Drug-induced eosinophilic lung disease

Clin Chest Med 25 (2004) 77 – 88

Drug-induced eosinophilic lung disease James N. Allen, MD Division of Pulmonary and Critical Care Medicine, The Ohio State University, 201 Heart Lung Institute Building, 473 West 12th Avenue, Columbus, OH 43210, USA

Many drugs can cause eosinophilic lung disease. In this article, the biology of the eosinophil and the mechanism of drug-induced eosinophilia is reviewed. The differential diagnosis of pulmonary eosinophilia is summarized. Drugs that commonly cause eosinophilic lung disease are reviewed. Finally, the clinical approach to drug-induced eosinophilic lung disease is discussed.

Eosinophil biology Eosinophil production in the bone marrow is regulated largely by T lymphocytes, especially the T-helper (Th) type 2 lymphocytes. Interleukin (IL)-5, produced by Th2 cells, is the main cytokine that regulates eosinophil production. Other cytokines that promote eosinophil production include IL-3 and granulocyte-macrophage colony stimulating factor (Fig. 1) [1]. Once produced, eosinophil release into the blood stream is regulated by cytokines, such as IL-5 and eotaxin. Most eosinophils rapidly migrate to the tissues, especially those with an environmental interface, such as the lung, gastrointestinal tract, and skin. Eosinophils can accumulate preferentially in the lung by local production of eosinophil chemotactic factors, such as IL-5 and eotaxin (Fig. 2). IL-5 is produced by lung Th2 lymphocytes that have been activated by antigen-presenting cells, such as alveolar macrophages. Eotaxin is produced by alveolar macrophages, pulmonary endothelial cells, airway smooth muscle cells, and alveolar epithelial cells, that, in turn, are regulated by T lymphocytes [2].

E-mail address: [email protected]

The eosinophil tissue:blood ratio is approximately 100:1 in most normal individuals. In these tissues, the eosinophil is involved in host defense, especially against parasites. The eosinophil granules contain several chemicals, such as major basic protein, eosinophil cationic protein, eosinophil-derived neurotoxin, and eosinophil peroxidase, all of which are toxic to invading micro-organisms. Release of these granule contents in response to a drug can result in tissue injury and the development of the clinical abnormalities that are associated with eosinophilic lung disease [1]. Depending on the disease process, eosinophils in the lung may be located predominately in the airway, alveoli, interstitium, or blood vessels. In the peripheral blood, eosinophils are young and have a typical appearance of a bilobed nucleus and abundant large cytoplasmic granules. As tissue eosinophils age and degranulate, they can develop additional nuclear lobes and may lack cytoplasmic granules. Thus, eosinophils that are seen in bronchoalveolar lavage (BAL) fluid or in histologic specimens can be mistaken for neutrophils. It is uncertain why eosinophils preferentially accumulate in the lung in drug-induced eosinophilic lung disease; it can be speculated that antigen-presenting cells, such as alveolar macrophages, consume drugs and then present them as antigens by way of their major histocompatibility complex (MHC) class 2 receptors. This antigen-receptor complex is recognized by the Th2 lymphocyte T-cell receptor in conjunction with the lymphocyte CD4 receptor. This, in turn, activates the Th2 lymphocyte to release IL-5 and results in eosinophil production, chemotaxis to the lung, and degranulation (Fig. 3). The abundance of macrophages in the lung could result in large local production of IL-5 and eotaxin with the consequence

0272-5231/04/$ – see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/S0272-5231(03)00141-2

78

J.N. Allen / Clin Chest Med 25 (2004) 77–88

Fig. 1. Regulation of eosinophil production and chemotaxis.

of preferential accumulation of eosinophils within the lung.

Diagnosis Eosinophilic lung disease can be identified by the finding of an increased number of eosinophils in the lung tissue or BAL fluid of a patient who has pulmonary symptoms or infiltrates on chest radiographs. When lung biopsy or BAL is not available, eosinophilic lung disease can be inferred by the finding of increased blood eosinophils in association

with pulmonary infiltrates on chest radiographs. A normal blood eosinophil count is less than 350/mL. Some diseases can present with prominent eosinophil infiltration in lung tissue without associated blood eosinophilia; therefore, a normal blood eosinophil count does not exclude the possibility of eosinophilic lung disease. Drug-induced lung disease can pose a challenge in diagnosis. Many patients who present with otherwise unexplained pulmonary eosinophilia may have underlying diseases that can be associated with eosinophilia or may be taking many medications that could cause eosinophilia. Although the most certain way to

Fig. 2. Regulation of eosinophil migration and degranulation in the lung.

J.N. Allen / Clin Chest Med 25 (2004) 77–88

79

Fig. 3. Alveolar macrophages present antigens (Ag) to Th2 lymphocytes by way of their MHC II. Lymphocyte T-cell receptors (TCR), in conjunction with CD4 receptors, recognize the antigen and stimulate the lymphocyte to produce IL-5, which activates eosinophils.

determine if a patient has drug-induced eosinophilic lung disease is for the eosinophilia to resolve after discontinuing the medication and then recur after rechallenging the patient, this can be risky and should be avoided in most cases. More practically, five criteria for eosinophilic drug-induced lung disease should be met. The patient should: (1) have no other likely cause of lung disease, (2) have symptoms consistent with the suspect drug, (3) have a time course compatible with drug-induced lung disease, (4) have tissue or BAL findings compatible with drug-induced lung disease, and (5) improve after the drug is discontinued. Patients who have all five of these criteria can be considered as having definite drug-induced eosinophilic lung disease, patients who meet four criteria can be considered as having probable disease, and those who meet three criteria are suspected to have the disease. Sometimes, entire classes of drugs should be considered as candidates for drug-induced eosinophilic lung disease. For example, many penicillins, cephalosporins, statins, and selective serotonin reuptake inhibitors have been associated with pulmonary eosinophilia. Therefore, a newly-released drug from one of these classes should be a suspected cause of eosinophilic lung disease, even if no reports yet exist that associate the new drug with lung disease. Lymphocyte transformation tests have been used to determine whether a given drug is responsible for eosinophilia. Although these studies may help to support the diagnosis, they can be negative even when

there is strong clinical evidence that a drug is responsible for eosinophilic lung disease; therefore their positive and negative predictive values are uncertain [3]. Moreover, these tests are not widely available for routine clinical use and are largely relegated to experimental use. In summary, the diagnosis of drug-induced eosinophilic lung disease is still established best based on the finding of increased blood or lung eosinophils and a clinical history of exposure to a drug that is known to cause eosinophilic lung disease.

