Amiodarone pulmonary toxicity

INTRODUCTION — Amiodarone is an iodinated benzofuran derivative that is used to suppress ventricular and supraventricular tachyarrhythmias. (See "Amiodarone: Clinical uses".)

Pulmonary toxicity is among the most serious adverse effects of amiodarone [1]. Several forms of pulmonary disease occur among patients treated with amiodarone, including interstitial pneumonitis, eosinophilic pneumonia, organizing pneumonia, acute respiratory distress syndrome (ARDS), diffuse alveolar hemorrhage (DAH), pulmonary nodules and solitary masses, and also (rarely) pleural effusion. Other adverse effects from amiodarone include photosensitivity, blue-gray discoloration of the skin, thyroid dysfunction, corneal deposits, abnormal liver function tests, and bone marrow suppression [2].

The types, pathogenesis, risk factors, diagnosis, and treatment of amiodarone pulmonary toxicity will be reviewed here. The other major side effects of amiodarone are discussed separately. (See "Amiodarone: Adverse effects, potential toxicities, and approach to monitoring".)

INTERSTITIAL PNEUMONITIS — Interstitial pneumonitis is the most common presentation of amiodarone-induced pulmonary disease. Interstitial pneumonitis usually presents after two or more months of therapy, especially in patients in whom the dose of amiodarone exceeds 400 mg per day [3]. The incidence of pulmonary toxicity from amiodarone is not precisely known; it is estimated to be 1 to 5 percent, depending on the dose of amiodarone [4-7]. The rate increases at higher doses of amiodarone, particularly long-term use of doses over 500 mg daily.

Pathogenesis — The mechanisms involved in amiodarone-induced interstitial pneumonitis are incompletely understood. Two major hypotheses have been suggested, a direct toxic injury to lung cells and an indirect immunologic reaction [8]. Genetic susceptibility of certain individuals may play an additional role in determining the type of injury that ensues. Whether these processes pertain to other forms of amiodarone-induced lung toxicity is not known.

●Cytotoxicity – A direct toxic reaction in the lung is supported by the following findings:

•Amiodarone has a long half-life and a high tissue affinity for the lung. In addition, an active metabolite of amiodarone, monodesethylamiodarone (DEA), exhibits cytotoxic activity and tends to accumulate in the lungs even more than amiodarone [9].

•Drug-phospholipid complexes accumulate in lung cells (eg, macrophages, interstitial cells) and interfere with normal cellular metabolic pathways, which ultimately leads to direct cell injury and death. Both apoptotic and necrotic cell death have been implicated [10]. The role of autophagy in the pathogenesis of amiodarone-induced lung injury is controversial. On one hand, induction of autophagy is considered to be a host-protective response to amiodarone-induced cellular toxicity [11]. On the other hand, amiodarone was found to induce autophagy-dependent apoptosis of type II alveolar epithelial cells [12]. The cellular injury is thought to cause chronic inflammation, which can ultimately lead to fibrosis.

•Amiodarone alters the phospholipid bilayer, which in turn disrupts cellular and organelle membrane function.

•Toxic oxygen species resulting in tissue injury are generated by amiodarone [8].

•Based on evidence from experimental animal models, stimulation of the angiotensin enzyme system may contribute to the pathogenesis of amiodarone-induced lung cell apoptosis, airway epithelial cell injury, mononuclear cell infiltration of the lamina propria, and interalveolar septal thickening [13]. Furthermore, concomitant administration of the angiotensin receptor blocker, olmesartan, in the amiodarone-exposed animals attenuated these histopathologic abnormalities [13]. In addition, retrospective analyses found that the prevalence of amiodarone-related lung disease was greater in patients not on an angiotensin-converting enzyme inhibitor (ACEi) or angiotensin receptor blocker (ARB), although this potential effect has not been sufficiently studied to guide their use to prevent amiodarone lung toxicity [13,14]. Another study implicates elevated levels of angiotensin II as increasing the risk for amiodarone-induced lung toxicity because repeated episodes of heart failure, which are associated with increased angiotensin II, appear to be a risk factor.

●Hypersensitivity – An immunologic reaction is suggested by several patients who presented with histopathologic features of a hypersensitivity pneumonitis:

•Lymphocytic infiltration, often with "intraalveolar buds" (organizing pneumonia-like reaction)

•CD8 T-cell lymphocytosis [15]

•Positive IgG immunofluorescence in the lung

Risk factors — Risk factors for amiodarone pulmonary toxicity are uncertain but may include a high cumulative dose, a daily dose greater than 400 mg/day, a duration of therapy exceeding two months, increased patient age, preexisting lung disease, thoracic or nonthoracic surgery, and pulmonary angiography. In a retrospective analysis of over 200 patients with amiodarone lung toxicity, patients over 60 years old and those on amiodarone for 6 to 12 months had the highest risk of toxicity [16].

●Cumulative dose – Pulmonary toxicity correlates more closely with the total cumulative dose than with serum drug levels. Consistent with this observation, pulmonary toxicity usually occurs several months to several years after the initiation of amiodarone therapy [3,17,18]. However, there exist anecdotal cases of severe pulmonary toxicity developing within two to three weeks of therapy with low cumulative doses [19,20].

