Bromide-associated lower airway disease: a retrospective study of seven cats

Abstract

Seven cats were presented for mild-to-moderate cough and/or dyspnoea after starting bromide (Br) therapy for neurological diseases. The thoracic auscultation was abnormal in three cats showing increased respiratory sounds and wheezes. Haematology revealed mild eosinophilia in one cat. The thoracic radiographs showed bronchial patterns with peribronchial cuffing in most of them. Bronchoalveolar lavage performed in two cats revealed neutrophilic and eosinophilic inflammation. Histopathology conducted in one cat showed endogenous lipid pneumonia (EnLP). All cats improved with steroid therapy after Br discontinuation. Five cats were completely weaned off steroids, with no recurrence of clinical signs. In one cat, the treatment was discontinued despite persistent clinical signs. The cat presenting with EnLP developed secondary pneumothorax and did not recover. Br-associated lower airway disease can appear in cats after months of treatment and clinical improvement occurs only after discontinuing Br therapy.

Introduction

Potassium bromide (KBr) and sodium bromide (NaBr) are chemical compounds that were widely used for their anticonvulsive and sedative effects in the late 19th and early 20th centuries in human medicine, but they have since been replaced by other anticonvulsive drugs, such as phenobarbital. ¹

In dogs, KBr has been reported to be a safe, effective and economical anticonvulsive treatment for idiopathic epilepsy, either administered alone or as a component of a multi-drug protocol.^2–4 The adverse effects commonly associated with bromide administration in dogs are usually mild and self-limiting.^3,5,6 Polyphagia, vomiting, polyuria, ataxia and sedation are the most frequently encountered.^3,6

Cats do not respond to KBr seizure management as well as dogs do, and the therapeutic range may be different.^2,3 Furthermore, KBr use has been associated with respiratory signs in cats. The literature describing this condition is scant. In a retrospective study, 6/17 cats receiving KBr therapy for seizure control developed coughs. ⁷ Another retrospective study described coughing in 11/26 cats treated for epilepsy with KBr, and two were taken or referred to an emergency medicine hospital. ⁸ The respiratory signs — primarily cough — develop within a variable latency period (from 7 weeks to 14 months).^7–9 Bronchial patterns are the primary finding from thoracic radiographs. Bronchoalveolar lavage (BAL) revealed severe eosinophilic inflammation in one cat and mixed neutrophilic and eosinophilic inflammation in another. ⁸ This disease may be reversible after drug discontinuation but has been reported to be lethal in some cases (1/6 and 2/11 in studies by Boothe and Wagner, respectively).^7,8 The airway disorders seem to be the result of a hypersensitivity or allergic reaction without any dose or time dependence.^7,9

Numerous medications causing ‘drug-induced lung disease’, such as bronchitis or pneumonia, have been identified in human medicine.^10,11 Some of these reactions have been well described, including those to amiodarone and nitrofurantoin.^12,13 Clinical signs, such as coughing, dyspnoea or fever, may appear days or years after initiating treatment, or even months or years after discontinuing drug therapy.^10,14 The exact mechanisms of drug-induced lung disease are unknown in the majority of cases. ¹⁵ Direct cytotoxic effects have been described for some compounds, but most drugs require some form of bioactivation before inducing inflammatory and fibrotic lung changes.^10,14,15 BAL findings may show characteristic changes, such as those in amiodarone-induced lung injuries, but these changes are not specific to most drug-induced lung diseases.^14,16 Similarly, the histopathological findings ascribed to drug-induced lung disease are generally not specific to these drug reactions. ¹⁶ Discontinuing the offending drug often results in improvement and even resolution of the condition but, in some cases, the lung injury may be irreversible.^15,17,18 To our knowledge, no case report on KBr-induced lung disease exists in the human medical literature.

The aim of this retrospective study was to describe the clinical, haematological and radiographic findings, treatments and outcomes in cats with bromide-associated lower airway disease.

Case selection

The medical records of the Frégis Veterinary Hospital Center (CHVF) from 2007 to 2010 were reviewed for cats that were referred for lower airway signs while on bromide therapy for neurological disease. The diagnosis of bromide-associated lower airway disease was based on developing lower respiratory signs while on bromide therapy and improvement after drug cessation. The available work-up was collected for each case. The work-up included the signalment, history, time interval between beginning the KBr therapy and the first respiratory signs, treatment prior to referral and clinical signs. The haematological, biochemical, radiographic, bronchoscopic, BAL and histopathological findings, treatment, responses to treatment and follow-up data were also evaluated. For the cases followed up at CHVF, the referring veterinary surgeons and/or owners were contacted by telephone to obtain the long-term follow-up information.