Differential diagnosis Several diseases exist in which the lung contains abnormally increased eosinophils; it is important to exclude these diseases during the clinical evaluation of suspected eosinophilic lung disease. The most common eosinophilic lung diseases are listed in Box 1. Because drug-induced eosinophilic lung disease can mimic several other causes of eosinophilic lung disease, a brief mention of some of these is warranted. A more detailed discussion of these diseases may be found in other publications [4]. Simple pulmonary eosinophilia, or Lo¨effler’s syndrome, causes minimal or no pulmonary symptoms and transient, often migratory, radiographic infiltrates. Chronic eosinophilic pneumonia presents with weeks or months of progressive respiratory symptoms, diffuse or peripheral radiographic infiltrates, and increased

80

J.N. Allen / Clin Chest Med 25 (2004) 77–88

Box 1. Common eosinophilic lung diseases Simple pulmonary eosinophilia Chronic eosinophilic pneumonia Acute eosinophilic pneumonia Churg-Strauss syndrome Idiopathic hypereosinophilic syndrome Allergic bronchopulmonary Aspergillosis Parasite-induced pulmonary eosinophilia Fungal-induced pulmonary eosinophilia Drug-induced pulmonary eosinophilia

blood eosinophils [5]. Acute eosinophilic pneumonia presents with sudden onset of respiratory failure and usually lacks blood eosinophilia [6]. Churg-Strauss syndrome presents with asthma, blood eosinophilia, and granulomatous vasculitis that involves the lung and other organs [7]. Idiopathic hypereosinophilic syndrome presents with profound blood eosinophilia (generally greater than 1500/mL) for more than 6 months; many organs, including the lung, may become infiltrated with eosinophils [8]. Other diseases that can present as eosinophilic lung disease include fungal infections (especially aspergillosis and coccidioidomycosis), parasitic infections, interstitial lung diseases, allergic bronchopulmonary aspergillosis, bronchocentric granulomatosis, and malignancy [4].

Causes of drug-induced eosinophilic lung disease Every year, as new medications are released, the list of drugs that are reported to cause eosinophilic lung disease grows. The clinician who is faced with a patient who has undifferentiated eosinophilic lung disease needs to inquire about prescription drug use, over-the-counter medications, herbal preparations, and illicit drugs, all of which can be responsible. For any given drug, there are varying degrees of certainty that it can be responsible for pulmonary eosinophilia. For example, a single case report of a drug in the medical literature provides less certainty than multiple case series of eosinophilic lung disease that are associated with the same drug. One of the best resources for clinicians is the internet web site www.pneumotox.com, which is maintained by Dr. Pascal Foucher, Dr. Philippe Camus, and the Groupe d’Etudes de la Pathologie Pulmonaire Iatroge`ne. This site grades evidence that a given drug is responsible for a specific lung disease

Box 2. Drugs that commonly are reported to cause pulmonary eosinophilia Amiodarone [9] Bleomycin [10] Captopril [11] Gold salts [12] Iodine, radiographic contrast media [13] L-tryptophan [14] Methotrexate [4] Nitrofurantoin [4] Phenytoin [15]

in four categories: one to five isolated case reports, approximately ten cases, twenty to one hundred cases, and more than one hundred cases. A drug that is used rarely (eg, a chemotherapy medication) has fewer opportunities to cause iatrogenic effects than a drug that is used frequently (eg, an antibiotic). At the time of preparation of this article, www.pneumotox. com listed 110 drugs that were reported to cause pulmonary infiltrates with eosinophilia. The following drugs are reported to cause eosinophilic lung disease (Boxes 2 – 5). For most of these drugs, many references exist in the medical literature but only one pertinent citation is referenced. Two epidemics of drug-induced eosinophilic lung disease warrant special mention because of historical significance: eosinophilia-myalgia syndrome and the toxic oil syndrome. Eosinophilia-myalgia syndrome occurred in the 1980s [105] and was caused by a contaminant found in L-tryptophan that was manufactured by a single company [106]. At the time, L-tryptophan was prescribed commonly for insomnia and depression; it also was available in the United States without prescription and often Box 3. Drugs that occasionally are reported to cause pulmonary eosinophilia Acetylsalicylic acid [16] Carbamazepine [17] Granulocyte-macrophage colony-stimulating factor [18] Minocycline [19] Nilutamide [20] Penicillamine [21] Propylthiouracil [22] Sulfa-containing antibiotics [23] Sulfasalazine [24]

J.N. Allen / Clin Chest Med 25 (2004) 77–88

Box 4. Drugs that rarely are reported to cause pulmonary eosinophilia Beclomethasone [25] Chloroquine [26] Cocaine [27] Dapsone [28] Desipramine [29] Diclofenac [30] Erythromycin [31] Heroin [32] Imipramine [33] IL-2 [34] Isoniazid [35] Isotretinoin [36] Mesalamine [37] Methylphenidate [38] Paclitaxel [39] Para-(4)-aminosalicylic acid [40] Penicillins [41] Phenylbutazone [42] Procarbazine [43] Propranolol [12] Ranitidine [44] Trimipramine [45] Zafirlukast [46]

was sold in health food stores. In one series, approximately one half of persons who ingested the contaminated drug developed acute peripheral blood eosinophilia accompanied by severe myalgias and multiorgan involvement [107]. Respiratory findings occurred in more than 50% of patients [108] and included chest radiograph infiltrates, pleural effusions, dyspnea, cough, pulmonary hypertension, and respiratory muscle weakness. The IgE and creatinine kinase levels were normal. Pulmonary function tests usually revealed a reduced diffusing capacity, often with restriction. Lung histology demonstrated interstitial infiltration by lymphocytes and eosinophils, alveolar eosinophils, and small- to medium-vessel vasculitis [109]. Although pulmonary involvement generally was self-limited, patients frequently had lasting abnormalities of the musculoskeletal system, skin, and the nervous system. The ‘‘toxic oil’’ syndrome occurred in 1981 and 1982 when approximately 20,000 cases of lung disease that was accompanied by peripheral blood eosinophilia developed in Spain [110,111]. This epidemic was traced to rapeseed oil that was contaminated with oleoanilide that was marketed fraudulently

