Bronchiectasis and Other Bronchial Abnormalities

Anatomy

With successive generations of bronchial branching, there is a progressive diminution of airway caliber; a lack of tapering is abnormal. The bronchial tree is composed of the large (bronchi) and small (bronchioles) airways. The traditional distinction is based on the presence or absence of cartilage in the wall; airways reinforced by cartilage and typically greater than 1 mm in diameter are termed bronchi , whereas those without cartilage and typically less than 1 mm in diameter are termed bronchioles . Although larger airways are conducive to maximum rate of airflow, a rapid increase in total cross-sectional diameter at the level of the small airways facilitates gas exchange across the alveolar-capillary membrane.

Pathology

The vicious circle hypothesis for the pathogenesis of bronchiectasis, first advocated by Cole and colleagues, is the most widely accepted. The theory is based on the premise that there is initial damage to or disruption of the process of normal mucociliary clearance. This may have been as a result of a viral respiratory infection or a genetic susceptibility, as is the case in patients with cystic fibrosis. The inability to clear mucus permits the “normal” oropharyngeal flora, including Haemophilus influenzae, Pseudomonas aeruginosa, and Streptococcus pneumoniae , to persist within the airways. The organisms lead to further epithelial damage, reducing mucociliary clearance and encouraging further colonization. An inflammatory host response also contributes to further airway damage and dilatation of the airways. With time, there is loss of the elastin layer in the bronchial wall and destruction of cartilage. Some data, based on serial observations by use of computed tomography (CT), suggest that the vicious circle theory may need some refinement, particularly in the pediatric population, because bronchial dilatation diagnosed by high-resolution CT is not always irreversible or progressive.

Three different patterns of bronchiectasis are recognized: cylindrical, varicose, and cystic. In cylindrical bronchiectasis, there is relatively uniform bronchial dilatation, and the implication is that this is perhaps the least severe morphologic type. This should be contrasted with varicose bronchiectasis, in which there are focal narrowings along otherwise dilated airways, and cystic bronchiectasis, which, as the term suggests, indicates grossly dilated airways that have a cystic appearance. Although these morphologic subtypes of bronchiectasis are generally identifiable on high-resolution CT, the clinical utility of making these distinctions is questionable.

Lung Function

The classic physiologic defect in bronchiectasis is airflow obstruction, which in some patients is partially reversible. The obstructive defect has been variably attributed to the presence of coexistent emphysema, airway hyperreactivity or asthma, constrictive bronchiolitis, and expiratory large airway collapse. However, CT studies suggest that areas of decreased attenuation, possibly reflecting constrictive bronchiolitis at the pathologic level, and severity of bronchial wall thickening are the best correlates of obstructive physiology. In some patients with bronchiectasis, there is an associated restrictive ventilatory defect, possibly related to patchy peribronchial fibrosis (presumably consequent on inflammation) and atelectasis.

Radiography

The chest radiograph is generally the first radiologic test requested by the pulmonologist when a diagnosis of bronchiectasis is suspected. However, radiography has low sensitivity and specificity in the diagnosis ( Fig. 57.1 ). However, before the advent of CT and before the conception of the high-resolution CT technique, pulmonologists and radiologists were necessarily reliant on chest radiography, followed in many cases by the then gold standard, but now obsolete, investigation of bronchography, in patients with suspected bronchiectasis ( Fig. 57.2 ). In the frequently quoted study by Gudbjerg, the conclusion was that a mere 10% of patients with bronchiectasis would have a normal chest radiograph. However, although this would have been true at the time, the study was published more than 50 years ago, and as discussed before, the clinical manifestations of bronchiectasis have changed. Moreover, many of the radiographic signs described by Gudbjerg (e.g., linear markings, bronchial wall thickening, and patchy or confluent shadows) are nonspecific. This was clearly illustrated by the later study from the Brompton Hospital that compared chest radiographic diagnoses of bronchiectasis with bronchography and confirmed the highly insensitive nature of chest radiography, also showing considerable disagreement between two experienced pulmonary radiologists simply for confirming the presence or absence of individual features of bronchiectasis.

When identified, bronchial dilatation is the key radiographic feature of bronchiectasis. Thus the observer may see poorly defined ring shadows (reflecting the dilated bronchus with peribronchial inflammation when it is seen end-on) or “tram-line” opacities (denoting the nontapering airway), depending on the orientation of the individual airway with respect to the x-ray beam ( Fig. 57.3 ). With more severe disease, obvious thin-walled cysts, with or without air-fluid levels, may be apparent ( Fig. 57.4 ).

