Acute Eosinophilic Pneumonia in a Patient with Inflammatory Bowel Disease
Corresponding author: Ms. Jacqueline N. Chu, Stanford University School of Medicine, Dept of Medicine, 300 Pasteur Dr, Room S102, MC, 5110, Stanford CA 94305, Tel: (703) 581-9224; E-mail: email@example.com
Pulmonary manifestations of inflammatory bowel disease(IBD) have been recognized as a rare complication since initially described in 1976, but are varied in presentation and pathology, making the diagnosis difficult. Acute eosinophilicpneumonia has been described a subtype of IBD lungdisease, but previous case reports are limited to IBD drugtherapy-induced eosinophilic pneumonia. Here we reporta case of acute eosinophilic pneumonia in a patient withulcerative colitis, independent of drug therapy, followed bya review of the literature on IBD-associated lung disease.
A 77-year-old Korean man with a history of ulcerativecolitis presented to our hospital with four days of a newcough, associated with dyspnea, weakness, and low-grade fevers. The cough was nonproductive and not associatedwith other upper respiratory illness symptoms, weight loss,night sweats, or hemoptysis. He did have a 6 year history ofa chronic productive cough, thought to be related to chronicsinusitis, but the patient felt this acute cough was moresevere. His past medical history was notable for chronicsinusitis, obstructive sleep apnea, ulcerative colitis (diagnosedand treated with total colectomy ten years prior),hypertension, hyperlipidemia, gout, benign prostatic hyperplasia,and duodenal ulcer. Medications included:ipratropium, fluticasone, losartan, atorvastatin, allopurinol,tamsulosin, finasteride, pantoprazole, and CPAP. Hedenied taking any anti-metabolite drugs such as sulfasalazine,mesalamine, or methotrexate. He had not startedany new medications recently. Family history was negativefor lung disease. Social history was negative for newexposures including significant dust, travel, or sick contacts.The patient was retired and did not have new or regularcontact with smoke, chemicals, or animals. He had neversmoked and did not drink alcohol.
On presentation, the patient was afebrile, with pulse of103, blood pressure 125/66, respiratory rate of 22, SpO293% on 2 liters per minute nasal cannula. He was a thin, well appearing man in no acute distress. Ears, nose, and throatexam revealed clear nares and oropharynx. Lung exam wasnotable for coarse breath sounds in all lung fields. Abdominalexam was notable for an intact ostomy with soft green-brownstool. The rest of the physical exam was unremarkable. Whiteblood cell count was 17.1 with a neutrophilic predominanceand no eosinophilia. Chest radiograph revealed bilateralpatchy infiltrates. The patient was diagnosed with community-acquired pneumonia and started on ceftriaxone and azithromycin.
However, over the next two days, the patient did not noteimprovement in his cough and continued to have coarsebreath sounds on exam and leukocytosis. On the third day of treatment, he began to spike new fevers to 38 degrees C,prompting re-imaging. His repeat CXR showed worsening ofthe bilateral infiltrates, and a chest CT revealed consolidativeand ground glass opacities in the dependent areas of the lung,suggestive of aspiration pneumonia. Interestingly, the patientdid not have known risk factors for aspiration; he did not haveneurologic deficits or an esophageal abnormality and was notan alcoholic. A swallow study was performed and was normal.However, the patient did have a chronic very productivecough and used a CPAP for obstructive sleep apnea; therefore,we suspected that he was aspirating on his chronic secretionswhile sleeping. His antibiotic therapy was thus broadened topiperacillin-tazobactam for anaerobic coverage, in addition tocontinuation of azithromycin for atypical coverage. Aspirationprecautions were initiated. A viral panel came back negative.
Over the next seven days, the patient’s acute cough persisted,his lung exam remained congested, and he continued to haveleukocytosis. Repeat CT on the eighth day of piperacillin-tazobactamshowed worsening of the bilateral infiltrates, again in a pattern consistent with aspiration. Bilateral pleural effusionswere now present. Given the patient’s lack of response to piperacillin-tazobactam and azithromycin, an unusual pathogensuch as tuberculosis or other fungal infection, or other lungdisease such as a vasculitis, sarcoidosis, or interstitial lungdisease were suspected. Quantiferon test, coccidiomycosiscomplement fixation and immunodiffusion tests were negative.The patient then underwent a bronchoalveolar lavage(BAL), which revealed 57% eosinophils. Cytology showedno malignant cells or viral inclusions and Gram stain wasnegative. Bacterial, fungal, and AFB cultures were negative.
The patient met diagnostic criteria for acute eosinophilicpneumonia: acute onset with febrile respiratory manifestations,bilateral infiltrates on imaging, oxygen saturation<90% on room air, lung eosinophilia with >25% eosinophilisin BAL fluid, and absence of causes such as infection,drugs, smoking, and inhaled dusts . He was started on oralprednisone 40mg daily with an improved lung exam,resolution of leukocytosis, and improved CXR within threedays. He was discharged from the hospital and kept onprednisone 40mg daily for six weeks, followed by a nineweek taper for a total course of 15 weeks with fullresolution of his disease.
