A 7-Year-Old With Persistent Fever and Cough

The patient is a 7-year-old, previously healthy girl who was admitted with a 7-week history of fevers and a worsening cough. She initially presented to her primary care physician in December 2019, before the coronavirus disease 2019 pandemic, after developing daily fevers as high as 102.6°F and a wet cough. She was treated with a 5-day course of azithromycin for suspected pneumonia but her fevers and cough persisted. She was seen again by the primary care physician, who diagnosed her with acute otitis media and prescribed amoxicillin-clavulanic acid, of which she completed 7 out of 10 days. Given her worsening fever curve and continued cough despite these antibiotic courses, the patient was subsequently admitted to an outside hospital (OSH) for additional workup. At that time, 12 days after the onset of her symptoms, she was noted to have persistent high fevers, a spasmodic cough, mild diarrhea (thought to be secondary to multiple antibiotic courses), decreased oral intake, and a 5-pound weight loss.

The patient’s past medical history was unremarkable, void of hospitalizations or chronic medical conditions, and her immunizations were up-to-date except for the influenza vaccine. She had no personal or family history of persistent fevers or recurrent infections. Her family history was otherwise void of malignancy or autoimmune disease except for Crohn disease in her paternal grandfather. She had no sick contacts but attended school regularly. The patient’s family lived in a home built on an old potato farm in the mid-Atlantic region. She had no exposure to individuals who had lived in homeless shelters in the past or were from foreign countries. She strictly drank well water and reported no new or raw food exposures. Her most recent travel included vacations to Naples, Florida in November 2019, where she notably fed giraffes; Bermuda in 2017; and Puerto Rico in 2014. She reported no recent rodent or insect exposures but did have a pet dog and parakeet.

After admission to the OSH, the patient was started on intravenous (IV) ampicillin-sulbactam, and both infectious disease and pulmonology were consulted. An extensive workup was performed; however, an underlying etiology was unable to be identified (Fig 1). The most notable laboratory findings were markedly elevated inflammatory markers (erythrocyte sedimentation rate [ESR] 75 mm/hr and C-reactive protein [CRP] 19.78 mg/dL). The patient’s fevers slowly improved, and she was discharged from the hospital to complete a 10-day course of cefdinir, of which she ultimately completed 7 days. Shortly after returning home, the patient once again began having daily fevers. After an emergency department visit with a reassuring workup and noted improvement in her CRP and white blood cell count, with a stably elevated ESR (Fig 1), she was referred to outpatient rheumatology for her ongoing fevers. Dr Mehta, what are your thoughts about this patient’s clinical course? Would a rheumatologic process be high on your differential?

At this point, the patient had been having ∼3 to 4 weeks of fevers without any stigmata of common rheumatologic conditions, such as those of systemic juvenile idiopathic arthritis (JIA), Kawasaki disease, or rheumatic fever. As mentioned, her laboratory tests were underwhelming from a rheumatology perspective, and by the time of this outpatient visit, her CRP appeared to be down-trending. Given all of this, I would agree with the outpatient rheumatologist’s assessment that there were no clear or obvious symptoms pointing toward an autoimmune disease and that a diagnosis of recurrent viral infections could explain her illness course.

The patient was also referred to pulmonology given the persistence of her barky cough. She was started on a 5-day course of prednisone, which resulted in significant improvement in and near resolution of her symptoms. The patient began to decline again after the completion of her steroid course, with return of her fevers and worsening of her cough. She was seen by her primary care physician and was given dexamethasone with transient improvement in symptoms. A lateral neck radiograph performed then was normal, and subsequent nasopharyngolaryngoscopy by otorhinolaryngology (ENT) was similarly unremarkable (Fig 1). After completion of this second steroid course, the patient’s high fevers resumed and her cough intensified. She had outpatient laboratory tests obtained, which revealed a worsening leukocytosis with left shift and, again, significantly elevated CRP and ESR (Fig 1), prompting presentation to the OSH emergency department. She was subsequently admitted with notable unremitting high fevers and a worsening barky cough. A computed tomography (CT) scan of the chest and neck revealed prominent hilar and mediastinal lymph nodes (Fig 1). A blood culture was obtained, and she was started on IV ampicillin-sulbactam and vancomycin. She was subsequently transferred to the authors’ institution for additional evaluation in February 2020. On arrival, the patient’s vital signs were notable for a fever to 40.1°C and tachycardia with heart rates in the 140s. In terms of growth metrics, she weighed 27.5 kg (72nd percentile, Z-score 0.58) and was 129 cm tall (69th percentile) with a BMI of 16.5 (66th percentile, Z-score 0.41). Her physical examination was notable for a tired, pale, and ill-appearing girl with mild stridor, a muffled voice, and shotty cervical lymphadenopathy. The infectious disease team was subsequently consulted.

