- Email: [email protected]
Multimodality imaging of pulmonary infarction
European Journal of Radiology 83 (2014) 2240–2254
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Review
Multimodality imaging of pulmonary infarction T.J.P. Bray a,1 , K.H. Mortensen a,b,∗ , D. Gopalan a,1 a
Department of Radiology, Papworth Hospital NHS Foundation Trust, Ermine Street, Papworth Everard, Cambridge CB23 3RE, United Kingdom University Department of Radiology, Addenbrookes Hospital, Cambridge University Hospitals NHS Foundation Trust, Hills Road, Box 318, Cambridge CB2 0QQ, United Kingdom b
a r t i c l e
i n f o
Article history: Received 20 March 2014 Received in revised form 16 June 2014 Accepted 20 July 2014 Keywords: Pulmonary infarction Arterial infarction Venous infarction Pulmonary embolus Thromboembolic disease Septic infarction In situ thrombus Fat embolism Vasculitic infarct Pulmonary torsion
a b s t r a c t The impact of absent pulmonary arterial and venous flow on the pulmonary parenchyma depends on a host of factors. These include location of the occlusive insult, the speed at which the occlusion develops and the ability of the normal dual arterial supply to compensate through increased bronchial arterial flow. Pulmonary infarction occurs when oxygenation is cut off secondary to sudden occlusion with lack of recruitment of the dual supply arterial system. Thromboembolic disease is the commonest cause of such an insult but a whole range of disease processes intrinsic and extrinsic to the pulmonary arterial and venous lumen may also result in infarcts. Recognition of the presence of infarction can be challenging as imaging manifestations often differ from the classically described wedge shaped defect and a number of weighty causes need consideration. This review highlights aetiologies and imaging appearances of pulmonary infarction, utilising cases to illustrate the essential role of a multimodality imaging approach in order to arrive at the appropriate diagnosis. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Pulmonary infarction is due to coagulative ischaemic necrosis [1]. This process occurs when parenchymal perfusion fails to meet the tissue requirements due to occlusive insults that are either intrinsic or extrinsic to the vascular lumen. Normally, dual arterial supply comprised of deoxygenated blood in the pulmonary arteries and oxygenated blood in the highly adaptable bronchial arteries protects against infarcts. The risk of ischaemic necrosis is especially increased in perturbations of blood flow that dually impede this system, be it congestive heart failure, pulmonary venous hypertension, or diminished bronchial arterial flow in systemic arterial hypotension [2]. Less commonly infarction may result from perturbations in pulmonary venous flow [3,4]. A ‘complete’ infarct is characterised by irreversible ischaemic injury, wherein incident capillary ischaemia increases vascular permeability with subsequent bronchial artery reperfusion leading to alveolar haemorrhage
∗ Corresponding author at: Department of Radiology, Papworth Hospital NHS Foundation Trust, Ermine Street, Papworth Everard, Cambridge CB23 3RE, United Kingdom. Tel.: +44 1480364562. E-mail addresses: [email protected] (T.J.P. Bray), [email protected] (K.H. Mortensen), [email protected] (D. Gopalan). 1 Tel.: +44 1480364562. http://dx.doi.org/10.1016/j.ejrad.2014.07.016 0720-048X/© 2014 Elsevier Ireland Ltd. All rights reserved.
and necrosis ensuing hereafter. The end-result is a fibrotic scar. An infarct presents macroscopically as a dark necrotic area with a narrow rim of hyperaemia and inflammation, presenting microscopically as a ‘ghost-like’ architecture with an outer perimeter of cellular infiltrate (Fig. 1) [1,5]. The infarct can also be ‘incomplete’ when there is transient haemorrhagic congestion but no other acute or chronic parenchymal sequelae – the parenchymal opacity resolves in a matter of days [6,7]. By virtue of the pathophysiological mechanism, arterial infarcts are peripheral, occurring in areas supplied by medium and small sized arteries [8]. The acute infarct is a wedge-shaped (less frequently nodular) pleurally based parenchymal opacity (‘Hampton hump’) that forms an obtuse angle with the visceral pleura, with the apex pointing towards the pulmonary hilum and centred on a bronchovascular bundle, whereas a pulmonary venous infarction is paraseptal in distribution (Fig. 2) [6,9,10]. Less typical features of arterial and venous infarcts are increasingly reported with a move to raised suspicion of infarction with a whole array of imaging appearances [11]. Infarcts may be solitary or multiple, temporally uniform or variegated much depending on the inciting event, and characteristically the airspace opacity should diminish in size over time (‘melting sign’) with complete resolution or leaving residual plate-like scarring and focal pleural thickening. Timely diagnosis of pulmonary infarction is important in order to relieve the inciting event and avert further respiratory and
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
vascular compromise. This pictorial review aims to demonstrate imaging features of pulmonary infarction using a multimodality approach in order to illustrate the spectrum of aetiologies that may result in infarction. Although computed tomography (CT) currently forms the mainstay of diagnosis the use of ancillary imaging modalities is often necessary. 2. Imaging modalities
Table 1 Salient imaging findings for pulmonary arterial infarcts by modality. Radiography
Usually non-specific parenchymal opacity Wedge-shaped pleurally based opacity: Hampton hump Pleural effusion (small, unilateral), elevated hemidiaphragm (volume loss), atelectasis
Computed tomography
Hampton hump Internal lucency in peripheral consolidation: bubbly consolidation Thickened vessels leading to apex Pleural effusion (localised or diffuse on Ipsilateral side) Cavitation (<10%)
2.1. Computed tomography The wedge-shaped, broad pleurally based ‘Hampton hump’ frequently has a truncated apex and a convex border with less common findings including internal air lucency, linear stranding from apex to hilum and thickened vessels leading to the apex of the opacity (Table 1) [12]. Contrast-enhanced CT pulmonary angiography (CTPA) may demonstrate decreased enhancement within consolidated lung (Fig. 3); this is present in 95% of infarcts but can also be seen in collapse, pneumonic consolidation and neoplasm [12,13]. Central lucency within peripheral consolidation is particularly indicative of pulmonary infarction (‘bubbly consolidation’) (Fig. 3) [14] but other causes for these appearances include cavitation, cysts and dilated airways [15]. Although cavitation within infarcts is often due to bland necrosis, the presence of superinfection of established bland infarction or primary septic emboli can give a similar appearance [16,17]. CT is particularly helpful for the detection of complications of infarcts, including potentially detrimental sterile or purulent involvement of the mediastinum, pleural space, remaining parenchyma and airways, or the thoracic cage [18–20]. Dual Energy CT lends promise for the visualisation of pulmonary infarcts and vascular occlusions [21]. Contrast-enhanced data sets are acquired at two energy spectra and reconstructed to form a conventional CTPA plus a static iodine map, the latter representing a surrogate measure of microvascular circulation and perfusion (Fig. 4) [22,23]. The functional iodine map increases diagnostic accuracy by detecting more peripheral occlusions than conventional CTPA [24]. In cases of pulmonary arterial occlusion without infarction the iodine map depicts relative oligaemia due to persistent perfusion via the bronchial arteries; infarction on the other hand is characterised by complete segmental absence of iodine due to absent perfusion and a corresponding morphological abnormality [25]. The segmental distribution aids in the differentiation of infarction from other opacities with low iodine content such as tumour, pneumonia, or even streak artefact. Dynamic Perfusion CT is another technique with potential future applications in the
2241
Dual energy computed tomography
Contrast enhanced: hypoenhancing centre Adjacent or ipsilateral pleural effusion Perfusion defect on iodine map with correlating parenchymal opacity
Ultrasound
Wedge-shaped, pleurally based area of reduced echogenicity Hyperechoic centre (reverberation artefact) Late contrast-enhancement
Magnetic resonance imaging
Hyperacute: hyperintense T2WI and hypointense T1WI Acute (up to 1 week): hyperintense T1WI Hyperintense signal on both T1WI and T2 WI compared with tumour Perfusion defect: visualised with static and/or dynamic contrast enhanced MRI
Ventilation/perfusion scintigraphy
Matched ventilation and perfusion defect with radiographic parenchymal opacity – ‘Triple-match’
Single positron emission tomography
Consolidation in peripheral interface between severely decreased and relatively preserved perfusion areas
Positron emission tomography
FDG-avid rim along the infarct periphery Relative central photopenia Alternative appearance: diffuse uptake
assessment of infarction but its use has hitherto remained limited due to radiation concerns [26]. At present CTPA remains the principal diagnostic tool with the dual energy technique potentially providing a future one-stop morphological and functional test. 2.2. Magnetic resonance imaging Imaging appearances of infarcts on magnetic resonance imaging (MRI) vary according to signal characteristics of the aging blood that has accumulated within the alveoli [27]. Acute infarcts
Fig. 1. Haematoxylin and eosin stain images of two different arterial infarcts. Left image (original magnification 50×) is an organised pulmonary arterial infarction (white star) with a chronic inflammatory infiltrate in the margin (white arrow). Right image (original magnification 16×) shows a peripheral infarct (white star) with pleural thickening (black arrow).
2242
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 2. CT pulmonary angiography on mediastinal (left image) and lung (middle image) windows demonstrates a wedge shaped pleurally based parenchymal opacity (‘Hampton hump’, arrow) with the apex pointing towards the pulmonary hilum and centred on a bronchovascular bundle. Far right image is a microscopic specimen showing a wedge shaped pulmonary arterial infarct (star) from a different patient (haematoxylin and eosin stain, original magnification 160×).
Fig. 3. Coronal (left image) and axial (right image) CT images on mediastinal and lung windows, respectively, depict a large left pulmonary embolus (thin arrow) and a peripheral infarct (block arrow) with bubbly consolidation, linear ‘stranding’ from apex to hilum and lack of parenchymal enhancement.
(<24 h) are T1 hypointense and T2 hyperintense with subacute infarction being T1 hyperintense (>24 h to 1 week) [28]. The subsequent temporal imaging characteristics of infarcts have not been assessed. Infarcts secondary to thromboembolic disease on
average measure 4 cm (ranging from 1 to 12 cm) whilst septic infarcts tend to be smaller but still detectable with the current spatial resolution of MRI [12,29,30,31]. Appearances with techniques such as dynamic contrast enhanced and diffusion weighted
Fig. 4. Dual energy CT on coronal reformats demonstrate a filling defect (arrows) in the left lower lobe segmental pulmonary artery (left image) and a corresponding perfusion defect on the iodine map (right image). Also note the bilateral pleural effusions and dependent atelectasis.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
MRI remain to be established but these techniques may help diagnosing alternative causes of parenchymal opacity in the setting of suspected pulmonary infarction [32]. The current role of MRI in imaging of infarcts is that of an ancillary, problem-solving tool, uniquely combining comprehensive anatomical and functional assessment of the thoracic vasculature with excellent soft tissue characterisation without the use of radiation [32,33]. In certain situations where CTPA is not available secondary to concerns regarding the use of radiation or iodinated contrast media, MRI may however have a primary role. 2.3. Ultrasound Infarcts are conspicuous on high-frequency ultrasound (typically 3.5–8 mHz) as pleurally based hypoechoic, triangular or infrequently round lesions [34,35]. Typically, infarcts are sharply demarcated. Early infarcts are often homogenously hypoechoic compared with the centrally hyperechoic older infarction [36]. When present, internal reverberation artefact represents a bronchiole and is characteristic of segmental involvement. Congested enlarged vessels leading to the apex of the lesion may also be evident [36]. Contrast-enhanced ultrasound displays increased time-to-enhancement when compared with infective consolidation, although this is not very specific [37]. The main role for the detection of infarction by thoracic ultrasound lies in facilitating the diagnosis of acute pulmonary embolus with bedside imaging, requiring detection of two or more infarcts [38]. Ultrasound predominantly has a role in the acute setting in the emergency room when alternative imaging modalities are not feasible. 2.4. Radiography The stereotypical appearance of pulmonary infarction as a peripheral wedge shaped pleurally based opacity was initially described on chest radiography, the ‘Hampton hump’ [6,7]. Temporally, approximately half of pulmonary infarcts leave residual radiographic changes that include linear scars, pleural diaphragmatic adhesions, and localised pleural thickening and retraction [39,40]. These residual findings are typically mild compared to the initial abnormality with a concomitant reduction in the initially detected perfusion abnormality on perfusion scintigraphy [39]. The presence of infarction on radiographs may increase awareness especially of pulmonary embolism in the acute setting. However, chest radiography may maintain a role as a general screening tool in cardiorespiratory symptoms but this modality adds little value when resolving the aetiology on an infarct.
2243
surrounding inflammatory reaction [46–49]. Unfortunately, the rim sign is not specific with many infarcts displaying homogenously increased tracer uptake [49,50].
3. Intravascular aetiology 3.1. Thromboembolic infarction Pulmonary thromboembolism from peripheral deep venous thrombosis is the most common cause of infarction [51], occurring in as many as a third of acute pulmonary emboli [2,52]. At autopsy, infarcts are more likely to occur and be larger in size with a higher embolic burden but recent CT studies have not shown similar correlation between clot burden and pleurally based opacities [53,54]. Infarction has a predilection for the lower zones due to a relatively increased vascularity. Because infarcts typically extend to the pleural surface, pleural effusion is more common in pulmonary embolism with than without infarction (29% versus 4%) [13]. Infarcts are the only more common pleuroparenchymal CT feature in patients with confirmed embolus in cohorts with suspected pulmonary thromboembolism [53,55]. Infarcts in chronic thromboembolic pulmonary disease can be particularly difficult to differentiate from malignancy and infection. There will, however, often be temporal variegation in the imaging appearances of infarcts due to the protracted nature of the disease, ranging from acute wedge-shaped opacities to plate-like thin fibrous scars (parenchymal bands) with focal areas of architectural distortion and volume loss (Fig. 5). Cavitation occurs in 11% of infarcts in chronic thromboembolic disease compared with 5% in acute thromboemboli [16,56]. Typically the cavity is solitary, arising from days 2 to 63 following the ischaemic insult and often on the backdrop of bland infarction that may become complicated by infection [57,58]. Cavitation is more common within mid and upper zones with the cavity wall generally being a poor discriminator of the underlying aetiology (Fig. 6) [16,57,58]. Superinfection with clostridium species has been reported to be especially common, resulting in a necrotising, cavitatory pneumonia [59]. Congenital heart disease can also predispose to infarct formation following in situ thrombosis and subsequent embolisation (Fig. 7). Other causes of sterile, intra-cardiac migration thrombus as a cause of infarction are coagulopathy, tricuspid valve stenosis, intracardiac tumour (surface thrombus) or myocardial infarction [60].