●Daily dose – Early reports (when patients were usually treated with amiodarone doses ≥400 mg/day) noted a 5 to 15 percent incidence of pulmonary toxicity [2,21]. The incidence appears to be lower (1 to 5 percent) with smaller maintenance doses [3,21,22]. As an example, one study of 573 patients demonstrated that patients who developed pulmonary toxicity took higher daily doses than those who did not develop pulmonary disease (517 versus 409 mg/day) [21]. In this study, there were no cases of pulmonary toxicity when the maintenance dose was below 305 mg/day [21].

Although the likelihood of pulmonary toxicity may be less in patients receiving smaller daily doses, pulmonary toxicity may still occur at low doses, albeit infrequently. In a meta-analysis of four trials (1465 patients) that randomly assigned patients to receive a low-dose amiodarone (150 to 330 mg/day) or placebo for a minimum of one year, the incidence of pulmonary toxicity with amiodarone was not statistically different from placebo (1.9 versus 0.7 percent) [22]. In a subsequent, eight-year retrospective study of over 6000 amiodarone users in Quebec, Canada, there was independent association of either lower dose amiodarone (≤200 mg/day) or higher dose amiodarone (>200 mg/day) with alveolar and interstitial lung disease [4]. Similarly, a retrospective series of 500 Japanese patients found that daily doses as low as approximately 140 mg per day were associated with a cumulative incidence of amiodarone lung toxicity of approximately 4 to 11 percent at one to five years, respectively [23]. Older age and higher maintenance dose of amiodarone were risk factors for lung toxicity, while pre-existing lung disease (eg, COPD, sarcoidosis), baseline diffusing capacity, and loading dose of amiodarone were not.  

●Preexisting lung disease – An association between preexisting lung disease and the development of amiodarone pulmonary toxicity has been reported in some series. However, it is possible that these patients had limited pulmonary reserve and thus become symptomatic earlier in their course than other individuals. The following two studies illustrate the conflicting literature:

One randomized trial of 519 patients with heart failure found no accelerated loss of diffusing capacity among patients with chronic obstructive pulmonary disease who received amiodarone, compared with those who received placebo [24]. In contrast, a separate randomized trial demonstrated that patients with preexisting pulmonary disease had an increased risk of amiodarone pulmonary toxicity [25]. However, amiodarone use did not increase pulmonary death or all-cause mortality.

Clinical manifestations — Interstitial pneumonitis due to amiodarone toxicity is characterized by the insidious onset of nonproductive cough and/or dyspnea, which are present in 50 to 75 percent at presentation [3,7]. Fever is present in 33 to 50 percent; and other symptoms, such as pleuritic pain, weight loss, and malaise, may also be reported [3]. The onset of symptoms is usually within 6 to 12 months of starting amiodarone, but may occur within two months or after several years of treatment [7].

The physical examination often reveals bilateral inspiratory crackles, while clubbing is not seen.

Peripheral blood findings are nonspecific but include elevations of the white blood cell count, serum lactate dehydrogenase (LDH) level, C-reactive protein, and erythrocyte sedimentation rate [3]. Eosinophilia and antinuclear antibodies are not typically seen. Amiodarone levels are usually within the normal range [3].

Evaluation — Amiodarone-induced interstitial pneumonitis is generally suspected on the basis of new onset of dyspnea and/or cough in a patient who is taking or has recently discontinued taking amiodarone. The purpose of the evaluation is to narrow the diagnostic possibilities, exclude alternate diagnoses, and assess the severity of respiratory impairment. A careful history of medication use, symptoms suggestive of rheumatic disease, and occupational and environmental exposures is an important component of the evaluation, as with any interstitial lung disease. Similarly, the physical examination should include a search for exacerbation of cardiovascular disease and the presence of rheumatic disease with a potential pulmonary component. (See "Approach to the adult with interstitial lung disease: Clinical evaluation".)

Laboratory studies — A white blood cell count with differential and plasma brain natriuretic peptide (BNP) or N-terminal pro-BNP (NT-proBNP) levels are obtained to evaluate for infection and heart failure. While the erythrocyte sedimentation rate and C-reactive protein levels are often elevated, these tests do not differentiate well among the diagnostic possibilities, and we do not usually obtain them. Serologic studies such as an antinuclear antibody test and rheumatoid factor are obtained based on the degree of suspicion for amiodarone lung toxicity versus rheumatic disease. (See "Approach to the adult with interstitial lung disease: Diagnostic testing", section on 'Laboratory tests'.)

Amiodarone levels are not predictive or diagnostic of pulmonary toxicity.

Measurement of serum concentrations of KL-6, a mucin-like high molecular weight glycoprotein secreted by proliferating type II pneumocytes, is a sensitive marker of disease activity in various interstitial lung diseases. Serum KL-6 has been proposed as a marker of amiodarone pulmonary toxicity, but lack of specificity limits its utility [23,26-28].

Chest imaging — The chest radiograph in interstitial pneumonitis due to amiodarone typically shows new, diffuse or localized, reticular, consolidative, or mixed opacities [29]. These changes may be migratory and can occur in the absence of symptoms (image 1) [30]. Pleural effusions are rare. (See 'Pleural disease' below.)

High-resolution computed tomography (HRCT) is obtained to clarify the radiographic pattern and distribution of abnormalities. Supine and prone HRCT images should be compared to exclude dependent changes, given the possibility of heart failure in these patients [18,31,32]. HRCT in patients with amiodarone-induced interstitial pneumonitis may show areas of high attenuation in the lungs, as well as the liver and spleen, due to the accumulation of iodinated amiodarone in tissue macrophages (image 2) [33]. Although this finding is specific for amiodarone use, it is not necessary to the diagnosis of amiodarone pneumonitis and may be seen in the absence of lung toxicity. Other HRCT findings related to amiodarone-induced interstitial pneumonitis include diffuse (usually bilateral) ground glass opacities and septal thickening; honeycombing and traction bronchiectasis can also be seen [18,34,35].