Box 5. Drugs that very rarely are reported to cause pulmonary eosinophilia Acetaminophen [47] Aminoglutethimide [48] Amitriptyline [49] Ampicillin [50] Azithromycin [51] Bicalutamide [52] Bucillamine [53] Cephalosporins [54] Chlorpromazine [55] Chlorpropamide [56] Cladribine [57] Clofibrate [58] Cromolyn [59] Cyproterone acetate [60] Diflunisal [61] Enalapril [62] Ethambutol [63] Febarbamate [64] Fenbufen [65] Fenoprofen [66] Fludarabine [67] Fosinopril [62] Furazolidone [68] Glafenine [69] Ibuprofen [70] Indomethacin [71] Labetalol [72] Levofloxacin [73] Lovastatin [74] Loxoprofen [75] Maprotiline [76] Mephenesin [77] Metronidazole [78] Montelukast [79] Nalidixic acid [80] Naproxen [81] Niflumic acid [82] Nomifensine [83] Oxaliplatin [84] Pentamidine [85] Perindopril [86] Piroxicam [87] Pranoprofen [88] Pyrimethamine [89] Roxithromycin [90] Serrapeptase [91] Sertraline [92] Sulindac [93] Tamoxifen [94]

81

82

J.N. Allen / Clin Chest Med 25 (2004) 77–88

Tenidap [95] Tetracycline [96] Tiaprofenic acid [97] Ticlopidine [98] Tolazamide [99] Tolfenamic acid [100] Tosufloxacin [101] Trazodone [102] Troleandomycin [103] Venlafaxine [104]

as olive oil, thus the name ‘‘toxic oil syndrome.’’ Symptoms typically included fever, respiratory distress, nausea, vomiting, pruritus, abdominal pain, headache, and cough. Signs included rash, pulmonary crackles, generalized lymphadenopathy, and hepatosplenomegaly. Chest radiographs showed interstitial or alveolar infiltrates, often with Kerley B lines. Corticosteroids usually were effective in the acute stage of the disease but not in the later stages. Pulmonary fibrosis and pulmonary hypertension were common late sequela; more than 300 fatalities were recorded. Although the causative agents for eosinophilia-myalgia syndrome and the toxic oil syndrome are no longer available commercially, these two episodes illustrate the speed and extent with which contaminants in medications or foods can result in epidemics of eosinophilic lung disease. It is likely that other large- or small-scale epidemics will occur as new medications and chemicals are produced.

Clinical patterns of drug-induced eosinophilic lung disease Drug induced eosinophilic lung disease can have several different clinical presentations, including simple pulmonary eosinophilia, chronic eosinophilic pneumonia, acute eosinophilic pneumonia, and ChurgStrauss syndrome. Isolated eosinophilic pleural effusions are rare causes of pulmonary manifestations of drug allergy; however, pleural effusions can occur as part of a generalized pulmonary reaction that includes parenchymal infiltrates, especially when drug-induced eosinophilic lung disease manifests as an acute eosinophilic pneumonia – like presentation. Drug-induced pleural effusions are reviewed elsewhere in this issue. Perhaps the most common form of drug-induced eosinophilic lung disease is simple pulmonary eosinophilia or Lo¨ffler’s syndrome. Patients may be

asymptomatic or may have mild cough or dyspnea. Chest radiographs typically show patchy segmental or subsegmental pulmonary infiltrates that often are migratory over time. Peripheral blood eosinophil counts usually are elevated. Signs and symptoms resolve with discontinuation of the drug and typically recur if it is resumed. Treatment with corticosteroids rarely is required. For many patients who have simple pulmonary eosinophilia, the cause cannot be identified. The most important cause of simple pulmonary eosinophilia worldwide is parasitic infection. In Loeffler’s original series [4], occult ascaris infection may have been responsible for most of the cases. In some patients, a subacute presentation can occur. These patients may present with cough and dyspnea that has lasted for weeks or months. In contrast to patients who have idiopathic chronic eosinophilic pneumonia who have symptoms that are limited to the lung, drug-induced eosinophilic lung disease is accompanied often by systemic symptoms, such as rash or fever. Peripheral blood eosinophilia usually is present. Infiltrates may be patchy or confluent, unilateral or bilateral. The main diagnostic consideration is idiopathic chronic eosinophilic pneumonia. Like chronic eosinophilic pneumonia, patients who present with this form of drug-induced eosinophilic lung disease improve promptly with corticosteroids. Unlike idiopathic chronic eosinophilic pneumonia, drug-induced eosinophilic lung disease resolves with discontinuation of the offending medication without the use of steroids. Patients who have idiopathic chronic eosinophilic pneumonia require months or years of steroids or relapse will occur. Conversely, patients who have drug-induced lung disease will not relapse after a brief course of steroids if the offending medication is discontinued. Many drugs have been reported to cause symptoms that present as acute eosinophilic pneumonia. Minocycline has been reported particularly frequently [112 – 114]. It is likely that drug allergy is just one kind of antigenic stimulation that culminates in acute eosinophilic pneumonia. New exposure to tobacco smoke also has been linked to acute eosinophilic pneumonia [115]. Patients present with just a few days of severe dyspnea and hypoxemia. Other symptoms, such as fever, myalgias, and pleuritic chest pain, are common. Patients who have idiopathic acute eosinophilic pneumonia have normal or only slightly elevated blood eosinophil counts, whereas those who have drug-induced acute eosinophilic pneumonia often have moderately or highly elevated blood eosinophil counts. An eosinophil count in excess of 1000/mL should heighten the suspicion of drug-induced, as opposed to idiopathic, acute eosinophilic

