INVESTIGATIONS
Imaging in chest disease
Indications for chest radiography (CXR)
Aarti Shah
Clinical settings in which CXR may be of value C Investigation of patients with symptoms/signs of respiratory disease C Initial evaluation in patients with suspected lung cancer C As part of ‘baseline’ tests in other chronic but non-malignant lung disease (e.g. sarcoidosis, hypersensitivity pneumonitis, bronchiectasis) C As a ‘triage’ tool in diagnostic algorithms for patients with suspected acute pulmonary embolism C Initial screening for patients with systemic disorders (e.g. rheumatoid arthritis, systemic sclerosis) in which lung disease occurs and contributes significantly to morbidity and mortality C Follow-up (e.g. after treatment or a period of observation) in patients with established acute or chronic lung disease Clinical settings in which CXR has less impact C Accurate morphological characterization and diagnosis of diffuse interstitial lung disease C Accurate diagnosis of airways disease (e.g. bronchiectasis, obliterative bronchiolitis) C Diagnosis of acute pulmonary embolism C Staging of lung cancer and other extrathoracic malignancies
Sujal R Desai
Abstract Chest radiography (CXR) and computed tomography (CT) are among the more commonly needed imaging investigations in patients with lung disease. The CXR is often the first test requested and has the advantage of a low radiation dose. However, the utility of CXR is limited e particularly for diffuse lung diseases e and CT is then generally required. Modern CT scanners can acquire images of the thorax in a single breath-hold and provide exquisite morphological detail but the added radiation burden should be a major consideration for physicians and radiologists alike.
Keywords airways disease; chest radiography; computed tomography (CT); consolidation; diffuse parenchymal lung disease; ground-glass opacification; lung cancer
Introduction Table 1
Clinicians regularly request imaging studies in patients with suspected or established lung disease and modern radiology departments offer a variety of tests. In routine practice, chest radiography (CXR) and computed tomography (CT) usually suffice but in specific circumstances, other tests (for instance, thoracic ultrasound to investigate a pleural effusion or 18fluorodeoxyglucose-positron emission tomography (18FDG-PET) in the patient with lung cancer) may be indicated. In this review, the basic principles of imaging tests are discussed; the advantages and limitations of each will be considered. Some of the common terms that appear in radiology reports are discussed and this is followed by a review of radiological features in a variety of common, and some less familiar lung diseases.
follow-up of patients with an established diagnosis (Table 1). However, CXR interpretation is not straightforward. The twodimensional format leads to anatomical superimposition and the characterization of radiological patterns can be difficult. Another issue with CXR is the limited contrast resolution (i.e. the ability to distinguish between tissues of differing physical density). Thus, in specific clinical scenarios (most notably diffuse parenchymal lung disease), the confidence and, more importantly, the accuracy of a CXR diagnosis is unacceptably low.1 The standard CXR projection is the postero-anterior or PA, meaning that the radiographic film cassette is in front of the patient and the X-ray source behind. The PA projection minimizes the magnification of the cardiac silhouette (an anteriorly positioned organ) caused by the divergent X-ray beam. The anterior-posterior (AP) projection, where the patient faces the X-ray source and the radiographic cassette is against the back, is most often used in immobile or critically-ill patients and image quality may be less than ideal. Other radiographic projections (e.g. the lateral, the lateral ‘shoot-through’ and lordotic views) are requested less frequently, in large measure because of the wider availability of and access to CT.
Imaging tests commonly requested in clinical medicine Chest radiography (CXR) Despite acknowledged limitations, the CXR is a mainstay of clinical investigation. In most institutions, digital technology has paved the move away from conventional hard-copy to soft-copy review using picture archiving and communications systems, more conveniently known by the acronym PACS. Because of the relatively low radiation burden, the CXR is a valuable test not only for confirming the presence of lung disease but also in the
Computed tomography (CT) The key components of all CT machines are the X-ray tube and detectors (both housed in the scanner gantry), computer hardware to reconstruct the image and a sliding table, which, in the newest generation of multidetector scanners, moves continuously during scanning. The essence of CT imaging is that, as the X-ray beam passes through tissues of varying physical density, there is differential attenuation; the X-ray source and detectors rotate around the patient and the transmitted X-ray energy, from all projections, is captured by the detector array.2 The final image
Aarti Shah MBBS FRCR is a Specialist Registrar in Radiology, King’s College London, King’s Health Partners, King’s College Hospital NHS Foundation Trust, London, UK. Conflicts of interest: none declared. Sujal R Desai MD FRCP FRCR is a Consultant Radiologist & Honorary Senior Lecturer, King’s College London, King’s Health Partners, King’s College Hospital NHS Foundation Trust, London, UK. Conflicts of interest: none declared.