While gallium scans showing increased lung uptake are a sensitive marker for the presence of amiodarone pneumonitis, gallium scans are generally not performed in patients with suspected amiodarone toxicity due to lack of specificity. Furthermore, gallium uptake in the lungs may remain elevated despite discontinuation of amiodarone and resolution of clinical disease [36].

Pulmonary function testing — Pulmonary function tests, including spirometry, lung volumes, DLCO, and pulse oxygen saturation (SpO₂) at rest and with exertion, are typically obtained to evaluate patients with dyspnea and cough. While amiodarone-induced pulmonary toxicity is often associated with a restrictive pattern (reduced forced vital capacity and total lung capacity) and a reduction in diffusing capacity, these findings are nonspecific. However, a documented decline in the DLCO of greater than 20 percent is useful in suggesting the need for closer monitoring or further diagnostic testing with chest imaging [7].

Bronchoalveolar lavage — Flexible bronchoscopy with bronchoalveolar lavage (BAL) is performed when the clinical diagnosis of amiodarone-induced interstitial pneumonitis is uncertain and is more helpful in the exclusion of alternative diagnoses (eg, infection, hemorrhage, malignancy) than in securing a diagnosis of amiodarone lung toxicity. Lavage samples should be sent for cell counts, microbial stains, culture, and cytologic evaluation. (See "Basic principles and technique of bronchoalveolar lavage" and "Role of bronchoalveolar lavage in diagnosis of interstitial lung disease" and 'Diffuse alveolar hemorrhage' below.)

The BAL pattern in amiodarone pneumonitis is highly variable, and no particular BAL cellular pattern appears predictive of outcome. The presence of "foam" cells (due to the accumulation of phospholipids in alveolar macrophages) is not pathognomonic of pulmonary toxicity because up to 50 percent of exposed patients without signs or symptoms of toxicity may have these findings. The absence of foam cells, however, makes the diagnosis of amiodarone pulmonary toxicity unlikely [3]. (See 'Histopathology' below.)

Reported findings in BAL fluid from patients with amiodarone pneumonitis include [37-39]:

●Lymphocytosis

●Neutrophilia

●Eosinophilia

●Normal BAL cellular counts

Among patients presenting with more acute forms of amiodarone-related respiratory disease, BAL may reveal alveolar hemorrhage with sequentially more hemorrhagic lavage returns and/or a high concentration of hemosiderin-laden macrophages [40]. (See 'Diffuse alveolar hemorrhage' below.)

Lung biopsy — Lung biopsy is usually not necessary to prove amiodarone pulmonary toxicity in patients with new-onset respiratory symptoms and compatible pulmonary function test and radiographic changes. Due to concerns regarding the development of acute respiratory distress syndrome (ARDS) following surgical procedures in patients taking amiodarone, open or thoracoscopic lung biopsies are usually reserved for patients in whom other efforts to diagnose the illness have been unsuccessful, including a trial of drug withdrawal and perhaps also systemic glucocorticoid administration, and those in whom an alternate diagnosis is suspected. Occasionally, when amiodarone is felt to be essential to the management of a patient's dysrhythmias and the diagnosis of amiodarone toxicity is uncertain, a biopsy may be indicated for confirmation. (See 'Acute respiratory distress syndrome' below.)

For patients who require a lung biopsy, biopsies obtained via video-assisted thoracoscopic surgery (VATS) or open thoracotomy are preferred over transbronchial biopsies as the latter are generally insufficient to make definitive diagnosis of amiodarone toxicity. On the other hand, it may be reasonable to obtain a transbronchial lung biopsy at the time of flexible bronchoscopy to help rule out infection or malignancy. (See "Role of lung biopsy in the diagnosis of interstitial lung disease", section on 'Transbronchial lung biopsy'.)

Histopathology — The histopathologic findings of amiodarone-induced interstitial pneumonitis include a nonspecific interstitial pneumonitis (with a mononuclear cell infiltrate), type II cell hyperplasia, interstitial edema, fibrosis, and lipid-laden alveolar macrophages [41].

The presence of numerous lipid-laden, "foamy" macrophages in the air spaces is a characteristic finding in all patients exposed to amiodarone (picture 1). The "foamy" appearance is due to amiodarone-phospholipid complexes (image 3) and is also seen in patients without lung toxicity who are taking amiodarone. Amiodarone can cause an accumulation of phospholipids within lysosomes in other lung cells, although this is less common [2,41]. Ultrastructural studies show myelinoid (lamellated) inclusion bodies in the affected tissue.

Various forms of lymphocytic infiltration have been described in amiodarone interstitial pneumonitis, including diffuse lymphoid hyperplasia, follicular bronchiolitis, and lymphoid interstitial pneumonia [42].

Occasionally, a component of organizing pneumonia is seen within the larger overall appearance of interstitial pneumonia [18].

Diagnosis and differential diagnosis — The diagnosis of amiodarone-induced lung disease (interstitial pneumonitis and other types) is essentially a diagnosis of exclusion. While a lung biopsy showing the characteristic histopathologic changes is the gold standard for the diagnosis of interstitial pneumonitis due to amiodarone (see 'Histopathology' above), a clinical diagnosis can often be made when the clinical features and evaluation are consistent, other possibilities have been excluded, and the patient improves with drug cessation with or without a trial of glucocorticoid therapy [7,43].