Dr Russo, after meeting this patient, where did an infectious etiology fall on your differential, given her protracted fever history, leukocytosis with left shift, and elevated inflammatory markers?

I thought an infectious process was highly unlikely given the patient’s time course and clinical picture. Typically, an infectious process will either self-resolve in days to weeks, or a specific treatment will eradicate the infection. Or, if neither of these options occurs, the infection will progress to some point at which the patient becomes significantly worse, with the time course usually being days to, perhaps, 2 weeks. This patient, on the other hand, was more or less the same for 2 months, a static course. If this was of bacterial or fungal etiology, we should have observed localizing symptoms or progressive sickening of the patient. I was extremely struck by the fact that the only thing the fever really responded to was her 2 courses of steroids and not the antibiotics. I was so convinced that this was not infectious that I strongly recommended discontinuation of the antibiotics started at the OSH, and they were stopped on hospital day (HD) 2.

The patient’s admission laboratory tests were also notable for hypoalbuminemia (3.4 g/dL; reference range 3.7–5.6 g/dL) and microcytic anemia, as well as elevated ferritin and fecal calprotectin (Fig 1). Her peripheral smear, uric acid, and lactate dehydrogenase were all within normal limits. Infectious study results, including blood and urine cultures, cytomegalovirus serologies, and Bartonella serologies were negative. An echocardiogram was normal. A second review by radiology of the patient’s OSH CT chest and neck scan that was obtained before transfer noted that a perihilar and precarinal lymph node appeared prominent but nonpathologic.

Dr Mehta, now that the patient had continued symptoms for an additional 3 to 4 weeks without significant improvement, did that change the likelihood of an underlying autoimmune process? What was your differential for the patient at this point, given her laboratory tests, imaging, and clinical findings?

A combination of 7 weeks of fever and cough, as well as new complaints by the patient of some mild ear and jaw pain during admission, leads one down a path of infectious, oncologic, and rheumatologic etiologies for such a fever of unknown origin. From a rheumatologic perspective, a croup-like cough can be associated with antineutrophilic cytoplasmic autoantibody (ANCA)-associated vasculitis, specifically, granulomatosis with polyangiitis affecting the upper airway and presenting as subglottic stenosis, which is much more common in children than adults. Systemic JIA is always on the differential with prolonged fever and systemic inflammation, although her normal ferritin and lactate dehydrogenase on admission made this diagnosis less likely. Cricoarytenoid joint involvement can occur in systemic JIA, causing airway stridor. Given the prominent lymph nodes seen on her chest CT scan, we also considered pulmonary processes such as sarcoidosis and relapsing polychondritis. However, that same CT scan revealed tracheal narrowing but no posterior tracheal sparing, putting relapsing polychondritis lower on the differential. You can also think of polyarteritis nodosa or inflammatory bowel disease (IBD), but she had no gastrointestinal (GI) symptoms on presentation to support the latter. Her 2 steroid courses may have contributed to this somewhat “relapsing” course before hospitalization.

Was there any additional workup you recommended at that time?

We recommended some additional laboratory tests (Table 1), which were generally unremarkable and helped to move ANCA-associated vasculitis and sarcoidosis lower on the differential diagnosis. Her normal admission echocardiogram and 4-extremity blood pressures also made Takayasu’s arteritis less likely. We recommended an ophthalmologic examination, which we do for every patient with concern for underlying an inflammatory process. This was extremely helpful because the examination revealed normal visual acuity but was notable for nonacute granulomatous bilateral anterior uveitis.