The other recognized features of bronchiectasis on chest radiographs include tubular and branching opacities caused by plugging of dilated airways with mucus, volume loss or hyperexpansion, and foci of subsegmental atelectasis. These are ancillary radiographic signs and must not be regarded as diagnostic of bronchiectasis.

Computed Tomography

In the era before the development of the thin-section CT technique, data on the value of CT in diagnosis of bronchiectasis were disappointing, with sensitivities in the range of 66% to 79%. However, with the realization that spatial resolution could be improved significantly by reducing slice collimation, the reported sensitivities for the CT diagnosis of bronchiectasis improved, with sensitivities approaching 100% using thin-section (1–1.5 mm) imaging.

A factor to be considered in review of the airways on thin-section CT is the window setting because, in the context of a diagnosis of bronchiectasis, this may have a significant influence on the assessment of luminal diameter. In their experimental study, Webb and colleagues showed that for airways surrounded by air, a window centered at −450 Hounsfield units (HUs) provided the most accurate measurement of wall thickness. Indeed, at this window mean, the window width had an imperceptible effect on the evaluation of wall thickness. When the window center was set lower than −450 HUs, wall thickness was overestimated, and the converse was true when the setting was higher.

Computed Tomographic Features of Bronchiectasis

CT signs of bronchiectasis were first described by Naidich and colleagues in 1982, albeit on non–high-resolution CT (10-mm collimation) images. The features identified by these investigators have, with some modifications, largely been validated in subsequent series. Although bronchial dilatation is the key morphologic abnormality, bronchial wall thickening, patchy areas of decreased parenchymal attenuation and vascularity (termed mosaic attenuation ), plugging of large and small airways, volume loss, crowding of airways, and, in some patients with idiopathic disease, thickening of interlobular septa can be seen in association with bronchiectasis.

Bronchial Dilatation.

Bronchiectasis is abnormal permanent dilatation of the airways, with or without associated thickening of the walls of the airways. In the normal state the internal diameter of the airway is roughly equal to the transverse diameter of the accompanying pulmonary artery; a bronchoarterial ratio greater than 1 should generally be regarded as abnormal. The CT appearance of bronchiectatic airways depends on the orientation of abnormal bronchi relative to the plane of section. Thus, for airways that are perpendicular to the imaging plane, the radiologist will look for a dilated airway next to a dot representing the normal adjacent pulmonary arterial branch, an appearance that has been likened to a signet ring ( Fig. 57.5 ). In contrast, for bronchi that lie in the plane of section, as in the middle lobe and lingula, the radiologist should look for an absence of normal tapering ( Fig. 57.6 ).

The identification of airway dilatation in patients with severe (varicose or cystic) bronchiectasis does not usually pose a challenge. However, the majority of patients with bronchiectasis now present with mild disease, and the recognition of less severe (cylindrical) bronchial dilatation is certainly problematic. A spurious increase in the bronchoarterial ratio may be seen because the accompanying pulmonary artery can divide before the airway. In this way, the bronchial diameter will appear to be increased compared with the homologous arterial branch. Another consideration is that a bronchoarterial ratio greater than 1 is not always abnormal because it may be seen in up to one-fifth of healthy subjects. Furthermore, it has been shown that there is a normal but progressive increase in the internal diameter of the airways with age, with a bronchoarterial ratio exceeding 1 in just more than 40% of asymptomatic subjects older than 65 years. Finally, altitude may have a bearing on the relationship between airways and their adjacent pulmonary arteries; 50% of healthy subjects had evidence of bronchial dilatation by conventional CT criteria in one study performed in Colorado. Another study confirmed higher bronchoarterial ratios at high altitude compared with sea level. Although the exact mechanisms are not clear, it is conceivable that the increase in bronchoarterial ratio at altitude might be due to vasoconstriction associated with hypoxia, coupled possibly with the phenomenon of hypoxic bronchodilatation.

Kim and colleagues looked for features on CT that would differentiate normal subjects from patients with surgically proven bronchiectasis. This study showed that, in contrast to an increased bronchoarterial ratio and lack of tapering, the visualization of airways within 1 cm of the costal or paravertebral pleura and the identification of airways abutting the mediastinal pleura were more discriminatory, being seen in 81% and 53%, respectively, of the patients with bronchiectasis but in none of the normal subjects.