Key features that support a clinical diagnosis of amiodarone-induced interstitial pneumonitis include the following:

●New or worsening dyspnea, cough, and weight loss in a patient taking ≥200 mg of amiodarone a day, particularly 6 to 12 months into therapy

●New ground glass or reticular opacities on chest radiograph and confirmatory findings on HRCT

●Negative evaluation for heart failure (eg, normal plasma brain natriuretic peptide, normal ventricular ejection fraction, absent or incomplete improvement with diuresis) or at least improving cardiac function despite worsening lung function

●Exclusion of lung infection

●Exclusion of other interstitial lung diseases (eg, hypersensitivity pneumonitis, occupational, rheumatic)

●Presence of foamy macrophages in the BAL (although not diagnostic, their absence makes the diagnosis unlikely) (see 'Bronchoalveolar lavage' above)

●Improvement in symptoms and radiographic manifestations following withdrawal of amiodarone (with or without glucocorticoid therapy)

The differential diagnosis of interstitial pneumonia due to amiodarone includes processes that can present similarly, such as heart failure, infectious pneumonia, intercurrent interstitial lung disease, and lung toxicity from another drug, as well as other types of pulmonary toxicity due to amiodarone described below. (See "Approach to the adult with interstitial lung disease: Clinical evaluation" and "Approach to the adult with interstitial lung disease: Diagnostic testing" and "Pulmonary disease induced by cardiovascular drugs" and "Nitrofurantoin-induced pulmonary injury".)

Screening — There are no adequate predictors of pulmonary toxicity due to amiodarone. Guidelines suggest obtaining a baseline and annual chest radiograph and baseline pulmonary function tests (including a DLCO) [43,44]. However, serial pulmonary function tests are not helpful because changes in DLCO are not specific for toxicity [45,46]. One study prospectively evaluated the usefulness of serial DLCO and spirometry measurements in 91 patients who were receiving amiodarone therapy for refractory arrhythmias [46]. Most of the asymptomatic patients whose DLCO decreased more than 20 percent did not develop pulmonary toxicity over the next year despite continued amiodarone therapy.

Treatment — Treatment of amiodarone interstitial pneumonitis consists primarily of stopping amiodarone and, in more symptomatic patients, initiating systemic glucocorticoids. Patients with mild symptoms and normal oxygenation can be observed off amiodarone without glucocorticoids. Due to the accumulation of amiodarone in fatty tissues and its long elimination half-life (approximately 45 days), pulmonary toxicity may progress initially despite drug discontinuation.

For the majority of patients with more than mild symptoms (eg, dyspnea on mild to moderate exertion) and evidence of respiratory impairment, we initiate systemic glucocorticoid therapy (eg, prednisone 40 to 60 mg per day). In a number of case reports, glucocorticoid therapy has been associated with dramatic improvement even in patients with severe disease [7,35]. After a clinical response is evident, glucocorticoid therapy is tapered slowly, as tolerated, over two to six months. If symptoms and signs recur, the glucocorticoid dose is returned to the last effective dose and further tapering is slower, over approximately 12 months. Recurrent pulmonary toxicity during tapering of glucocorticoids may be due to greater body stores of amiodarone associated with greater body mass indices and/or amounts of body fat [47].

We suggest NOT resuming amiodarone in patients who have recovered from amiodarone-induced pulmonary toxicity due to the risk of recurrent disease and progressive pulmonary fibrosis. Alternative medications and procedures are available to treat supraventricular arrhythmias, and the implantable cardioverter-defibrillator may be an alternative for potentially life-threatening ventricular arrhythmias. (See "Implantable cardioverter-defibrillators: Overview of indications, components, and functions".)

Dronedarone, an antiarrhythmic with similar structure to amiodarone, appears to have lower risk of lung toxicity than amiodarone, but its use in patients with prior amiodarone toxicity has not been formally examined. There have been three case reports in which dronedarone was implicated as the cause of lung toxicity – one with diffuse alveolar damage and two with organizing pneumonia [48-50]. The association of dronedarone with increased mortality in patients with advanced heart failure will likely limit its use [51]. (See "Clinical uses of dronedarone".)

While preliminary studies suggest a protective effect of ACEi and ARB [14], the use of these agents solely for the reduction of amiodarone lung toxicity cannot be recommended without additional studies. (See 'Pathogenesis' above.)

Prognosis — The prognosis of amiodarone interstitial pneumonitis is generally favorable. In one study, for example, three-fourths of patients stabilized or improved after withdrawal of the drug with or without glucocorticoid treatment [38]. Death attributable to amiodarone pneumonitis occurred in 10 percent of cases reported in the literature, although the actual mortality in clinical practice is probably less [18].

EOSINOPHILIC PNEUMONIA — A few cases of eosinophilic pneumonia have been reported in association with amiodarone [42]. Presentations consistent with acute or chronic eosinophilic pneumonia have been described, although the acute presentation appears more common.

●Clinical features – Acute eosinophilic pneumonia typically presents with less than one month and often less than one week of fever, nonproductive cough, and dyspnea, while chronic eosinophilic pneumonia presents as a subacute illness with more than a month of cough, fever, progressive breathlessness, weight loss, wheezing, and night sweats. (See "Idiopathic acute eosinophilic pneumonia", section on 'Clinical features' and "Overview of pulmonary eosinophilia", section on 'Chronic eosinophilic pneumonia'.)

