a disease” [3] and is therefore not limited to forceps biopsy for the purposes of histological examination. Tissue can be obtained at bronchoscopy for the histological and cytological diagnosis of both benign and malignant disease, and can be very important in the diagnosis of infection when the appropriate microbiological samples are taken. This chapter covers the basic- and intermediate-level techniques for obtaining tissue via the bronchoscope, with more advanced techniques discussed elsewhere in this Monograph [4].

General considerations

There are a number of general considerations when one is considering sample collection at bronchoscopy. Why do you want samples: are you suspicious of malignancy, infection, parenchymal disease? How are you hoping to process those samples: histology, cytology, microbiology? Then consider where in the lung the pathology in question is: peripheral or central; in the airways, parenchyma or nodes? What comorbidities does the patient have, are they on anticoagulants or antiplatelets and can they cope with the complications of sampling (e.g. pneumothorax after TBB)?

Training and assessment of competence in bronchoscopy is covered in more depth elsewhere in this Monograph [5], but every bronchoscopic sampling technique, however apparently simple, will present challenges and a learning curve, and each bronchoscopist will progress along that curve at a different rate [6]. Therefore, it is essential that tuition and supervision are provided during the learning phase for each and every technique, and that records are kept to demonstrate the maintenance of competence.

In this chapter, sampling techniques have been split into: 1) basic techniques: endobronchial biopsy, bronchial brushings and bronchial washings (i.e. those that are most commonly used and easiest to learn) and 2) intermediate techniques: TBB, BAL and TBNA (generally variations on basic techniques that require more practice and skill, but should still be part of the basic armoury of any independent bronchoscopist).

Most of the yield and sensitivity data for bronchial sampling techniques have been collected in the context of malignant airways disease (also see MYERS and LAM [7]) and although this is the focus of most of the evidence presented here, benign disease is discussed where appropriate, especially for intermediate techniques. Also discussed where relevant is how to combine techniques to ensure the highest possible yields.

Basic techniques

Endobronchial biopsy

The predominant aim of bronchial biopsy is to obtain a small piece of tissue that contains all the relevant structures and cells for appropriate histological analysis. However, in cases where there is the suspicion that endobronchial abnormalities result from infection (e.g. tuberculosis and fungal disease), biopsy specimens can also be sent for tissue culture. The indication for bronchial biopsy is usually to sample abnormal tissue visualised in the airways to confirm or refute the diagnosis of suspected malignancy (primary or secondary), but is also useful in the diagnosis of benign airway pathology, such as conditions of an infiltrative or inflammatory nature (e.g. sarcoidosis, amyloidosis and tracheobronchopathia osteochondroplastica). Biopsy can also be used to confirm the presence of normal tissue, such as at previous resection sites or when monitoring areas of previous dysplasia or carcinoma in situ.

The vast majority of endobronchial biopsies, however, are for the diagnosis of suspected malignancy and endobronchial biopsy is the most reliable technique for obtaining a diagnosis in cases where a lesion is visible at bronchoscopy. Location, lesion accessibility, operator experience and tumour type are all factors in determining yield. The overall yield quoted for forceps biopsy in the literature is predominantly within the range of 74–85% [8–14], although that from exophytic tumours in the central airways has been documented at >90% [15–17]. Yield can be increased further with the addition of brushings and washings for cytological examination [10, 18–20].

Once a visible lesion has been identified, the bronchoscopist passes the forceps through the working channel of the bronchoscope. The forceps are opened and then advanced onto (or into) the lesion and closed when sufficient tissue is felt to be within the jaws of the forceps. The forceps are removed from the working channel and the sample placed in the desired transport medium for transfer to the laboratory for processing. Guidelines suggest that at least five biopsies are taken from visible tumours, such that enough tissue is retrieved to ensure diagnostic material is obtained, and that further molecular and genetic testing can be performed as appropriate [21]. Ideally a subepithelial area of at least 0.3–0.5 mm² is required; it may be necessary to take more biopsies if the mass appears necrotic.

Greater than 50% of lung cancers involve the central airways [22], with >80% of some tumour subtypes visible at bronchoscopy [14]. However, there has been a shift in histological subtypes of lung cancer over the last 40 years or so, with a greater proportion of adenocarcinomas and a decreasing proportion of squamous and small cell lung cancers, at least partly due to changing smoking habits (figure 1) [23]. Adenocarcinomas have a greater predilection for the peripheral lung, and squamous and small cell cancers for the central airways, which has led to a greater number of lesions invisible to the standard bronchoscope. For these tumours, the yield is lower and more advanced techniques need to be employed, some of which are discussed later in this chapter and elsewhere in this Monograph [4].