How did the uveitis change your differential?

This finding increased the likelihood of an underlying systemic or chronic inflammatory process and made an infectious etiology even less likely. It also helped to rule out specific rheumatologic disorders, such as systemic JIA, which is almost never associated with uveitis. This would also have been unusual for atypical Kawasaki disease. Spondyloarthritis can present with uveitis, but the eye involvement is typically symptomatic, and our patient had no reported eye pain. The uveitis seen in sarcoidosis is granulomatous, which was not the case for our patient. Relapsing polychondritis can have uveitis; however, the pattern of tracheal involvement on the CT scan was not consistent with it. ANCA-negative associated vasculitis could still be in play. I think IBD also rose a bit on the differential.

Dr Russo, what did you make of the eye findings?

In my mind, this solidified the fact that the patient’s disease process was unlikely to be infectious. Posterior uveitis can happen in the setting of congenital infections or certain hematogenously spread infections that would invade the retina, such as disseminated candidiasis in premature infants. However, for the anterior uvea to be affected by an infectious process, either the infection source would have to be external to the eye or the infection would need to be so significant that it overwhelmed the posterior eye and spread to the anterior chamber by direct extension. Thus, this isolated anterior uveitis was more likely to be related to a systemic inflammatory or rheumatologic process.

In the days after the patient’s hospitalization, her cough continued to worsen despite trials of ipratropium, racemic epinephrine, albuterol, and lorazepam. A repeat nasopharyngolaryngoscopy on HD 4 was unremarkable. Dr Muego, given the significant respiratory component of the patient’s symptoms, was a primary pulmonary process high on your differential for her?

For illnesses presenting with prolonged fevers and cough, we think about more indolent pulmonary infections, such as mycobacteria and various fungi including Cryptococcus and Histoplasma. However, on the initial chest CT scan, the patient’s lungs looked clear, with nothing in the lower airways suggesting an infection like those mentioned. Everything about the chest CT scan pointed to an upper airway pathology; even the supposed lymphadenopathy, again read at our institution as nonpathologic, was located around the upper airway. When I see a radiograph and CT scan that are without focal findings, it makes me less likely to think of a more severe pulmonary infection. When we meet children with such prolonged fevers, we often think of a rheumatologic diagnosis as opposed to a primary pulmonary process. Of course, there could be a superimposed respiratory infection on top of a chronic process, but, based on the history, I felt this to be primarily rheumatologic or inflammatory in nature.

On HD 5, the patient was noted to have new biphasic stridor with tripoding and neck extension, but no drooling or difficulty handling secretions. A lateral neck radiograph revealed worsening stenosis of the cervical trachea, which was confirmed on repeat CT neck scans (Fig 2). Given her respiratory decompensation, she was transferred to the ICU, where she was eventually emergently intubated in the operating room with ENT on HD 6.

Dr Borek, given the patient’s previously normal nasopharyngolaryngoscopic studies, what were your thoughts about the patient’s progression?

As noted previously, before this respiratory decompensation, neither her CT neck scan nor 2 normal nasopharyngolaryngoscopic examinations demonstrated an obstructive lesion, a source of her cough, or any sign of impending respiratory failure. However, after examination of the patient in the ICU, and after a review of the CT neck scan from that day, there was a heightened concern for extensive subglottic narrowing and immediate airway collapse. The patient had loud biphasic stridor and was becoming aphonic, so the concern of an obstructive laryngeal lesion or subglottic lesion, such as a web, was newly high on the differential. She had no surgical or intubation history to provide a mechanism for acquired subglottic stenosis, and, as a result, our working differential from most to least likely was infectious, oncologic, and then idiopathic causes. Based on her change in presentation, it was vital that her airway was secured at that moment.