Peripheral blood eosinophilia may be present in both acute and chronic eosinophilic pneumonia. Chest computed tomography (CT) findings include ground glass opacities, diffuse reticular changes, and masses; in the chronic form, the radiographic changes may have a peripheral distribution [42]. (See "Overview of pulmonary eosinophilia", section on 'Idiopathic acute eosinophilic pneumonia' and "Overview of pulmonary eosinophilia", section on 'Chronic eosinophilic pneumonia'.)

●Diagnosis – In both acute and chronic eosinophilic pneumonia due to amiodarone, the diagnosis is supported by bronchoalveolar lavage (BAL) fluid showing abundant eosinophils, often >25 percent, and abundant foamy macrophages [42,52-54], but requires exclusion of other causes of eosinophilic pneumonia. The differential diagnosis of eosinophilic pneumonia includes fungal and parasitic infection, vasculitis, and eosinophilic pneumonia induced by other drugs. These processes must be excluded by appropriate history (eg, medication use, travel or residence in areas with endemic fungi or parasites) and serologic studies (eg, antineutrophil cytoplasmic antibody, rheumatoid factor, enzyme-linked immunoassays for specific coccidioidal IgM and IgG). (See "Overview of pulmonary eosinophilia", section on 'Diagnostic approach'.)

●Treatment – In addition to cessation of amiodarone, the majority of patients with eosinophilic pneumonia associated with amiodarone are treated with systemic glucocorticoids, following the approaches for idiopathic acute and chronic eosinophilic pneumonias, depending on which presentation the individual patient displays. (See "Idiopathic acute eosinophilic pneumonia", section on 'Treatment' and "Chronic eosinophilic pneumonia", section on 'Treatment'.)

ORGANIZING PNEUMONIA — Organizing pneumonia (OP), formerly called bronchiolitis obliterans organizing pneumonia or BOOP, occurs in approximately 25 percent of cases of amiodarone pulmonary toxicity [29,55-57]. (See "Cryptogenic organizing pneumonia".)

The typical pathologic findings in OP include excessive proliferation of granulation tissue, consisting of loose collagen-embedded fibroblasts and myofibroblasts, and involving alveolar ducts and alveoli, with or without bronchiolar intraluminal polyps. In addition to these findings, lymphoid hyperplasia can rarely be seen on lung biopsies obtained from patients with OP associated with amiodarone [42]. In eight of these patients, lymphoid hyperplasia manifested as diffuse lymphoid hyperplasia, follicular bronchiolitis, lymphoid interstitial pneumonia, or lymphocytic perivascular cuffing [42].

●Clinical features – The presentation of OP due to amiodarone is typically more acute in onset than that of chronic interstitial pneumonitis and occurs over a few weeks to months. It is characterized by a nonproductive cough, pleuritic chest pain, fever, and dyspnea, which often mimic an infectious pneumonitis [58]. (See "Cryptogenic organizing pneumonia", section on 'Clinical features'.)

Pulmonary auscultation typically reveals crackles, which may be focal, and sometimes a pleural rub.

The chest radiograph usually demonstrates patchy areas of consolidation, sometimes with an air bronchogram. High-resolution computed tomography (HRCT) usually confirms patchy and sometimes nodular areas of consolidative opacity. When patients have a prominence of lymphoid hyperplasia, additional chest computed tomography (CT) findings include septal thickening, ground glass nodules, and masses [42].

●Diagnosis – As the clinical presentation may initially suggest infection, the diagnosis of OP is often suspected after a lack of response to antibiotics administered for suspected bacterial pneumonia. In the majority of patients, a biopsy demonstrating organizing pneumonia is needed to secure a diagnosis. In order to have enough tissue for the pathologist to exclude other processes, we prefer to obtain a lung biopsy via video-assisted thoracoscopic surgery (VATS) or open thoracotomy rather than transbronchial biopsy. The histopathologic features of organizing pneumonia are described separately. (See "Cryptogenic organizing pneumonia", section on 'Diagnosis' and "Cryptogenic organizing pneumonia", section on 'Histopathologic diagnosis of organizing pneumonia'.)

●Treatment – Treatment of OP in the context of amiodarone therapy requires cessation of amiodarone and generally also a course of systemic glucocorticoid therapy [59]. Similar to the treatment of cryptogenic OP, the decision to implement glucocorticoid therapy, the choice of initial dose, and length of the tapered dose regimen depend on the severity of respiratory involvement. (See "Cryptogenic organizing pneumonia", section on 'Treatment'.)

ACUTE RESPIRATORY DISTRESS SYNDROME — Acute respiratory distress syndrome (ARDS) is a rare but potentially fatal form of amiodarone pulmonary toxicity and is usually characterized by diffuse alveolar damage with hyaline membranes (picture 2) [29]. Rarely, a histopathologic pattern of diffuse alveolar damage plus eosinophilic infiltration has been attributed to amiodarone lung toxicity [42]. (See 'Eosinophilic pneumonia' above.)

