Although airway disease associated with inflammatory bowel disease is uncommon, its involvement may have severe clinical consequences. This article reviews the breadth of pathologic processes that can be expected in the various sizes of airways, and provides a differential diagnosis from other airway diseases that can be found in association with inflammatory bowel disease. It also makes suggestions as to how airway disease can be best differentiated by using appropriate special stains.
Inflammatory bowel disease (IBD), including both ulcerative colitis (UC) and Crohn disease (CD), are considered to be systemic conditions with a wide variety of extraintestinal manifestations. The frequency of these manifestations appears to vary between the 2 diseases1 but may be present in as many as 40% of cases. While overt pulmonary involvement is relatively rare, affecting 0.4% of 1400 patients in one study,2 it can have severe clinical consequences. Major patterns of lung disease include upper airway disease with stenosis, tracheal and large airway diseases with inflammation and bronchiectasis, small airway diseases, and parenchymal lung diseases.3 While clinical and radiologic findings are the mainstay in differentiating between lung disease due to the effects of therapy and disease induced directly by IBD, these may not be sufficient, and a lung biopsy or autopsy specimen may result. For the pathologist, diagnosis of large and small airway diseases and their appropriate attribution to IBD are the most problematic. This article is therefore directed toward the diagnosis and differential diagnosis of large and small airway pathology in IBD.
LARGE AIRWAY INVOLVEMENT
Bronchiectasis is the most common pulmonary manifestation associated with IBD, but in a study that involved 1258 patients from 7 European cohorts without cystic fibrosis, bronchiectasis was ascribed to IBD in only 1.9% of patients.4 In comparison, this was only slightly more frequent than bronchiectasis related to ciliary dysfunction and approximately a third of the cases due to immunodeficiency. Interestingly, while bronchiectasis is usually associated with active IBD, it has also been reported to develop after colectomy.5 There is also a single report describing diffuse panbronchiolitis with symptoms that preceded those of UC, and improvement in the respiratory disease after colectomy.6
Bronchiectasis results from airway injury, which can be caused by a wide variety of insults. It is defined as a progressive large airway disease characterized by an irreversible dilatation of the bronchial tree, which can be identified both radiologically and pathologically.7 The pathologic process of bronchiectasis in IBD may be nonspecific with mixed inflammatory infiltrate and epithelial ulcerations. Granulomatous inflammation may be found with loosely formed aggregates of multinucleated giant cells. For the pathologist, in the absence of granulomatous inflammation, it would be very difficult to ascribe bronchiectasis solely to IBD. However, the possibility of this association should still be considered in a patient who has bronchiectasis and IBD.
SMALL AIRWAY INVOLVEMENT
Small airway involvement in IBD is clinically uncommon, but high-resolution computed tomography has been able to identify alterations of these airways in patients with IBD. This is true even when pulmonary function test findings are normal, and computed tomography abnormalities have been identified in a significant proportion of patients before the onset of gastrointestinal symptoms (reviewed in Black et al8). There are 2 different patterns of small airway involvement in IBD: bronchiolitis and obliterative bronchiolitis (see below). In both UC and CD, there can be bronchiolitis with a mild to moderate infiltrate of mononuclear cells in the airway wall. This is a nonspecific finding with a broad differential diagnosis including infection, aspiration, and drug reaction. A pattern that favors Crohn disease–associated bronchiolitis is characterized by nonnecrotizing granulomatous inflammation in the bronchiolar wall, sometimes highlighted by numerous giant cells9 (Figure 1). This pattern requires ruling out aspiration and infection, so additional examination using polarized light and special stains for acid-fast bacilli and fungi is important. This is especially imperative with patients receiving anti–tumor necrosis factor (TNF) therapy, as opportunistic infections are an important consideration (see below for discussion). Another pattern of IBD-associated bronchiolitis is characterized by intense acute inflammation with necrosis and microabscess formation, usually in the adjacent lung parenchyma9 (Figure 2).
Obliterative bronchiolitis has 2 forms, polypoid and constrictive, which can be present in isolation or can exist together in the same lung or even in the same airway. Polypoid (proliferative) obliterative bronchiolitis with or without associated organizing pneumonia is relatively nonspecific with the same differential diagnosis as described above for other patterns of bronchiolitis. Polypoid obliterative bronchiolitis is characterized by intraluminal polyps of granulation tissue usually without multinucleated giant cells. In contrast, the pathologic findings of constrictive bronchiolitis can be easy to miss, in both mild disease when findings are subtle and even in severe disease when the airway is obliterated. When the bronchiole is only mildly affected, there is an increase in subepithelial fibrous tissue, which can be highlighted by a fibrous tissue stain such as Masson trichrome or Movat pentachrome (Figure 3, A and B). Pathologists must therefore remember that in the normal small bronchioles, the epithelial layer is generally in close proximity to the underlying muscular layer, and that any increase in the subepithelial fibrous tissue compartment is abnormal. Since the degree of airway resistance is related to 1/radius,4 even a tiny degree of narrowing will have a significant effect on pulmonary function. In comparison, when the airway is completely obliterated, an elastic stain may be necessary to show that an area of fibrosis adjacent to a pulmonary artery is actually an airway remnant (Figure 4, A and B).