MEDICINE 40:4
177
Ó 2012 Published by Elsevier Ltd.
INVESTIGATIONS
Figure 1 Three CT images from a single acquisition, photographed at identical anatomical levels but displayed on three different window settings: (a) lung, (b) soft-tissue and (c) bone.
comprises a matrix of picture elements (pixels) with each pixel assigned a numerical value (ranging from 1000 through zero to þ1000 Hounsfield units, HU) proportional to the attenuation in that specific volume (voxel) of tissue. Based on these Hounsfield values, each pixel is assigned a grey-scale value reflecting the tissue: bone is highly attenuating and is represented as white whereas air is represented by pixels that are near black. All other tissues and fluid have intermediate values between these two extremes. One of the advantages of CT is that the observer may, by choosing the appropriate ‘window’ settings, use the digitized data to focus on different tissues (Figure 1): typical settings for analysing the lung parenchyma would be a window centred at around 550 HU with a width of 1500 HU. The technique of high-resolution CT (HRCT) was an important advance in lung imaging (Table 2). Narrow beam collimation, the use of a dedicated (mathematical) image reconstruction algorithm and a small field-of-view are the essence of HRCT imaging. However, a fourth element of HRCT, sometimes forgotten in the era of multidetector CT (MDCT), is the gap (typically 10e20 mm) between image slices. The logic is that in diffuse lung disease the whole lung does not need to be ‘sampled’; the interspaces between images has the welcome benefit of significantly reducing dose.3 The thin collimation reduces partial volume averaging and, together with the highspatial-frequency reconstruction, gives HRCT images of the lung their characteristic ‘sharp’ appearance (Figure 2).
The development of spiral or helical CT scanning was a significant advance, because of the greater acquisition speed and the consequent reduction of misregistration artefacts caused by inconsistent breath holds between scanning
Indications for high-resolution CT C
C
C
C
C
In patients with a strong clinical suspicion of diffuse parenchymal lung disease but a normal CXR To establish a histospecific diagnosis of a diffuse parenchymal lung disease in patients with an abnormal CXR To clarify the morphological basis of sometimes complex lung functional impairment profiles (e.g. emphysema co-existing with lung fibrosis) For ‘staging’ the extent/severity of lung disease (e.g. prior to commencing treatment or in clinical trials/research) For the diagnosis and follow-up of patients with airways disease (e.g. bronchiectasis, obliterative bronchiolitis)
Figure 2 Two CT images at the level of the carina in a young patient being investigated for a rare cystic lung disease. (a) 5-mm collimated image with a standard reconstruction algorithm compared to (b) 2-mm section collimation and processed with a high-spatial-frequency (bone) reconstruction algorithm.
Table 2
MEDICINE 40:4
178
Ó 2012 Published by Elsevier Ltd.
INVESTIGATIONS
segments.4,5 The addition of further detectors in MDCT machines led to significant reductions in scan acquisition times. Modern-day (64-channel and greater) machines can acquire images of the whole chest during a single breath hold, while the very thin collimation of images on MDCT machines now generates lung images of exceptional detail. MDCT images can be reconstructed in any plane desired, thereby matching the capabilities that were previously the domain of magnetic resonance imaging.