3.2. Malignant infarction 2.5. Nuclear medicine Acute infarcts appear on ventilation perfusion scintigraphy (V/Q scintigraphy) as matched perfusion and ventilation defects with corresponding radiographic opacities – the ‘triple match sign’ [41]. Typically this is seen in a mid-to-upper zone location, with completely absent perfusion and involvement of 25–50% of a lung zone [41–43]. Single positron emission CT combined with unenhanced CT can differentiate infarcts from infective consolidation; parenchymal opacity at the peripheral lung interface between severely decreased and adjacent relatively preserved perfusion areas within relatively large and severely decreased perfusion defects favours infarction whilst infective consolidation is located in the proximal aspect of often smaller perfusion defects [44,45]. Positron emission tomography CT (FDG-PET CT) imaging of infarcts has identified the ‘rim sign’, demonstrating relative central photopenia with a rim of high tracer uptake along the border of the parenchymal opacity, corresponding to central necrosis and
Malignancy associated infarcts mainly occur with haematogenous dissemination of larger tumour emboli that instantaneously occlude the pulmonary vascular lumen, creating an opacity that arises at a speed disjointed from the tumour doubling time (Fig. 8). The appearance is dissimilar from classical haematogenous metastasis where microemboli lodge within smaller peripheral pulmonary arteries and give rise to peribronchovascular nodules that enlarge according to the expected growth rate. Some tumours such as hepatocellular and renal cell carcinoma are especially prone to producing larger intravascular tumour fragments with a potential to cause infarction [60]. When a parenchymal opacity suddenly arises in the setting known or suspected malignancy the differential must therefore alongside infarction include disease progression and pneumonitis [61,62]. This distinction is important since erroneous interpretation may cause faulty tumour burden upstaging and histology may be required for confirmation when imaging is equivocal [49].
2244
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 5. Axial CT in 3 different patients showing the temporal variegation of infarcts on lung windows. Left image is an acute wedge-shaped infarct (block arrow) and plate atelectasis (thin arrow). Middle image is a chronic infarct with focal architectural distortion and volume loss (arrowhead). Right image is a right lower lobe thick walled cavitating infarct (black arrow) with background mosaic attenuation due to chronic thromboembolic disease.
Infarction due to local invasion or compression by tumour growth is less frequent because the pulmonary vasculature will have time to adapt with increments in bronchial artery collateral supply when flow in an invaded pulmonary artery is restricted [63,64]. Due to the contiguous nature of the infarction and adjacent tumour MRI can be used for differentiation with infarcts, being T1 and T2 hyperintense relative to tumour [62]. Of note, infarcts are generally non-enhancing although necrotic tumour may show similar features [65]. Primary pulmonary artery sarcoma is a rare cause of infarction that needs to be kept in mind when assessing for malignant infarction, wherein interval growth may be misdiagnosed as thromboembolic disease (Fig. 9) [66].
3.3. Septic infarction Embolic infarction secondary to septic clots in the pulmonary arteries is particularly frequent in tricuspid valve and pacemaker wire endocarditis, indwelling central venous catheters, septic thrombophlebitis, infected arteriovenous shunts, and infections of bone, skin or soft tissues (Fig. 10) [17,65,67]. Staphylococci are the most common pathogen [68]. Temporally separated embolic showers can result in diffuse bilateral opacities, ranging from solid nodules to cavitation [69,70]. Even though CT features such as larger nodule size, cavitation and infarction may suggest a gram-positive organism, no imaging feature is specific to a pathogen [71]. In the
Fig. 6. Axial CT (top left panel) shows a rounded well-defined peripheral soft tissue opacity containing central lucency (block arrow) in the right upper lobe. Differential diagnoses include malignancy, infection and infarction. The presence of mosaic attenuation (thin arrow) suggests underlying chronic thromboembolic disease. Subsequent PET-CT (remaining images) demonstrates the lung lesion (block arrow; bottom left panel) to be ‘cold’, making infarction more likely than malignancy.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
2245
Fig. 7. A 50-year-old male with known atrial septal defect and haemoptysis. (a) Chest radiograph shows cardiomegaly and enlarged proximal pulmonary arteries (white thin arrow) with an ill-defined opacity in the left apex (white block arrow). Also note the presence of a ‘Reveal’ device (black arrow). (b) Axial contrast enhanced CT demonstrates an enlarged main pulmonary artery (star) with large load of bilateral proximal thrombus (notched arrows; left panel), a sinus venosus type atrial septal defect (thin arrow; middle panel) and a mycetoma complicating a long-standing pulmonary infarct cavity (block arrow; right panel).
setting of known septic embolus, classical wedge shaped opacity is seen in 17% ranking in prevalence after multiple bilateral nodules, cavitation, non-specific infiltrates and pleural effusions with the “feeding vessel sign” often present in relation to both nodules and opacities [17,70]. Ancillary features from complications such as cavitation and empyema may support sepsis as a cause of infarction (Fig. 11). With progressive disease, appearances of the primary infected focus such as endocarditic vegetation, osteomyelitis, or indwelling catheters with adherent soft tissue may be dwarfed by manifestations from metastatic septic deposits (Fig. 12).
Imaging findings overlap considerably between septic and sterile infarcts, though cavitation is more common in septic embolic disease where the parenchymal opacity is, on balance, smaller [12,30,71]. The cavity may rapidly enlarge, causing rupture of the surrounding wall with extension of the cavity and its contents into the pleural cavity leading to pyopneumothorax; the time to cavitation is longer in aseptic compared with septic infarcts (weeks rather than days) with a risk of complications such as pneumothorax and bronchopleural fistula being present with both aetiologies [58,72,73]. The presence of a smooth outline, increasing size or
Fig. 8. Contrast enhanced CT in a 57-year-old male (left panel) demonstrates a right hilar bronchogenic carcinoma with occlusion of the upper lobe artery (thin arrow) and vein. The spiculated mass in the right apex (notched arrow; top right image) was much larger on a previous CT from 3 weeks before (bottom right image), making an infarction more likely than metastases.
2246
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 9. Pulmonary artery sarcoma and infarcts. Axial contrast enhanced CT shows an expansile soft tissue mass in the main and right pulmonary arteries (white star), destruction of vascular integrity in the posterior aspect and resultant mediastinal involvement (left panel). The lung windows (right panel) show an ill-defined pulmonary infarction in the right lower lobe (block arrow).
internal fluid levels all favour primary septic over bland infarction. Absent perfusion within areas of non-cavitating consolidation and a pulmonary arterial filling defect proximal to the opacity are also helpful in distinguishing cavitating infarction from primary necrotising pneumonia [74]. Of note, the risk of superinfection of susceptible necrotic parenchyma is not limited to bacterial pathogens; patients with established cavitatory infarcts also face an increased risk of fungal or mycobacterial infection (Fig. 7). Septic infarction secondary to invasive pulmonary aspergillosis differs in pathogenesis from intravascular dissemination of infected embolus. Most commonly caused by Aspergillus fumigatus this potentially life-threatening infection is seen in severely immunocompromised individuals, often with severe neutropenia. Fungus
invades medium-sized pulmonary arteries with fungal hyphae obstructing blood flow and causing haemorrhagic infarction [75]. CT features include solid nodules surrounded by haemorrhagic ground glass attenuation (‘halo sign’) or consolidation in a patchy, segmental/subsegmental or peribronchiolar pattern (Fig. 13) [76–79]. The areas of consolidation may be pleurally based in some instances [80]. During the convalescence stage, cavitation occurs in more than half of patients [79]. This is evident as an air crescent surrounding the nodular lesion (‘air crescent sign’) and/or cavitation within areas of confluent consolidation. Less severely immunocompromised patients can develop a more indolent and milder form semi-invasive aspergillosis [68]. CT signs in the early stages of this variant are typically upper lobar presenting in some
Fig. 10. A 22-year-old male intravenous drug abuser presented with dyspnoea, chest pain and fever. (a) Axial contrast enhanced CT (top left panel) shows a cavitating soft tissue lesion in the left lower lobe (white block arrow). Transoesophageal (top middle panel) and 3-dimensionsional echocardiography (top right panel) demonstrate a large vegetation (thin arrow) of the tricuspid valve. Retrospective analysis of the non-gated CT after the echocardiography confirmed the presence of an ill-defined low attenuation lesion intimately related to the tricuspid valve (black block arrow). (b) MRI of the spine performed for back pain 2 weeks after the initial diagnosis of infective endocarditis shows a large ill-defined area of consolidation in the right lower lobe (thin arrow) on the coronal localiser image (bottom left panel). Subsequent contrast enhanced CT (bottom middle and left panels) in axial and coronal reformats demonstrate interim reduction in the size of the left lower lobe infarction whilst a new large partially cavitating infarct has developed in the right lower lobe.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
2247
Fig. 11. A 65-year-old male developed fever, chest pain, dyspnoea and weight loss 2 weeks after right knee arthroscopy. Contrast enhanced CT demonstrates lobar pulmonary emboli (arrowheads; left panel) in a large feeding vessel at the apex of a right lower lobe infarct (block arrow; left panel) with an enhancing right pleural collection also noted (thin white arrow; left panel). Careful analysis of the valves showed a large vegetation in the pulmonary valve (notched arrows; right panel).
instances as bilateral segmental consolidation or nodules larger than 1 cm, often with pleural thickening alongside surrounding architectural distortion and volume loss from fibrotic reaction [81–83]. 3.4. Bland infarction Fat embolism can cause microvascular pulmonary infarction; adipose microglobules embolise the pulmonary capillaries together with erythrocytes and thrombocytes, causing diffuse peripheral microinfarction compounded by increased vascular permeability secondary to the toxic effects of fatty acids and inflammatory
cells [84]. Bone trauma is the commonest culprit [85], with a time lag of 12–24 h from symptoms to onset of heterogeneous diffuse airspace opacity that occurs in the absence of features of pulmonary venous congestion (Fig. 14). Imaging findings are often indistinguishable from acute respiratory distress syndrome [85]. Fat embolism can also complicate pancreatitis, hepatitis or other conditions where there may be intravascular fat mobilisation [85]. Other less frequent causes of pulmonary embolism, and even more so infarction, include air, amniotic fluid, talc and heavy metal (in intravenous drug users), hydatid, cement (in vertebroplasty) and other materials that may embolise to the pulmonary circulation [85,86].
Fig. 12. Temporal evolution of cavitating infarcts in a 45-year-old male smoker with history of intravenous drug abuse. Top deck: left panel shows a thick walled cavity in right upper lobe (block arrow) and a tiny cavity abutting the oblique fissure in left lung (thin arrow). There is a large thrombus (black block arrow) in the dilated right ventricle (right panel). Note the bilateral pleural effusions (white star). Bottom deck: following 6 weeks of antibiotic treatment and anticoagulation, the cavities have changed in size and morphology. The right ventricular thrombus persists but has reduced in size.
2248
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 13. A 35-year-old immunosuppressed female with a focal consolidation (block arrow; left image) surrounded by ground glass change (‘halo sign’) in the left upper lobe as evident on axial CT. Tracheal aspirate was positive for Aspergillus fumigatus. Image on the right is a few weeks following antifungal treatment and demonstrates a pleurally based opacity with mild architectural distortion (notched block arrow).
3.5. In situ occlusive infarction 3.5.1. Sickle cell disease The ‘acute chest syndrome’ in erythrocyte sickling disease is defined by a triad of new pulmonary infiltrate, fever and respiratory symptoms, caused by pulmonary infection and infarcts that may be present in isolation or work concert to cause respiratory compromise [87]. Pulmonary arterial occlusions in sickle cell disease are the result of in situ erythrocyte sickling, embolus from fat and necrotic bone marrow in osteonecrosis, or thromboembolic occlusion occurring as a part of a generally procoagulant status [88]. A stereotypical imaging feature is subsegmental atelectasis with or without consolidation, with ancillary features of sickle cell disease often present such as osteonecrosis with vertebral deformity, extramedullary haematopoiesis, or cardiomegaly with or without pulmonary arterial dilatation (Fig. 15). Due to the chronicity of the disorder, infarcts of varying ages may be encountered. Despite a propensity for multi-organ involvement secondary to systemic sickling of erythrocytes, imaging appearances can however be subtle. It is important to detect these features as treatments vary according to the source of occlusion and the ‘acute chest syndrome’ remains a major cause of death in this cohort [87].
Fig. 14. Coronal reformat of contrast enhanced CT demonstrates bilateral heterogeneous diffuse airspace opacity due to fat embolism in a 48-year-old male following bilateral leg amputation as a consequence of a road traffic accident. Also note the normal sized cardiac silhouette and extracorporeal membrane oxygenation cannula (thin arrows).
3.5.2. Vasculitis Primary idiopathic large vessel vasculitis (Takayasu arteritis, giant cell arteritis) is the most common cause of pulmonary involvement in vasculitic disease followed by small-vessel vasculitidis (e.g., granulomatous pulmonary polyangitis) [89]. Pulmonary involvement is less frequent in vasculitis of the medium sized vessels (Polyarteritis nodosa, Kawasaki disease), primary immune-complex vasculitis (e.g., Goodpasture syndrome), and secondary vasculitides (e.g., rheumatoid arthritis). Radiological manifestations vary, and include arterial wall thickening, nodular or cavitatory lesions, ground-glass opacities and consolidation (Fig. 16) [90]. Consolidation tends to be peripheral, pleurally based and of a configuration that varies from classical wedge shape to more poorly defined nodularity, which is the result of in situ occlusion with infarction, haemorrhage or atelectasis. There may be accompanying ground glass changes, representing alveolar haemorrhage, and in the longer term, repeated haemorrhage and disease progression can result in interstitial fibrosis. There will often be ancillary features to suggest the diagnosis of a specific form of vasculitis such as mural thickening of the thoracic aorta in Takayasu arteritis, sinonasal involvement of ANCA-positive granulomatous polyangitis, or renal atrophy following glomerulonephritis in Goodpasture syndrome.