Ancillary Computed Tomographic Features of Bronchiectasis

Although dilatation of bronchi is the cardinal morphologic manifestation of bronchiectasis, there are a number of ancillary signs on CT. A diagnosis of bronchiectasis cannot be made on the basis of these ancillary CT features in the absence of bronchial dilatation. These additional signs of bronchiectasis are discussed briefly.

Bronchial Wall Thickening.

Thickening of the bronchial wall commonly accompanies bronchiectasis on CT and is likely to reflect an inflammatory component. As has already been mentioned, appreciation of wall thickening is subject to technical factors; window settings that are too narrow can lead to an overestimation of bronchial wall thickness. Interpretation of bronchial wall thickening is also hampered by the lack of consensus on what constitutes normal wall thickness at CT. In an attempt to address this, Remy-Jardin considered bronchi to be thick walled if the wall of any individual airway was at least twice as thick as that of a normal airway. However, the obvious drawback is the assumption that a normal comparable bronchus will be visible; naturally, this cannot be guaranteed when bronchiectasis is severe and widespread. An alternative definition of wall thickening is internal diameter of the airway lumen less than 80% of its external diameter. This definition is useful but only when bronchial dilatation is relatively mild; plainly, if the dilatation is marked (as in cystic bronchiectasis), wall thickening will be underestimated.

Mosaic Attenuation.

Patchy areas of decreased attenuation, within which there is a reduction in the number or caliber of pulmonary vessels, are not only common but also a functionally important finding on CT in patients with bronchiectasis ( Fig. 57.7 ). The extent of areas of decreased attenuation on expiratory CT does not correlate with indices of gas transfer such as diffusing lung capacity for carbon monoxide, suggesting that areas of mosaic attenuation are due to associated constrictive bronchiolitis.

Regions of air-trapping on expiratory CT tend to be most prevalent in lobes with either extensive or cystic bronchiectasis or in association with large mucous plugs and plugging within the centrilobular bronchioles. However, an additional intriguing finding is the presence of areas of decreased attenuation in some lobes without overt bronchiectasis ( Fig. 57.8 ). On this basis, it has been hypothesized that constrictive bronchiolitis may be the key initial pathologic lesion in bronchiectasis.

Airway Plugging.

Inflammatory mucous secretions may fill the ectatic bronchi, resulting in tubular and nodular opacities representing bronchi seen along their long axis and short axis, respectively. Care must be taken not to miss bronchiectasis in these cases ( Fig. 57.9 ). Mucous secretions within the distal airways and inflammatory thickening of the bronchiolar walls may produce a tree-in-bud pattern, which manifests as Y - or V -shaped opacities or a linear branching structure ( Fig. 57.10 ), depending on the orientation of the airway relative to the imaging plane.

Volume Loss.

Crowding of the airways and volume loss (including, in some patients, complete collapse of a lobe) are recognized features of bronchiectasis that are readily appreciated at CT ( Fig. 57.11 ). It is reasonable to hypothesize that these features of bronchiectasis are a consequence of inflammation and fibrosis around the airways. However, this is not to be confused with the entity of traction bronchiectasis; as the term suggests, this refers to tractional dilatation of airways from adjacent fibrotic interstitial lung diseases (see later).

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) has a limited role in the diagnosis of bronchiectasis ( Fig. 57.12 ), and the clinical utility of such tests for routine clinical management of patients with bronchiectasis remains debatable.

Imaging Algorithms

Despite its constraints, chest radiography is often the first imaging modality obtained for the evaluation of patients with bronchiectasis. However, given the limited sensitivity and the likelihood that the majority of patients now present with less severe disease, many will require thin-section CT imaging.

Etiology, Prevalence, and Epidemiology

Cystic fibrosis (CF) is an autosomal-recessive hereditary disease. It is the most common lethal genetically transmitted disease in white individuals; the estimated incidence is about 1 per 2000 to 3500 live births. CF is uncommon in nonwhites. There is no sex preponderance. The diagnosis is made before age 5 years in about 80% of cases, during adolescence in 10%, and during adulthood in 10% of cases. The fundamental abnormality consists of the production of abnormal secretions from exocrine glands, such as the salivary and sweat glands, and the pancreas, large bowel, and tracheobronchial tree. The major clinical manifestations are obstructive pulmonary disease, which is found in varying degrees of severity in almost all patients, and pancreatic insufficiency, present in 80% to 90%. CF is the most common cause of pulmonary insufficiency in the first 3 decades of life.