Ultrasound (US) A key attraction of US (in which a real-time grey scale image is generated from high-frequency sound waves reflected at anatomic surfaces with differing acoustic impedance) is the absence of ionizing radiation. However, the indications for US in chest diseases are relatively limited. This is because sound waves are completely reflected at soft tissueeair interfaces. Therefore, more often than not, US only has a ‘problem-solving’ role in answering specific clinical questions. A common indication for US is in the evaluation of a pleural effusion, including those patients in whom intervention (e.g. diagnostic aspiration or therapeutic drainage) is contemplated. US is also indicated for the investigation and biopsy of chest wall lesions and, in specialist hands, for the in utero diagnosis of congenital lung abnormalities.
18
Fluorodeoxyglucose-positron emission tomography (18FDG-PET) After chest radiography and CT, 18FDG-PET has become an important imaging test. Clinical PET imaging is based on the principle that metabolically active tissues need glucose. For the majority of clinical indications, glucose is coupled with a radioactive positron-emitting nuclide of fluorine (18F), the combination being termed 18fluorodeoxyglucose (18FDG). A positron-emitting isotope is one that is unstable by being neutron deficient. The instability causes the transmutation of a proton into a neutron, releasing a positron and a neutrino. The reaction between the positron and an electron creates two g ray photons that travel at 180 to each other, and it is these that are captured by the PET detectors; thus, PET scanners build a ‘map’ of metabolically active tissues. One drawback of PET images alone is their limited spatial resolution but this has been overcome in ‘hybrid’ PET-CT machines, which fuse (i.e. co-register) the superior anatomical detail from CT with the functional information provided by 18FDGPET. The most obvious indications for PET are the differentiation between malignant and benign disease e the former tending to have higher standardized uptake values e and the staging/ follow-up of patients with known lung cancer.6 False-negative PET in malignancy is well recognized when lesions are small, and in adenocarcinoma and carcinoid tumours. Conversely, very high FDG uptake may be a feature of benign (especially infective) pathologies.
Magnetic resonance imaging (MRI) Despite the initial enthusiasm, MRI has found few, if any, routine roles in investigating chest diseases. The main constraints to MRI imaging7e10 of the lungs are: the low proton density (giving a poor signal-to-noise ratio) the degradation of images because of movement related to breathing and cardiac pulsation the rapid decay of the MR signal caused by widely differing magnetic susceptibilities at airetissue interfaces. In contrast to lung imaging, there is a more established role for MRI in the investigation of mediastinal pathology, including aortic diseases. The availability of many dedicated imaging sequences means that questions about the composition and anatomical relationships of mediastinal lesions are readily answered by MRI; the latter may be of particular value when neurogenic masses are being investigated because the proximity to the cord/nerve roots is shown to advantage. A guide to interpretation When reviewing chest radiographs and CT it is important to have a systematic approach. One of the first steps for the observer should be to identify the predominant radiological pattern or patterns. Typical patterns of lung disease include consolidation, ground-glass opacification, a reticular pattern e with or without honeycombing e and thickening of interlobular septa (see Terms commonly encountered in radiology reports). An appreciation of the dominant pattern(s) will inform the observer about the predominant site of disease at an anatomical/pathological level: for instance, a pattern of consolidation generally indicates that there is pathology in the air spaces. The distribution of the abnormality (upper versus lower zone, central versus peripheral or random) may be of diagnostic importance. It is known that some interstitial lung diseases have a predilection for the upper zones (e.g. sarcoidosis, Langerhans’ cell histiocytosis) whereas others are more pronounced in the lower zones (idiopathic pulmonary fibrosis, asbestosis). A review of serial changes on CXR and, particularly, the rate of change may also be useful in making a diagnosis: air space opacities that appear and resolve relatively quickly (over a period of hours or, at most, days) are more suggestive of oedema fluid than any other pathology. Similarly areas of consolidation that appear to migrate should make the observer think about an eosinophilic lung disease or organizing pneumonia. Note should be made of presence or
Ventilation-perfusion (V/Q) scintigraphy V/Q scintigraphy, a test that is most frequently requested for the diagnosis or exclusion of acute pulmonary embolism (PE), is a radionuclide technique that demonstrates the distribution of pulmonary ventilation and perfusion. The ventilation phase of imaging is performed using a radioactive 99mTechnetium-labelled aerosol (e.g. 99mTc-DTPA) or a more expensive inert gas (e.g. 81m krypton or 133xenon). A map of pulmonary perfusion is then produced after injection of 99mTc-labelled microspheres which are trapped in vessels occluded by emboli thereby creating “photonpoor” regions on the perfusion image. Two or more mismatched (i.e. abnormal perfusion but preserved ventilation) segmental defects indicate a high probability of acute pulmonary embolism. The corollary is that a normal V/Q study, in a patient with a low clinical pre-test probability, virtually rules out the diagnosis. The almost wholesale inclusion of CT pulmonary angiography in many diagnostic algorithms is contrasted with the apparent reluctance nowadays of clinicians to request V/Q scans. This shift in clinical behaviour might, in part, be explained by the high rate of intermediate-probability studies and also the difficulties in performing V/Q outside of normal working hours.