Fig. 15. Chest radiograph in a 27-year-old female shows a wedge shaped peripheral infarct (notched arrow) in the left lower lobe. Also note the ‘H’ shaped vertebral bodies consistent with osteonecrosis (arrow), a classical feature of sickle cell disease.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
2249
Fig. 16. A 25-year-old female with pulmonary vasculitis secondary to Takayasu arteritis. (a) Top deck: there is concentric thickening of the left pulmonary artery and an occluded right pulmonary artery on CT (arrowheads; left image) and MRI (block arrow; right image). Bottom deck: there are multiple infarcts in the right lung; a thick walled cavity in right upper lobe (notched arrow; left image), small thin walled cavity (white arrow; right image) and a ground glass opacity (black arrow; right image) in right lower lobe. (b) FDG PET-CT shows intense uptake in the thick walled cavity (notched arrows; top deck), very low-level tracer uptake in the ground glass opacity (block arrows; bottom deck) and no uptake in the small thin walled cavity (thin arrows; bottom deck). Subsequent biopsy showed the upper lobe cavity to be infected with Mycobacterium tuberculosis.
3.6. Miscalaneous causes 3.6.1. Iatrogenic injury In rare instances iatrogenic injury to the vessel lumen causes infarction, as is seen in radiofrequency ablation of the left atrium complicated by pulmonary venous stenosis or in situ thrombosis
(Fig. 17) [4,91,92]. Depending of the location of the occlusive disease, the venous infarct affects an entire lobe or lung as opposed to the typically multiple scattered peripheral arterial infarcts [9]. The venous infarction is furthermore paraseptal in distribution, and there will be imaging findings of pulmonary venous congestion such as smooth interlobular septal thickening and alveolar oedema
2250
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 17. Pulmonary infarction secondary to vascular anastomotic injury following single lung transplant for interstitial fibrosis. (a) Contrast enhanced axial CT demonstrates stenosis of pulmonary arterial anastomosis (block arrow; far left image) and narrowing of pulmonary venous anastomosis (block arrow, middle image) with proximal consolidation at the left hilum (thin arrow; middle image). Multifocal peripheral ground glass opacities in the left (block arrow, left image) represent early infarcts. (b) Follow up CT performed 3 weeks after the previous study shows evolution of the left upper lobe infarct (thin arrow; left image) and loss of volume in left lower lobe due to the development of fibrotic scarring (block arrow, right image).
within areas affected by the venous occlusion [9]. Venous infarcts related to in situ thrombosis as may also be seen with trauma, infection and malignancy [3,63,93] should be separated from those secondary to fibrosing mediastinitis, congenital pulmonary venous stenosis or left atrial pathology such as myxoma or clot [10]. Pulmonary veno-occlusive disease is another distinct entity in which fibrosis of the pulmonary veins causes similar paraseptal venous infarcts that co-exist with imaging signs of pulmonary arterial hypertension and a normal sized left atrium [94]. Iatrogenic causes of arterial infarction include mal-positioning of Swann-Ganz catheters, either from wedging of the catheter or due to clot formation on the catheter surface [95,96]. Infarcts have also been reported as a rare complication of transarterial chemoembolisation of hepatocellular carcinoma; in one case the tumour’s collateral blood supply communicated directly with a major pulmonary artery, allowing embolic material to enter the pulmonary circulation [97]. 3.6.2. Lung transplantation and surgery Anastomotic stenosis as a cause of parenchymal infarction is a significant issue following lung transplantation, partly because of an absent collateral bronchial artery circulation [98]. This may occur not just secondary to a thrombotic event but also due to progressive anastomotic narrowing due to fibrosis, with the latter seen in as many as 4% of lung transplants (Fig. 18) [99]. Pulmonary venous infarction is a further, life-threatening complication following lung transplantation, producing segmental and/or lobar opacities on chest radiography and a dramatic reduction of perfusion on scintigraphy [100]. Pulmonary infarction is also a known complication of pneumonectomy [101].
4. Extraluminal aetiology 4.1. Torsion Pulmonary torsion is rare post-surgical complication that has primarily been described after pulmonary lobectomy (in up to 0.4%) [51,102], but this severe complication should be remembered as an unusual cause of pulmonary infarction, not just following lobectomy but also after other thoracic surgery, trauma or even as a spontaneous event [103–105]. Rotation of the bronchovascular pedicle results in dramatic vascular compromise with haemorrhagic lobar infarction due to a simultaneous pulmonary arterial and venous insult. Often, a collapsed or consolidated lobe is seen in an unusual position; other imaging features include ground glass attenuation, interlobular septal thickening, and poor pulmonary vascular enhancement (Fig. 19) [105,106]. 4.2. Compression The pulmonary arteries can also be compressed by nearby structures, including aortic aneurysms (Fig. 20) and soft tissue masses such as lymphoma [107]. Neoplasm may similarly occlude pulmonary arteries (as described above) [108]. In these cases it can be difficult to differentiate areas of infarction from metastases, though serial imaging and MRI may be helpful to distinguish. Pulmonary venous infarction due to fibrosing mediastinitis is another recognised cause of extrinsic compression, albeit rare [10]. Overall, compressive aetiology of infarcts is not common due to rapidity of the compression often allowing time for the bronchial artery collateral supply to adapt.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
2251
Fig. 18. 42-year-old female with progressive dyspnoea 3 months after pulmonary vein ablation for paroxysmal atrial fibrillation. (a) A chest radiograph shows marked asymmetry in the proximal pulmonary vasculature (block arrow) and an ill-defined peripheral consolidation in the left lower zone (thin arrow). (b) Mediastinal (top images) and lung (bottom images) windows from a contrast enhanced CT examination. There are homogenous filing defects in the left superior and inferior pulmonary veins (block arrows) with peripheral infarcts (thin arrows). Also note the thin rim of pleural fluid on the left and good opacification of right superior pulmonary vein (notched arrows, upper panels). (c) Ventilation perfusion scintigraphy demonstrates complete absence of perfusion in the left lung. (d) Catheter pulmonary venography illustrates a proximal thrombus in the stenosed left superior pulmonary vein (block arrow; left image) and establishment of flow after removal of the thrombus and placement of a stent (notched arrow, right image).
2252
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 19. Right upper lobe torsion in a 57-year-old woman who had undergone right middle lobectomy for bronchioloalveolar carcinoma. The post-surgical rotation of the bronchovascular pedicle resulted in vascular compromise and haemorrhagic lobar infarction. (a) 1 day postoperative chest radiograph (left image) shows satisfactory appearances with good expansion of the remaining lung on the right. Chest radiograph 4 days later (right image) demonstrates rapidly developing consolidation with a bulging neo-fissure (arrow), indicating volume increase within the right upper lobe. (b) Contrast enhanced CT images (top panel: axial images on lung and mediastinal windows; bottom panel: coronal reformats using maximum intensity projection and volume rendered technique). There is consolidation with increased volume (thin white arrow points to bulging fissure), ground-glass attenuation, interlobular septal thickening (notched black arrow), and poor parenchymal and pulmonary vascular enhancement (black white arrow points to the tapered obliteration of the proximal pulmonary artery and vein) in the torted upper lobe. Also note the thickening around the bronchus on the right side (white arrowhead).
Fig. 20. Coronal contrast enhanced CT (left image) demonstrates severe extrinsic compression of the main pulmonary artery (thin arrow by a large ascending aortic aneurysm (AA)). Also note the overall reduced perfusion of the right lung. There were resultant multifocal infarcts within the right lung, one of which is illustrated (black arrow; right image).
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
5. Conclusion Pulmonary infarctions have a multifactorial aetiology. The classical imaging signs may not always be present and the detection of supplementary features can be instrumental in narrowing the differential diagnosis. Contrast enhanced CT is the mainstay in identifying the presence and cause of pulmonary infarction but a multimodality imaging approach is often necessary. As an ancillary imaging modality, MRI is an especially valuable problem-solving tool with roles for nuclear imaging and ultrasound predominantly confined to certain patients subsets. Disclosure The authors have nothing to disclose. Conflict of interest The authors declare that there are no conflict of interest. Acknowledgements K.H.M. has research support from the NIHR Cambridge Biomedical Research Centre. Histological images (Figs. 1 and 2) have been contributed by Dr Doris Rassl, Consultant Histopathologist, Papworth Hospital NHS Foundation Trust, Cambridge, United Kingdom. References [1] Yousem SA. The surgical pathology of pulmonary infarcts: diagnostic confusion with granulomatous disease, vasculitis, and neoplasia. Mod Pathol 2009;22:679–85. [2] Tsao MS, Schraufnagel D, Wang NS. Pathogenesis of pulmonary infarction. Am J Med 1982;72:599–606. [3] Nelson E, Klein JS. Pulmonary infarction resulting from metastatic osteogenic sarcoma with pulmonary venous tumor thrombus. AJR – Am J Roentgenol 2000;174:531–3. [4] Braun S, Platzek I, Zophel K, et al. Haemoptysis due to pulmonary venous stenosis. Eur Respir Rev: Off J Eur Respir Soc 2014;23:170–9. [5] Frazier AA, Galvin JR, Franks TJ, Rosado-De-Christenson ML. From the archives of the AFIP: pulmonary vasculature: hypertension and infarction. Radiographics 2000;20:491–524, quiz 530-491, 532. [6] Hampton AO, Castleman B. Correlation of postmortem chest teleroentgenograms with autopsy findings. Am J Roentgenol Radium Ther 1940;34:305–26. [7] Smith MJ. Roentgenographic aspects of complete and incomplete pulmonary infarction. Dis Chest 1953;23:532–46. [8] Webb WR, Higgins CB. Thoracic imaging: pulmonary and cardiovascular radiology, North American edition. Philadelphia: Lippincott Williams & Wilkins; 2012. [9] Gleeson TG, Cheyne I, English JC, Quadri SM, Leipsic JA. A 50-year-old woman with a history of atrial fibrillation presents with acute dyspnea and pleuritic chest pain. Chest 2010;137:1225–30. [10] Williamson WA, Tronic BS, Levitan N, Webb-Johnson DC, Shahian DM, Ellis Jr FH. Pulmonary venous infarction. Chest 1992;102:937–40. [11] Balakrishnan J, Meziane MA, Siegelman SS, Fishman EK. Pulmonary infarction: CT appearance with pathologic correlation. J Comput Assist Tomogr 1989;13:941–5. [12] He H, Stein MW, Zalta B, Haramati LB. Pulmonary infarction: spectrum of findings on multidetector helical CT. J Thorac Imaging 2006;21:1–7. [13] Johnson PT, Wechsler RJ, Salazar AM, Fisher AM, Nazarian LN, Steiner RM. Spiral CT of acute pulmonary thromboembolism: evaluation of pleuroparenchymal abnormalities. J Comput Assist Tomogr 1999;23:369–73. [14] Revel MP, Triki R, Chatellier G, et al. Is it possible to recognize pulmonary infarction on multisection CT images? Radiology 2007;244:875–82. [15] Grabowski GA, Leslie N, Wenstrup R. Enzyme therapy for gaucher disease: the first 5 years. Blood Rev 1998;12:115–33. [16] Harris H, Barraclough R, Davies C, Armstrong I, Kiely DG, van Beek Jr E. Cavitating lung lesions in chronic thromboembolic pulmonary hypertension. J Radiol Case Rep 2008;2:11–21. [17] Ye R, Zhao L, Wang C, Wu X, Yan H. Clinical characteristics of septic pulmonary embolism in adults: a systematic review. Respir Med 2014;108:1–8. [18] Hall FM, Salzman EW, Ellis I, Kurland GS. Pneumothorax complicating aseptic cavitating pulmonary infarction. Chest 1977;72:232–4.