Clinical summary

A summary of the signalment, clinical signs, diagnostic tests and long-term follow-up is provided in Table 1.

Table 1.

Signalment, clinical signs, diagnostic tests and long-term follow-up

Cat	1	2	3	4	5	6	7
Breed	DSH	Chartreux	DSH	DSH	DSH	DSH	DSH
Age (years)	2	4	3	8	2	1	3
Sex	MN	MN	MN	FN	FN	MN	MN
Duration of KBr therapy prior to onset of clinical signs (months)	10	2	8	96	8	1	2
Cough	+	+	+	+	+	+	+
Dyspnoea	-	-	+	+	-	-	-
Tachypnoea	-	-	-	+	+	-	-
Blood work	CBC	CBC,BC, FIV/FeLV	BC	CBC	BC	ND	ND
Thoracic radiographs	+	+	+	+	+	+	+
Bronchoscopy	ND	+	ND	ND	ND	ND	ND
BAL	ND	+	ND	ND	+	ND	ND
Histopathology	ND	ND	ND	ND	ND	ND	+
Duration of follow-up (months)	13	24	19	19	21	9	Died
Duration of bronchitis treatment (months)	2	24	4	4	5	2	2

MN = male neutered, FN = female neutered, DSH = domestic shorthair, CBC = complete blood count, FIV = feline immunodeficiency virus, FeLV = feline leukemia virus, BC = biochemistry, + = observed/realised, - = not observed, ND = not done

Signs

Seven cats fulfilled the inclusion criteria. All of the cats were referred for respiratory signs appearing during bromide treatment for neurological diseases. The onset of respiratory signs while on bromide therapy varied between 1 month and 8 years.

The age at presentation ranged from 1 to 8 years. All of the cats had received KBr (Crisax; TVM) — with the exception of cat 1 which had received NaBr (Nervicanis; Véto-Centre) — for various neurological diseases: idiopathic epilepsy in cats 2 and 4–7, seizures secondary to head trauma in cat 3 and cerebellar ataxia likely secondary to intrauterine panleukopenia virus infection in cat 1.

Clinical signs and physical examinations

All of the cats presented with coughing, and cats 3 and 4 were also dyspnoeic. The clinical signs were generally mild-to-moderate (with the exception of cat 3, which presented with severe respiratory distress). Diminished appetite was observed in cat 4.

The mean duration of respiratory signs prior to referral was 5.1 months (ranging from 1–14 months). The following treatments were given prior to arriving at CHVF: amoxicillin/clavulanic acid (Synulox; Pfizer) in cats 3 and 5; marbofloxacin (Marbocyl; Vétoquinol) and cefovecin (Convenia; Pfizer) in cat 2; doxycycline (Ronaxan; Merial) in cat 7; prednisolone (Megasolone; Merial or Dermipred; Sepval) in cats 2, 3, 6 and 7; sustained-release injectable methylprednisolone acetate (Vetacortyl; Vétoquinol) in cat 4; and butopiprine (Felitussyl; Sepval) in cats 2 and 6. No significant improvements had been observed with these treatments.

Severe restrictive dyspnoea (cat 3), moderate expiratory dyspnoea (cat 4), tachypnoea (cats 4 and 5), increased respiratory sounds (cats 4, 6 and 7) and wheezing (cat 6) were observed at the physical examination. The physical examination was unremarkable in cats 1 and 2.

Diagnostic test results

The routine haematology conducted in cats 1, 2 and 4 revealed mild eosinophilia (2270 cells/μl; reference interval: 100–1300 cells/μl) in cat 1 and was unremarkable in cats 2 and 4. Biochemistry profiles were performed in cats 2, 3 and 5, and the results were within normal limits. Cat 2 was tested for feline immunodeficiency virus and feline leukemia virus; the result was negative. The thoracic radiographs showed moderate-to-severe generalised bronchial patterns with peribronchial cuffing in cats 1–4, 6 and 7. A micronodular pattern was observed in cat 5. Multiple nodular lesions (4–20 mm) were found in cat 7 (Figure 1). Alveolar patterns, loss of lung lobe volume associated with mediastinal shifting, rib cage asymmetry and hemidiaphragm displacement consistent with atelectasis were noted in cats 3–5 and 7. Hyperinflated lung fields were found in cats 3–5. Finally, patchy pulmonary mineralisation was observed in cat 2 and a loss of normal tapering of the bronchial wall consistent with bronchiectasis was found in cat 3.