J.N. Allen / Clin Chest Med 25 (2004) 77–88

pneumonia. Chest radiograph findings in idiopathic and drug-induced acute eosinophilic pneumonia are similar [116]. The earliest findings are patchy interstitial infiltrates with Kerley B lines which progress to diffuse alveolar infiltrates with small- or moderatesized pleural effusions. Patients can improve simply by discontinuing the offending drug [112], but steroids usually are given to accelerate resolution of hypoxemia and dyspnea. Patients who have idiopathic acute eosinophilic pneumonia resolve completely following a brief course of corticosteroids; however, those who have drug-induced acute eosinophilic pneumonia will relapse if the offending drug is readministered [113]. Churg-Strauss syndrome has been linked to leukotriene inhibitors, including zifirlukast [117,118] and monolukast [79,119]. The clinical presentation of leukotriene inhibitor – associated Churg-Strauss syndrome cannot be distinguished from idiopathic ChurgStrauss syndrome. Patients present with antecedent asthma and allergic rhinitis followed by high blood eosinophil counts and vasculitis that involves many organs, including the skin, peripheral nerves, heart, kidney, and gastrointestinal system. It is unknown whether leukotriene inhibitors cause Churg-Strauss syndrome or if their use, as steroid-sparing agents in patients who have asthma, allows the development of symptoms of occult Churg-Strauss syndrome that were controlled previously by the steroids [120]. Treatment requires long-term corticosteroids plus cytotoxic medications, such as cyclophosphamide.

Clinical approach The history and physical examination are the most important tools in the diagnosis of druginduced eosinophilic lung disease. Patients must be questioned carefully because they may not be forthcoming about nonprescription drugs, herbal preparations, and street drugs. Many over-the-counter medications and herbal preparations have multiple ingredients and identification of the individual components can be difficult. Patients may be taking more than one drug that is known to cause eosinophilic lung disease; this can make incrimination of a single drug difficult. In outpatients, the temporal association of eosinophilia and pulmonary abnormalities with initiation of a given medication can be an important clue; inpatients frequently are started on many medications over a short period of time, which eliminates these clues. Often, drug-induced lung disease can be diagnosed only after excluding other causes of eosinophi-

83

lic lung disease. Thus, a history of asthma should raise the possibility of allergic bronchopulmonary aspergillosis, bronchocentric granulomatosis, or Churg-Strauss syndrome. Travel history may suggest parasitic infestation. Travel to relevant areas of the world should raise the possibility of filarial infection (tropical pulmonary eosinophilia), Schistosoma, or Paragonimus westermani. In the United States, Ascaris, Strongyloides, Toxacara, and Ancylostoma are the most common culprits [4]. Strongyloides is a human intestinal worm that can be particularly difficult to diagnose; it can exist for years or decades before causing respiratory symptoms, pulmonary infiltrates, and blood eosinophilia. It can be diagnosed by finding larvae in sputum, BAL, or stool. Stool examination can be negative, however, so diagnosis often requires the use of anti-Strongyloides antibodies in the blood. Ascaris infection causes pulmonary infiltrates and high blood eosinophil counts; skin rash is common. The stool examination usually is negative at the time of pulmonary symptoms and does not reveal ova until the worms mature in the intestine approximately 8 weeks later. T canis (‘‘visceral larval migrans’’) is the dog roundworm; adult worms do not develop in humans. It causes marked blood eosinophilia with pulmonary infiltrates as the larvae migrate through the internal organs. Stool examinations are negative and definitive diagnosis usually requires serology. Ancylostoma brasiliense is the dog hookworm and in humans causes ‘‘cutaneous larval migrans’’ also known as ‘‘creeping eruption.’’ Adult worms do not develop in humans so stool examinations are negative. The diagnosis is established by finding the characteristic serpiginous, red rash on the extremities in association with blood eosinophilia, and, occasionally, pulmonary infiltrates. Two fungal infections can present as eosinophilic lung disease; these are particularly important to recognize because administration of steroids for presumptive drug-induced lung disease can result in progressive infection and even death. Coccidioidomycosis may occur after travel to the southwestern United States and can result in pulmonary infiltrates with peripheral blood eosinophilia [121]. The diagnosis can be suggested by the presence of antibodies against Coccidioides immitis or by growth of Coccidioides in sputum or BAL cultures. Aspergillosis can occur after exposure to compost or decomposing organic materials [122]; it can be more difficult to identify than C immitis because it tends to be a more tissue invasive fungus and does not always grow in sputum or BAL cultures. Furthermore, many normal persons may have airway colonization with Aspergil-

84

J.N. Allen / Clin Chest Med 25 (2004) 77–88

lus species; therefore, growth of Aspergillus from respiratory specimens does not indicate true Aspergillus infection. Serologic studies can be suggestive but are rarely diagnostic. Lung biopsy may be required in difficult cases. Laboratory evaluation rarely is diagnostic for druginduced eosinophilic lung disease but may be useful for other diseases in the differential diagnosis. The blood eosinophil count frequently is elevated in druginduced eosinophilic lung disease, but also is elevated in many other forms of eosinophilic lung disease. The IgE level is nonspecific; however, a markedly elevated IgE (greater than 2000 ng/mL) is seen mainly with allergic bronchopulmonary aspergillosis and some cases of Churg-Strauss syndrome. Chest radiograph and CT findings are not specific. In a study of 111 patients who had eosinophilic lung disease who underwent thin section CT, the radiologist correctly diagnosed only 3 of the 11 patients who had druginduced lung disease [123]. Radiographs are useful to diagnose other lung diseases that may be confused with drug-induced lung disease. For example, bronchiectasis in the setting of increased blood eosinophils suggests allergic bronchopulmonary aspergillosis. BAL is performed often to exclude infection or other lung diseases. Although bacterial pneumonias are not associated with increases in either blood or BAL eosinophil counts, such increases can be seen in the setting of pneumonia if the patient has a coexistent drug allergy, for example, to an antibiotic. In the intensive care unit, the differentiation between druginduced lung disease and nosocomial pneumonia can be particularly vexing because allergy to drugs that are prescribed in the intensive care unit can develop coincident with pulmonary infiltrates, for a variety of other reasons, including pneumonia. BAL also can be useful to help exclude fungal or parasitic infection as a cause of pulmonary eosinophilia. BAL in drug-induced eosinophilic lung disease demonstrates elevated percentages of eosinophils, usually in association with elevated percentages of lymphocytes. In a study of 19 patients who had druginduced lung disease, increased lymphocytes were found in 95% of patients, but increased eosinophils were noted in 42% of patients [124]. Conversely, drug allergy is a common cause of increased BAL eosinophils. In a series of 48 patients who had more than 5% BAL eosinophils, drugs were responsible in 12% of patients [94]. Lung biopsy is not required to diagnose druginduced eosinophilic lung disease; however, biopsy may be required to diagnose other lung diseases, such as Churg-Strauss syndrome, Aspergillus infection, or interstitial lung disease. When performed in drug-

induced eosinophilic lung disease, lung biopsy is nonspecific and typically shows lymphocyte and eosinophil infiltration in the interstitium and alveoli. In some drug exposures (eg, L-tryptophan), vasculitis may be seen [125].