2253
[19] Narita JI, Ito S, Terada M, et al. Pulmonary artery involvement in takayasu’s arteritis with lung infarction and pulmonary aspergillosis. J Clin Rheumatol 2002;8:260–4. [20] Wick MR, Ritter JH, Schuller D. Ruptured pulmonary infarction: a rare, fatal complication of thromboembolic disease. Mayo Clin Proc 2000;75:639–42. [21] Geyer LL, Scherr M, Korner M, et al. Imaging of acute pulmonary embolism using a dual energy CT system with rapid KVP switching: initial results. Eur J Radiol 2012;81:3711–8. [22] Hansmann J, Apfaltrer P, Zoellner FG, et al. Correlation analysis of dual-energy CT iodine maps with quantitative pulmonary perfusion MRI. World J Radiol 2013;5:202–7. [23] Johnson TR, Krauss B, Sedlmair M, et al. Material differentiation by dual energy CT: initial experience. Eur Radiol 2007;17:1510–7. [24] Thieme SF, Graute V, Nikolaou K, et al. Dual energy CT lung perfusion imaging—correlation with spect/CT. Eur J Radiol 2012;81:360–5. [25] Chai X, Zhang LJ, Yeh BM, Zhao YE, Hu XB, Lu GM. Acute and subacute dual energy CT findings of pulmonary embolism in rabbits: correlation with histopathology. Br J Radiol 2012;85:613–22. [26] van Beek EJ, Hoffman EA. Functional imaging: CT and MRI. Clin Chest Med 2008;29:195–216, vii. [27] Huber DJ, Kobzik L, Solorzano C, Melanson G, Adams DF. Nuclear magnetic resonance spectroscopy of acute and evolving pulmonary hemorrhage. An in vitro study. Invest Radiol 1987;22:632–7. [28] Kessler R, Fraisse P, Krause D, Veillon F, Vandevenne A. Magnetic resonance imaging in the diagnosis of pulmonary infarction. Chest 1991;99:298–300. [29] Iwasaki Y, Nagata K, Nakanishi M, et al. Spiral CT findings in septic pulmonary emboli. Eur J Radiol 2001;37:190–4. [30] Kuhlman JE, Fishman EK, Teigen C. Pulmonary septic emboli: diagnosis with CT. Radiology 1990;174:211–3. [31] Puderbach M, Hintze C, Ley S, Eichinger M, Kauczor HU, Biederer J. Mr imaging of the chest: a practical approach at 1.5 t. Eur J Radiol 2007;64:345–55. [32] Barreto MM, Rafful PP, Rodrigues RS, et al. Correlation between computed tomographic and magnetic resonance imaging findings of parenchymal lung diseases. Eur J Radiol 2013;82:e492–501. [33] Kluge A, Mueller C, Strunk J, Lange U, Bachmann G. Experience in 207 combined MRI examinations for acute pulmonary embolism and deep vein thrombosis. AJR – Am J Roentgenol 2006;186:1686–96. [34] Mathis G, Metzler J, Fussenegger D, Sutterlutti G, Feurstein M, Fritzsche H. Sonographic observation of pulmonary infarction and early infarctions by pulmonary embolism. Eur Heart J 1993;14:804–8. [35] Reissig A, Heyne JP, Kroegel C. Sonography of lung and pleura in pulmonary embolism: sonomorphologic characterization and comparison with spiral CT scanning. Chest 2001;120:1977–83. [36] Mathis G, Dirschmid K. Pulmonary infarction: sonographic appearance with pathologic correlation. Eur J Radiol 1993;17:170–4. [37] Gorg C, Bert T, Gorg K. Contrast-enhanced sonography for differential diagnosis of pleurisy and focal pleural lesions of unknown cause. Chest 2005;128:3894–9. [38] Kreuter M, Mathis G. Emergency ultrasound of the chest. Respiration 2014;87:89–97. [39] McGoldrick PJ, Rudd TG, Figley MM, Wilhelm JP. What becomes of pulmonary infarcts? AJR – Am J Roentgenol 1979;133:1039–45. [40] Worsley DF, Alavi A, Aronchick JM, Chen JT, Greenspan RH, Ravin CE. Chest radiographic findings in patients with acute pulmonary embolism: observations from the pioped study. Radiology 1993;189:133–6. [41] Worsley DF, Kim CK, Alavi A, Palevsky HI. Detailed analysis of patients with matched ventilation–perfusion defects and chest radiographic opacities. J Nucl Med: Off Publ Soc Nucl Med 1993;34:1851–3. [42] Gottschalk A, Stein PD, Henry JW, Relyea B. Matched ventilation, perfusion and chest radiographic abnormalities in acute pulmonary embolism. J Nucl Med: Off Publ Soc Nucl Med 1996;37:1636–8. [43] Kim CK, Worsley DF, Alavi A. Ventilation–perfusion-chest radiography match is less likely to represent pulmonary embolism if perfusion is decreased rather than absent. Clin Nucl Med 2000;25:665–9. [44] Zaki M, Suga K, Kawakami Y, et al. Preferential location of acute pulmonary thromboembolism induced consolidative opacities: assessment with respiratory gated perfusion spect-CT fusion images. Nucl Med Commun 2005;26:465–74. [45] Suga K. Pulmonary function-morphologic relationships assessed by spect-CT fusion images. Ann Nucl Med 2012;26:298–310. [46] Goethals I, Smeets P, De Winter O, Noens L. Focally enhanced f-18 fluorodeoxyglucose (FDG) uptake in incidentally detected pulmonary embolism on PET/CT scanning. Clin Nucl Med 2006;31:497–8. [47] Soussan M, Rust E, Pop G, Morere JF, Brillet PY, Eder V. The rim sign: FDGPET/CT pattern of pulmonary infarction. Insights Imaging 2012;3:629–33. [48] Badr A, Joyce JM, Durick J. Rim of FDG uptake around a pulmonary infarct on PET/CT in a patient with unsuspected pulmonary embolism. Clin Nucl Med 2009;34:285–6. [49] Kamel EM, McKee TA, Calcagni ML, et al. Occult lung infarction may induce false interpretation of 18F-FDG PET in primary staging of pulmonary malignancies. Eur J Nucl Med Mol Imaging 2005;32:641–6. [50] Wittram C, Scott JA. 18F-FDG PET of pulmonary embolism. AJR – Am J Roentgenol 2007;189:171–6. [51] Parambil JG, Savci CD, Tazelaar HD, Ryu JH. Causes and presenting features of pulmonary infarctions in 43 cases identified by surgical lung biopsy. Chest 2005;127:1178–83.
2254
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
[52] Mandelli V, Schmid C, Zogno C, Morpurgo M. False negatives and false positives in acute pulmonary embolism: a clinical-postmortem comparison. Cardiologia (Rome, Italy) 1997;42:205–10. [53] Karabulut N, Kiroglu Y. Relationship of parenchymal and pleural abnormalities with acute pulmonary embolism: CT findings in patients with and without embolism. Diagn Interv Radiol 2008;14:189–96. [54] Schraufnagel DE, Ming-Soung T, Yao YT, Nai-San W. Factors associated with pulmonary infarction. A discriminant analysis study. Am J Clin Pathol 1985;84:15–8. [55] Shah AA, Davis SD, Gamsu G, Intriere L. Parenchymal and pleural findings in patients with and patients without acute pulmonary embolism detected at spiral CT. Radiology 1999;211:147–53. [56] Morgenthaler TI, Ryu JH, Utz JP. Cavitary pulmonary infarct in immunocompromised hosts. Mayo Clin Proc 1995;70:66–8. [57] Wilson AG, Joseph AE, Butland RJ. The radiology of aseptic cavitation in pulmonary infarction. Clin Radiol 1986;37:327–33. [58] Libby LS, King TE, LaForce FM, Schwarz MI. Pulmonary cavitation following pulmonary infarction. Medicine (Baltimore) 1985;64:342–8. [59] Cannon JW. Necrotizing clostridial pneumonia: a case report and review of the literature. Mil Med 2002;167:85–6. [60] Mortensen KH, Gopalan D, Balan A. Atrial masses on multidetector computed tomography. Clin Radiol 2013;68:e164–75. [61] Marriott AE, Weisbrod G. Bronchogenic carcinoma associated with pulmonary infarction. Radiology 1982;145:593–7. [62] Nomori H, Horio H, Morinaga S, Suemasu K. Multiple pulmonary infarctions associated with lung cancer. Jpn J Clin Oncol 2000;30:40–2. [63] Ishiguro T, Kasahara K, Matsumoto I, et al. Primary pulmonary artery sarcoma detected with a pulmonary infarction. Intern Med (Tokyo, Japan) 2007;46:601–4. [64] Takahashi M, Murakami Y, Nitta N, et al. Pulmonary infarction associated with bronchogenic carcinoma. Radiat Med 2008;26:76–80. [65] Engelke C, Schaefer-Prokop C, Schirg E, Freihorst J, Grubnic S, Prokop M. Highresolution CT and CT angiography of peripheral pulmonary vascular disorders. Radiographics 2002;22:739–64. [66] Cox JE, Chiles C, Aquino SL, Savage P, Oaks T. Pulmonary artery sarcomas: a review of clinical and radiologic features. J Comput Assist Tomogr 1997;21:750–5. [67] Rodrigues ACT, Kay FU, Ogawa A, et al. Pulmonary complications in right sided endocarditis. Circulation 2013:128. [68] Gadkowski LB, Stout JE. Cavitary pulmonary disease. Clin Microbiol Rev 2008;21:305–33, table of contents. [69] Han D, Lee KS, Franquet T, et al. Thrombotic and nonthrombotic pulmonary arterial embolism: spectrum of imaging findings. Radiographics 2003;23:1521–39. [70] Huang RM, Naidich DP, Lubat E, Schinella R, Garay SM, McCauley DI. Septic pulmonary emboli: CT-radiographic correlation. AJR – Am J Roentgenol 1989;153:41–5. [71] Kwon WJ, Jeong YJ, Kim KI, et al. Computed tomographic features of pulmonary septic emboli: comparison of causative microorganisms. J Comput Assist Tomogr 2007;31:390–4. [72] Rajagopala S, Devaraj U, D’Souza G. Infected cavitating pulmonary infarction. Respir Care 2011;56:707–9. [73] Scharf J, Nahir AM, Munk J, Lichtig C. Aseptic cavitation in pulmonary infarction. Chest 1971;59:456–8. [74] Djordjevic I, Pejcic T, Nastasijevic-Borovac D, Radjenovic-Petkovic T. Cavitary pulmonary infarct or necrotizing pneumonia: the differential diagnostic dilemma. Chest 2012:142. [75] Denning DW. Invasive aspergillosis. Clin Infect Dis: Off Publ Infect Dis Soc Am 1998;26:781–803, quiz 804-785. [76] Brown MJ, Worthy SA, Flint JD, Muller NL. Invasive aspergillosis in the immunocompromised host: utility of computed tomography and bronchoalveolar lavage. Clin Radiol 1998;53:255–7. [77] Blum U, Windfuhr M, Buitrago-Tellez C, Sigmund G, Herbst EW, Langer M. Invasive pulmonary aspergillosis. MRI, CT, and plain radiographic findings and their contribution for early diagnosis. Chest 1994;106: 1156–61. [78] Won HJ, Lee KS, Cheon JE, et al. Invasive pulmonary aspergillosis: prediction at thin-section CT in patients with neutropenia—a prospective study. Radiology 1998;208:777–82. [79] Horger M, Hebart H, Einsele H, et al. Initial CT manifestations of invasive pulmonary aspergillosis in 45 non-HIV immunocompromised patients: association with patient outcome? Eur J Radiol 2005;55:437–44.
[80] Franquet T, Muller NL, Gimenez A, Guembe P, de La Torre J, Bague S. Spectrum of pulmonary aspergillosis: histologic, clinical, and radiologic findings. Radiographics 2001;21:825–37. [81] Logan PM, Muller NL. CT manifestations of pulmonary aspergillosis. Crit Rev Diagn Imaging 1996;37:1–37. [82] Franquet T, Muller NL, Gimenez A, Domingo P, Plaza V, Bordes R. Semiinvasive pulmonary aspergillosis in chronic obstructive pulmonary disease: radiologic and pathologic findings in nine patients. AJR – Am J Roentgenol 2000;174:51–6. [83] Kim SY, Lee KS, Han J, et al. Semiinvasive pulmonary aspergillosis: CT and pathologic findings in six patients. AJR – Am J Roentgenol 2000;174: 795–8. [84] Montagnana M, Cervellin G, Franchini M, Lippi G. Pathophysiology, clinics and diagnostics of non-thrombotic pulmonary embolism. J Thromb Thrombolysis 2011;31:436–44. [85] Jorens PG, Van Marck E, Snoeckx A, Parizel PM. Nonthrombotic pulmonary embolism. Eur Respir J 2009;34:452–74. [86] Bach AG, Restrepo CS, Abbas J, et al. Imaging of nonthrombotic pulmonary embolism: biological materials, nonbiological materials, and foreign bodies. Eur J Radiol 2013;82:e120–41. [87] Vichinsky EP, Neumayr LD, Earles AN, et al. Causes and outcomes of the acute chest syndrome in sickle cell disease. National acute chest syndrome study group. N Engl J Med 2000;342:1855–65. [88] Hawley DA, McCarthy LJ. Sickle cell disease: two fatalities due to bone marrow emboli in patients with acute chest syndrome. Am J Forensic Med Pathol 2009;30:69–71. [89] Brown KK. Pulmonary vasculitis. Proc Am Thorac Soc 2006;3:48–57. [90] Castaner E, Alguersuari A, Gallardo X, et al. When to suspect pulmonary vasculitis: radiologic and clinical clues. Radiographics 2010;30:33–53. [91] Yataco J, Stoller JK. Pulmonary venous thrombosis and infarction complicating pulmonary venous stenosis following radiofrequency ablation. Respir Care 2004;49:1525–7. [92] Duehmke RM, Fynn SP, Gopalan D. Pulmonary venous infarction following pulmonary vein isolation. Eur Heart J 2008;29:2893. [93] Lee H, Wanless A. Pulmonary vein thrombosis: a case report and literature review. J Hosp Med 2011;6:S207–8. [94] Frazier AA, Franks TJ, Mohammed TL, Ozbudak IH, Galvin JR. From the archives of the AFIP: pulmonary veno-occlusive disease and pulmonary capillary hemangiomatosis. Radiographics 2007;27:867–82. [95] Reinke RT, Higgins CB. Pulmonary infarction complicating the use of swanganz catheters. Br J Radiol 1975;48:885–8. [96] McLoud TC, Putman CE. Radiology of the swan-ganz catheter and associated pulmonary complications. Radiology 1975;116:19–22. [97] Shackles C, Rundback JH, Herman K. Pulmonary infarct after transarterial chemoembolization. J Vasc Interv Radiol – JVIR 2013;24:1928–30. [98] Krivokuca I, van de Graaf EA, van Kessel DA, van den Bosch JM, Grutters JC, Kwakkel-van Erp JM. Pulmonary embolism and pulmonary infarction after lung transplantation. Clin Appl Thromb Hemost 2011;17:421–4. [99] Clark SC, Levine AJ, Hasan A, Hilton CJ, Forty J, Dark JH. Vascular complications of lung transplantation. Ann Thorac Surg 1996;61:1079–82. [100] Choong CK, Hu DZ, Huddleston CB. Pulmonary segmental venous infarction after living-donor lobar lung transplantation. J Thorac Cardiovasc Surg 2005;130:919–21. [101] Haraguchi S, Koizumi K, Hirata T, et al. Surgical results of completion pneumonectomy. Ann Thorac Cardiovasc Surg 2011;17:24–8. [102] Cable DG, Deschamps C, Allen MS, et al. Lobar torsion after pulmonary resection: presentation and outcome. J Thorac Cardiovasc Surg 2001;122:1091–3. [103] Niekel MC, Horsch AD, Ven M, Reijnen MM, Joosten FB. Right middle lobe torsion: evaluation with CT angiography. Emerg Radiol 2009;16: 387–9. [104] Oliveira C, Zamakhshary M, Abdallah MR, et al. Lung torsion after tracheoesophageal fistula repair: a case report and review of literature. J Pediatr Surg 2007;42:E5–9. [105] Felson B. Lung torsion: radiographic findings in nine cases. Radiology 1987;162:631–8. [106] Moser Jr ES, Proto AV. Lung torsion: case report and literature review. Radiology 1987;162:639–43. [107] Robinson T, Lynch J, Grech E. Non-hodgkin’s lymphoma causing extrinsic pulmonary artery compression. Eur J Echocardiogr 2008;9:577–8. [108] Held BT, Siegelman SS. Pulmonary infarction secondary to bronchogenic carcinoma. Am J Roentgenol Radium Ther Nucl Med 1974;120:145–50.
