Radiologic findings of drug-induced lung disease

Radiologic findings of drug-induced lung disease Poster No.: P-0115 Congress: ESTI 2015 Type: Educational Poster Authors: A. I. C. Santos, A. F. Roque, R. Mamede, L. Oliveira, T. Saldanha; Lisbon/PT Keywords: Toxicity, Drugs / Reactions, Diagnostic procedure, CT-High Resolution, CT, Conventional radiography, Thorax, Lung DOI: 10.1594/esti2015/P-0115 Any information contained in this pdf file is automatically generated from digital material submitted to EPOS by third parties in the form of scientific presentations. References to any names, marks, products, or services of third parties or hypertext links to thirdparty sites or information are provided solely as a convenience to you and do not in any way constitute or imply ECR's endorsement, sponsorship or recommendation of the third party, information, product or service. ECR is not responsible for the content of these pages and does not make any representations regarding the content or accuracy of material in this file. As per copyright regulations, any unauthorised use of the material or parts thereof as well as commercial reproduction or multiple distribution by any traditional or electronically based reproduction/publication method ist strictly prohibited. You agree to defend, indemnify, and hold ECR harmless from and against any and all claims, damages, costs, and expenses, including attorneys' fees, arising from or related to your use of these pages. Please note: Links to movies, ppt slideshows and any other multimedia files are not available in the pdf version of presentations. www.myesr.org Page 1 of 28

Learning objectives To describe and illustrate the most common chest radiography (CR) and Computed Tomography (CT) manifestations of drug-induced lung disease (DILD) and the most frequently drugs involved; To discuss the correlation between radiologic findings of DILD and their histopathological patterns; To review its pathogenesis, presentation, diagnostic approach, management and prognosis. Background Drug-induced lung toxicity is an increasingly frequent cause of morbidity and it is becoming more commonly diagnosed as a cause of acute and chronic lung disease. More than 380 medications are now recognized as being implicated in respiratory diseases of various types, involving the airways, lung parenchyma, mediastinum, pleura, pulmonary vasculature, and/or the neuromuscular system. The most common form of pulmonary drug toxicity is drug-induced interstitial lung disease. The exact frequency of DILD is unknown, although is estimated that 2.5-3% of cases of interstitial lung disease are drug-induced. Causative agents DILD can be caused by several therapeutic agents, including: cytotoxic agents; non-cytotoxic drugs, such as antibiotics, antiarrhythmic and immunosuppressive agents. Cytotoxic drugs constitute the most important group associated with DILD, with examples in this therapeutic class such as bleomycin, busulfan, cyclosphophamide, carmustine and methotrexate. However, with the recent advances in cancer chemotherapy, there are now several molecularly targeted agents that have different mechanisms of action Page 2 of 28

compared to standard cytotoxic agents and their toxicity is also different in both clinical and radiological manifestations. Some of the most frequently encountered non-cytotoxic drugs that cause pulmonary toxicity are nitrofurantoin, amiodarone, sulfasalazine and aspirin. Pathogenesis Even though all types of immunological reactions have been described, there are two main mechanisms responsible for DILD, which may be implicated independently or in combination: direct, cytotoxic injury to pneumocytes or the alveolar capillary endothelium; immune-mediated, mostly T cell-mediated. Direct effects consist of a toxic reaction of the drug or one of its metabolites causing lung injury through the release of free oxygen radicals, reduction in deactivation of metabolites of the lung or impairment of alveolar repair mechanisms, with subsequent release of cytokines and recruitment of inflammatory cells. These reactions lead to an inflammatory process with alveolitis and edema, which can resolve or may progress to chronic inflammation and eventually lead to fibrosis (Fig. 1 on page 7). Immune mechanisms may also activate an inflammatory response leading to pulmonary edema and interstitial lung disease. An immune cascade is created after contact to a particular drug in some individuals. For instance, drugs are recognized as potential antigens (or haptens) and bind to drug-specific T cells or drug-specific antibodies causing the deposition of immunogenic complexes that ultimately can lead to immune-mediated lung toxicity (Fig. 1 on page 7). Page 3 of 28