During the laryngoscopy and bronchoscopy, we noted extensive diffuse edema and copious secretions extending from the supraglottis to the carina (Fig 3A). We were surprised to see microabscesses along the epiglottic petiole and false vocal folds. The subglottis was noted to have a grade 1 circumferential narrowing with edema and mucopurulent exudate. The trachea was noted to have a diffuse pseudomembranous exudate and sloughing of infected appearing mucosa throughout, concerning for tracheitis. Cultures were taken of both the supraglottic abscesses, as well as the tracheal exudates.

The finding of severe purulent tracheitis prompted the initiation of vancomycin, piperacillin-tazobactam, and fluconazole with eventual transition to ceftriaxone and clindamycin. Despite airway compromise and concern for an autoinflammatory process, steroids were deferred given concern for indiscriminate immunosuppression and the potential for resultant exacerbation of presumed infectious tracheitis in the setting of a still unclear diagnosis. She completed a full antibiotic course on HD 13 while throat cultures remained pending. Dr Russo, after the initial discovery of what seemed like bacterial tracheitis, how did your thinking change regarding the patient’s overall clinical picture and where infection fell on your diagnosis?

I did not foresee this at all and was confused for several reasons. First, acute bacterial tracheitis usually follows a preceding illness, which I did not feel she necessarily had, as mentioned. Second, acute bacterial tracheitis is rare; I have seen only 1 true case of it in the past year. Moreover, she was just on a significant amount of antibiotics, so it did not make sense to have florid tracheitis, especially because she would have had residual antibiotics in her system due to the medications’ known half-lives. In reality, bacteria do not just suddenly rebound. The cadence was, thus, odd. Even with the bacterial tracheitis, there is no way this explains her whole illness course. I found myself asking what could be missing; could this only be the complication of something else? Could this be a partially treated pathogen leading to such a course? Equivocally, no.

Dr Mehta, did this finding support one of the previously discussed inflammatory conditions?

The purulent tracheitis was just weird. It is not something we have seen in our rheumatologic patients. Our conditions can cause neutrophilic inflammation, but never to the point at which it causes frank pus. Could this be a secondary infection? Maybe, but in any event, the waters were muddied at this point.

At this time, additional workup and interim laboratory monitoring were significant for severe anemia requiring packed red blood cell transfusion, continued elevated inflammatory markers, a significantly elevated fecal calprotectin of 2530 μg/g noted on HD 12, and a negative large volume blood culture result. A bone marrow biopsy was performed, which revealed no evidence of hematolymphoid neoplasm. Rheumatoid factor, HLA antigen B27, HLA antigen B51, antinuclear antibody, and antistreptolysin O tests were sent, and the results were all negative. An MRI of the brain and orbit revealed mild edema within the subcutaneous soft tissues lateral to the orbital rims bilaterally. A whole-body MRI revealed only wall-thickening and edema of the subglottic airway with resultant luminal narrowing, as well as multifocal muscle edema involving the lateral abdominal wall musculature, gluteal musculature, and the right vastus lateralis muscle.

The patient soon defervesced after intubation and antibiotic administration, with repeat bronchoscopy with ENT on HD 10 demonstrating significant interval improvement. The patient was eventually extubated on HD 12 to room air and repeat bronchoscopy after discharge revealed significant improvement of her tracheal inflammation (Fig 3B). Her tracheal culture results remained negative.

After her clinical improvement, the patient was still noted to have an improved but elevated fecal calprotectin level (156 μg/g) on HD 17. Dr Kelly, can you please speak to the fecal calprotectin trends noted throughout the hospitalization in the context of her clinical scenario?

Fecal calprotectin is a marker of neutrophilic inflammation within the GI tract. Calprotectin is nonspecific and can be elevated in association with IBD, bacterial or viral enteritis, juvenile polyps, medication use, such as nonsteroidal antiinflammatory drugs, food allergies and other allergic conditions affecting the GI tract, immunologic disorders, including celiac disease, and lymphoma. The patient’s other findings of anemia, hypoalbuminemia, and elevated inflammatory markers are nonspecific but certainly can be seen with IBD. Although IBD remained on the differential given family history of Crohn disease with the above constellation of laboratory findings in addition to bilateral uveitis, other etiologies outside the GI tract were still being considered. Our team felt that, at this time, her symptoms of fever, tracheitis, and bilateral uveitis were most likely representative of a hyperinflammatory state, and that, in the absence of prominent GI symptoms, the elevated calprotectin was partly due to inflammation of the GI tract but not the primary source of it.^1,2 

If the GI tract was not thought to be the primary source, what could have accounted for the significantly elevated fecal calprotectin?