Contents lists available at ScienceDirect
European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad
Review
Multimodality imaging of pulmonary infarction T.J.P. Bray a,1 , K.H. Mortensen a,b,∗ , D. Gopalan a,1 a
Department of Radiology, Papworth Hospital NHS Foundation Trust, Ermine Street, Papworth Everard, Cambridge CB23 3RE, United Kingdom University Department of Radiology, Addenbrookes Hospital, Cambridge University Hospitals NHS Foundation Trust, Hills Road, Box 318, Cambridge CB2 0QQ, United Kingdom b
a r t i c l e
i n f o
Article history: Received 20 March 2014 Received in revised form 16 June 2014 Accepted 20 July 2014 Keywords: Pulmonary infarction Arterial infarction Venous infarction Pulmonary embolus Thromboembolic disease Septic infarction In situ thrombus Fat embolism Vasculitic infarct Pulmonary torsion
a b s t r a c t The impact of absent pulmonary arterial and venous flow on the pulmonary parenchyma depends on a host of factors. These include location of the occlusive insult, the speed at which the occlusion develops and the ability of the normal dual arterial supply to compensate through increased bronchial arterial flow. Pulmonary infarction occurs when oxygenation is cut off secondary to sudden occlusion with lack of recruitment of the dual supply arterial system. Thromboembolic disease is the commonest cause of such an insult but a whole range of disease processes intrinsic and extrinsic to the pulmonary arterial and venous lumen may also result in infarcts. Recognition of the presence of infarction can be challenging as imaging manifestations often differ from the classically described wedge shaped defect and a number of weighty causes need consideration. This review highlights aetiologies and imaging appearances of pulmonary infarction, utilising cases to illustrate the essential role of a multimodality imaging approach in order to arrive at the appropriate diagnosis. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Pulmonary infarction is due to coagulative ischaemic necrosis [1]. This process occurs when parenchymal perfusion fails to meet the tissue requirements due to occlusive insults that are either intrinsic or extrinsic to the vascular lumen. Normally, dual arterial supply comprised of deoxygenated blood in the pulmonary arteries and oxygenated blood in the highly adaptable bronchial arteries protects against infarcts. The risk of ischaemic necrosis is especially increased in perturbations of blood flow that dually impede this system, be it congestive heart failure, pulmonary venous hypertension, or diminished bronchial arterial flow in systemic arterial hypotension [2]. Less commonly infarction may result from perturbations in pulmonary venous flow [3,4]. A ‘complete’ infarct is characterised by irreversible ischaemic injury, wherein incident capillary ischaemia increases vascular permeability with subsequent bronchial artery reperfusion leading to alveolar haemorrhage
∗ Corresponding author at: Department of Radiology, Papworth Hospital NHS Foundation Trust, Ermine Street, Papworth Everard, Cambridge CB23 3RE, United Kingdom. Tel.: +44 1480364562. E-mail addresses: [email protected] (T.J.P. Bray), [email protected] (K.H. Mortensen), [email protected] (D. Gopalan). 1 Tel.: +44 1480364562. http://dx.doi.org/10.1016/j.ejrad.2014.07.016 0720-048X/© 2014 Elsevier Ireland Ltd. All rights reserved.
and necrosis ensuing hereafter. The end-result is a fibrotic scar. An infarct presents macroscopically as a dark necrotic area with a narrow rim of hyperaemia and inflammation, presenting microscopically as a ‘ghost-like’ architecture with an outer perimeter of cellular infiltrate (Fig. 1) [1,5]. The infarct can also be ‘incomplete’ when there is transient haemorrhagic congestion but no other acute or chronic parenchymal sequelae – the parenchymal opacity resolves in a matter of days [6,7]. By virtue of the pathophysiological mechanism, arterial infarcts are peripheral, occurring in areas supplied by medium and small sized arteries [8]. The acute infarct is a wedge-shaped (less frequently nodular) pleurally based parenchymal opacity (‘Hampton hump’) that forms an obtuse angle with the visceral pleura, with the apex pointing towards the pulmonary hilum and centred on a bronchovascular bundle, whereas a pulmonary venous infarction is paraseptal in distribution (Fig. 2) [6,9,10]. Less typical features of arterial and venous infarcts are increasingly reported with a move to raised suspicion of infarction with a whole array of imaging appearances [11]. Infarcts may be solitary or multiple, temporally uniform or variegated much depending on the inciting event, and characteristically the airspace opacity should diminish in size over time (‘melting sign’) with complete resolution or leaving residual plate-like scarring and focal pleural thickening. Timely diagnosis of pulmonary infarction is important in order to relieve the inciting event and avert further respiratory and
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
vascular compromise. This pictorial review aims to demonstrate imaging features of pulmonary infarction using a multimodality approach in order to illustrate the spectrum of aetiologies that may result in infarction. Although computed tomography (CT) currently forms the mainstay of diagnosis the use of ancillary imaging modalities is often necessary. 2. Imaging modalities
Table 1 Salient imaging findings for pulmonary arterial infarcts by modality. Radiography
Usually non-specific parenchymal opacity Wedge-shaped pleurally based opacity: Hampton hump Pleural effusion (small, unilateral), elevated hemidiaphragm (volume loss), atelectasis
Computed tomography
Hampton hump Internal lucency in peripheral consolidation: bubbly consolidation Thickened vessels leading to apex Pleural effusion (localised or diffuse on Ipsilateral side) Cavitation (<10%)
2.1. Computed tomography The wedge-shaped, broad pleurally based ‘Hampton hump’ frequently has a truncated apex and a convex border with less common findings including internal air lucency, linear stranding from apex to hilum and thickened vessels leading to the apex of the opacity (Table 1) [12]. Contrast-enhanced CT pulmonary angiography (CTPA) may demonstrate decreased enhancement within consolidated lung (Fig. 3); this is present in 95% of infarcts but can also be seen in collapse, pneumonic consolidation and neoplasm [12,13]. Central lucency within peripheral consolidation is particularly indicative of pulmonary infarction (‘bubbly consolidation’) (Fig. 3) [14] but other causes for these appearances include cavitation, cysts and dilated airways [15]. Although cavitation within infarcts is often due to bland necrosis, the presence of superinfection of established bland infarction or primary septic emboli can give a similar appearance [16,17]. CT is particularly helpful for the detection of complications of infarcts, including potentially detrimental sterile or purulent involvement of the mediastinum, pleural space, remaining parenchyma and airways, or the thoracic cage [18–20]. Dual Energy CT lends promise for the visualisation of pulmonary infarcts and vascular occlusions [21]. Contrast-enhanced data sets are acquired at two energy spectra and reconstructed to form a conventional CTPA plus a static iodine map, the latter representing a surrogate measure of microvascular circulation and perfusion (Fig. 4) [22,23]. The functional iodine map increases diagnostic accuracy by detecting more peripheral occlusions than conventional CTPA [24]. In cases of pulmonary arterial occlusion without infarction the iodine map depicts relative oligaemia due to persistent perfusion via the bronchial arteries; infarction on the other hand is characterised by complete segmental absence of iodine due to absent perfusion and a corresponding morphological abnormality [25]. The segmental distribution aids in the differentiation of infarction from other opacities with low iodine content such as tumour, pneumonia, or even streak artefact. Dynamic Perfusion CT is another technique with potential future applications in the
2241
Dual energy computed tomography
Contrast enhanced: hypoenhancing centre Adjacent or ipsilateral pleural effusion Perfusion defect on iodine map with correlating parenchymal opacity
Ultrasound
Wedge-shaped, pleurally based area of reduced echogenicity Hyperechoic centre (reverberation artefact) Late contrast-enhancement
Magnetic resonance imaging
Hyperacute: hyperintense T2WI and hypointense T1WI Acute (up to 1 week): hyperintense T1WI Hyperintense signal on both T1WI and T2 WI compared with tumour Perfusion defect: visualised with static and/or dynamic contrast enhanced MRI
Ventilation/perfusion scintigraphy
Matched ventilation and perfusion defect with radiographic parenchymal opacity – ‘Triple-match’
Single positron emission tomography
Consolidation in peripheral interface between severely decreased and relatively preserved perfusion areas
Positron emission tomography
FDG-avid rim along the infarct periphery Relative central photopenia Alternative appearance: diffuse uptake
assessment of infarction but its use has hitherto remained limited due to radiation concerns [26]. At present CTPA remains the principal diagnostic tool with the dual energy technique potentially providing a future one-stop morphological and functional test. 2.2. Magnetic resonance imaging Imaging appearances of infarcts on magnetic resonance imaging (MRI) vary according to signal characteristics of the aging blood that has accumulated within the alveoli [27]. Acute infarcts
Fig. 1. Haematoxylin and eosin stain images of two different arterial infarcts. Left image (original magnification 50×) is an organised pulmonary arterial infarction (white star) with a chronic inflammatory infiltrate in the margin (white arrow). Right image (original magnification 16×) shows a peripheral infarct (white star) with pleural thickening (black arrow).
2242
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 2. CT pulmonary angiography on mediastinal (left image) and lung (middle image) windows demonstrates a wedge shaped pleurally based parenchymal opacity (‘Hampton hump’, arrow) with the apex pointing towards the pulmonary hilum and centred on a bronchovascular bundle. Far right image is a microscopic specimen showing a wedge shaped pulmonary arterial infarct (star) from a different patient (haematoxylin and eosin stain, original magnification 160×).
Fig. 3. Coronal (left image) and axial (right image) CT images on mediastinal and lung windows, respectively, depict a large left pulmonary embolus (thin arrow) and a peripheral infarct (block arrow) with bubbly consolidation, linear ‘stranding’ from apex to hilum and lack of parenchymal enhancement.
(<24 h) are T1 hypointense and T2 hyperintense with subacute infarction being T1 hyperintense (>24 h to 1 week) [28]. The subsequent temporal imaging characteristics of infarcts have not been assessed. Infarcts secondary to thromboembolic disease on
average measure 4 cm (ranging from 1 to 12 cm) whilst septic infarcts tend to be smaller but still detectable with the current spatial resolution of MRI [12,29,30,31]. Appearances with techniques such as dynamic contrast enhanced and diffusion weighted
Fig. 4. Dual energy CT on coronal reformats demonstrate a filling defect (arrows) in the left lower lobe segmental pulmonary artery (left image) and a corresponding perfusion defect on the iodine map (right image). Also note the bilateral pleural effusions and dependent atelectasis.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
MRI remain to be established but these techniques may help diagnosing alternative causes of parenchymal opacity in the setting of suspected pulmonary infarction [32]. The current role of MRI in imaging of infarcts is that of an ancillary, problem-solving tool, uniquely combining comprehensive anatomical and functional assessment of the thoracic vasculature with excellent soft tissue characterisation without the use of radiation [32,33]. In certain situations where CTPA is not available secondary to concerns regarding the use of radiation or iodinated contrast media, MRI may however have a primary role. 2.3. Ultrasound Infarcts are conspicuous on high-frequency ultrasound (typically 3.5–8 mHz) as pleurally based hypoechoic, triangular or infrequently round lesions [34,35]. Typically, infarcts are sharply demarcated. Early infarcts are often homogenously hypoechoic compared with the centrally hyperechoic older infarction [36]. When present, internal reverberation artefact represents a bronchiole and is characteristic of segmental involvement. Congested enlarged vessels leading to the apex of the lesion may also be evident [36]. Contrast-enhanced ultrasound displays increased time-to-enhancement when compared with infective consolidation, although this is not very specific [37]. The main role for the detection of infarction by thoracic ultrasound lies in facilitating the diagnosis of acute pulmonary embolus with bedside imaging, requiring detection of two or more infarcts [38]. Ultrasound predominantly has a role in the acute setting in the emergency room when alternative imaging modalities are not feasible. 2.4. Radiography The stereotypical appearance of pulmonary infarction as a peripheral wedge shaped pleurally based opacity was initially described on chest radiography, the ‘Hampton hump’ [6,7]. Temporally, approximately half of pulmonary infarcts leave residual radiographic changes that include linear scars, pleural diaphragmatic adhesions, and localised pleural thickening and retraction [39,40]. These residual findings are typically mild compared to the initial abnormality with a concomitant reduction in the initially detected perfusion abnormality on perfusion scintigraphy [39]. The presence of infarction on radiographs may increase awareness especially of pulmonary embolism in the acute setting. However, chest radiography may maintain a role as a general screening tool in cardiorespiratory symptoms but this modality adds little value when resolving the aetiology on an infarct.
2243
surrounding inflammatory reaction [46–49]. Unfortunately, the rim sign is not specific with many infarcts displaying homogenously increased tracer uptake [49,50].
3. Intravascular aetiology 3.1. Thromboembolic infarction Pulmonary thromboembolism from peripheral deep venous thrombosis is the most common cause of infarction [51], occurring in as many as a third of acute pulmonary emboli [2,52]. At autopsy, infarcts are more likely to occur and be larger in size with a higher embolic burden but recent CT studies have not shown similar correlation between clot burden and pleurally based opacities [53,54]. Infarction has a predilection for the lower zones due to a relatively increased vascularity. Because infarcts typically extend to the pleural surface, pleural effusion is more common in pulmonary embolism with than without infarction (29% versus 4%) [13]. Infarcts are the only more common pleuroparenchymal CT feature in patients with confirmed embolus in cohorts with suspected pulmonary thromboembolism [53,55]. Infarcts in chronic thromboembolic pulmonary disease can be particularly difficult to differentiate from malignancy and infection. There will, however, often be temporal variegation in the imaging appearances of infarcts due to the protracted nature of the disease, ranging from acute wedge-shaped opacities to plate-like thin fibrous scars (parenchymal bands) with focal areas of architectural distortion and volume loss (Fig. 5). Cavitation occurs in 11% of infarcts in chronic thromboembolic disease compared with 5% in acute thromboemboli [16,56]. Typically the cavity is solitary, arising from days 2 to 63 following the ischaemic insult and often on the backdrop of bland infarction that may become complicated by infection [57,58]. Cavitation is more common within mid and upper zones with the cavity wall generally being a poor discriminator of the underlying aetiology (Fig. 6) [16,57,58]. Superinfection with clostridium species has been reported to be especially common, resulting in a necrotising, cavitatory pneumonia [59]. Congenital heart disease can also predispose to infarct formation following in situ thrombosis and subsequent embolisation (Fig. 7). Other causes of sterile, intra-cardiac migration thrombus as a cause of infarction are coagulopathy, tricuspid valve stenosis, intracardiac tumour (surface thrombus) or myocardial infarction [60].