●Clinical presentation – Amiodarone-associated ARDS has been reported in patients who have undergone thoracic surgery or pulmonary angiography; such incidences have occurred during chronic treatment with amiodarone or when the amiodarone was initiated in the peri-procedural period [60-63]. However, the role of amiodarone in acute periprocedural ARDS is controversial, and a review of amiodarone pulmonary toxicity cites evidence against such an association [62,64-66]. In reports describing surgical patients, ARDS developed within one to four days after extubation [60,61]. In comparison, the report of two patients with fatal ARDS following pulmonary angiography described respiratory deterioration within 30 minutes of the procedure [62]. While some surgical literature suggests that use of amiodarone is safe after lung resection [64,66,67], others have voiced concern about lung toxicity in such circumstances and call for clinical studies to determine the true risks and benefits of peri-operative amiodarone [65].  

●Diagnosis – The diagnosis of ARDS is based on clinical criteria: acute onset of symptoms (≤1 week), bilateral opacities consistent with pulmonary edema on chest computed tomography (CT), absence of cardiac failure or fluid overload, and a moderate to severe impairment of oxygenation. The differential diagnosis of ARDS includes cardiogenic pulmonary edema, diffuse alveolar hemorrhage (see below), acute interstitial pneumonia, acute eosinophilic pneumonia, and cancer. Usually, when ARDS is suspected, bronchoalveolar lavage (BAL) is performed to exclude lung infection, hemorrhage, eosinophilic pneumonia, and malignancy. Once these processes have been excluded, other causes of ARDS, such as sepsis, aspiration, transfusion, and drug toxicity (other than amiodarone), need to be excluded. (See 'Diffuse alveolar hemorrhage' below and "Acute respiratory distress syndrome: Clinical features, diagnosis, and complications in adults" and "Acute respiratory distress syndrome: Epidemiology, pathophysiology, pathology, and etiology in adults" and "Basic principles and technique of bronchoalveolar lavage".)

Acute fibrinous and organizing pneumonia (AFOP) is a separate process that is in the differential diagnosis of ARDS as it presents similarly. At least one case of AFOP associated with amiodarone has been reported [68]. AFOP has a different histopathologic pattern from ARDS, characterized by the presence of intra-alveolar fibrin (fibrin "balls") and organizing pneumonia, but without the hyaline membranes associated with ARDS. (See "Interpretation of lung biopsy results in interstitial lung disease", section on 'Rare histopathologic interstitial pneumonia patterns'.)

●Treatment – Management of ARDS associated with amiodarone toxicity should include cessation of amiodarone, implementation of supportive care for the critically ill patient, and use of mechanical ventilation with strategies to minimize supplemental oxygen and also reduce barotrauma and volutrauma. (See "Acute respiratory distress syndrome: Fluid management, pharmacotherapy, and supportive care in adults" and "Acute respiratory distress syndrome: Ventilator management strategies for adults".)

While systemic glucocorticoids are not part of routine care in ARDS and have not been formally evaluated in ARDS associated with amiodarone, most patients with amiodarone-associated ARDS are treated empirically with systemic glucocorticoids (eg, methylprednisolone 500 to 1000 mg/day intravenously, followed by prednisone 0.5 mg/kg daily) after infection has been excluded [7,69]. This practice is based on the response to glucocorticoids observed in other forms of amiodarone lung toxicity and in a small number of patients with amiodarone-associated ARDS [7,69]. Often, patients will need to be treated empirically for more than one possibility (eg, infection and drug-induced lung toxicity) while studies are pending. (See "Cryptogenic organizing pneumonia", section on 'Treatment'.)

In one patient with life-threatening amiodarone lung toxicity manifested by diffuse radiographic opacities and respiratory failure that were unresponsive to glucocorticoids, hemoperfusion with a polymyxin B-immobilized fiber column dramatically improved oxygenation with concomitant reduction in serum amiodarone and desethylamiodarone levels [70]. Further study is needed to determine the efficacy of this treatment.

Patients in whom ARDS develops due to amiodarone have a mortality rate of approximately 50 percent [71].

●Prevention – Given the apparent association of amiodarone-associated ARDS with surgery or medical procedures, it has been hypothesized that amiodarone may sensitize patients to high concentrations of inspired oxygen or to iodinated contrast media. Some experts advise close monitoring of the pulse oxygen saturation (SpO₂) and/or partial pressure of arterial oxygen (PaO₂) to minimize supplemental oxygen, although data to support this practice are lacking. While caution should be used when considering patients on amiodarone for surgery, especially in the setting of preoperative pulmonary dysfunction, the concerns about performing surgery on such patients must be weighed against the alternative concern that withdrawal of amiodarone before surgery may delay the operation for several weeks (because amiodarone has a very long half-life) and put the patient at increased risk of malignant dysrhythmias.

DIFFUSE ALVEOLAR HEMORRHAGE — Diffuse alveolar hemorrhage (DAH) is a rare complication of amiodarone [40,72-75]. It may occur abruptly in the first few days or months (average six months) after initiation of the drug [29,73,74,76-78].

●Clinical features – The patients commonly present with an acute onset of cough, shortness of breath, fever, and sometimes hemoptysis. Many of the patients have pre-existing chronic lung disease [74]. Laboratory studies often show a decrease in hemoglobin. Imaging studies show bilateral diffuse ground glass opacities and/or consolidation.

●Diagnosis – The diagnosis of DAH due to amiodarone is usually suspected based on the clinical presentation (eg, acute onset, hemoptysis) and imaging studies (diffuse opacities). Confirmation is usually obtained by sequential bronchoalveolar lavages (BALs) in an area of radiographic opacities that reveal progressively more hemorrhagic returns [40]. Hemosiderin-laden macrophages, which may be demonstrated by Prussian blue staining, are also characteristic of DAH. While hemosiderin-laden macrophages can be seen in cardiogenic pulmonary edema, they are typically present in lower numbers than in pulmonary hemorrhage due to amiodarone. A lung biopsy is generally not indicated. (See "The diffuse alveolar hemorrhage syndromes", section on 'Bronchoalveolar lavage'.)