Fig. 1: Mechanisms responsible for DILD. References: A. I. C. Santos et al. ESTI 2015 EPOS(TM) These immune reactions are influenced by various factors (Fig. 2 on page 8). In fact, the person-to-person variability in drug response is multifactorial. Genetic and intrinsic factors affecting the disposition of certain drugs (absorption, distribution, metabolism, and excretion), as well as extrinsic factors (environmental aspects) can explain why some patients display high susceptibility while others don't. Page 4 of 28

References: A. I. C. Santos et al. ESTI 2015 EPOS(TM) Multiple routes of drug administration have been implicated in DILD, but oral and parenteral are the most commonly referred. Risk factors Although the likelihood of developing DILD is unpredictable and idiosyncratic, there are several factors known to increase the risk of DILD (Table 1 on page 9). Table 1: Risk factors for DILD. Table created from data in Schwaiblmair, M. et al. (2012) Drug Induced Interstitial Lung Disease. Open Respir Med J. 6: 63-74 References: A. I. C. Santos et al. ESTI 2015 EPOS(TM) Diagnosis, management and prognosis There is no specific clinical, radiologic or histological pattern of DILD, so the diagnosis is often difficult and essentially depends on a definite temporal association between an exposure to a causative agent known to result in lung disease and the development of pulmonary abnormalities. Page 6 of 28

A detailed recent drug history is fundamental and other lung injury causes, such as infections, malignancy, pulmonary edema or connective tissue disease, should be excluded. In this setting, the suspicion of DILD rises if there is association between: nonspecific signs and symptoms, such as dry cough, progressive dyspnea, fever, pleuritic chest pain, hypoxemia; a restrictive pattern and decreased diffusing capacity for carbon monoxide in pulmonary function tests; suggestive radiological findings. Physical and laboratory tests are also nonspecific; however, bronchoscopy with bronchoalveolar lavage can be useful in this context. The treatment consists in discontinuation of the drug, avoidance of further exposure and systemic corticosteroids in more severe cases. DILD prompt recognition is important since it may completely resolve if the drug is discontinued or appropriate therapy is instituted in early stages. Besides, the identification of the causative agent is indispensable to avoid secondary reactions. Failure to recognize DILD can lead to pulmonary fibrosis and respiratory failure requiring mechanical ventilation, causing significant morbidity and mortality. Images for this section: Page 7 of 28

Fig. 1: Mechanisms responsible for DILD. Page 8 of 28

Fig. 2: Factors determining immune-mediated DILD. Figure created from data in Matsuno, O. (2012) Drug-induced interstitial lung disease: mechanisms and best diagnostic approaches. Respir Res 13:39 Table 1: Risk factors for DILD. Table created from data in Schwaiblmair, M. et al. (2012) Drug Induced Interstitial Lung Disease. Open Respir Med J. 6: 63-74 Page 9 of 28

Imaging findings OR Procedure details DILD has an unpredictable onset and time course and may be mild to progressive, manifesting with diverse patterns ranging from benign infiltrates to the potentially fatal acute respiratory distress syndrome (ARDS). The radiologic features of DILD, although heterogeneous and nonspecific, usually reflect the elemental histopathologic lesions. Therefore, it is reasonable to approach the radiological manifestations of DILD based on the underlying histological pattern. Principal presentations include: Diffuse alveolar damage (DAD) Chronic interstitial pneumonitis with fibrosis (comprising usual interstitial pneumonia, UIP, and nonspecific interstitial pneumonia, NSIP) Organizing pneumonia (OP) Eosinophilic pneumonia (EP) and hypersensivity reactions Pulmonary hemorrhage (PH) Pulmonary edema (PE). Each pattern typically is related to a different group of drugs, but the same therapeutic agent can cause multiple manifestations. CT is currently the best non-invasive method to evaluate the presence of DILD and is recommended on its first suspicion. This technique allows a precise assessment of the presence, pattern and distribution of parenchymal and airway abnormalities with a higher sensitivity than CR. In fact, CT may reveal abnormalities in symptomatic patients with a normal CR, especially in early stages of disease. It also identifies findings suggesting an alternative diagnosis and has the potential to monitor therapeutic response. Diffuse alveolar damage DAD with ARDS is a common manifestation of DILD and can be induced by several agents, most typically cytotoxic drugs, particularly cyclophosphamide, busulfan, bleomycin and carmustine, as well as nitrofurantoin and amiodarone. The onset is generally sudden, within a few days after starting chemotherapy. DAD reflects necrosis of type II pneumocytes and alveolar endothelial cells and histopathologically can be divided into: Page 10 of 28