Summary For most patients who have suspected drug-induced eosinophilic lung disease, the history provides a presumptive diagnosis that can be confirmed by pulmonary findings and eosinophilia after cessation of the drug. As new drugs are developed and released for clinical use, many will result in eosinophilic lung disease in susceptible patients. Therefore, development of pulmonary abnormalities in conjunction with blood or lung eosinophilia after prescription of a newly released medication should raise the possibility of drug-induced lung disease, even if that medication has not yet been reported to cause eosinophilic lung disease. In all patients, the diagnosis requires exclusion of other causes of eosinophilic lung disease by history, and, if necessary, laboratory testing or lung biopsy.

References [1] Allen JN, Davis WB. Eosinophils and the lung. In: Crystal RG, West J, Weibel E, Barnes PL, editors. The lung: scientific foundations. New York: Raven Press, Ltd.; 1996. p. 905 – 16. [2] Conroy DM, Williams TJ. Eotaxin and the attraction of eosinophils to the asthmatic lung. Respir Res 2001; 2:150 – 6. [3] Primeau MN, Adkinson Jr NF. Recent advances in the diagnosis of drug allergy. Curr Opin Allergy Clin Immunol 2001;1:337 – 41. [4] Allen JN, Davis WB. State of the art: the eosinophilic lung diseases. Am J Respir Crit Care Med 1994;150: 1423 – 38. [5] Marchand E, Reynaud-Gaubert M, Lauque D, et al. Idiopathic chronic eosinophilic pneumonia: a clinical and follow up study of 62 cases. Medicine 1998;77: 299 – 312. [6] Pope-Harman AL, Davis WB, Christoforidis AJ, et al. Acute eosinophilic pneumonia: a summary of fifteen cases and review of the literature. Medicine 1966;75: 334 – 42. [7] Guillevin L, Cohen P, Gayraud M, et al. ChurgStrauss syndrome: clinical study and long-term follow-up of 96 patients. Medicine 1999;78:26 – 37. [8] Winn RE, Kollef MH, Meyer JI. Pulmonary involve-

J.N. Allen / Clin Chest Med 25 (2004) 77–88

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17] [18]

[19]

[20]

[21] [22]

[23]

[24]

[25]

ment in the hypereosinophilic syndrome. Chest 1994; 105:656 – 60. Darmanata JI, Van Zandwijk N, Duren DR, et al. Amiodarone pneumonitis: three further cases with a review of published reports. Thorax 1984;39:57 – 64. White DA, Kris MG, Stover DE. Bronchoalveolar lavage cell populations in bleomycin lung toxicity. Thorax 1987;42:551 – 2. Schatz PH, Mesologites D, Hyun J, et al. Captoprilinduced hypersensitivity lung disease. Chest 1989;95: 685 – 7. Akoun GM, Cadranel JL, Milleron BJ, et al. Bronchoalveolar lavage cell data in 19 patients with drugassociated pneumonitis (except amiodarone). Chest 1991;99:98 – 104. Jennings CA, Deveikis J, Azumi N, et al. Eosinophilic pneumonia associated with reaction to radiographic contrast medium. South Med J 1991;84: 92 – 5. Kaufman LD, Seidman RJ, Gruber BL. L-tryptophanassociated eosinophilic perimyositis, neuritis, and fasciitis. A clinicopathologic and laboratory study of 25 patients. Medicine 1990;69:187 – 99. Mahatma M, Haponik EF, Nelson S, et al. Phenytoininduced acute respiratory failure with pulmonary eosinophilia. Am J Med 1989;87:93 – 4. Schatz M, Wasserman S, Patterson R. Eosinophils and immunologic lung disease. Med Clin North Am 1981;65:1055 – 71. Cullinan SA, Bower GC. Acute pulmonary hypersensitivity to carbamazepine. Chest 1975;68:580 – 1. Gonzales-Chambers R, Rosenfeld C, Winkelstein A, et al. Eosinophilia resulting from administration of recombinant granulocyte-macrophage colony-stimulating factor (rhGM-CSF) in a patinet with T-g lymphoproliferative disease. Am J Hematol 1991;36: 157 – 9. Sitbon O, Bidel N, Dussopt C, et al. Minocyclineinduced eosinophilic pneumonitis: an analysis in 5 patients. Am Rev Respir Dis 1993;147:A76. Pfitzenmeyer P, Foucher P, Piard F, et al. Nilutamide pneumonitis: a report on eight patients. Thorax 1992; 47:622 – 7. Davies D, Lloyd-Jones J. Pulmonary eosinophilia caused by penicillamine. Thorax 1980;35:957 – 8. Middleton K, Santella R, Couser JIJ. Eosinophilic pleuritis due to propylthiouracil. Chest 1993;103: 955 – 6. Fiegenberg DS, Weiss H, Kirshman H. Migratory pneumonia with eosinophilia associated with sulfamide administration. Arch Intern Med 1967;120: 85 – 9. Sullivan SN. Sulfasalazine lung. Desensitization to sulfasalazine and treatment with acrylic coated 5-ASA and azodisalicylate. J Clin Gastroenterol 1987;9:461 – 3. Mollura JL, Bernstein R, Fine SR, et al. Pulmonary eosinophilia in a patient receiving beclomethasone dipropionate. Ann Allergy 1979;42:326 – 9.