3.2. Malignant infarction 2.5. Nuclear medicine Acute infarcts appear on ventilation perfusion scintigraphy (V/Q scintigraphy) as matched perfusion and ventilation defects with corresponding radiographic opacities – the ‘triple match sign’ [41]. Typically this is seen in a mid-to-upper zone location, with completely absent perfusion and involvement of 25–50% of a lung zone [41–43]. Single positron emission CT combined with unenhanced CT can differentiate infarcts from infective consolidation; parenchymal opacity at the peripheral lung interface between severely decreased and adjacent relatively preserved perfusion areas within relatively large and severely decreased perfusion defects favours infarction whilst infective consolidation is located in the proximal aspect of often smaller perfusion defects [44,45]. Positron emission tomography CT (FDG-PET CT) imaging of infarcts has identified the ‘rim sign’, demonstrating relative central photopenia with a rim of high tracer uptake along the border of the parenchymal opacity, corresponding to central necrosis and
Malignancy associated infarcts mainly occur with haematogenous dissemination of larger tumour emboli that instantaneously occlude the pulmonary vascular lumen, creating an opacity that arises at a speed disjointed from the tumour doubling time (Fig. 8). The appearance is dissimilar from classical haematogenous metastasis where microemboli lodge within smaller peripheral pulmonary arteries and give rise to peribronchovascular nodules that enlarge according to the expected growth rate. Some tumours such as hepatocellular and renal cell carcinoma are especially prone to producing larger intravascular tumour fragments with a potential to cause infarction [60]. When a parenchymal opacity suddenly arises in the setting known or suspected malignancy the differential must therefore alongside infarction include disease progression and pneumonitis [61,62]. This distinction is important since erroneous interpretation may cause faulty tumour burden upstaging and histology may be required for confirmation when imaging is equivocal [49].
2244
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 5. Axial CT in 3 different patients showing the temporal variegation of infarcts on lung windows. Left image is an acute wedge-shaped infarct (block arrow) and plate atelectasis (thin arrow). Middle image is a chronic infarct with focal architectural distortion and volume loss (arrowhead). Right image is a right lower lobe thick walled cavitating infarct (black arrow) with background mosaic attenuation due to chronic thromboembolic disease.
Infarction due to local invasion or compression by tumour growth is less frequent because the pulmonary vasculature will have time to adapt with increments in bronchial artery collateral supply when flow in an invaded pulmonary artery is restricted [63,64]. Due to the contiguous nature of the infarction and adjacent tumour MRI can be used for differentiation with infarcts, being T1 and T2 hyperintense relative to tumour [62]. Of note, infarcts are generally non-enhancing although necrotic tumour may show similar features [65]. Primary pulmonary artery sarcoma is a rare cause of infarction that needs to be kept in mind when assessing for malignant infarction, wherein interval growth may be misdiagnosed as thromboembolic disease (Fig. 9) [66].
3.3. Septic infarction Embolic infarction secondary to septic clots in the pulmonary arteries is particularly frequent in tricuspid valve and pacemaker wire endocarditis, indwelling central venous catheters, septic thrombophlebitis, infected arteriovenous shunts, and infections of bone, skin or soft tissues (Fig. 10) [17,65,67]. Staphylococci are the most common pathogen [68]. Temporally separated embolic showers can result in diffuse bilateral opacities, ranging from solid nodules to cavitation [69,70]. Even though CT features such as larger nodule size, cavitation and infarction may suggest a gram-positive organism, no imaging feature is specific to a pathogen [71]. In the
Fig. 6. Axial CT (top left panel) shows a rounded well-defined peripheral soft tissue opacity containing central lucency (block arrow) in the right upper lobe. Differential diagnoses include malignancy, infection and infarction. The presence of mosaic attenuation (thin arrow) suggests underlying chronic thromboembolic disease. Subsequent PET-CT (remaining images) demonstrates the lung lesion (block arrow; bottom left panel) to be ‘cold’, making infarction more likely than malignancy.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
2245
Fig. 7. A 50-year-old male with known atrial septal defect and haemoptysis. (a) Chest radiograph shows cardiomegaly and enlarged proximal pulmonary arteries (white thin arrow) with an ill-defined opacity in the left apex (white block arrow). Also note the presence of a ‘Reveal’ device (black arrow). (b) Axial contrast enhanced CT demonstrates an enlarged main pulmonary artery (star) with large load of bilateral proximal thrombus (notched arrows; left panel), a sinus venosus type atrial septal defect (thin arrow; middle panel) and a mycetoma complicating a long-standing pulmonary infarct cavity (block arrow; right panel).
setting of known septic embolus, classical wedge shaped opacity is seen in 17% ranking in prevalence after multiple bilateral nodules, cavitation, non-specific infiltrates and pleural effusions with the “feeding vessel sign” often present in relation to both nodules and opacities [17,70]. Ancillary features from complications such as cavitation and empyema may support sepsis as a cause of infarction (Fig. 11). With progressive disease, appearances of the primary infected focus such as endocarditic vegetation, osteomyelitis, or indwelling catheters with adherent soft tissue may be dwarfed by manifestations from metastatic septic deposits (Fig. 12).
Imaging findings overlap considerably between septic and sterile infarcts, though cavitation is more common in septic embolic disease where the parenchymal opacity is, on balance, smaller [12,30,71]. The cavity may rapidly enlarge, causing rupture of the surrounding wall with extension of the cavity and its contents into the pleural cavity leading to pyopneumothorax; the time to cavitation is longer in aseptic compared with septic infarcts (weeks rather than days) with a risk of complications such as pneumothorax and bronchopleural fistula being present with both aetiologies [58,72,73]. The presence of a smooth outline, increasing size or
Fig. 8. Contrast enhanced CT in a 57-year-old male (left panel) demonstrates a right hilar bronchogenic carcinoma with occlusion of the upper lobe artery (thin arrow) and vein. The spiculated mass in the right apex (notched arrow; top right image) was much larger on a previous CT from 3 weeks before (bottom right image), making an infarction more likely than metastases.
2246
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 9. Pulmonary artery sarcoma and infarcts. Axial contrast enhanced CT shows an expansile soft tissue mass in the main and right pulmonary arteries (white star), destruction of vascular integrity in the posterior aspect and resultant mediastinal involvement (left panel). The lung windows (right panel) show an ill-defined pulmonary infarction in the right lower lobe (block arrow).
internal fluid levels all favour primary septic over bland infarction. Absent perfusion within areas of non-cavitating consolidation and a pulmonary arterial filling defect proximal to the opacity are also helpful in distinguishing cavitating infarction from primary necrotising pneumonia [74]. Of note, the risk of superinfection of susceptible necrotic parenchyma is not limited to bacterial pathogens; patients with established cavitatory infarcts also face an increased risk of fungal or mycobacterial infection (Fig. 7). Septic infarction secondary to invasive pulmonary aspergillosis differs in pathogenesis from intravascular dissemination of infected embolus. Most commonly caused by Aspergillus fumigatus this potentially life-threatening infection is seen in severely immunocompromised individuals, often with severe neutropenia. Fungus
invades medium-sized pulmonary arteries with fungal hyphae obstructing blood flow and causing haemorrhagic infarction [75]. CT features include solid nodules surrounded by haemorrhagic ground glass attenuation (‘halo sign’) or consolidation in a patchy, segmental/subsegmental or peribronchiolar pattern (Fig. 13) [76–79]. The areas of consolidation may be pleurally based in some instances [80]. During the convalescence stage, cavitation occurs in more than half of patients [79]. This is evident as an air crescent surrounding the nodular lesion (‘air crescent sign’) and/or cavitation within areas of confluent consolidation. Less severely immunocompromised patients can develop a more indolent and milder form semi-invasive aspergillosis [68]. CT signs in the early stages of this variant are typically upper lobar presenting in some
Fig. 10. A 22-year-old male intravenous drug abuser presented with dyspnoea, chest pain and fever. (a) Axial contrast enhanced CT (top left panel) shows a cavitating soft tissue lesion in the left lower lobe (white block arrow). Transoesophageal (top middle panel) and 3-dimensionsional echocardiography (top right panel) demonstrate a large vegetation (thin arrow) of the tricuspid valve. Retrospective analysis of the non-gated CT after the echocardiography confirmed the presence of an ill-defined low attenuation lesion intimately related to the tricuspid valve (black block arrow). (b) MRI of the spine performed for back pain 2 weeks after the initial diagnosis of infective endocarditis shows a large ill-defined area of consolidation in the right lower lobe (thin arrow) on the coronal localiser image (bottom left panel). Subsequent contrast enhanced CT (bottom middle and left panels) in axial and coronal reformats demonstrate interim reduction in the size of the left lower lobe infarction whilst a new large partially cavitating infarct has developed in the right lower lobe.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
2247
Fig. 11. A 65-year-old male developed fever, chest pain, dyspnoea and weight loss 2 weeks after right knee arthroscopy. Contrast enhanced CT demonstrates lobar pulmonary emboli (arrowheads; left panel) in a large feeding vessel at the apex of a right lower lobe infarct (block arrow; left panel) with an enhancing right pleural collection also noted (thin white arrow; left panel). Careful analysis of the valves showed a large vegetation in the pulmonary valve (notched arrows; right panel).
instances as bilateral segmental consolidation or nodules larger than 1 cm, often with pleural thickening alongside surrounding architectural distortion and volume loss from fibrotic reaction [81–83]. 3.4. Bland infarction Fat embolism can cause microvascular pulmonary infarction; adipose microglobules embolise the pulmonary capillaries together with erythrocytes and thrombocytes, causing diffuse peripheral microinfarction compounded by increased vascular permeability secondary to the toxic effects of fatty acids and inflammatory
cells [84]. Bone trauma is the commonest culprit [85], with a time lag of 12–24 h from symptoms to onset of heterogeneous diffuse airspace opacity that occurs in the absence of features of pulmonary venous congestion (Fig. 14). Imaging findings are often indistinguishable from acute respiratory distress syndrome [85]. Fat embolism can also complicate pancreatitis, hepatitis or other conditions where there may be intravascular fat mobilisation [85]. Other less frequent causes of pulmonary embolism, and even more so infarction, include air, amniotic fluid, talc and heavy metal (in intravenous drug users), hydatid, cement (in vertebroplasty) and other materials that may embolise to the pulmonary circulation [85,86].
Fig. 12. Temporal evolution of cavitating infarcts in a 45-year-old male smoker with history of intravenous drug abuse. Top deck: left panel shows a thick walled cavity in right upper lobe (block arrow) and a tiny cavity abutting the oblique fissure in left lung (thin arrow). There is a large thrombus (black block arrow) in the dilated right ventricle (right panel). Note the bilateral pleural effusions (white star). Bottom deck: following 6 weeks of antibiotic treatment and anticoagulation, the cavities have changed in size and morphology. The right ventricular thrombus persists but has reduced in size.
2248
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 13. A 35-year-old immunosuppressed female with a focal consolidation (block arrow; left image) surrounded by ground glass change (‘halo sign’) in the left upper lobe as evident on axial CT. Tracheal aspirate was positive for Aspergillus fumigatus. Image on the right is a few weeks following antifungal treatment and demonstrates a pleurally based opacity with mild architectural distortion (notched block arrow).
3.5. In situ occlusive infarction 3.5.1. Sickle cell disease The ‘acute chest syndrome’ in erythrocyte sickling disease is defined by a triad of new pulmonary infiltrate, fever and respiratory symptoms, caused by pulmonary infection and infarcts that may be present in isolation or work concert to cause respiratory compromise [87]. Pulmonary arterial occlusions in sickle cell disease are the result of in situ erythrocyte sickling, embolus from fat and necrotic bone marrow in osteonecrosis, or thromboembolic occlusion occurring as a part of a generally procoagulant status [88]. A stereotypical imaging feature is subsegmental atelectasis with or without consolidation, with ancillary features of sickle cell disease often present such as osteonecrosis with vertebral deformity, extramedullary haematopoiesis, or cardiomegaly with or without pulmonary arterial dilatation (Fig. 15). Due to the chronicity of the disorder, infarcts of varying ages may be encountered. Despite a propensity for multi-organ involvement secondary to systemic sickling of erythrocytes, imaging appearances can however be subtle. It is important to detect these features as treatments vary according to the source of occlusion and the ‘acute chest syndrome’ remains a major cause of death in this cohort [87].
Fig. 14. Coronal reformat of contrast enhanced CT demonstrates bilateral heterogeneous diffuse airspace opacity due to fat embolism in a 48-year-old male following bilateral leg amputation as a consequence of a road traffic accident. Also note the normal sized cardiac silhouette and extracorporeal membrane oxygenation cannula (thin arrows).
3.5.2. Vasculitis Primary idiopathic large vessel vasculitis (Takayasu arteritis, giant cell arteritis) is the most common cause of pulmonary involvement in vasculitic disease followed by small-vessel vasculitidis (e.g., granulomatous pulmonary polyangitis) [89]. Pulmonary involvement is less frequent in vasculitis of the medium sized vessels (Polyarteritis nodosa, Kawasaki disease), primary immune-complex vasculitis (e.g., Goodpasture syndrome), and secondary vasculitides (e.g., rheumatoid arthritis). Radiological manifestations vary, and include arterial wall thickening, nodular or cavitatory lesions, ground-glass opacities and consolidation (Fig. 16) [90]. Consolidation tends to be peripheral, pleurally based and of a configuration that varies from classical wedge shape to more poorly defined nodularity, which is the result of in situ occlusion with infarction, haemorrhage or atelectasis. There may be accompanying ground glass changes, representing alveolar haemorrhage, and in the longer term, repeated haemorrhage and disease progression can result in interstitial fibrosis. There will often be ancillary features to suggest the diagnosis of a specific form of vasculitis such as mural thickening of the thoracic aorta in Takayasu arteritis, sinonasal involvement of ANCA-positive granulomatous polyangitis, or renal atrophy following glomerulonephritis in Goodpasture syndrome.