MEDICINE 40:4
179
Ó 2012 Published by Elsevier Ltd.
INVESTIGATIONS
Some causes of common radiological patterns on CXR and CT Consolidation Infection Infarction/haemorrhage/contusion Tumours (adenocarcinoma, lymphoma) Organizing pneumonia Ground-glass opacification Oedema (cardiogenic, non-cardiogenic) Infection Haemorrhage Interstitial lung diseases (e.g. non-specific interstitial pneumonia, subacute hypersensitivity pneumonitis, respiratory bronchiolitis) Tumours (adenocarcinoma, lymphoma) Organizing pneumonia Eosinophilic lung disease Reticular pattern Interstitial fibrosis (e.g. idiopathic pulmonary fibrosis, sarcoidosis) Infection (e.g. viral, mycoplasma) Lymphangitis carcinomatosa Oedema Nodules/masses Lung cancer Infection (e.g. granulomatous (e.g. tuberculous) abscess) Metastases Organizing pneumonia Intrapulmonary lymph node Table 3
absence of any ancillary findings (e.g. enlargement of intrathoracic lymph nodes in a patient with suspected tuberculosis, pleural plaques in conjunction with signs of basal interstitial fibrosis). Naturally, the importance of considering the clinical background, in concert with the imaging findings, cannot be overemphasized. A history of smoking, an underlying connective tissue disease or treatment with drugs known to cause lung disease may be the all-important clinical clue when interpreting CXR and CT.
Figure 3 Lingular consolidation on (a) targeted CXR and (b) HRCT in a 48-year-old female patient with neutropenic sepsis. The dense opacification on both CXR and HRCT obscures vessels and bronchial markings. An air bronchogram/bronchiologram is well-shown on the targeted CXR (arrows) and CT (arrowheads).
Terms commonly encountered in radiology reports Ground-glass opacification This refers to the hazy increase in lung density that, on CXR, obscures the margins of vessels. By contrast, on CT, the vessels and walls of airways are readily visible in areas of increased parenchymal density (Figure 4). Regardless of the specific cause, this radiological sign is seen when any process causes partial displacement of air. This includes an inadequate inspiratory effort during image acquisition.
Diseases of the chest can produce varied radiological signs (Table 3). However, certain common radiological terms appear regularly in radiologists’ reports and are described below11: Consolidation Consolidation is the homogenous increase in lung density that, on CXR and CT, obscures the margins of vessels and the walls of airways (Figure 3). An air bronchogram/bronchiologram may or may not be seen. At a pathological level, consolidation is a sign of any process that replaces air in the alveoli. This pattern is a common sign of infection, but is also seen in lung infarction, malignancy (specifically, adenocarcinoma and lymphoma) and in organizing pneumonia.
MEDICINE 40:4
Reticular pattern or reticulation This term refers to the presence of innumerable interlacing lines on CXR and CT, which create the impression of a ‘net’. The appearance is often seen in conditions where there is thickening
180
Ó 2012 Published by Elsevier Ltd.
INVESTIGATIONS
Figure 5 Reticular pattern in idiopathic pulmonary fibrosis on (a) CXR and, (b) on targeted HRCT (in another patient). On the CXR there are multiple intersecting lines in the mid and lower zones. The targeted HRCT through the left lung shows a subpleural reticular pattern with microscopic honeycombing (arrows).