an acute or exudative phase (alveolar and interstitial edema, hemorrhage and hyaline membrane formation - Fig. 3 on page 18; in the first week after lung injury); a late reparative or proliferative phase (cellular hyperplasia and fibrosis; after 1 or 2 weeks). DAD presents radiographically as extensive bilateral patchy or homogeneous opacities involving middle and lower lung zones and on CT as scattered or diffuse areas of ground-glass opacity predominantly in the dependent lung regions (Fig. 4 on page 18). Depending on the severity of the lung injury, abnormalities can regress, remain stable, or progress to fibrosis. In early stages, these manifestations may not be apparent on CR, but as the disease progresses, marked architectural distortion and honeycombing can take place and be detected. Fig. 4: DAD in a patient with multiple myeloma treated with bortezomib (proteasome inhibitor) who presented with ARDS. DAD has been considered a characteristic side Page 11 of 28

effect of bortezomib. A-B: CR showing rapid progression of DAD (A. was performed 24h before B.) with extensive, bilateral and diffuse ill-defined opacities and air-space consolidation. C: CT shows diffuse areas of ground-glass opacity and consolidation mainly basal and in dependent regions. D: Follow-up CT 9 months after systemic corticotherapy revealed significant regression of lesions. References: Department of Radiology, Centro Hospitalar de Lisboa Ocidental, E.P.E./ Portugal 2015 Chronic interstitial pneumonitis with fibrosis Both UIP and NSIP are associated with cytotoxic chemotherapeutic agents such as bleomycin, busulfan, methotrexate, doxorubicin and carmustine, but can also be caused by noncytotoxic drugs, most commonly nitrofurantoin and amiodarone. NSIP is one of the most common forms of drug-associated pneumonitis, usually occuring within several months after initiating therapy and corresponding histologically to interstitial fibrosis and infiltration of the alveolar septa and the peribronchial spaces by lymphocytes, lymphoid aggregates and plasma cells. NSIP is typically more homogeneous and more cellular (Fig. 5 on page 19) than that seen in cases of UIP (Fig. 6) and include a cellular and a fibrotic pattern. Radiologic findings in patients with chronic pneumonitis and fibrosis are bilateral and symmetric, with a basal distribution, frequently with a peripheral and subpleural distribution. CR typically show diffuse heterogeneous opacities, with CT most frequent pattern comprising irregular reticular opacities, honeycombing (Fig. 7 on page 20), architectural distortion, traction bronchiectasis (Fig. 8 on page 21) and eventually consolidation. Early CT scans show scattered or diffuse ground-glass opacity. Page 12 of 28

Fig. 7: UIP caused by amiodarone. A: CR shows bilateral, diffuse heterogeneous opacities with a basal distribution; B: CT with fine reticular opacities and honeycombing visible at lung bases in peripheral, subpleural regions. References: Department of Radiology, Centro Hospitalar de Lisboa Ocidental, E.P.E./ Portugal 2015 Fig. 8: Fibrotic NSIP in a patient treated with methotrexate: bilateral reticular opacities and ground-glass opacity, as well as architectural distortion and traction bronchiectasis indicating fibrosis. NSIP is the most common manifestation of methotrexate-induced lung disease. References: Department of Radiology, Centro Hospitalar de Lisboa Ocidental, E.P.E./ Portugal 2015 Organizing pneumonia OP (also known as bronchiolitis obliterans organizing pneumonia, BOOP) has been reported most frequently in association with bleomycin, cyclophosphamide, gold salts and methotrexate. Amiodarone, nitrofurantoin, penicillamine, and sulfasalazine represent Page 13 of 28

less common causes. It is characterized by proliferation of immature fibroblastic plugs within the respiratory bronchioles, alveolar ducts and alveoli (Fig. 9 on page 22). There is no traction bronchiectasis and no histological honeycombing. OP responds well to discontinuation of the causative agent. CR demonstrates bilateral, scattered, hetero- and homogeneous peripheral opacities equally distributed between upper and lower lobes. CT often shows consolidation or ground-glass opacity that may have a patchy or nodular distribution and may predominate in a peribronchial or subpleural location (Fig. 10 on page 22). Lung nodules or masses, which may be irregular in shape, eventually associated with the "atoll sign", may be present. Fig. 10: CT findings of OP in a patient treated with nitrofurantoin: patchy ground-glass opacities with peribronchial distribution; poorly defined nodular areas of consolidation associated with centrilobular nodules and branching linear opacities ("tree-in-bud opacities"), predominating in subpleural location, and bronchial dilatation. References: Department of Radiology, Centro Hospitalar de Lisboa Ocidental, E.P.E./ Portugal 2015 Page 14 of 28

