Clarithromycin and pulmonary infiltration with eosinophilia

Pulmonary diseases induced by drugs include bronchial asthma, pulmonary infiltration with eosinophilia, diffuse fibrosing alveolitis, vasculitis, and pleural diseases.^1–4 Most such diseases recede when the drug is withdrawn, although on rare occasions the pulmonary damage is irreversible and progressive.⁴,⁵ We describe a patient with asthma referred to our respiratory diseases clinic who twice developed fever and pulmonary infiltration with eosinophilia after taking antibiotics.

Case report

A 17 year old white man who has had bronchial asthma since childhood was referred to our clinic in January 2002. The patient also reported sinusitis and allergic rhinitis. Results of earlier prick tests and radioallergosorbent tests were positive for wall pellitory (Parietaria judaica) and grasses, and the tests resulted in a mild increase in peripheral blood eosinophil counts (0.6-0.7 × 10⁹/l (6-7%)). His general practitioner had prescribed salbutamol as a rescue treatment. The patient did not report any allergy to drugs.

In December 2001 he had reported fever (38°C), accompanied by mucopurulent nasal secretion and pain in his forehead. X ray pictures of the paranasal sinuses showed maxillary sinusitis on the right side and hypertrophy of the turbinates. The general practitioner prescribed combined amoxicillin (875 mg) and clavulanic acid (125 mg) twice daily for seven days, followed by clarithromycin (500 mg) twice daily for a further seven days. Figure 1 shows the patient's course of treatment. At the end of this treatment period the patient reported dry cough and mild dyspnoea. Chest x ray pictures showed pulmonary consolidations localised at the right apex.

Fig 1.

Temporal relation between the patient's course of treatment and pulmonary infiltration with eosinophilia (PIE)

A few days later he had a new episode of fever (38.5°C) and was admitted to the infectious diseases department of a hospital. His blood eosinophil count was 1.43 × 10⁹/l (14%) and his erythrocyte sedimentation rate was 52 mm in first hour. Results of faecal examination (parasites), throat culture (swab specimen), haemoculture, and the purified protein derivative tuberculin test were negative. He was treated with cefotaxime (1 g as intravenous injection) three times a day and clarithromycin (500 mg) twice daily. After a week of treatment he no longer had fever, his asthmatic symptoms improved mildly, and his peripheral eosinophil count was 3.74 × 10⁹/l (29%). Twelve days later, while he was still taking antibiotics, fever (39°C) returned, accompanied by dyspnoea and wheezing. His eosinophil count worsened, to 5.17 × 10⁹/l (35%). Computerised tomography of his chest showed bilateral peripheral pulmonary infiltrates (figure 2).

Fig 2.

Computerised tomography of lungs: in January 2002 (top), showing bilateral peripheral pulmonary infiltrates, multiple pulmonary consolidations peripherally localised at both superior and middle lobes, lymphadenopathy, and mild pleural effusion at the right side; two weeks later (middle), showing complete resolution of pulmonary consolidations, lymphadenopathy, and pleural effusion; and in April 2002 (bottom), showing recurrence of bilateral apical pulmonary infiltrates

A sputum test indicated eosinophilia, and serum IgE levels for parasites and aspergillus were negative. Bone marrow biopsy confirmed eosinophilia. The patient was referred to our lung unit, where antibiotic treatment was immediately withdrawn. Prednisone 25 mg twice daily, fluticasone propionate 250 μg from a metered dose inhaler (two puffs twice daily), formoterol 12 μg from a metered dose inhaler (one puff twice daily), and ranitidine 150 mg taken orally once daily were prescribed instead. A few days later his eosinophil count reduced dramatically to 0.27 × 10⁹/l (1%).

Two weeks later his eosinophil count was 0.92 × 10⁹/l (8%). Lung function tests showed a moderate obstructive ventilatory defect, accompanied by a reduction in carbon monoxide diffusing capacity: forced expiratory volume in 1 second (FEV₁) 59.9% of predicted value, ratio of FEV₁ to forced vital capacity 69.3%, carbon monoxide diffusing capacity (DLCO) 56.3%, and ratio of DLCO to alveolar volume 68.5%. Diffuse wheezing was detected in both lungs.