Figure 4 Ground-glass opacification on (a) CXR and (b) HRCT in an HIV-positive patient with Pneumocystis jiroveci pneumonia. On CXR there is increased density which obscures vessel markings in the mid/lower zones bilaterally. By contrast, on HRCT, vessels and walls of airways are still clearly visible.
Interlobular septal thickening The interlobular septa marginate individual pulmonary lobules and contain connective tissue, venules and lymphatics. Thickening of the interlobular septa is seen on CXR in the lower zones as thin linear opacities extending to and contacting the lateral pleural surface. Normal interlobular septa may be seen on HRCT images. Pathological conditions in which thickened interlobular septa are seen include pulmonary oedema and lymphangitis carcinomatosa.
of inter- and intralobular septa e a classic example of this is idiopathic pulmonary fibrosis (Figure 5). A net-like pattern caused by interlobular septal thickening may also be seen in lymphangitis carcinomatosa and in pulmonary oedema (due to venous congestion). A ‘delicate’ reticular pattern comprised of residual interlobular septa is seen in panacinar emphysema. Nodules and masses On CXR, a nodule is a rounded opacity that is well or poorly defined and has a diameter of up to 3 cm. On CT a nodule may be rounded or irregular, but again, well or poorly defined and measuring up to 3 cm. On CT, nodules may be further characterized based on size (a nodule of less than 3 mm diameter being termed a micronodule) and attenuation characteristics (solid, non-solid or part-solid). A mass refers to any lesion that measures more than 3 cm in diameter.
MEDICINE 40:4
Traction bronchiectasis/bronchiolectasis This term refers to the irregular dilatation of bronchi and bronchioles caused by surrounding lung fibrosis; on HRCT these may be seen as non-tapering airways or, depending on the orientation of the dilated airway relative to the imaging plane, variably-sized ‘cysts’. Dilated airways seen within regions of ground-glass opacification indicate that the latter represents fine fibrosis that is below the resolution limits of HRCT.
181
Ó 2012 Published by Elsevier Ltd.
INVESTIGATIONS
broad groupings: granulomatous (e.g. sarcoidosis), DPLDs of known cause (e.g. drug-induced, connective tissue diseaserelated), idiopathic interstitial pneumonias and a miscellaneous group. A CXR is one of the initial tests requested but most patients will, at some stage, require HRCT. In some patients, a histospecific diagnosis will be provided by HRCT, obviating surgical biopsy. However, HRCT interpretation is far from straightforward. The CT appearances may be atypical or nonspecific and observer experience influences the diagnostic performance of HRCT. This emphasizes the importance of the multidisciplinary approach. The imaging features of some DPLDs are discussed below.
Figure 6 Tree-in-bud pattern in a patient with rhinovirus and parainfluenza infection after haematopoietic stem cell transplantation. There are small nodules (thin arrows) and a branching peripheral opacity (arrowhead) in the right lower lobe.
Idiopathic pulmonary fibrosis (IPF)/usual interstitial pneumonia (UIP): idiopathic pulmonary fibrosis (IPF) is a chronic progressive fibrosing interstitial pneumonia, usually seen in older adults. IPF is associated with the features of a usual interstitial pneumonia (UIP) pattern on histopathological examination and/or HRCT. In the typical patient, the cardinal HRCT finding is a bilateral, roughly symmetrical subpleural and basal reticular pattern with honeycombing. There may be associated ground-glass opacification but this is generally less extensive than the reticular pattern. Tractional dilatation of segmental and subsegmental airways (traction bronchiectasis/bronchiolectasis) may be seen. The presence of these typical features allows experienced radiologists to make a confident and accurate HRCT diagnosis of a UIP pattern. Atypical CT features e sometimes overlapping with NSIP (see below) e are present in between
Tree-in-bud pattern Centrilobular branching structures denoting endo- and peribronchiolar pathology (e.g. mucoid impaction, inflammation and/or fibrosis). This pattern is often seen with endobronchial spread of mycobacterial infection and in patients with cystic fibrosis (Figure 6).