Figure 1.

A right lateral thoracic radiograph of cat 7, showing generalised peribronchial cuffing, nodular lesions (black arrows) and homogeneous increase in density in the caudal part of the left cranial lung lobe (white arrows)

Blind and bronchoscopy-guided BALs were performed in cats 2 and 5, respectively. The cytological examination revealed an apparently high degree of cellularity, which consisted of 80% neutrophils with 10% macrophages in cat 5 and 70% eosinophils in cat 2. Escherichia coli grew on a bacterial culture from cat 2. A control-blind BAL was conducted in both cats 1 month after ceasing KBr treatment and while on steroid therapy. An apparently moderate degree of cellularity, which consisted of 50% neutrophils and 10% macrophages, was noted in cat 5; an apparently high degree of cellularity, which consisted of 70% macrophages and 10% neutrophils, was found in cat 2.

In cat 7, ultrasound-guided fine needle aspirations of the nodular lesions observed on the thoracic radiographs were performed; cytology revealed degenerated neutrophils.

Treatment and outcome

Bromide treatment was discontinued in all of the cats and replaced with phenobarbital (Gardénal; Rhône-Poulenc Rorer), mean dose 2.8 mg/kg q12h orally (PO) in all of the cats except for cat 1, resulting in adequate seizure control. Cat 1 continued to present with mild signs of cerebellar ataxia but no treatment was recommended.

Cats 1–4 and 6: Positive outcome

All of the cats were discharged on the day of presentation, with the exception of cat 3, which required hospitalisation for 3 days. Cat 3 received oxygen therapy, dexamethasone (Dexadreson; Intervet), at a dose of 0.2 mg/kg q24h subcutaneously (SC) and terbutaline (Bricanyl; Astra) at a dose of 0.01 mg/kg q12h SC and was discharged on dexamethasone (Dexoral; Virbac) at a dose of 0.1 mg/kg q24h PO. The other cats were discharged on prednisolone (Megasolone; Merial) at a mean dose of 1.5 mg/kg q24h PO. Cats 1, 2 and 4 were also treated with inhaled fluticasone (Flixotide; GlaxoSmithKline) at a dose of 110 µg q12h for cat 1 and 220 µg q12h for cats 2 and 4. Antibiotics were given to two cats; cat 6 received amoxicillin/clavulanic acid (Synulox; Pfizer) at a dose of 20 mg/kg q12h PO and cat 3 received enrofloxacin (Baytril; Bayer) at a dose of 5.2 mg/kg q24h PO. Inhaled salbutamol (Ventoline; GlaxoSmithKline) at a mean dose of 100 µg q12h was used for 4 weeks in cat 3. Significant clinical improvements were observed in all of the cases after ceasing bromide therapy. The antibiotic therapy was discontinued in all of the cats after 2 weeks. The steroid dosage was progressively decreased; the mean duration of systemic steroid treatment was 3 months (range 2–4 months). All cats were successfully weaned off steroids, with complete resolution of clinical signs. Radiographic improvements (Figure 2a, b) were noted in cats 1 and 3. The severity of the bronchial patterns and peribronchial cuffing decreased. In cat 2, pulmonary mineralisation decreased but persisted for 20 months after the diagnosis. No radiographic monitoring was performed in cats 4 and 6.

Figure 2.

A severe bronchial pattern with peribronchial cuffing in cat 3 at the initial evaluation (a) and persistent moderate lesions after 7 months of treatment and bromide therapy cessation (b)

Cat 5: Persistent respiratory signs

Cat 5 was discharged on prednisolone (Megasolone; Merial) at a dose of 1.4 mg/kg q24h PO and enrofloxacin (Baytril; Bayer) at a dose of 5 mg/kg q24h PO. Significant clinical improvement was noted initially, but the owners discontinued steroid therapy after 5 months, despite persistent intermittent coughing and mild exercise intolerance. Six months after the diagnosis, radiological signs, including left cranial lobe atelectasis, were still present in cat 5. At the time of writing, the cat is coughing twice a day and not receiving treatment.