Figure 1. Changes in relative incidences of lung cancer cell types 1973–2010: a) male and b) female (all races). Reproduced and modified from [23] with permission.

Several types of forceps are available (figure 2). Forceps with a central spike can assist with anchoring to a smooth, hard, tumour where other forceps may simply slide off as they are closed. However, the choice of forceps is usually down to operator preference and there is no good evidence that yield or performance is affected by cup design. Some forceps are equipped with a swing-jaw mechanism, which can allow for tangential biopsies to be taken.

Figure 2. Examples of different biopsy forceps: a) alligator cup forceps, b) alligator cup forceps with spike, c) oval cup forceps and d) oval cup forceps with spike. Reproduced with kind permission of Olympus (Southend-on-Sea, UK).

Biopsies are immediately fixed in formalin in the bronchoscopy suite and embedded in paraffin wax blocks on arrival at the laboratory. Sections are then cut onto glass slides and stained with haematoxylin/eosin, although other targeted stains can also be used in certain situations (e.g. Masson’s trichrome for the collagenous connective tissue of pulmonary fibrosis). Additional sections can then be cut for immunohistochemistry and molecular testing as appropriate.

Complications from simple forceps biopsy of visible lesions are rare. Statistics for bleeding rates are difficult to determine, as the definition of what constitutes a “significant” bleed is variable, and accurately measuring blood loss can be impossible owing to the admixture of secretions, saline and other instilled liquids, but clinically significant bleeding following endobronchial biopsy is seen in <0.5% of cases [24]. Bleeding requiring intervention is usually the preserve of therapeutic rather than diagnostic procedures. Several patient factors have been associated with an increased risk of bleeding at bronchoscopy, including renal failure [25] and the use of anticoagulant or antiplatelet medication [26], and these should be taken into account when deciding whether to undertake a biopsy. Thrombocytopenia has been associated with bleeding following BAL [22], but there are no robust statistics available for the risk following endobronchial biopsy.

Bronchial brushings

Bronchial brushings involve the collection of cells and material from the airway epithelium, tumour or other abnormality using a catheter-based flexible bush. The technique is simple and relatively cheap, but only provides cytology and therefore cannot give information about invasion. Covered brushes can also be used to obtain microbiology samples, predominantly for fungi or viruses. More recently, brushings have been utilised for the assessment of the microbiome and microenvironment of the airways in health and disease [27–30]. The lesion in question is identified either under direct vision or at fluoroscopy and the brush passed through the working channel to make contact with the lesion. The brush is then agitated against the lesion. Processing of the brush sample varies between institutes, and close liaison is required between the bronchoscopy team and the pathologists to ensure optimum sample processing. At the author’s institute, specimens are first cytospun onto slides and stained with Papanicolaou stain for reporting. If immunocytochemistry is needed, formalin is added to fix the cells, which are then centrifuged and agar added to make a pellet that is processed as a histology specimen to make paraffin wax blocks and slides.

Reported sensitivities for bronchial brushings vary greatly depending on location and tumour type. Diagnostic yields of 44–82% have been reported for endoscopically visible tumours [14, 31–34]. Samples from the main bronchi and bronchus intermedius outperform those from lobar airways, and when a tumour is not visible at bronchoscopy, yields are as low as 19–40% [14].

Although common sense dictates that when there is visible tumour within the airway, bronchial brushings are unlikely to add much to forceps biopsy, which provides a far greater amount of information for the pathologist and crucially provides information about invasion, there is evidence of increased yields over biopsy alone [14, 35]. As sampling relies on the adherence of material to the bristles of the brush, yields can be affected by other material present in the airway, e.g. blood and mucus, and the order in which samples are taken does appears to be important, with pre-biopsy brushings providing higher yields that post-biopsy brushings in one study [36]. In contrast to endobronchial biopsy, brushings may perform better when faced with a necrotic tumour and at least one study has shown brushings to outperform biopsies in the diagnosis of small cell lung cancer as there is no crush artefact to obscure cell morphology [37].

Bronchial washings

Bronchial washings are used to clear the airways of debris and secretions, and where indicated are intended to gather cells and microorganisms washed from the airway walls to be analysed for the presence of disease states. Aliquots of 10–20 mL of 0.9% saline are instilled into the airways and suctioned into a specimen collection pot. Washings can be particularly useful in the diagnosis of infectious states and are the test of choice for the detection of nontuberculous mycobacteria. Washings for the diagnosis of Pneumocystis jirovecii pneumonia (PJP) first came to prominence in the early years of HIV/AIDS in the USA, and have even been shown to outperform tissue biopsy and bronchial brushings [38, 39].