Eosinophilic pneumonia and hypersensivity reactions Hypersensitivity reactions may have features of simple eosinophilia, chronic eosinophilic pneumonia or acute EP and can result from a diversity of drugs. The most common are methotrexate, nitrofurantoin, penicillamine, sulfasalazine, cyclophosphamide and nonsteroidal anti-inflammatory drugs. Peripheral eosinophilia is present in up to 40% and elevated IgE levels are common. EP is characterized histologically by the accumulation of eosinophils and macrophages in the alveolar airspaces. Alveolar septa are thickened and infiltrated by eosinophils, lymphocytes and plasma cells. CR and CT demonstrate bilateral patchy areas of consolidation or ground-glass opacity typically involving the peripheral lung regions and the upper lobes, which may be chronic or relatively acute, transient and fleeting (Fig. 11 on page 23). Gradual-onset pulmonary fibrosis, although rare, can develop with features of NSIP or UIP with fibrosis. Fig. 11: Azathioprine-induced EP in a patient with ulcerative colitis. CT shows peripheral areas of alveolar consolidation, mostly in the right upper lobe. Page 15 of 28

References: Department of Radiology, Centro Hospitalar de Lisboa Ocidental, E.P.E./ Portugal 2015 Pulmonary hemorrhage Diffuse PH (Fig. 12 on page 24) has potentially significant morbidity and mortality and is induced by anticoagulants, high-dose cyclophosphamide, amphotericin B and penicillamine. Patients may present with ARDS or hemoptysis. CR findings are bilateral patchy opacities and less commonly focal consolidation; CT shows bilateral, scattered, or diffuse areas of ground-glass opacity or consolidation (Fig. 13 on page 25). Pleural effusion characteristically does not occur. Page 16 of 28

Fig. 13: CT showing bilateral, scattered, areas of ground-glass opacity in a patient with PH induced by cyclophosphamide. References: Department of Radiology, Centro Hospitalar de Lisboa Ocidental, E.P.E./ Portugal 2015 Pulmonary edema Increased permeability PE results in the characteristic manifestations of PE such as interlobular septal thickening (Kerley's lines), ground-glass opacity and eventually consolidation and can resolve with appropriate treatment. Interleukin-2, aspirin, nitrofurantoin and cytotoxic agents (such as methotrexate) are implicated in this pattern. Page 17 of 28

Drugs affecting the heart or systemic vasculature can lead to hydrostatic PE. Radiologic features are the same as in other causes and pleural effusion may occur. Images for this section: Fig. 3: DAD. Photomicrograph (hematoxylin-eosin stain, x100) showing acute exsudative phase with hyaline membranes production (arrow). Courtesy of Dr. Sância Ramos, Department of Anatomic Pathology, Centro Hospitalar de Lisboa Ocidental, E.P.E. Page 18 of 28

Fig. 4: DAD in a patient with multiple myeloma treated with bortezomib (proteasome inhibitor) who presented with ARDS. DAD has been considered a characteristic side effect of bortezomib. A-B: CR showing rapid progression of DAD (A. was performed 24h before B.) with extensive, bilateral and diffuse ill-defined opacities and air-space consolidation. C: CT shows diffuse areas of ground-glass opacity and consolidation mainly basal and in dependent regions. D: Follow-up CT 9 months after systemic corticotherapy revealed significant regression of lesions. Page 19 of 28