DIFFERENTIAL FROM LUNG DISEASES ASSOCIATED WITH IBD
Association With Asthma
There are 2 ways that this association can occur depending upon whether IBD is a primary or secondary diagnostic event. A nationwide cohort study in Taiwan10 found that there was an increased risk of asthma in patients with UC and CD, with an adjusted hazard ratio of 1.46 and 1.50, respectively. A population-based study of 8072 cases of IBD in Manitoba, Canada,11 has shown that asthma is a common comorbid disease, with a prevalence of 7.9% and 7.1% in UC and CD, respectively. These authors also found that patients with asthma had an age-adjusted increased risk of 1.56 and 1.38 of being diagnosed with either UC or CD, respectively, with no sex predilection. Similar findings were present in a Quebec-based study,12 which included 136 178 subjects with asthma and 143 904 subjects with chronic obstructive pulmonary disease (COPD). The authors found that the incidence of CD in asthma and COPD was 27% and 55% greater, respectively, than that of the general population, while the incidence of UC was 30% greater in COPD patients than in the general population.
For the pathologist, the abovementioned information is important when ascribing abnormalities of the airways to IBD. Pathologically, there is not much overlap between the 2 diseases. In the large airways, asthma will have a thickened basement membrane and extensive goblet cell metaplasia of the bronchial mucosal epithelium, while the bronchiectasis of IBD is inflammatory. The small airways in asthma may or may not demonstrate basement membrane thickening, but will have goblet cell metaplasia and varying amounts of smooth muscle hyperplasia, which are not features of IBD-associated disease. The presence of subepithelial fibrosis is not a diagnostic discriminator but in general, this is mild in degree in asthmatic airways as opposed to the definite constrictive and/or proliferative/polypoid obliterative bronchiolitis of IBD. The inflammatory infiltrate in asthma tends to be mild to moderate with eosinophils and lymphocytes, while active obliterative bronchiolitis of IBD is lymphoplasmacytic and may have giant cells. Adventitial fibrosis can be found in both conditions.
Association With Sarcoidosis
There is now a considerable wealth of literature documenting the rare coexistence of sarcoidosis and both UC and CD.8,13,14 This may have a genetic basis, as sarcoidosis and IBD share “risk loci” 15 and human leukocyte antigen haplotypes (reviewed in Black et al8). Although both may respond to anti-inflammatory drug therapies, it is important to remember that the 2 diseases follow independent clinical courses, and it is necessary to make the appropriate individual diagnoses.
For the pathologist, if only dealing with a bronchial biopsy specimen, it might be very difficult to differentiate between sarcoid granuloma and CD bronchiolitis, as the granulomata are relatively similar. However, a wedge biopsy would highlight the lymphatic distribution of sarcoidosis, which is not present in CD. Furthermore, granulomatous lymphadenitis in an endobronchial ultrasound-guided biopsy of the mediastinal lymph nodes would strongly favor sarcoid.
Association With Infections
When considering CD-associated airway disease with the presence of granulomatous inflammation, it is essential to rule out infections by culture and special stains; clinical correlations are extremely important in these cases. This is especially true when the patient is receiving anti–TNF-α monoclonal antibody therapy owing to the association with opportunistic infections, including tuberculosis and infections caused by Pneumocystis jirovecii and various dimorphic fungi (see Schleriermacher and Hoffmann16 for details).
No definitive pathogenic mechanism has been identified to explain either IBD or its associated airway diseases. The fact that both the lung and the gastrointestinal tract arise from the primitive foregut, and both have mucosa-associated lymphoid tissue, at least provides a commonality between the 2 systems. It is postulated that there may be a failure of the gastrointestinal mucosal barrier with increased intestinal permeability to fecal antigens that then migrate to the lung, following which there is activation of the acquired immune system. Upregulation of this system would result in activation of inflammatory cells and inflammatory mediators. It would also explain the association between IBD and asthma. It is also possible that bowel or bacterial antigens cross-react with airway antigens, resulting in an autoantibody type of immunologic reaction. Finally, epigenetic or genetic influences cannot be discounted (see Majewski and Piotrowski17 for discussion). Although there is some evidence that would support these theories, they remain highly speculative at the present time.
Most lung abnormalities in patients with IBD are parenchymal based, and are due to drug reactions or the secondary effects of therapy; clinical history is extremely important in differentiating these lesions from those primarily due to IBD itself. Inflammatory bowel disease–associated airway disease can affect any part of the respiratory tree and it can be very difficult to definitively ascribe pathologic changes to IBD. In the small airways, pathologic changes may be very subtle and may require special stains to demonstrate increased subepithelial collagen. Crohn disease may show granulomatous inflammation, in which case, opportunistic infections must be ruled out by culture and special stains, especially when the patient is receiving anti–TNF-α inhibitors. Granulomatous inflammation attributed to anti-TNF therapy directly has been described for patients being treated for inflammatory arthritis (after infection has been ruled out of course),18 and although this has not been characterized in patients treated for IBD to the best of our knowledge, it should be considered as part of the differential diagnosis. Finally, airway disease due to IBD must be differentiated from diseases that themselves are associated with IBD, and which have a prominent airway component as a part of their individual pathologic processes.
The authors have no relevant financial interest in the products or companies described in this article.