Acute eosinophilic pneumonia (AEP) is an acute febrilerespiratory illness that often masquerades as infectiouspneumonia or acute respiratory distress syndrome . Diagnosticfeatures of AEP include: acute onset (< 1 month, morecommonly < 7 days) with febrile respiratory manifestations,bilateral diffuse infiltrates on imaging, severe hypoxia (PaO2on room air ≤ 50 mmHg, or PaO2/FiO2 ≤ 300 mmHg, or oxygensaturation on room air < 90%), and lung eosinophilia with≥ 25% eosinophils in BAL fluid . Causes of AEP includeinfection (especially parasitic), many drugs (most notablyNSAIDs; antibiotics, especially minocycline and nitrofurantoin;captopril, carbamazepine, and GM-CSF), and significant dustexposure [1-3]. AEP is otherwise idiopathic in etiology. Whilewe initially included AEP in our differential, our patient didnot have the exposures listed above. Moreover, idiopathicAEP typically occurs in previously healthy young adults (averageage of 30 years) ; our patient was 77 years old andhad multiple comorbidities. Two-thirds of patients withidiopathic AEP are smokers ; our patient was a never-smoker.Idiopathic AEP frequently causes acute respiratory failurerequiring intensive care unit admission and mechanicalventilation ; though our patient was significantly hypoxicwith oxygen saturation of about 90% on room air, he did not require intensive care. Given these discrepancies, weperformed a literature search and discovered the rarelyreported association between IBD and pulmonary disease, including acute eosinophilic pneumonia.
Pulmonary manifestations of inflammatory bowel diseasewere first described in 1976 in a case series of six patients .Since then, there have been more accounts of IBD-associated lung disease, more commonly in ulcerative colitis than Crohn’sdisease , though the diagnosis continues to be rare. One contributing factor to the infrequency of the diagnosis is thatsuch lung disease is difficult to diagnose. According to Camus et al., lung disease presents on average 9 years following onsetof inflammatory bowel disease . Additionally, about 30% ofpatients are post-colectomy, and 90% have inactive disease. Because the relationship between the patient’s IBD andtheir acute lung disease is so remote, clinicians rarely considerthe diagnosis of IBD-related lung disease, and the prevalenceis likely under-recognized. Another challenging aspect of diagnosisis the varied pathology of IBD-associated lung disease.Roughly 60% of cases are respiratory disease involving theairways and parenchyma; the majority of these are bronchiec tasis, chronic bronchitis, and interstitial lung disease . About 40% of cases are serositis, such as myocarditis, pericarditis, and pleuropericarditis . Regardless of the specific pathology, patients respond very well to steroids, either systemic or inhaled (for airway disease) . There are several theories as to pathogenesis of lung disease in IBD. One theory is cross-reactivity between autoantibodies against bowel antigens and antigens in the lungs; the common embryologic origin of the gastrointestinal tract and the lungs make both organs similarly susceptible to the inflammatory process underlying IBD [6,7]. Another theory suggests that inflammatory cells, induced by chronic inflammation in the bowel in IBD, aberrantly home to the lungs and cause respiratory disease . Genetic susceptibility may also play a role, with high concordance between twins in extra-intestinal manifestations of IBD in general .
AEP has been described in association with IBD, though almost always as a drug reaction from anti-metabolites for IBD (specifically mesalamine, sulfasalazine, and methotrexate) . Though reviews report occurrences of eosinophilic pneumonia in patients not taking such drugs [9,10], we could not find specific case reports of AEP in the literature. We therefore believe that this is the first case report of a patient with IBD developing AEP as a direct result of his inflammatory disease, rather than as a drug side effect. Our patient did not fit the clinical picture of idiopathic AEP; rather, he fit the profile of IBD-associated lung disease as described above: our patient had ulcerative colitis rather than Crohn’s disease, he developed his lung disease 10 years after the onset of his ulcerative colitis, and he had inactive disease following a total colectomy. Additionally, IBD-associated lung disease is more indolent and subtle ; though our patient presented with an acute eosinophilic pneumonia, his course was much more benign than usual cases of idiopathic AEP, which may require mechanical ventilation.
Based on the similarity of our patient’s clinical presentation to that of other cases of IBD-associated lung disease, we suspect that this case represents a novel example of acute eosinophilic pneumonia associated with IBD itself, independent of a drug effect.
This case was presented at the Society of General Internal Medicine California-Hawaii 2014 Meeting on January 31, 2014.
The authors have no conflicts of interest.
8.Birring SS, Brightling CE, Symon FA, Barlow SG, Wardlaw AJ et al. Idiopathic chronic cough: association with organ specific autoimmune disease and bronchoalveolar lymphocytosis. Thorax. 2003, 58(12): 1066-1070.