Fig. 5: NSIP. Photomicrograph (hematoxylin-eosin stain, x400) showing cellular (A) and fibrotic (B) patterns of NSIP. A: Inflamatory infiltrate and type II pneumocytes hyperplasia; B: thickening of interalveolar septa and collagen deposition. Courtesy of Dr. Sância Ramos, Department of Anatomic Pathology, Centro Hospitalar de Lisboa Ocidental, E.P.E. Fig. 6: UIP. Photomicrograph (hematoxylin-eosin stain, x400) representing the heterogeneous pattern of UIP with honeycombing (A) and interstitial fibrosis (B). Courtesy of Dr. Sância Ramos, Department of Anatomic Pathology, Centro Hospitalar de Lisboa Ocidental, E.P.E. Page 20 of 28

Fig. 8: Fibrotic NSIP in a patient treated with methotrexate: bilateral reticular opacities and ground-glass opacity, as well as architectural distortion and traction bronchiectasis indicating fibrosis. NSIP is the most common manifestation of methotrexate-induced lung disease. Fig. 9: Histopathological pattern of OP (hematoxylin-eosin stain, x100): airspaces partially occupied by fibroblastic proliferations with collagen production. Courtesy of Dr. Sância Ramos, Department of Anatomic Pathology, Centro Hospitalar de Lisboa Ocidental, E.P.E. Page 22 of 28

Fig. 10: CT findings of OP in a patient treated with nitrofurantoin: patchy ground-glass opacities with peribronchial distribution; poorly defined nodular areas of consolidation associated with centrilobular nodules and branching linear opacities ("tree-in-bud opacities"), predominating in subpleural location, and bronchial dilatation. Page 23 of 28

Fig. 11: Azathioprine-induced EP in a patient with ulcerative colitis. CT shows peripheral areas of alveolar consolidation, mostly in the right upper lobe. Page 24 of 28

Fig. 12: PH (hematoxylin-eosin stain, x100): alveolar spaces occupied by fibrin mixed with red blood cells. Courtesy of Dr. Sância Ramos, Department of Anatomic Pathology, Centro Hospitalar de Lisboa Ocidental, E.P.E. Page 25 of 28

Fig. 13: CT showing bilateral, scattered, areas of ground-glass opacity in a patient with PH induced by cyclophosphamide. Page 26 of 28

Conclusion Clinical, radiological and histological findings of DILD are nonspecific. Therefore, recognizing its common radiologic manifestations and the drugs most likely to cause pulmonary toxicity is crucial to a timely and accurate diagnosis, allowing the discontinuance of the offending agent and institution of appropriate treatment, which are determining to ensure a favorable outcome. References Schwaiblmair, Martin et al. Drug Induced Interstitial Lung Disease. The Open Respiratory Medicine Journal 2012; 6:63-74 Matsuno, Osamu. Drug-induced interstitial lung disease: mechanisms and best diagnostic approaches. Respiratory Research 2012; 13:39 Rossi, Santiago E. et al. Drug Toxicity: Radiologic and Pathologic Manifestations. RadioGraphics 2000; 20:1245-1259 Webb, W. R. and Higgins, C. B. Thoracic Imaging - Pulmonary and Cardiovascular Radiology, 2th edition, 2011, Lippincott Williams & Wilkins, Philadelphia, pp. 492-503 Camus, P., Kudoh, S. and Ebina, M. Interstitial lung disease associated with drug therapy. British Journal of Cancer 2004; 91(Suppl 2), S18 - S23 Mellot, F. and Scherrer, A. Imagerie des pneumopathies médicamenteuses iatrogènes. J Radiol 2005; 86:550-7 Akihiko, Gemma. Drug-Induced Interstitial Lung Diseases Associated with Molecular-Targeted Anticancer Agents. J Nippon Med Sch 2009; 76: 4#8 Torrisi, Jean M. et al. CT Findings of Chemotherapy-induced Toxicity: What Radiologists Need to Know about the Clinical and Radiologic Manifestations of Chemotherapy Toxicity. Radiology 2011; 258(1):41-56 Müller, N. L. et al. Diagnosis and management of drug-associated interstitial lung disease. British Journal of Cancer 2004; 91(Suppl 2), S24 - S30 Personal Information 1 1 1 2 A. I. C. Santos, A. F. Roque, R. Mamede, L. Oliveira, T. Saldanha 1 1 2 Department of Radiology and Pulmonology, Centro Hospitalar de Lisboa Ocidental, E. P. E., Lisbon, Portugal Page 27 of 28

E-mail: radiologia.chlo@gmail.com Page 28 of 28