85

[26] Coetmeur D, Guivarch G, Briens E, et al. Acute eosinophilic pneumonia. Possible role of chloroquine. Rev Mal Respir 1998;15:657 – 60. [27] Oh PI, Balter MS. Cocaine induced eosinophilic lung disease. Thorax 1992;47:478 – 9. [28] Jaffuel D, Lebel B, Hillaire-Buys D, et al. Eosinophilic pneumonia induced by dapsone. BMJ 1998; 317:181. [29] Panuska JR, King TR, Korenblat PE, et al. Hypersensitivity reaction to desipramine. J Allergy Clin Immunol 1987;80:18 – 23. [30] Khalil H, Molinary E, Stoller JK. Diclofenac (Voltaren)-induced eosinophilic pneumonitis. Case report and review of the literature. Arch Intern Med 1993; 153:1649 – 52. [31] Abramov LA, Yust IC, Fierstater ED, et al. Acute respiratory distress caused by erythromycin hypersensitivity. Arch Intern Med 1978;138:1156 – 8. [32] Brander PE, Tukiainen P. Acute eosinophilic pneumonia in a heroin smoker. Eur Respir J 1993;6: 750 – 2. [33] Wilson IC, Gambill JM, Sandifer MG. Loeffler’s syndrome occurring during imipramine therapy. Am J Psychiatry 1963;119:892 – 3. [34] Gonzales-Chambers R, Rosenfeld C, Winkelstein A, et al. Eosinophilia resulting from administration of recombinant granulocyte-macrophage colony-stimulating factor (rhGM-CSF) in a patient with T-g lymphoproliferative disease. Am J Hematol 1991;36: 157 – 9. [35] Perreau P, Fresneau M, Boumard B, et al. L’infiltrat pulmonaire avec e´osinophilie au cours du traitement par l’isoniazide [Pulmonary infiltrates with eosinophilia in the course of treatment with isoniazide]. Press Med 1955;63:1454 – 6 [in French]. [36] Oliviero G, Constans P, Caby I, et al. Lung disease induced by isotretinoin. Rev Mal Respir 1995;12: 631 – 3. [37] Tanigawa K, Sugiyama K, Matsuyama H, et al. Mesalazine-induced eosinophilic pneumonia. Respiration (Herrlisheim) 1999;66:69 – 72. [38] Wolf J, Fein A, Fehrenbacher L. Eosinophilic syndrome with methylphenidate abuse. Ann Intern Med 1978;89:224 – 5. [39] Fujimori K, Yokoyama A, Kurita Y, et al. Paclitaxelinduced cell-mediated hypersensitivity pneumonitis. Diagnosis using leukocyte migration test, bronchoalveolar lavage and transbronchial lung biopsy. Oncology 1998;55:340 – 4. [40] Cuthbert RJ. Lo¨ffler’s syndrome occurring during streptomycin and PAS therapy. BMJ 1954;2:398 – 9. [41] Riechlin S, Loveless MH, Kane EG. Loeffler’s syndrome following penicillin therapy. Ann Intern Med 1953;38:113 – 20. [42] Thuston JGB, Marks P, Trapnell D. Lung changes associated with phenylbutazone treatment. BMJ 1976; 2:1422 – 3. [43] Ecker MD, Jay B, Keohane MF. Procarbazine lung. AJR Am J Roentgenol 1978;131:527 – 8.

86

J.N. Allen / Clin Chest Med 25 (2004) 77–88

[44] Andreu V, Bataller R, Caballeria J, et al. Acute eosinophilic pneumonia associated with ranitidine. J Clin Gastroenterol 1996;23:160 – 2. [45] Hatzinger M, Stohler R, Hoslboer-Trachsler E. Eosinophile pleuritis und hepatopathie unter trimipramintherapie. [Eosinophilic pleuritis and hepatopathy under trimipramine therapy.] Schweiz Med Wochenschr 1991;121:910 – 2. [46] Knoell DL, Allen JN. Churg-Strauss syndrome associated with zafirlukast. Chest 1998;114:332 – 3. [47] Kondo K, Inoue Y, Hamada H, et al. Acetaminopheninduced eosinophilic pneumonia. Chest 1993;104: 291 – 2. [48] Bell SC, Anderson EG. Pulmonary eosinophilia associated with aminoglutethimide. Aust N Z J Med 1998;28:670 – 1. [49] Noh H, Lee YK, Kan SW, et al. Acute eosinophilic pneumonia associated with amitriptyline in a hemodialysis patient. Yonsei Med J 2001;42:357 – 9. [50] Poe RH, Condemi JJ, Weinstein SS, et al. Adult respiratory distress syndrome related to ampicillin sensitivity. Chest 1980;77:449 – 51. [51] Dietz A, Hubner C, Andrassy K. Macrolide antibioticinduced vasculitis (Churg-Strauss syndrome). Laryngorhinootologie 1998;77:111 – 4. [52] Wong P, Macris N, Difabrizio L, et al. Eosinophilic lung disease induced by bicalutamide. A case report and review of the medical literature. Chest 1998;113: 548 – 50. [53] Ogawa H, Fujimura M, Heki U, et al. Eosinophilic bronchitis presenting with only severe dry cough due to bucillamine. Respir Med 1995;89:219 – 21. [54] Felman RH, Sutherland DB, Conklin JL, et al. Eosinophilic cholecystitis, appendiceal inflammation, pericarditis, and cephalosporin-associated eosinophilia. Dig Dis Sci 1994;39:418 – 22. [55] Shear MK. Chlorpromazine-induced PIE syndrome. Am J Psychiatry 1978;135:492 – 3. [56] Bell RJ. Pulmonary infiltration with eosinophils caused by chlorpropamide. Lancet 1964;42: 1249 – 50. [57] Robak T, Blasins-Kamorawiec M, Krykowski E, et al. 2-chlorodeoxyadenosine (cladribine)-related eosinophilia in patients with lymphoproliferative diseases. Eur J Haematol 1997;59:216 – 20. [58] Hendrickson RM, Simpson F. Clofibrate and eosinophilic pneumonia. JAMA 1982;247:3082. [59] Repo UK, Nieminen P. Pulmonary infiltrates with eosinophilia and urinary symptoms during disodium cromoglycate treatment: a case report. Scand J Resp Dis 1976;57:1 – 4. [60] Similowski T, Orcel B, Derenne JP. CD8+ lymphocytic pneumonitis in a patient receiving cyproterone acetate. South Med J 1997;90:1048 – 9. [61] Rich MW, Thomas RA. A case of eosinophilic pneumonia and vasculitis induced by diflunisal. Chest 1997;111:1767 – 9. [62] Benzaquen-Forner H, Dournovo P, Tandjaoui-Lambiotte H, et al. Hypoxic eosinophilic pneumonia in