●Differential diagnosis – The differential diagnosis of DAH includes pulmonary edema (cardiogenic and noncardiogenic), systemic lupus, vasculitis, and hemorrhage due to another drug (eg, cocaine, propylthiouracil) or due to therapy with anticoagulants or platelet glycoprotein IIB/IIIA inhibitors. These possibilities can be narrowed by studies, such as a platelet count, coagulation tests, urinalysis, drug screening for cocaine, plasma brain natriuretic peptide (BNP), serologic tests for antinuclear antibodies, antineutrophil cytoplasmic antibodies (ANCA) and antiglomerular basement membrane antibodies, and an echocardiogram. (See 'Acute respiratory distress syndrome' above and "The diffuse alveolar hemorrhage syndromes", section on 'Clues to a specific etiology'.)

●Treatment – As with other forms of amiodarone toxicity, prompt discontinuation of the medication is indicated. Most patients are treated with systemic glucocorticoids following the treatment approach for other causes of DAH, although the value and dosing in this setting are uncertain. A reasonable initial dose of glucocorticoids, is the equivalent of methylprednisolone 500 to 1000 mg intravenously in divided doses, daily for up to five days followed by gradual tapering and then maintenance for weeks to months on an oral preparation. (See "The diffuse alveolar hemorrhage syndromes", section on 'Treatment'.)

One patient, who required continuation of amiodarone to prevent life-threatening arrhythmias, was successfully treated with glucocorticoids and a reduction in the dose of amiodarone to half of the prior dose [74].

PULMONARY NODULES OR MASS — Solitary and multiple pulmonary nodules and masses have been attributed to amiodarone toxicity in a number of case reports and case series [79,80].

Among four such patients who had histopathologic evaluation, one had a single mass with a background of increased reticular markings, while the others had two or more nodules [80]. Increased F-18-fluorodeoxyglucose-positron emission tomography (FDG-PET) uptake was noted in two patients. Pathologic examination revealed areas of necrosis within solid inflammation [80]. Surrounding the necrosis were sheets of vacuolated histiocytes and aggregates of neutrophils. Patches of organizing pneumonia were noted peripheral to some nodules. The appearance was somewhat similar to granulomatosis with polyangiitis, but palisading histiocytes, vasculitis, and granulomatous inflammation were not present. The patient with multiple nodules had improvement in symptoms and resolution of the nodules after cessation of amiodarone; the others had excisions of the lesions due to suspicion for malignancy.

In a separate report, a solitary pulmonary mass simulating a pulmonary malignancy and associated with additional smaller, peripheral nodules was described as a complication of amiodarone therapy [79]. Increased uptake was noted on FDG-PET scan in support of possible malignancy. However, needle aspirates yielded foamy macrophages and type II cells, and no evidence of malignancy. With cessation of amiodarone and treatment with prednisone, the lung abnormalities resolved over 12 months. Dual energy computed tomography (CT) demonstrated high iodine content in a mass-like lung consolidation of a patient treated with amiodarone, which was confirmed to be due to amiodarone by lung biopsy [81].

PLEURAL DISEASE — A few case reports have described exudative pleural effusions associated with amiodarone therapy either isolated or in combination with pneumonitis [82-85]. In at least one case, the patient presented with features of drug-induced lupus [84]. (See "Drug-induced lupus".)

Pleuroparenchymal fibroelastosis was described in one patient after five years of amiodarone therapy [86]; no other cause of PPFE was identified.

SUMMARY AND RECOMMENDATIONS

●Amiodarone is associated with several forms of pulmonary toxicity including interstitial pneumonitis (the most common presentation), eosinophilic pneumonia, organizing pneumonia, acute respiratory distress syndrome (ARDS), diffuse alveolar hemorrhage (DAH), pulmonary nodules and masses, and rarely pleural effusions. The incidence of pulmonary toxicity from amiodarone is not precisely known; it is estimated to be 1 to 5 percent, depending on the dose of amiodarone. (See 'Introduction' above and 'Interstitial pneumonitis' above.)

●Risk factors for amiodarone-induced pulmonary toxicity include a daily dose ≥400 mg/day, duration of therapy exceeding two months, patient age >60 years, preexisting lung disease, surgery, and pulmonary angiography. Pulmonary toxicity can also occur at lower daily doses. (See 'Risk factors' above.)

●Permanent discontinuation of amiodarone is the primary therapy for amiodarone pulmonary toxicity. (See 'Treatment' above.)

●Interstitial pneumonitis – Interstitial pneumonitis due to amiodarone toxicity is characterized by the insidious onset of nonproductive cough and/or dyspnea. Fever, pleuritic pain, weight loss, and malaise can also occur. The onset of symptoms is usually 6 to 12 months of amiodarone therapy, but may occur within two months or after several years of treatment. (See 'Clinical manifestations' above.)

•The chest radiograph typically reveals new focal or diffuse reticular or ground glass opacities. High-resolution computed tomography (HRCT) of the chest and upper abdomen usually shows ground glass and reticular opacities and also increased attenuation in the lungs, liver, and spleen. Pulmonary function tests typically show restriction and a reduced diffusing capacity (DLCO). (See 'Laboratory studies' above.)