Imaging findings in specific lung diseases Diffuse parenchymal lung diseases The diffuse parenchymal lung diseases (DPLDs) are an important and challenging group of pulmonary disorders.12,13 The term encompasses a range of disorders which are categorized by four
Figure 7 CXR and HRCT in three patients with sarcoidosis. (a) Stage I sarcoidosis: symmetrical bilateral hilar and right paratracheal (arrowheads) lymph node enlargement. There is no overt lung infiltration. (b) Stage II disease: there is nodal enlargement and parenchymal disease. Note the thickening of the horizontal fissure (arrows). (c) HRCT through the upper lobes demonstrates multiple small bronchocentric nodules.
MEDICINE 40:4
182
Ó 2012 Published by Elsevier Ltd.
INVESTIGATIONS
one-third to one-half of patients and, in this group, biopsy confirmation may be needed.
bilateral peripheral consolidation, generally in the lower zones and changing over time, is the classical finding. However, organizing pneumonia may have varied CT appearances: a distinctive peri-lobular pattern or bronchocentric distribution, lesions with the reverse-halo or atoll pattern, a solitary ‘mass’ (simulating lung cancer), multiple small nodules and linear bands have all been reported.
Non-specific interstitial pneumonia: in NSIP, the HRCT appearances can be more variable. However, numerous studies have shown that ground-glass opacification is the key finding. This is typically bilateral and symmetrical and most pronounced in the lower zones. There may be some superimposed reticulation that may remain stable or increase over time. On serial CT examination, a pattern similar to UIP may develop in patients with NSIP although honeycombing tends to be less conspicuous.
Sarcoidosis: in sarcoidosis, the characteristic histopathological lesion is the non-caseating granuloma distributed along the lymphatic pathways; the axial (bronchovascular) and subpleural interstitium are typically affected. Most patients have an abnormality on CXR at some time during the course of their disease. Enlargement of intrathoracic lymph nodes is seen in
Organizing pneumonia: this is a pattern of interstitial pneumonia that predominantly affects the air spaces. On HRCT,
Figure 8 Lung cancer in a 65-year-old smoker on (a) CXR, (b) CT and (c) 18FDG- PET/CT. The CXR shows a well-defined rounded left upper lobe mass. On CT there are characteristics spicules radiating from the mass; extensive emphysema is present. The fused PET-CT image shows increased metabolic activity in the left upper lobe mass.
MEDICINE 40:4
183
Ó 2012 Published by Elsevier Ltd.
INVESTIGATIONS
up to 80% of subjects (Figure 7). The classical finding is of bilateral, symmetrical hilar lymph node enlargement with either right or bilateral paratracheal nodal disease; asymmetrical or unilateral hilar nodal enlargement is far less common. Calcification may also be seen in lymph nodes. Parenchymal disease e best characterized on HRCT e manifests as nodular infiltrate (reflecting conglomerate granulomata) surrounding the bronchovascular structures. Subpleural nodularity is also commonly seen. In later stages of the disease there may be obvious signs of established lung fibrosis with upper zone volume loss, parenchymal distortion and traction bronchiectasis.