[63]

[64]

[65] [66] [67]

[68]

[69]

[70]

[71]

[72]

[73]

[74]

[75]

[76]

[77]

[78] [79] [80]

two patients treated with ACE inhibitors. Rev Mal Respir 1998;15:804 – 10. Wong PC, Yew WW, Wong CF, et al. Ethambutolinduced pulmonary infiltrates with eosinophilia and skin involvement. Eur Respir J 1995;8:866 – 8. Gali JM, Vilanova JL, Mayos M, et al. Febarbamateinduced pulmonary eosinophilia: a case report. Respiration (Herrlisheim) 1986;49:231 – 4. Burton GH. Rash and pulmonary eosinophilia associated with fenbufen. BMJ 1990;300:82 – 3. Barbandi F, Quenzer F, Rapp K. Pulmonary hypersensitivity to nalfon. Ann Allergy 1986;57:205 – 7. Trojan A, Meier R, Licht A, et al. Eosinophilic pneumonia after administration of fludarabine for the treatment of non-Hodgkin’s lymphoma. Ann Hematol 2002;81:535 – 7. Cortez LM, Pankey GA. Acute pulmonary hypersensitivity to furazolidone. Am Rev Respir Dis 1972; 105:823 – 6. Gheysens B, Van Mieghem W. Pulmonary infiltrates with eosinophilia due to glafenine. Eur J Respir Dis 1984;65:456 – 9. Goodwin SD, Glenny RW. Nonsteroidal anti-inflammatory drug-associated pulmonary infiltrates with eosinophilia. Review of the literature and Food and Drug Administration Adverse Drug Reaction reports. Arch Intern Med 1992;152:1521 – 4. Oishi Y, Sando Y, Tajima S, et al. Indomethacin induced bulky lymphadenopathy and eosinophilic pneumonia. Respirology 2001;6:57 – 60. Bonniaud P, Sgro C, Foucher P, et al. Labetalol, pneumopathie a` e´osinophiles et se´rologie lupique positive. Rev Mal Respir 1998;15(Suppl 1):1S103. Fujimori K, Shimatsu Y, Suzuki E, et al. Levofloxacin-induced eosinophilic pneumonia complicated by bronchial asthma. Nihon Kokyuki Gakkai Zasshi 2000;38:385 – 90. Liebhaber MI, Wright RS, Gelberg HJ, et al. Polymyalgia, hypersensitivity pneumonitis and other reactions in patients receiving HMG-CoA reductase inhibitors: a report of ten cases. Chest 1999;115:886 – 9. Li T, Doutsu Y, Ashitani J. A case of loxoprofeninduced pulmonary eosinophilia. Nihon Kyobu Shikkan Gakkai Zasshi 1992;30:926 – 9. Gaudenz R, Hartmann K, Reinhart WH, et al. Extensive eosinophilic pulmonary infiltrates in a depressive patient treated with maprotiline. Schweiz Rundsch Med Prax 1999;88:1047 – 51. Rodman T, Fraimow W, Myerson RM. Lo¨ffler’s syndrome: report of a case associated with administration of mephenesin carbamate (Tolseram). Ann Intern Med 1958;48:668 – 74. Kristenson M, Fryden A. Pneumonitis caused by metronidazole. JAMA 1988;260:184. Franco J, Artes MJ. Pulmonary eosinophilia associated with montelukast. Thorax 1999;54:558 – 60. Dan M, Aderka D, Topilsky M, et al. Hypersensitivity pneumonitis induced by nalidixic acid. Arch Intern Med 1986;146:1423 – 4.

J.N. Allen / Clin Chest Med 25 (2004) 77–88 [81] Buscaglia AJ, Cowden FE, Brill H. Pulmonary infiltrates associated with naproxen. JAMA 1984;251: 65 – 6. [82] Dan M, Aderka D, Topilsky M, et al. Hypersensitivity pneumonitis induced by nalidixic acid. Arch Intern Med 1986;146:1423 – 4. [83] Hamm H, Aumiller J, Bohmer R, et al. Alveolitis associated with nomifensine. Lancet 1985;1:1328 – 9. [84] Gagnadoux F, Roiron C, Carrie E, et al. Eosinophilic lung disease under chemotherapy with oxaliplatin for colorectal cancer. Am J Clin Oncol 2002;25: 388 – 90. [85] Dupon M, Malou M, Rogues AM, et al. Acute eosinophilic pneumonia induced by inhaled pentamidine isethionate. BMJ 1993;306:109. [86] Melloni B, Vergnenegre A, Bonnaud F, et al. Pneumopathie induite par le perindopril. A propos d’un cas. [Pneumopathy induced by prindopril. A case report.] Rev Fr Mal Respir 1994;11:308 – 11 [in French]. [87] Pfitzenmeyer P, Meier M, Zuck P, et al. Piroxicam induced pulmonary infiltrates and eosinophilia. J Rheumatol 1994;21:1573 – 7. [88] Fujimori K, Shimatsu Y, Suzuki E, et al. Pranoprofeninduced lung injury manifesting as acute eosinophilic pneumonia. Nihon Kokyuki Gakkai Zasshi 1999;37: 401 – 5. [89] Davidson AC, Bateman C, Shovlin C, et al. Pulmonary toxicity of malaria prophylaxis. BMJ 1988;297: 1240 – 1. [90] Perez-Castrillon JL, Jimenez-Garcia R, Martin-Escudero JC, et al. Roxithromycin-induced eosinophilic pneumonia. Ann Pharmacother 2002;36:1808 – 9. [91] Sasaki S, Kawanami R, Motizuki Y, et al. Serrapeptase-induced lung injury manifesting as acute eosinophilic pneumonia. Nihon Kokyuki Gakkai Zasshi 2000;38:540 – 4. [92] Barnes MT, Bascunana J, Garcia B, et al. Acute eosinophilic pneumonia associated with antidepressant agents. Pharm World Sci 1999;21:241 – 2. [93] Fein M. Sulindac and pneumonitis. Ann Intern Med 1981;95:245. [94] Allen JN, Davis WB, Pacht ER. Diagnostic significance of increased bronchoalveolar lavage fluid eosinophils. Am Rev Respir Dis 1990;142:642 – 7. [95] Middleton K, Santella R, Couser JIJ. Eosinophilicpleuritis due to propylthiouracil. Chest 1993;103: 955 – 6. [96] Ho D, Tashkin DP, Bein ME, et al. Pulmonary infiltrates with eosinophilia associated with tetracycline. Chest 1979;76:33 – 6. [97] Cauchie M, Lafitte JJ, Fortin F, et al. Poumon a` e´osinophiles et prise d’acide tiaprofe´nique. [Lung eosinophilia due to medications.] Rev Mal Respir 1989; 6:R131 [in French]. [98] Persoz CF, Cornella F, Kaeser P, et al. Ticlopidineinduced interstitial pulmonary disease: a case report. Chest 2001;119:1963 – 5. [99] Bondi E, Slater S. Tolazamide-induced chronic eosinophilic pneumonia. Chest 1981;80:652.