Cat 7: Negative outcome

Cat 7 was discharged on prednisolone (Megasolone; Merial) at a dose of 1.3 mg/kg q12h PO. The cat improved initially but developed severe restrictive dyspnoea secondary to a pneumothorax 2 months afterwards. After stabilising with oxygen therapy and thoracocentesis, the pneumothorax recurred and a thoracoscopy was performed. The thoracoscopy revealed multiple nodular lesions (Figure 3) in all of the pulmonary lobes. The nodules in the right caudal lobe were large (between 10 and 15 mm in diameter), and a rupture was observed in one nodule. A lobectomy of the caudal right lobe was performed. The cat did not adequately recover from the surgery and was euthanased. The histopathology revealed severe and extensive diffuse parenchymal granulomatous inflammatory infiltrates, including macrophages associated with multinucleated giant cells, cholesterol clefts and necrosis. The bronchial lumen was dilated and filled with necrotic material and degenerated neutrophils (Figure 4). These changes suggested endogenous lipid pneumonia (EnLP) associated with suppurative bronchitis. No extracellular or intracellular pathogens were detected. Periodic acid-Schiff, Fite-Faraco stainings and bacterial cultures were all negative.

Figure 3.

Multiple nodular lesions on the lung parenchyma (white arrows) observed during thoracoscopy in cat 7

Figure 4.

A granulomatous infiltrate associated with multinucleated giant cells (white arrow) and multiple cholesterol clefts (black arrow). Haematoxylin and eosin stain, magnification 20×

Discussion

There have been few reports in the veterinary literature describing bromide-associated lower airway disease in cats.^7–9 Potentially severe respiratory adverse effects have been described but they were considered generally reversible.^7,8 Long-term follow-up description of cats with bromide associated lower airways diseases are sparse.

All of the cats in previous studies had received bromide as KBr. In our study, one cat received bromide as NaBr.

The clinical presentation of the bromide-associated lower airway disease in the present study showed similarities and differences with the disease described in the literature.^7,8 The time intervals for developing respiratory signs after beginning bromide therapy in this study were similar to those previously reported (generally 1–2 years maximum).^7,8 However, one cat in this study developed respiratory signs 8 years after beginning therapy, suggesting that coughing may appear after several years of bromide therapy.

The cats were referred for coughing and dyspnoea. None of the cats had experienced respiratory signs before using bromides. The clinical signs were generally mild-to-moderate, as described previously, but some of the cats developed more severe clinical signs.^7,8 Cat 3 required initial hospitalisation for oxygen therapy. Cat 7 developed a spontaneous pneumothorax and did not recover from the thoracoscopy. In a previous report, 2 of 26 cats died secondary to bromide-associated lower airway disease. ⁸ In our study, all of the cats improved after being treated and ceasing bromide therapy, but only four of the cats could be weaned off steroid treatment without the respiratory signs recurring. In a case study by Wagner, 8/9 surviving cats had complete resolution of clinical signs, but one cat had persistent occasional coughing 7 months after being weaned off KBr. ⁸ In another study, 2/5 surviving cats were weaned off KBr and showed resolution of clinical signs. ⁷ These results suggest that bromide administration may lead to chronic inflammatory changes that require long-term steroid therapy, even after ceasing bromide treatment. This situation has also been described in human drug-associated lung disease (from amiodarone). ¹⁰

The diagnostic test results were generally non-specific. Bronchial patterns with peribronchial cuffing were the most common findings from the thoracic radiographs, and the few available BAL findings in our study revealed mixed inflammation, similar to previous reports. ⁸ The clinical presentation of cat 7 was different from the other cases in this study and from those reported in the literature. The thoracic radiographs revealed generalised bronchial patterns with peribronchial cuffing and the presence of nodular lesions. The histopathology revealed diffuse lipid pneumonia associated with suppurative bronchitis. Lipid pneumonia may be exogenous or endogenous.^19,20 Exogenous lipid pneumonia was excluded, as no mineral oil had been administered to this cat. Endogenous lipid pneumonia (EnLP) results from damage to pneumocytes that causes cell degeneration and the subsequent release of cholesterol and cholesterol esters into the alveolar spaces. ²⁰ These lipids provoke an inflammatory reaction and lipid phagocytosis by the alveolar macrophages. In humans, EnLP occurs primarily as a postobstructive bronchial condition associated with neoplastic-, infectious-, inflammatory-, toxin-, drug (amiodarone)- and metabolic-induced lung disease, but may also be idiopathic.^12,19–21 In cats, EnLP is frequently associated with non-infectious inflammatory airway diseases and pulmonary arterial thrombosis. EnLP has also been associated with neoplastic and infectious diseases. ¹⁹ In addition, pulmonary nodules in radiographs have been found to be associated with EnLP in cats. ¹⁹ The nodular lesions were severe in cat 7. An underlying infectious disease was thought to be unlikely, as the specific stainings and bacterial cultures did not reveal any pathogenic organisms. Feline infectious peritonitis and neoplastic infiltration were not diagnosed based on the histopathology. ²² Bromide-associated lower airway disease was considered to be the most likely cause of the EnLP. This presentation, combined with the inability to wean some of the cats off steroid therapy and the mortality seen in three cats from previously described cases, suggests that bromide may induce severe and irreversible lower airway disease in some cases. Therefore, using bromide for seizure control in cats cannot be recommended.