The reported yield of bronchial washings for the diagnosis of cancer varies greatly, and can be dependent on the location and type of tumour, with a range of 14–63% [14, 31, 40–43], although yields can be increased by preparing cell blocks [44, 45]. However, while guidelines continue to recommend them [21], the utility of bronchial washings in cases where the tumour is bronchoscopically visible is questionable if biopsies and brushings have also been obtained, and several studies have shown washings not to be cost-effective or significantly increase yield [11, 40, 46, 47]. Unlike with bronchial brushings, the order in which samples are obtained does not appear to alter yield [20], except perhaps in submucosal lesions where post-biopsy washings perform best, probably owing to the exposure of cancer cells with the removal of overlying tissue [48].

Intermediate techniques

Transbronchial biopsy

TBLB is performed when the target lesion is in the peripheral airways and not visible at bronchoscopy. The bronchoscope is first wedged in the (sub)segmental bronchus of interest and forceps are passed through the working channel of the bronchoscope. The forceps are advanced beyond view until resistance is encountered and then withdrawn by 1–2 cm. The jaws of the forceps are then opened, the forceps advanced and the jaws closed to obtain the biopsy sample. Some operators advocate the coordination of forceps advancement and closure with the patient’s respiration, where the forceps are opened and advanced in inspiration and the biopsy taken at end-expiration. The sample is placed into formalin (histology) or saline (microbiology) and processed in the same manner as for endobronchial biopsy.

As with BAL, TBLB performs best in the diagnosis of diffuse lung diseases, including ILD, infection and malignancy [49–51]. In solitary lesions, yield is heavily influenced by lesion size and also distance from the hilum, with very poor yields for lesions <2 cm in diameter in the outer third of the lung [7, 49, 52–53]. When targeting a discrete or focal lesion, the use of fluoroscopic guidance has been shown to significantly improve the diagnostic yield [54, 55], although it appears not to do so in diffuse disease [55, 56]. The use of fluoroscopic guidance is, however, being superseded by modern navigation techniques in the investigation of discrete pulmonary lesions and these are described in detail elsewhere in this Monograph [57].

Although BAL and CT together can be diagnostic in some cases of ILD, TBB is still sought in many cases for histological confirmation where the multidisciplinary team is not confident about the underlying process(es). In a review of 164 TBLB cases, HANSON et al. [58] reported diagnostic accuracies of 62%, 64% and 67% in infectious, interstitial and malignant lung diseases, and another early report by SMITH et al. [59] of TBLB via the flexible bronchoscope reported diagnostic material in 34 out of 40 patients (85%). Yields for sarcoidosis are variable, ranging from 50% to 91% [56, 60–63]. TBLB is also used in the surveillance of rejection in lung and heart–lung transplant patients [64, 65], and has been shown to be useful in the diagnosis of peripheral lung amyloidosis [66].

The usefulness of TBLB in the diagnosis of usual interstitial pneumonia or idiopathic pulmonary fibrosis has been hotly debated, particularly in the era of high-resolution cross-sectional imaging [67, 68], and is probably most useful for excluding other interstitial processes [69]. However, especially in early disease or where other confounding factors exist, such as the presence of rheumatological disease, it is still desirable to obtain diagnostic tissue. Although some studies have suggested that TBLB may have a role to play in the diagnosis of usual interstitial pneumonia [70, 71], surgical lung biopsy is often recommended by guidelines [72–74]. Nonetheless, recent advances in the examination of pathological specimens, such as machine learning and genetic analysis [75], have the potential to increase the sensitivity of TBLB in this setting, and the advent of transbronchial cryobiopsy may allow surgical biopsy to be avoided in many cases [76, 77]. Transbronchial cryobiopsy is discussed in detail elsewhere in this Monograph [78].

The sensitivity of TBLB reported in the setting of suspected malignancy varies greatly, ranging from 17% to 80% [14, 16, 17, 79–82], and depends on the population studied, whether primary or secondary disease and on lesion size (as discussed earlier). TBLB is, however, the investigation of choice when considering lymphangitis carcinomatosis, with yields as high as 100%, albeit in small series [51, 83]. There is also recent evidence that TBLB is helpful for rebiopsy in relapsed nonsmall cell carcinoma for the purposes of mutational analysis [84].

There appears to be little or no difference in the size or quality of biopsies obtained with different types of biopsy forceps [85–87] and although larger forceps appear to obtain larger specimens, this does not necessarily translate into significantly higher diagnostic yields [88, 89].