87

[100] Stro¨mberg C, Palva E, Alhava E, et al. Pulmonary infiltrates induced by tolfenamic acid. Lancet 1987; 2:685. [101] Kimura N, Miyazaki E, Matsuno O, et al. Druginduced pneumonitis with eosinophilic infiltration due to tosufloxacin tosilate. Nihon Kokyuki Gakkai Zasshi 1998;36:618 – 22. [102] Salerno SM, Strong JS, Roth BJ, et al. Eosinophilic pneumonia and respiratory failure associated with a trazodone over-dose. Am J Respir Crit Care Med 1995;152:2170 – 2. [103] Ducolone A, Vandevenne A, Ringwald P, et al. Poumon e´osinophile d’origine me´dicamenteuse. [Lung eosinophilia due to medications.] Semin Hop Paris 1986;62:1107 – 10 [in French]. [104] Fleisch MC, Blauer F, Gubler JG, et al. Eosinophilic pneumonia and respiratory failure associated with venlafaxine treatment. Eur Respir J 2001;15: 205 – 8. [105] Kaufman LD, Seidman RJ, Gruber BL. L-tryptophanassociated eosinophilic perimyositis, neuritis, and fasciitis. A clinicopathologic and laboratory study of 25 patients. Medicine 1990;69:187 – 99. [106] Philen RM, Hill RH, Flanders WD, et al. Tryptophan contaminants associated with eosinophilia-myalgia syndrome. Am J Epidemiol 1993;138:154 – 9. [107] Kamb ML, Murphy JJ, Jones JL, et al. Eosinophiliamyalgia syndrome in L-tryptophan-exposed patients. JAMA 1992;267:77 – 82. [108] Swygert LA, Maes EF, Sewell LE, et al. Eosinophiliamyalgia syndrome: results of national surveillance. JAMA 1990;264:1698 – 703. [109] Tazelaar HD, Myers JL, Drage CW, et al. Pulmonary disease associated with L-tryptophan-induced eosinophilic myalgia syndrome. Clinical and pathologic features. Chest 1990;97:1032 – 6. [110] Kilbourne EM, Rigau-Perez JG, Heath CW, et al. Clinical epidemiology of toxic oil syndrome: manifestations of a new illness. N Engl J Med 1983;309: 1408 – 14. [111] Alonso-Ruiz A, Calabozo M, Perez-Ruiz F, et al. Toxic oil syndrome. A long-term follow-up of a cohort of 332 patients. Medicine 1993;72:285 – 95. [112] Toyoshima M, Sato A, Hayakawa H, et al. A clinical study of minocycline-induced pneumonitis. Intern Med 1996;35:176 – 9. [113] Dussopt C, Mornex JF, Cordier JF, et al. Acute eosinophilic lung after a course of minocycline. Rev Mal Respir 1994;11:67 – 70. [114] Liegeon MN, De Blay F, Jaeger A, et al. A cause of respiratory distress: eosinophilic pneumopathy due to minocycline. Rev Mal Respir 1996;13:517 – 9. [115] Shintani H, Fujimura M, Ishiura Y, et al. A case of cigarette smoking-induced acute eosinophilic pneumonia showing tolerance. Chest 2000;117:277 – 9. [116] Yokoyama A, Mizushima Y, Suzuki H, et al. Acute eosinophilic pneumonia induced by minocycline: prominent Kerley B lines as a feature of positive re-challenge test. Jpn J Med 1990;29:195 – 8.

88

J.N. Allen / Clin Chest Med 25 (2004) 77–88

[117] Knoell DL, Lucas J, Allen JN. Churg-Strauss syndrome associated with zafirlukast. Chest 1998;114: 332 – 4. [118] Wechsler ME, Garpestad E, Flier SR, et al. Pulmonary infiltrates, eosinophilia, and cardiomyopathy following corticosteroid withdrawal in patients with asthma receiving zafirlukast. JAMA 1998;279:455 – 7. [119] Guilpain P, Viallard JF, Lagarde P, et al. ChurgStrauss syndrome in two patients receiving montelukast. Rheumatol 2002;41:535 – 9. [120] Weller PF, Plaut M, Taggart V, et al. The relationship of asthma therapy and Churg-Strauss syndrome: NIH workshop summary report. J Allergy Clin Immunol 2001;108:175 – 83. [121] Lombard CM, Tazelaar MD, Krasne DL. Pulmonary eosinophilia in coccidioidal infections. Chest 1987; 91:734 – 6.

[122] Ricker DH, Taylor SR, Gartner Jr JC, et al. Fatal pulmonary aspergillosis presenting as acute eosinophilic pneumonia in a previously healthy child. Chest 1991;100:875 – 7. [123] Johkoh T, Muller NL, Akira M, et al. Eosinophilic lung diseases: diagnostic accuracy of thin-section CT in 111 patients. Radiology 2000;216:773 – 80. [124] Akoun GM, Cadranel JL, Milleron BJ, et al. Bronchoalveolar lavage cell data in 19 patients with drugassociated pneumonitis (except amiodarone). Chest 1991;99:98 – 104. [125] Tazelaar HD, Myers JL, Strickler JG, et al. Tryptophan-induced lung disease: an immunophenotypic, immunofluorescent, and electron microscopic study. Mod Pathol 1993;6:56 – 60.