•Flexible bronchoscopy with bronchoalveolar lavage (BAL) is performed to exclude alternative diagnoses (eg, infection, hemorrhage, malignancy). Samples should be sent for cell counts, culture, and cytologic evaluation. The presence of "foam" cells (alveolar macrophages full of amiodarone-phospholipid complexes) is a characteristic of, but not pathognomonic for, pulmonary toxicity (picture 1). (See 'Bronchoalveolar lavage' above and 'Histopathology' above.)

•A clinical diagnosis of amiodarone-induced interstitial pneumonitis can often be made when the clinical features are consistent; other possibilities (eg, infection, heart failure) have been excluded; and the patient improves with drug cessation and, possibly, a trial of glucocorticoid therapy. Lung biopsy is usually deferred unless the diagnosis remains uncertain after a trial of drug cessation. (See 'Diagnosis and differential diagnosis' above.)

•In addition to cessation of amiodarone, for the majority of patients with more than mild symptoms of interstitial pneumonitis due to amiodarone, we suggest initiation of systemic glucocorticoid therapy (Grade 2C). The usual dose is the equivalent of oral prednisone 40 to 60 mg/day. Due to the long elimination half-life (approximately 45 days) of amiodarone, pulmonary toxicity may initially progress despite drug discontinuation and may recur upon glucocorticoid withdrawal. (See 'Treatment' above.)

•For screening, guidelines suggest a baseline and annual chest radiograph and baseline pulmonary function tests for patients on long-term amiodarone therapy. However, serial pulmonary function tests are not helpful in predicting amiodarone toxicity. (See 'Screening' above.)

●Eosinophilic pneumonia – Acute and chronic eosinophilic pneumonia are rarely associated with amiodarone. Chest computed tomography (CT) features include ground glass opacities, diffuse reticular changes, and masses. Peripheral blood eosinophilia may be present. In both acute and chronic eosinophilic pneumonia, the diagnosis is supported by BAL fluid that often contains >25 percent eosinophils. Other causes of pulmonary eosinophilia (eg, fungal or parasitic infection, pulmonary eosinophilia due to another drug, vasculitis) must be excluded. (See 'Eosinophilic pneumonia' above and "Overview of pulmonary eosinophilia".)

Treatment includes cessation of amiodarone and administration of systemic glucocorticoids, following the approaches for acute or chronic eosinophilic pneumonia, depending on the pattern manifested by the individual patient. (See "Idiopathic acute eosinophilic pneumonia", section on 'Treatment' and "Chronic eosinophilic pneumonia", section on 'Treatment'.)

●Organizing pneumonia – Organizing pneumonia (such as that seen in cryptogenic organizing pneumonia) is present in approximately 25 percent of cases of amiodarone pulmonary toxicity. Suggestive features include a subacute or acute onset mimicking an infectious pneumonia, patchy or nodular consolidative opacities on imaging, and no response to antimicrobial therapy. A lung biopsy is needed for diagnosis. Treatment requires cessation of amiodarone; addition of systemic glucocorticoids is based on severity of disease and follows that for cryptogenic organizing pneumonia. (See 'Organizing pneumonia' above and "Cryptogenic organizing pneumonia".)

●Acute respiratory distress syndrome – ARDS is a rare but potentially fatal form of amiodarone pulmonary toxicity that typically occurs shortly after a pulmonary angiogram or surgery. (See 'Acute respiratory distress syndrome' above.)

•Patients present with rapidly progressive respiratory failure, impaired oxygenation, and diffuse ground glass opacities on radiographic imaging. BAL is performed promptly to exclude infection, hemorrhage, and eosinophilic pneumonia. Other common causes of ARDS (eg, sepsis, aspiration) are also investigated.

•Management includes amiodarone cessation, supportive care, and mechanical ventilation. In addition, we suggest systemic glucocorticoid therapy (Grade 2C). A reasonable dose (after the exclusion of infection) is methylprednisolone 500 to 1000 mg/day intravenously, followed by prednisone 0.25 to 0.5 mg/kg daily. (See 'Acute respiratory distress syndrome' above.)

●Diffuse alveolar hemorrhage – DAH due to amiodarone has an acute presentation of dyspnea, cough, and sometimes hemoptysis. Imaging shows diffuse or patchy ground glass or consolidative opacities. The diagnosis is usually made by sequential BAL sampling that shows progressively more hemorrhagic effluent; other causes of DAH must be excluded. In addition to cessation of amiodarone, we suggest systemic glucocorticoid therapy (Grade 2C). A reasonable dose (after the exclusion of infection) is methylprednisolone 500 to 1000 mg/day, intravenously in divided doses for up to five days, followed by gradual tapering. (See 'Diffuse alveolar hemorrhage' above and "The diffuse alveolar hemorrhage syndromes", section on 'Glucocorticoids'.)

●Pulmonary nodules and masses – Solitary and multiple pulmonary nodules and masses due to amiodarone toxicity can mimic pulmonary malignancy in radiographic appearance and F-18-fluorodeoxyglucose-positron emission tomography (FDG-PET) uptake. A pathologic diagnosis is needed to exclude malignancy. (See 'Pulmonary nodules or mass' above.)

●Pleural effusion – Exudative pleural effusions are a rare manifestation of amiodarone toxicity; they may be isolated or occur in association with interstitial pneumonitis. (See 'Pleural disease' above.)