Owing to the inherent inaccuracies of the technique, guidelines recommend taking four to six samples in diffuse lung disease and seven or eight samples in focal lung disease to maximise the chances of obtaining diagnostic material [21], although there is a balance to be struck between the need for diagnostic tissue and the risk of complications.

Complications are more common with TBB than endobronchial biopsy, although they are rarely serious. Where reported, pneumothorax and clinically significant bleeding are the two main significant complications, occurring in 1–10% and up to 9%, respectively [49–51, 55, 56, 58, 63, 90]. Pleuritic pain during the procedure can be a feature and is important as it suggests the forceps have been advanced too far: the only pain-sensitive structure encountered is the parietal pleura, supplied by the intercostal and phrenic nerves (the visceral pleura is innervated by the autonomic nervous system and therefore has no sensory innervation). Other adverse events are rare, but include mediastinal and subcutaneous emphysema [91, 92]. Although death is very rare, it has been reported [93, 94], occurring in ∼0.2% of cases [90].

While there is a risk of pneumothorax, and a chest radiograph is often requested after the procedure, there is little or no evidence that routine imaging is required. One study at the University of Virginia (Charlottesville, VA, USA) did not find a single case of unsuspected pneumothorax in over 300 patients across nearly 6 years [95], suggesting that imaging should only be requested if there is clinical concern. The study only included patients undergoing TBLB with fluoroscopic guidance, but, perhaps surprisingly, there does not appear to be any difference in pneumothorax rates between procedures performed with and without fluoroscopy [55, 56, 96].

More than with any of the other techniques described in this chapter, attention to the patient’s coagulation status is vitally important. Peripheral lung biopsies very often contain vascular structures and while rare, catastrophic haemorrhage has been reported [90]. Each institution will have its own safety protocols in this regard, but the following seem practical and reasonable: international normalised ratio <1.5; low-molecular-weight heparin discontinued for at least 24 h; clopidogrel and other platelet ADP receptor blockers discontinued for 5–7 days. The duration of action of the novel oral anticoagulants (e.g. rivaroxaban and dabigatran) varies and advice on the duration of cessation should be sought where appropriate.

Owing to the potential increased risks with TBBs over other bronchoscopic biopsy techniques, there are a number of relative and absolute contraindications which largely relate to the ability of the patient to cope with either a pneumothorax or significant haemorrhage. Absolute contraindications include medical instability, severe hypoxia, life-threatening arrhythmia, massive haemoptysis and uncorrectable bleeding diathesis. Caution should be exercised if there is thrombocytopenia <50 000 μL^–1, pulmonary hypertension and uraemia, where there is a danger of serious haemorrhage even with normal coagulation. Desmopressin and cryoprecipitate can be effective in controlling haemorrhage in such situations [97, 98].

Bronchoalveolar lavage

BAL is a technique designed to sample the peripheral rather than central airways, and although used in the diagnosis of peripheral lung cancers, has a greater role to play in nonmalignant disease. BAL has been used for almost 100 years as a means of irrigating the lungs [99, 100], but began to be used more widely for the diagnosis of respiratory disease in the 1970s and 1980s. First, the bronchoscope is passed as distal as possible towards the target area of lung and “wedged” into an airway to provide as good a seal as possible. Aliquots of 0.9% saline are then instilled via the working channel of the bronchoscope and aspirated. In general, 30–60 mL is instilled at a time, with a total of between 100 and 300 mL being used dependent on institutional protocols and indication. Return is variable, and depends on the underlying condition(s) and the location of the lavage. The resultant aspirate contains cells, proteins, organisms and other material from the airways and epithelium. The American Thoracic Society clinical practice guidelines on the clinical utility of BAL in ILDs provides a comprehensive overview of the history and application of BAL [101].

When used for nonmalignant disease, BAL is important in determining the nature of a variety of pulmonary diseases and aims to sample the epithelial lining fluid (ELF). ELF provides information about the immunological, inflammatory and infectious processes taking place at the alveolar level, and is perhaps most often used in the investigation of ILDs and for the diagnosis of infection in the immunocompromised (e.g. chemotherapy and immunosuppressant drugs, haematological malignancies, and HIV). Approximately 1 mL of ELF is recovered for each 100 mL of saline instilled.

BAL fluid is usually sent for a differential cell count, which can give important guidance as to the nature of the underlying interstitial process when used in conjunction with the clinical history and radiology. Table 1 provides an overview of the diseases associated with various differential cell counts. Microbiology samples can be sent for standard bacterial and mycobacterial cultures, but also for PJP, viral studies and to determine levels of proteins associated with various atypical infections, such as galactomannan for invasive aspergillosis [102, 103] and β-glucan for PJP [104–106].

Stay updated, free articles. Join our Telegram channel

Join