Fig. 15. Chest radiograph in a 27-year-old female shows a wedge shaped peripheral infarct (notched arrow) in the left lower lobe. Also note the ‘H’ shaped vertebral bodies consistent with osteonecrosis (arrow), a classical feature of sickle cell disease.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
2249
Fig. 16. A 25-year-old female with pulmonary vasculitis secondary to Takayasu arteritis. (a) Top deck: there is concentric thickening of the left pulmonary artery and an occluded right pulmonary artery on CT (arrowheads; left image) and MRI (block arrow; right image). Bottom deck: there are multiple infarcts in the right lung; a thick walled cavity in right upper lobe (notched arrow; left image), small thin walled cavity (white arrow; right image) and a ground glass opacity (black arrow; right image) in right lower lobe. (b) FDG PET-CT shows intense uptake in the thick walled cavity (notched arrows; top deck), very low-level tracer uptake in the ground glass opacity (block arrows; bottom deck) and no uptake in the small thin walled cavity (thin arrows; bottom deck). Subsequent biopsy showed the upper lobe cavity to be infected with Mycobacterium tuberculosis.
3.6. Miscalaneous causes 3.6.1. Iatrogenic injury In rare instances iatrogenic injury to the vessel lumen causes infarction, as is seen in radiofrequency ablation of the left atrium complicated by pulmonary venous stenosis or in situ thrombosis
(Fig. 17) [4,91,92]. Depending of the location of the occlusive disease, the venous infarct affects an entire lobe or lung as opposed to the typically multiple scattered peripheral arterial infarcts [9]. The venous infarction is furthermore paraseptal in distribution, and there will be imaging findings of pulmonary venous congestion such as smooth interlobular septal thickening and alveolar oedema
2250
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 17. Pulmonary infarction secondary to vascular anastomotic injury following single lung transplant for interstitial fibrosis. (a) Contrast enhanced axial CT demonstrates stenosis of pulmonary arterial anastomosis (block arrow; far left image) and narrowing of pulmonary venous anastomosis (block arrow, middle image) with proximal consolidation at the left hilum (thin arrow; middle image). Multifocal peripheral ground glass opacities in the left (block arrow, left image) represent early infarcts. (b) Follow up CT performed 3 weeks after the previous study shows evolution of the left upper lobe infarct (thin arrow; left image) and loss of volume in left lower lobe due to the development of fibrotic scarring (block arrow, right image).
within areas affected by the venous occlusion [9]. Venous infarcts related to in situ thrombosis as may also be seen with trauma, infection and malignancy [3,63,93] should be separated from those secondary to fibrosing mediastinitis, congenital pulmonary venous stenosis or left atrial pathology such as myxoma or clot [10]. Pulmonary veno-occlusive disease is another distinct entity in which fibrosis of the pulmonary veins causes similar paraseptal venous infarcts that co-exist with imaging signs of pulmonary arterial hypertension and a normal sized left atrium [94]. Iatrogenic causes of arterial infarction include mal-positioning of Swann-Ganz catheters, either from wedging of the catheter or due to clot formation on the catheter surface [95,96]. Infarcts have also been reported as a rare complication of transarterial chemoembolisation of hepatocellular carcinoma; in one case the tumour’s collateral blood supply communicated directly with a major pulmonary artery, allowing embolic material to enter the pulmonary circulation [97]. 3.6.2. Lung transplantation and surgery Anastomotic stenosis as a cause of parenchymal infarction is a significant issue following lung transplantation, partly because of an absent collateral bronchial artery circulation [98]. This may occur not just secondary to a thrombotic event but also due to progressive anastomotic narrowing due to fibrosis, with the latter seen in as many as 4% of lung transplants (Fig. 18) [99]. Pulmonary venous infarction is a further, life-threatening complication following lung transplantation, producing segmental and/or lobar opacities on chest radiography and a dramatic reduction of perfusion on scintigraphy [100]. Pulmonary infarction is also a known complication of pneumonectomy [101].
4. Extraluminal aetiology 4.1. Torsion Pulmonary torsion is rare post-surgical complication that has primarily been described after pulmonary lobectomy (in up to 0.4%) [51,102], but this severe complication should be remembered as an unusual cause of pulmonary infarction, not just following lobectomy but also after other thoracic surgery, trauma or even as a spontaneous event [103–105]. Rotation of the bronchovascular pedicle results in dramatic vascular compromise with haemorrhagic lobar infarction due to a simultaneous pulmonary arterial and venous insult. Often, a collapsed or consolidated lobe is seen in an unusual position; other imaging features include ground glass attenuation, interlobular septal thickening, and poor pulmonary vascular enhancement (Fig. 19) [105,106]. 4.2. Compression The pulmonary arteries can also be compressed by nearby structures, including aortic aneurysms (Fig. 20) and soft tissue masses such as lymphoma [107]. Neoplasm may similarly occlude pulmonary arteries (as described above) [108]. In these cases it can be difficult to differentiate areas of infarction from metastases, though serial imaging and MRI may be helpful to distinguish. Pulmonary venous infarction due to fibrosing mediastinitis is another recognised cause of extrinsic compression, albeit rare [10]. Overall, compressive aetiology of infarcts is not common due to rapidity of the compression often allowing time for the bronchial artery collateral supply to adapt.
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
2251
Fig. 18. 42-year-old female with progressive dyspnoea 3 months after pulmonary vein ablation for paroxysmal atrial fibrillation. (a) A chest radiograph shows marked asymmetry in the proximal pulmonary vasculature (block arrow) and an ill-defined peripheral consolidation in the left lower zone (thin arrow). (b) Mediastinal (top images) and lung (bottom images) windows from a contrast enhanced CT examination. There are homogenous filing defects in the left superior and inferior pulmonary veins (block arrows) with peripheral infarcts (thin arrows). Also note the thin rim of pleural fluid on the left and good opacification of right superior pulmonary vein (notched arrows, upper panels). (c) Ventilation perfusion scintigraphy demonstrates complete absence of perfusion in the left lung. (d) Catheter pulmonary venography illustrates a proximal thrombus in the stenosed left superior pulmonary vein (block arrow; left image) and establishment of flow after removal of the thrombus and placement of a stent (notched arrow, right image).
2252
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
Fig. 19. Right upper lobe torsion in a 57-year-old woman who had undergone right middle lobectomy for bronchioloalveolar carcinoma. The post-surgical rotation of the bronchovascular pedicle resulted in vascular compromise and haemorrhagic lobar infarction. (a) 1 day postoperative chest radiograph (left image) shows satisfactory appearances with good expansion of the remaining lung on the right. Chest radiograph 4 days later (right image) demonstrates rapidly developing consolidation with a bulging neo-fissure (arrow), indicating volume increase within the right upper lobe. (b) Contrast enhanced CT images (top panel: axial images on lung and mediastinal windows; bottom panel: coronal reformats using maximum intensity projection and volume rendered technique). There is consolidation with increased volume (thin white arrow points to bulging fissure), ground-glass attenuation, interlobular septal thickening (notched black arrow), and poor parenchymal and pulmonary vascular enhancement (black white arrow points to the tapered obliteration of the proximal pulmonary artery and vein) in the torted upper lobe. Also note the thickening around the bronchus on the right side (white arrowhead).
Fig. 20. Coronal contrast enhanced CT (left image) demonstrates severe extrinsic compression of the main pulmonary artery (thin arrow by a large ascending aortic aneurysm (AA)). Also note the overall reduced perfusion of the right lung. There were resultant multifocal infarcts within the right lung, one of which is illustrated (black arrow; right image).
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
5. Conclusion Pulmonary infarctions have a multifactorial aetiology. The classical imaging signs may not always be present and the detection of supplementary features can be instrumental in narrowing the differential diagnosis. Contrast enhanced CT is the mainstay in identifying the presence and cause of pulmonary infarction but a multimodality imaging approach is often necessary. As an ancillary imaging modality, MRI is an especially valuable problem-solving tool with roles for nuclear imaging and ultrasound predominantly confined to certain patients subsets. Disclosure The authors have nothing to disclose. Conflict of interest The authors declare that there are no conflict of interest. Acknowledgements K.H.M. has research support from the NIHR Cambridge Biomedical Research Centre. Histological images (Figs. 1 and 2) have been contributed by Dr Doris Rassl, Consultant Histopathologist, Papworth Hospital NHS Foundation Trust, Cambridge, United Kingdom. References [1] Yousem SA. The surgical pathology of pulmonary infarcts: diagnostic confusion with granulomatous disease, vasculitis, and neoplasia. Mod Pathol 2009;22:679–85. [2] Tsao MS, Schraufnagel D, Wang NS. Pathogenesis of pulmonary infarction. Am J Med 1982;72:599–606. [3] Nelson E, Klein JS. Pulmonary infarction resulting from metastatic osteogenic sarcoma with pulmonary venous tumor thrombus. AJR – Am J Roentgenol 2000;174:531–3. [4] Braun S, Platzek I, Zophel K, et al. Haemoptysis due to pulmonary venous stenosis. Eur Respir Rev: Off J Eur Respir Soc 2014;23:170–9. [5] Frazier AA, Galvin JR, Franks TJ, Rosado-De-Christenson ML. From the archives of the AFIP: pulmonary vasculature: hypertension and infarction. Radiographics 2000;20:491–524, quiz 530-491, 532. [6] Hampton AO, Castleman B. Correlation of postmortem chest teleroentgenograms with autopsy findings. Am J Roentgenol Radium Ther 1940;34:305–26. [7] Smith MJ. Roentgenographic aspects of complete and incomplete pulmonary infarction. Dis Chest 1953;23:532–46. [8] Webb WR, Higgins CB. Thoracic imaging: pulmonary and cardiovascular radiology, North American edition. Philadelphia: Lippincott Williams & Wilkins; 2012. [9] Gleeson TG, Cheyne I, English JC, Quadri SM, Leipsic JA. A 50-year-old woman with a history of atrial fibrillation presents with acute dyspnea and pleuritic chest pain. Chest 2010;137:1225–30. [10] Williamson WA, Tronic BS, Levitan N, Webb-Johnson DC, Shahian DM, Ellis Jr FH. Pulmonary venous infarction. Chest 1992;102:937–40. [11] Balakrishnan J, Meziane MA, Siegelman SS, Fishman EK. Pulmonary infarction: CT appearance with pathologic correlation. J Comput Assist Tomogr 1989;13:941–5. [12] He H, Stein MW, Zalta B, Haramati LB. Pulmonary infarction: spectrum of findings on multidetector helical CT. J Thorac Imaging 2006;21:1–7. [13] Johnson PT, Wechsler RJ, Salazar AM, Fisher AM, Nazarian LN, Steiner RM. Spiral CT of acute pulmonary thromboembolism: evaluation of pleuroparenchymal abnormalities. J Comput Assist Tomogr 1999;23:369–73. [14] Revel MP, Triki R, Chatellier G, et al. Is it possible to recognize pulmonary infarction on multisection CT images? Radiology 2007;244:875–82. [15] Grabowski GA, Leslie N, Wenstrup R. Enzyme therapy for gaucher disease: the first 5 years. Blood Rev 1998;12:115–33. [16] Harris H, Barraclough R, Davies C, Armstrong I, Kiely DG, van Beek Jr E. Cavitating lung lesions in chronic thromboembolic pulmonary hypertension. J Radiol Case Rep 2008;2:11–21. [17] Ye R, Zhao L, Wang C, Wu X, Yan H. Clinical characteristics of septic pulmonary embolism in adults: a systematic review. Respir Med 2014;108:1–8. [18] Hall FM, Salzman EW, Ellis I, Kurland GS. Pneumothorax complicating aseptic cavitating pulmonary infarction. Chest 1977;72:232–4.