The clinical findings in the cats with bromide-associated lower airway disease are similar to those observed in cats with other feline bronchial diseases.^7,8,23 However, the history, exclusion of other respiratory diseases (infectious, neoplastic) and treatment responses supported the diagnosis. Steroid therapy was essentially unsuccessful when the bromide therapy was maintained. Significant improvement, up to resolution of the clinical signs, occurred as soon as the drug was discontinued.^7,8 This pattern has also been observed in human drug-associated pulmonary disease.^10,15,17,18 Re-challenging with the possible causative drug and inducing recurring clinical signs is one of the diagnostic criteria for drug-induced lung disease in human medicine, but such re-challenging is rarely performed. ¹⁴ For obvious ethical reasons, re-challenging was not conducted on the cats in the present study, but a re-challenge with the drug by the owner of cat 1 led to clinical sign recurrence.

The duration of bromide therapy appeared not to affect the outcomes in this study. The five cats weaned off steroid therapy without recurrence of clinical signs received a bromide course that ranged from 1 month to 8 years, and the other cats (cats 5 and 7) received bromide treatment for 1–8 months. Bromide is metabolised by kidneys, so renal insufficiency increases the risk of bromide toxicity.^5,7,24 Renal function was not evaluated in all of the cases, but no clinical history or signs associated with renal disease were observed.

The pathophysiology of drug-induced lung disease in humans has not been elucidated. ¹⁵ A combination of the direct cytotoxic effects of the offending drug or its metabolite- and immune-mediated reactions is suspected in the majority of cases. In humans, bromide administration has been associated with eosinophilic reactions, such as pruritic cutaneous allergic reactions (bromoderma). ²⁵ Alterations in bronchial secretions, mucociliary function and cytokine stimulation have also been described. ²⁶ Bromination of a tyrosine residue leading to increased eosinophil peroxidase activity may induce an eosinophil influx, causing respiratory tissue damage. ²⁵ Some adverse drug reactions in veterinary medicine have been commonly associated with eosinophilia in the peripheral blood or in BAL findings as part of a hypersensitivity reaction. ⁹ In cats, bromide-induced bronchitis is suspected to be the result of a hypersensitivity or allergic reaction without any dose or time dependence.^7,9 Serum bromide measurements were not performed in the present study. In this report, mild eosinophilia was observed in the haematology results of cat 1. The BAL findings revealed mixed inflammatory infiltrates, with eosinophils being the predominant cells in one of the two cats. These findings are similar to those described in the literature and may support the hypersensitivity reaction hypothesis. ⁸

A bias may be present in our study because the referral character of our practice may have favoured the selection of severe respiratory cases. The limited number of animals and available BAL findings, the histopathology results and the lack of other diagnostic imaging techniques, such as computed tomography (CT), are other limitations of this study. In people, some drug-induced lung diseases may show a typical CT scan or histopathological patterns for some drugs.^17,18,27 BAL is primarily used in human medicine to confirm inflammatory changes and exclude other diseases, such as infection or neoplasia, because no BAL findings are specific for any drug-induced lung disease. ¹⁶

In conclusion, bromide-associated lower airway disease was diagnosed in seven cats based on their histories, clinical and pathological findings and treatment responses. Clinical improvement from steroid therapy occurred only after discontinuing the bromide therapy. Nodular pulmonary lesions and EnLP with a secondary pneumothorax were observed in one cat. Bromide-associated lower airway disease should be included in the differential diagnosis for EnLP in cats. Although clinical signs can resolve, long-term steroid treatment may be required for managing bromide-associated lower airway disease in certain individuals. Further studies using systematic BAL and (ideally) pulmonary biopsies are needed to further elucidate the pathogenesis of bromide-associated lower airway disease and describe the typical cytological and histopathological findings, if any exist.