airway-based diseases, with relative characteristic findings on CXR and CT, are presented below. Bronchiectasis: the CXR signs of bronchiectasis include prominent ring-shaped (indicating abnormally dilated airways that are seen ‘end-on’) and tramline opacities. There may be plugging of airways and signs of hyperexpansion because of airflow obstruction in generalized bronchiectasis or, indeed, volume loss. Making a confident diagnosis of bronchiectasis on CXR, particularly in mild disease, is challenging; CT is certainly more accurate.14e16 The most reliable morphological sign for diagnosing bronchiectasis is the absence of normal tapering of airways, a feature more readily identified from a volumetric thin-
Lung cancer In the West, lung cancer is the commonest cause of death due to cancer. Around 40,000 new cases are diagnosed each year in UK. The CXR is often the first test that alerts the clinician to the diagnosis but CT is recommended in management pathways. The direct signs of lung cancer include an intrapulmonary nodule or mass (Figure 8), often with classical spiculated or lobulated margins or, less commonly, the tumour may be well defined. There may be a tag extending towards the pleural surface from the tumour. Cavitation occurs in less than 20% of cases and calcification e usually indicating benignity e is seen in a small proportion. In some patients the primary tumour is not demonstrated. Instead, indirect signs (e.g. lobar/lung collapse, distal ‘obstructive pneumonia’) will be seen. For staging, CT and 18FDG-PET are of particular importance. CT provides information not only about the site and size of the tumour but also its density characteristics (i.e. central low attenuation may indicate necrosis). The presence or absence of additional nodules in the same lobe, ipsilateral or contralateral lung e all of which may have staging significance e will be shown on CT. Additional findings such as rib destruction, pleural effusion and signs indicating metastasis to extrathoracic sites (e.g. adrenal glands, liver) will be detected on CT. An area of difficulty in lung cancer is the diagnosis of nodal metastases. In this context, one of the key problems with CT evaluation is the reliance on size criteria: a short axis nodal diameter of greater than 1 cm has been traditional yardstick for abnormal enlargement of intrathoracic lymph nodes. However, size alone may be an inaccurate discriminator. There is a range of normality: short axis diameters may vary from 0.7 mm (for hilar nodes) up to 1.5 cm (in the lower paratracheal and subcarinal regions). Moreover, micrometastases (particularly from central adenocarcinomas) may cause no enlargement whereas non-malignant diseases are associated with enlarged nodes. In recent years, there has been a growing dependence on 18 FDG-PET, which frequently changes the preliminary staging of lung cancer based on CT findings. However, false-negative (tumours with low metabolic activity) and false-positive (caused by non-malignant but metabolically active disease) results are well recognized, meaning that 18FDG-PET cannot replace invasive nodal sampling for pre-operative staging. Figure 9 Bronchiectasis on (a) axial HRCT showing dilated airways in both upper lobes. There is associated mosaicism (best seen in the left upper lobe) caused by obliterative bronchiolitis. (b) Coronal image in another patient with localized bronchiectasis in the right lower lobe. Note the nontapering airways.
Airways diseases Radiological tests are commonly requested for the investigation of diseases affecting the central (trachea and main bronchi) down to the peripheral small airways. The radiological features of some
MEDICINE 40:4
184
Ó 2012 Published by Elsevier Ltd.
INVESTIGATIONS
Imaging tests are integral to the management of patients with suspected or established lung disease. In clinical practice, CXR and CT are the most frequently requested tests. CXR has the benefit of low radiation dose but interpretation is hampered by anatomical superimposition and low contrast resolution. On the other hand, CT provides greater morphological detail and accuracy but with the drawback of a higher radiation burden. A
6 von Schulthess GK, Steinert HC, Hany TF. Integrated PET/CT: current applications and future directions. Radiol 2006 Feb 1; 238: 405e22. 7 Ailion DC. Introduction to magnetic resonance imaging techniques used in the lung. In: Cutillo AG, ed. Application of magnetic resonance to the study of the lung. 1st edn. New York: Futura, 1996; 33e47. 8 Cutillo A, Ganesan K, Ailion DC, et al. Alveolar air-tissue interface and nuclear magnetic resonance behaviour of lung. J Appl Physiol 1991; 70: 2145e54. 9 Cutillo A, Goodrich KC, Ganesan K, et al. Alveolar air/tissue interface and nuclear magnetic resonance behaviour of normal and edematous lungs. Am J Respir Crit Care Med 1995; 151: 1018e26. 10 Mayo JR. Thoracic magnetic resonance imaging: physics and pulse sequences. J Thorac Imaging 1993; 8: 1e11. 11 Hansell DM, Bankier AA, MacMahon H, McLoud TC, Muller NL, Remy J. Fleischner society: glossary of terms for thoracic imaging. Radiol 2008 Mar 1; 246: 697e722. 12 Wells AU, Hirani N. Interstitial lung disease guideline. Thorax 2008; 63: v1e58. 13 Travis WD, King Jr TE, The Multidisciplinary Core Panel. American Thoracic Society/European Respiratory Society international multidisciplinary consensus classification of idiopathic interstitial pneumonias. Am J Respir Crit Care Med 2002; 165: 277e304. 14 Naidich DP, McCauley DI, Khouri NF, Stitik FP, Siegelman SS. Computed tomography of bronchiectasis. J Comput Assist Tomogr 1982; 6: 437e44. 15 Munro NC, Cooke JC, Currie DC, Strickland B, Cole PJ. Comparison of thin section computed tomography with bronchography for identifying bronchiectatic segments in patients with chronic sputum production. Thorax 1990; 45: 135e9. €ller NL. Bronchiectasis: comparison of preop16 Kang EY, Miller RR, Mu erative thin-section CT and pathological findings in resected specimens. Radiol 1995; 195: 649e54. 17 Hansell DM. Small airways diseases: detection and insights with computed tomography. Eur Respir J 2001; 17: 1294e313.