2253
[19] Narita JI, Ito S, Terada M, et al. Pulmonary artery involvement in takayasu’s arteritis with lung infarction and pulmonary aspergillosis. J Clin Rheumatol 2002;8:260–4. [20] Wick MR, Ritter JH, Schuller D. Ruptured pulmonary infarction: a rare, fatal complication of thromboembolic disease. Mayo Clin Proc 2000;75:639–42. [21] Geyer LL, Scherr M, Korner M, et al. Imaging of acute pulmonary embolism using a dual energy CT system with rapid KVP switching: initial results. Eur J Radiol 2012;81:3711–8. [22] Hansmann J, Apfaltrer P, Zoellner FG, et al. Correlation analysis of dual-energy CT iodine maps with quantitative pulmonary perfusion MRI. World J Radiol 2013;5:202–7. [23] Johnson TR, Krauss B, Sedlmair M, et al. Material differentiation by dual energy CT: initial experience. Eur Radiol 2007;17:1510–7. [24] Thieme SF, Graute V, Nikolaou K, et al. Dual energy CT lung perfusion imaging—correlation with spect/CT. Eur J Radiol 2012;81:360–5. [25] Chai X, Zhang LJ, Yeh BM, Zhao YE, Hu XB, Lu GM. Acute and subacute dual energy CT findings of pulmonary embolism in rabbits: correlation with histopathology. Br J Radiol 2012;85:613–22. [26] van Beek EJ, Hoffman EA. Functional imaging: CT and MRI. Clin Chest Med 2008;29:195–216, vii. [27] Huber DJ, Kobzik L, Solorzano C, Melanson G, Adams DF. Nuclear magnetic resonance spectroscopy of acute and evolving pulmonary hemorrhage. An in vitro study. Invest Radiol 1987;22:632–7. [28] Kessler R, Fraisse P, Krause D, Veillon F, Vandevenne A. Magnetic resonance imaging in the diagnosis of pulmonary infarction. Chest 1991;99:298–300. [29] Iwasaki Y, Nagata K, Nakanishi M, et al. Spiral CT findings in septic pulmonary emboli. Eur J Radiol 2001;37:190–4. [30] Kuhlman JE, Fishman EK, Teigen C. Pulmonary septic emboli: diagnosis with CT. Radiology 1990;174:211–3. [31] Puderbach M, Hintze C, Ley S, Eichinger M, Kauczor HU, Biederer J. Mr imaging of the chest: a practical approach at 1.5 t. Eur J Radiol 2007;64:345–55. [32] Barreto MM, Rafful PP, Rodrigues RS, et al. Correlation between computed tomographic and magnetic resonance imaging findings of parenchymal lung diseases. Eur J Radiol 2013;82:e492–501. [33] Kluge A, Mueller C, Strunk J, Lange U, Bachmann G. Experience in 207 combined MRI examinations for acute pulmonary embolism and deep vein thrombosis. AJR – Am J Roentgenol 2006;186:1686–96. [34] Mathis G, Metzler J, Fussenegger D, Sutterlutti G, Feurstein M, Fritzsche H. Sonographic observation of pulmonary infarction and early infarctions by pulmonary embolism. Eur Heart J 1993;14:804–8. [35] Reissig A, Heyne JP, Kroegel C. Sonography of lung and pleura in pulmonary embolism: sonomorphologic characterization and comparison with spiral CT scanning. Chest 2001;120:1977–83. [36] Mathis G, Dirschmid K. Pulmonary infarction: sonographic appearance with pathologic correlation. Eur J Radiol 1993;17:170–4. [37] Gorg C, Bert T, Gorg K. Contrast-enhanced sonography for differential diagnosis of pleurisy and focal pleural lesions of unknown cause. Chest 2005;128:3894–9. [38] Kreuter M, Mathis G. Emergency ultrasound of the chest. Respiration 2014;87:89–97. [39] McGoldrick PJ, Rudd TG, Figley MM, Wilhelm JP. What becomes of pulmonary infarcts? AJR – Am J Roentgenol 1979;133:1039–45. [40] Worsley DF, Alavi A, Aronchick JM, Chen JT, Greenspan RH, Ravin CE. Chest radiographic findings in patients with acute pulmonary embolism: observations from the pioped study. Radiology 1993;189:133–6. [41] Worsley DF, Kim CK, Alavi A, Palevsky HI. Detailed analysis of patients with matched ventilation–perfusion defects and chest radiographic opacities. J Nucl Med: Off Publ Soc Nucl Med 1993;34:1851–3. [42] Gottschalk A, Stein PD, Henry JW, Relyea B. Matched ventilation, perfusion and chest radiographic abnormalities in acute pulmonary embolism. J Nucl Med: Off Publ Soc Nucl Med 1996;37:1636–8. [43] Kim CK, Worsley DF, Alavi A. Ventilation–perfusion-chest radiography match is less likely to represent pulmonary embolism if perfusion is decreased rather than absent. Clin Nucl Med 2000;25:665–9. [44] Zaki M, Suga K, Kawakami Y, et al. Preferential location of acute pulmonary thromboembolism induced consolidative opacities: assessment with respiratory gated perfusion spect-CT fusion images. Nucl Med Commun 2005;26:465–74. [45] Suga K. Pulmonary function-morphologic relationships assessed by spect-CT fusion images. Ann Nucl Med 2012;26:298–310. [46] Goethals I, Smeets P, De Winter O, Noens L. Focally enhanced f-18 fluorodeoxyglucose (FDG) uptake in incidentally detected pulmonary embolism on PET/CT scanning. Clin Nucl Med 2006;31:497–8. [47] Soussan M, Rust E, Pop G, Morere JF, Brillet PY, Eder V. The rim sign: FDGPET/CT pattern of pulmonary infarction. Insights Imaging 2012;3:629–33. [48] Badr A, Joyce JM, Durick J. Rim of FDG uptake around a pulmonary infarct on PET/CT in a patient with unsuspected pulmonary embolism. Clin Nucl Med 2009;34:285–6. [49] Kamel EM, McKee TA, Calcagni ML, et al. Occult lung infarction may induce false interpretation of 18F-FDG PET in primary staging of pulmonary malignancies. Eur J Nucl Med Mol Imaging 2005;32:641–6. [50] Wittram C, Scott JA. 18F-FDG PET of pulmonary embolism. AJR – Am J Roentgenol 2007;189:171–6. [51] Parambil JG, Savci CD, Tazelaar HD, Ryu JH. Causes and presenting features of pulmonary infarctions in 43 cases identified by surgical lung biopsy. Chest 2005;127:1178–83.
2254
T.J.P. Bray et al. / European Journal of Radiology 83 (2014) 2240–2254
[52] Mandelli V, Schmid C, Zogno C, Morpurgo M. False negatives and false positives in acute pulmonary embolism: a clinical-postmortem comparison. Cardiologia (Rome, Italy) 1997;42:205–10. [53] Karabulut N, Kiroglu Y. Relationship of parenchymal and pleural abnormalities with acute pulmonary embolism: CT findings in patients with and without embolism. Diagn Interv Radiol 2008;14:189–96. [54] Schraufnagel DE, Ming-Soung T, Yao YT, Nai-San W. Factors associated with pulmonary infarction. A discriminant analysis study. Am J Clin Pathol 1985;84:15–8. [55] Shah AA, Davis SD, Gamsu G, Intriere L. Parenchymal and pleural findings in patients with and patients without acute pulmonary embolism detected at spiral CT. Radiology 1999;211:147–53. [56] Morgenthaler TI, Ryu JH, Utz JP. Cavitary pulmonary infarct in immunocompromised hosts. Mayo Clin Proc 1995;70:66–8. [57] Wilson AG, Joseph AE, Butland RJ. The radiology of aseptic cavitation in pulmonary infarction. Clin Radiol 1986;37:327–33. [58] Libby LS, King TE, LaForce FM, Schwarz MI. Pulmonary cavitation following pulmonary infarction. Medicine (Baltimore) 1985;64:342–8. [59] Cannon JW. Necrotizing clostridial pneumonia: a case report and review of the literature. Mil Med 2002;167:85–6. [60] Mortensen KH, Gopalan D, Balan A. Atrial masses on multidetector computed tomography. Clin Radiol 2013;68:e164–75. [61] Marriott AE, Weisbrod G. Bronchogenic carcinoma associated with pulmonary infarction. Radiology 1982;145:593–7. [62] Nomori H, Horio H, Morinaga S, Suemasu K. Multiple pulmonary infarctions associated with lung cancer. Jpn J Clin Oncol 2000;30:40–2. [63] Ishiguro T, Kasahara K, Matsumoto I, et al. Primary pulmonary artery sarcoma detected with a pulmonary infarction. Intern Med (Tokyo, Japan) 2007;46:601–4. [64] Takahashi M, Murakami Y, Nitta N, et al. Pulmonary infarction associated with bronchogenic carcinoma. Radiat Med 2008;26:76–80. [65] Engelke C, Schaefer-Prokop C, Schirg E, Freihorst J, Grubnic S, Prokop M. Highresolution CT and CT angiography of peripheral pulmonary vascular disorders. Radiographics 2002;22:739–64. [66] Cox JE, Chiles C, Aquino SL, Savage P, Oaks T. Pulmonary artery sarcomas: a review of clinical and radiologic features. J Comput Assist Tomogr 1997;21:750–5. [67] Rodrigues ACT, Kay FU, Ogawa A, et al. Pulmonary complications in right sided endocarditis. Circulation 2013:128. [68] Gadkowski LB, Stout JE. Cavitary pulmonary disease. Clin Microbiol Rev 2008;21:305–33, table of contents. [69] Han D, Lee KS, Franquet T, et al. Thrombotic and nonthrombotic pulmonary arterial embolism: spectrum of imaging findings. Radiographics 2003;23:1521–39. [70] Huang RM, Naidich DP, Lubat E, Schinella R, Garay SM, McCauley DI. Septic pulmonary emboli: CT-radiographic correlation. AJR – Am J Roentgenol 1989;153:41–5. [71] Kwon WJ, Jeong YJ, Kim KI, et al. Computed tomographic features of pulmonary septic emboli: comparison of causative microorganisms. J Comput Assist Tomogr 2007;31:390–4. [72] Rajagopala S, Devaraj U, D’Souza G. Infected cavitating pulmonary infarction. Respir Care 2011;56:707–9. [73] Scharf J, Nahir AM, Munk J, Lichtig C. Aseptic cavitation in pulmonary infarction. Chest 1971;59:456–8. [74] Djordjevic I, Pejcic T, Nastasijevic-Borovac D, Radjenovic-Petkovic T. Cavitary pulmonary infarct or necrotizing pneumonia: the differential diagnostic dilemma. Chest 2012:142. [75] Denning DW. Invasive aspergillosis. Clin Infect Dis: Off Publ Infect Dis Soc Am 1998;26:781–803, quiz 804-785. [76] Brown MJ, Worthy SA, Flint JD, Muller NL. Invasive aspergillosis in the immunocompromised host: utility of computed tomography and bronchoalveolar lavage. Clin Radiol 1998;53:255–7. [77] Blum U, Windfuhr M, Buitrago-Tellez C, Sigmund G, Herbst EW, Langer M. Invasive pulmonary aspergillosis. MRI, CT, and plain radiographic findings and their contribution for early diagnosis. Chest 1994;106: 1156–61. [78] Won HJ, Lee KS, Cheon JE, et al. Invasive pulmonary aspergillosis: prediction at thin-section CT in patients with neutropenia—a prospective study. Radiology 1998;208:777–82. [79] Horger M, Hebart H, Einsele H, et al. Initial CT manifestations of invasive pulmonary aspergillosis in 45 non-HIV immunocompromised patients: association with patient outcome? Eur J Radiol 2005;55:437–44.
[80] Franquet T, Muller NL, Gimenez A, Guembe P, de La Torre J, Bague S. Spectrum of pulmonary aspergillosis: histologic, clinical, and radiologic findings. Radiographics 2001;21:825–37. [81] Logan PM, Muller NL. CT manifestations of pulmonary aspergillosis. Crit Rev Diagn Imaging 1996;37:1–37. [82] Franquet T, Muller NL, Gimenez A, Domingo P, Plaza V, Bordes R. Semiinvasive pulmonary aspergillosis in chronic obstructive pulmonary disease: radiologic and pathologic findings in nine patients. AJR – Am J Roentgenol 2000;174:51–6. [83] Kim SY, Lee KS, Han J, et al. Semiinvasive pulmonary aspergillosis: CT and pathologic findings in six patients. AJR – Am J Roentgenol 2000;174: 795–8. [84] Montagnana M, Cervellin G, Franchini M, Lippi G. Pathophysiology, clinics and diagnostics of non-thrombotic pulmonary embolism. J Thromb Thrombolysis 2011;31:436–44. [85] Jorens PG, Van Marck E, Snoeckx A, Parizel PM. Nonthrombotic pulmonary embolism. Eur Respir J 2009;34:452–74. [86] Bach AG, Restrepo CS, Abbas J, et al. Imaging of nonthrombotic pulmonary embolism: biological materials, nonbiological materials, and foreign bodies. Eur J Radiol 2013;82:e120–41. [87] Vichinsky EP, Neumayr LD, Earles AN, et al. Causes and outcomes of the acute chest syndrome in sickle cell disease. National acute chest syndrome study group. N Engl J Med 2000;342:1855–65. [88] Hawley DA, McCarthy LJ. Sickle cell disease: two fatalities due to bone marrow emboli in patients with acute chest syndrome. Am J Forensic Med Pathol 2009;30:69–71. [89] Brown KK. Pulmonary vasculitis. Proc Am Thorac Soc 2006;3:48–57. [90] Castaner E, Alguersuari A, Gallardo X, et al. When to suspect pulmonary vasculitis: radiologic and clinical clues. Radiographics 2010;30:33–53. [91] Yataco J, Stoller JK. Pulmonary venous thrombosis and infarction complicating pulmonary venous stenosis following radiofrequency ablation. Respir Care 2004;49:1525–7. [92] Duehmke RM, Fynn SP, Gopalan D. Pulmonary venous infarction following pulmonary vein isolation. Eur Heart J 2008;29:2893. [93] Lee H, Wanless A. Pulmonary vein thrombosis: a case report and literature review. J Hosp Med 2011;6:S207–8. [94] Frazier AA, Franks TJ, Mohammed TL, Ozbudak IH, Galvin JR. From the archives of the AFIP: pulmonary veno-occlusive disease and pulmonary capillary hemangiomatosis. Radiographics 2007;27:867–82. [95] Reinke RT, Higgins CB. Pulmonary infarction complicating the use of swanganz catheters. Br J Radiol 1975;48:885–8. [96] McLoud TC, Putman CE. Radiology of the swan-ganz catheter and associated pulmonary complications. Radiology 1975;116:19–22. [97] Shackles C, Rundback JH, Herman K. Pulmonary infarct after transarterial chemoembolization. J Vasc Interv Radiol – JVIR 2013;24:1928–30. [98] Krivokuca I, van de Graaf EA, van Kessel DA, van den Bosch JM, Grutters JC, Kwakkel-van Erp JM. Pulmonary embolism and pulmonary infarction after lung transplantation. Clin Appl Thromb Hemost 2011;17:421–4. [99] Clark SC, Levine AJ, Hasan A, Hilton CJ, Forty J, Dark JH. Vascular complications of lung transplantation. Ann Thorac Surg 1996;61:1079–82. [100] Choong CK, Hu DZ, Huddleston CB. Pulmonary segmental venous infarction after living-donor lobar lung transplantation. J Thorac Cardiovasc Surg 2005;130:919–21. [101] Haraguchi S, Koizumi K, Hirata T, et al. Surgical results of completion pneumonectomy. Ann Thorac Cardiovasc Surg 2011;17:24–8. [102] Cable DG, Deschamps C, Allen MS, et al. Lobar torsion after pulmonary resection: presentation and outcome. J Thorac Cardiovasc Surg 2001;122:1091–3. [103] Niekel MC, Horsch AD, Ven M, Reijnen MM, Joosten FB. Right middle lobe torsion: evaluation with CT angiography. Emerg Radiol 2009;16: 387–9. [104] Oliveira C, Zamakhshary M, Abdallah MR, et al. Lung torsion after tracheoesophageal fistula repair: a case report and review of literature. J Pediatr Surg 2007;42:E5–9. [105] Felson B. Lung torsion: radiographic findings in nine cases. Radiology 1987;162:631–8. [106] Moser Jr ES, Proto AV. Lung torsion: case report and literature review. Radiology 1987;162:639–43. [107] Robinson T, Lynch J, Grech E. Non-hodgkin’s lymphoma causing extrinsic pulmonary artery compression. Eur J Echocardiogr 2008;9:577–8. [108] Held BT, Siegelman SS. Pulmonary infarction secondary to bronchogenic carcinoma. Am J Roentgenol Radium Ther Nucl Med 1974;120:145–50.