Clinical Presentation

The pulmonary manifestations of CF include recurrent respiratory infections associated with productive cough, wheezing, and dyspnea. The infection is predominantly caused by bacteria, such as Pseudomonas aeruginosa , Staphylococcus aureus , and Haemophilus influenzae , although viruses, mycoplasma, and fungi are occasionally responsible. The presence of Burkholderia cepacia is usually associated with advanced disease. Common complications include hemoptysis and pneumothorax. Most patients eventually progress to respiratory insufficiency accompanied by pulmonary arterial hypertension and cor pulmonale.

Pathophysiology

Manifestations of the Disease

Radiography

The earliest manifestations of CF consist of round or poorly defined linear opacities measuring 3 to 5 mm in diameter that are located within 2 to 3 cm of the pleura. Less common early manifestations include thickened bronchial walls without bronchial dilation, usually seen as ring shadows, and mild hyperinflation. Progression of disease is characterized by increases in bronchial diameter, bronchial wall thickness, lung volume, and number and size of peripheral nodular opacities and by the development of mucoid impaction and focal areas of consolidation ( Figs. 57.14 and 57.15 ). Bronchiectasis, bronchial wall thickening, and mucous plugging are particularly frequent and are evident in almost all adult patients. The chest radiographic abnormalities have been incorporated into numerous semiquantitative scoring schemes that are believed to be of value in predicting prognosis and directing therapy.

Recurrent foci of consolidation occur in most patients, and lobar or segmental atelectasis occurs in many. Hilar enlargement may be due to lymph node enlargement or dilation of the central pulmonary arteries secondary to pulmonary arterial hypertension (see Fig. 57.15 ). Pneumothorax occurs in 3% to 19% of patients.

Computed Tomography

The main manifestation of CF is bronchiectasis, which is present in virtually all adult patients ( Fig. 57.16 ; see Fig. 57.14 ). Bronchiectasis usually involves all lobes but tends to be most severe in the upper lobes. The bronchiectasis may be cylindrical, varicose, or cystic. Other common findings are bronchial wall thickening, peribronchial interstitial thickening, and mucous plugging. Consolidation or atelectasis can be seen in 80% of cases. Cystic or bullous lung lesions also can be seen and typically predominate in the subpleural regions of the upper lobes (see Fig. 57.16 ).

Branching or nodular centrilobular opacities (tree-in-bud pattern) are frequently present and may be an early sign of disease. They reflect the presence of bronchiolar dilation with associated mucous impaction, infection, or peribronchiolar inflammation. Focal areas of decreased attenuation and vascularity on inspiratory CT and air-trapping on expiratory CT are common.

Hilar or mediastinal lymph node enlargement and pleural abnormalities may be present, largely reflecting chronic infection. Pulmonary artery dilation resulting from pulmonary hypertension is common in patients who have long-standing disease.

Magnetic Resonance Imaging

MRI may be used as an alternative tool to assess parenchymal and functional abnormalities and to monitor treatment in patients with CF. Although MRI has lower spatial resolution than CT and plays a limited role in the assessment of airway disease, it has the advantage of no radiation exposure, an important consideration in these young patients. MRI may be useful in the follow-up of patients with CF ( Fig. 57.17 ) and could be used to screen for early liver disease, given the increased life expectancy of CF patients in recent years.

Differential Diagnosis

The radiologic findings of CF may mimic those of other causes of bronchiectasis, such as allergic bronchopulmonary aspergillosis and postinfectious bronchiectasis. The most characteristic features of bronchiectasis in CF are bilateral symmetric distribution and upper lobe predominance. In one study, bronchiectasis in CF involved mainly the upper lobes in 79% of patients and was usually bilateral and fairly symmetric, whereas the bronchiectasis in the other conditions, such as prior tuberculosis, was most frequently unilateral or asymmetric.

Although the diagnosis of CF may be suggested by a positive family history, persistent respiratory disease, clinical evidence of pancreatic insufficiency, or fatty (or less common cystic) replacement of the pancreas on imaging, confirmation requires a positive sweat test or identification of two abnormal copies of the CF gene. Incorporation of molecular biologic techniques into rapid, cost-efficient, and specific diagnostic tests for most CF genotypes is now possible by using the multiplex polymerase chain reaction.