REFERENCES €ller NL. Chronic diffuse infil1 Mathieson JR, Mayo JR, Staples CA, Mu trative lung disease: comparison of diagnostic accuracy of CT and chest radiography. Radiol 1989; 171: 111e6. 2 Hounsfield GN. Computerized transverse axial scanning (tomography): Part 1. Description of system. Br J Radiol 1973; 46: 1016e22. €ller NL. High-resolution CT of the chest: 3 Mayo JR, Jackson SA, Mu radiation dose. Am J Roentgenol 1993; 160: 479e81. 4 Vock P, Soucek M, Daepp M, Kalender WA. Lung: spiral volumetric CT with single-breath-hold technique. Radiol 1990; 176: 864e7. 5 Kalender WA, Seissler W, Klotz E, Vock P. Spiral volumetric CT with single-breath-hold technique, continuous transport, and continuous scanner rotation. Radiol 1990; 176: 181e3.
FURTHER READING Goldstraw P, et al. The IASLC lung cancer staging project: proposals for the revision of the TNM stage groupings in the forthcoming (seventh) edition of the TNM classification of malignant tumours. J Thorac Oncol 2007; 2: 706e14. Hansell DM, Lynch DA, McAdams HP, Bankier AA, eds. Imaging of diseases of the chest. 5th edn. Mosby-Elsevier, 2010. Mayo JR, et al. Radiation dose at chest CT: a statement of the Fleischner Society. Radiol 2003; 228: 15e21. Remy-Jardin M, et al. Management of suspected pulmonary embolism in the era of CT angiography: a statement from the Fleischner society. Radiol 2007; 245: 315e26. €ller NL, Naidich DP, eds. HRCT of the lung. 4th edn. Wolters Webb WR, Mu Kluwer/Lippincott Williams & Wilkins, 2008.
section acquisition (as opposed to interspaced HRCT) (Figure 9). For airways running perpendicular to the axial CT imaging plane, a bronchus whose internal luminal diameter exceeds the dimensions of its accompanying pulmonary artery (the signet ring sign) is considered abnormally dilated. Bronchiectatic airways running in the plane of the section appear as ‘tramlines’, equivalent to those seen on CXR. Ancillary CT signs of bronchiectasis include mucus plugging, bronchial crowding, volume loss and a mosaic attenuation pattern. Obliterative bronchiolitis (OB): this pathological pattern (also known as bronchiolitis obliterans and constrictive obliterative bronchiolitis) is characterized by bronchiolar and peribronchial inflammation and eventual fibrous obliteration.17 OB is seen in clinical scenarios as diverse as post-viral infection, following solid organ or haematopoietic stem cell transplantation, and in patients €gren’s syndrome. The CXR signs of with rheumatoid arthritis or Sjo OB (hyperinflation and paucity of vessels) are non-specific. On CT, the cardinal sign is the mosaic attenuation pattern. In regions of decreased attenuation there is a reduction in the number and/or calibre of pulmonary vessels. In some patients, these patchy areas of black and grey lung are almost imperceptible and images obtained at end-expiration may highlight the density differences. Mosaicism is not specific for OB; similar appearances are seen in certain infiltrative diseases and chronic thromboembolism.
Conclusions
MEDICINE 40:4
185
Ó 2012 Published by Elsevier Ltd.