After a month his general condition improved. The dosage of prednisone was reduced to a maintenance dosage of 5 mg twice a day. His eosinophil count fell to 0.53 × 10⁹/l (7%). Computed tomography of his chest showed a complete recovery of the pulmonary consolidations (figure 2). Lung function tests showed an improvement of the obstructive defect, accompanied by a normal carbon monoxide transfer factor (FEV₁ 89.1%, ratio of FEV₁ to forced vital capacity 89.5%, DLCO 78.1%, and ratio of DLCO to alveolar volume 86.8% of predicted values). A bronchodilator (salbutamol) response test showed an increase in FEV₁ of 770 ml, a 22% increase from the baseline value.

In April 2002 his general practitioner prescribed clarithromycin 500 mg twice daily for the onset of rhinorrhoea accompanied by headache. Three days after the start of treatment the patient developed fever (37.2°C) and mild dyspnoea with eosinophilia (eosinophil count 1.84 × 10⁹/l (17%)). He was again referred to our lung unit. Computed tomography of the chest showed apical infiltrates in both lungs (figure 2).

Antibiotic treatment was immediately withdrawn, and the patient quickly recovered. Corticosteroid treatment was continued until his asthma symptoms were controlled and his peripheral blood eosinophil count was acceptable (0.44 × 10⁹/l (7%)). At follow up at six months his eosinophil count was 0.45 × 10⁹/l, and a chest x ray picture showed no abnormalities.

Discussion

As there was a temporal relation between our patient's symptoms and his antibiotic treatment, we quickly diagnosed drug induced pulmonary infiltration with eosinophilia. At first we suspected amoxicillin, given the number of reports in the international literature of reactions to penicillins.^6–9 But the second episode of pulmonary infiltration with eosinophilia after treatment with clarithromycin focused our attention on this drug. The unexpected rechallenge test when the patient's general practitioner prescribed clarithromycin a second time, followed by a new episode of pulmonary infiltration with eosinophilia, led to our revised diagnosis. This was aided by the fact that we weren't able to clarify the role of particular drugs in the first course of treatment, which included cephalosporin as well as macrolides. Studies describing the anti-inflammatory role of macrolides in patients with asthma have documented a reduction in pulmonary eosinophilia.¹⁰,¹¹ However, in rare cases macrolides have been shown to cause adverse events, such as eosinophilia, vasculitis, Churg-Strauss syndrome, colangitis, and hepatitis.^12–14 Eosinophilia as a reaction to clarithromycin has been reported in a patient with atopy and asthma¹²; but ours is the first case that we are aware of of pulmonary infiltration induced by clarithromycin.

Differential diagnoses include helmintic infestations, aspergillosis, tuberculosis, sarcoidosis, Hodgkin's disease, and other lymphoproliferative disturbances such as pulmonary eosinophilic granuloma, interstitial pneumonia, and vascular collagen disturbances.²,³,⁵,¹⁵ Appropriate laboratory tests include a blood cell count to determine the eosinophil count.¹⁶ Other important tests include plasma IgE, erythrocyte sedimentation rate, eosinophilia in sputum, faecal test for helminths, skin allergy tests, and plasma parasite evaluation.²,³,¹⁷ Radiographic characteristics of drug induced pulmonary infiltration with eosinophilia include focal alveolar opacity, diffuse alveolar opacity, diffuse interstitial opacity, and pulmonary nodules.⁵ Generally, lung function tests in the acute phase of pulmonary infiltration with eosinophilia show a restrictive ventilatory defect, which is accompanied by a reduction in carbon monoxide transfer factor. However, when asthma is present an obstructive ventilatory defect can appear during recovery, as in our patient.¹⁷,¹⁸

Often a diagnosis of drug induced pulmonary infiltration with eosinophilia is made on the basis of a temporal relation between drug treatment and the onset of symptoms.⁵,¹⁹ Rarely, when the clinical picture is not specific for the condition, diagnosis can be confirmed by a rechallenge test (which would be unethical) or a lymphocyte stimulation test with the suspected drug (which has low sensitivity).⁶,²⁰