Calprotectin is a heterodimer formed by 2 S100 calcium-binding cytosolic proteins, S100A8 and S100A9, found in some cells, most notably neutrophils. The heterodimer binds to calcium, which forms a stable compound that is not broken down in the intestines. There can be falsely elevated fecal calprotectin levels in the setting of respiratory tract infections because of the white blood cells in swallowed sputum. In fact, recent research has explored calprotectin’s role in the pathogenesis of respiratory diseases.³  At the time, in the absence of other symptoms and presence of concern for respiratory tract infection, we felt that this likely explained the patient’s findings.

The patient was ultimately discharged to inpatient rehabilitation after her extended hospitalization, with continued improvement in her fever curve and inflammatory markers (Fig 4). Approximately 2 weeks after discharge, the patient was seen by gastroenterology as an outpatient, at which time the possibility of IBD was again discussed. The patient had no intestinal symptoms, including diarrhea, abdominal pain, melena or hematochezia, or weight loss outside the setting of antibiotic exposure. She was noted to have a new oral ulcer. An outpatient esophagogastroduodenoscopy and colonoscopy were performed in conjunction with microlaryngoscopy and bronchoscopy with tracheal biopsies. The esophagogastroduodenoscopy was grossly normal, whereas the colonoscopy revealed a normal terminal ileum but discontinuous nonbleeding ulcers throughout the entire colon, with the worst appearance in the descending colon (Fig 5). The resultant Paris classification was A1a, L2, G₀ based on patient age, colonic involvement, and parental report of normal linear growth (in the absence of access to previous growth charts), respectively. GI biopsies revealed granulomatous panenteric Crohn disease with involvement of the stomach, duodenum, ileum, and colon (Fig 6). Soon thereafter, the tracheal biopsies resulted, revealing noncaseating granulomatous tracheitis. Given these findings, it was felt that the unifying diagnosis of IBD could explain the patient’s entire 2-month illness course.

Dr Kelly, can you speak to your initial thoughts after arriving at the final diagnosis of IBD? How unusual, in your experience, was such an initial presentation of IBD?

It is highly unusual for Crohn to first present with pulmonary involvement followed by GI involvement. The pathogenesis of pulmonary parenchymal disease and serosis in IBD is unknown. Airway inflammatory changes are thought to represent the same type of inflammatory changes as are seen in the bowel with a 2:1 female predominance and 3:1 female/male predominance for bronchial involvement.⁴  Bronchopulmonary involvement follows the onset of GI symptoms in IBD in 80% to 85% of cases, occurs concomitantly in 5% to 10% of cases, and precedes the diagnosis in ∼10% to 15% of cases.⁴  Parenchymal lung disease is often seen in the setting of active bowel disease, nearly up to 40%.^4–8 

The patient’s Pediatric Crohn Disease Activity Index value was 27.5 at diagnosis, indicating mild disease activity. She was subsequently started on infliximab and began to demonstrate sustained symptom improvement. Dr Kelly, patients with mild disease activity are typically started on oral steroids as induction therapy followed by azathioprine as maintenance. Why was an antitumor necrosis factor medication started instead in this case?

The patient was started on infliximab because, oftentimes, particularly in the adult gastroenterology community, patients with more severe, new-onset IBD presentations requiring hospitalization are started on biologics as a more aggressive top-down approach. We considered this patient’s presentation, despite the Pediatric Crohn Disease Activity Index value, as severe, given the pulmonary manifestations. Recent studies have revealed top-down therapy as a viable alternative to step-up therapy. Specifically, it has been shown that infliximab can better achieve short-term clinical and endoscopic remission while maintaining clinical remission better than conventional treatments such as azathioprine monotherapy.^9,10 

Going back to her clinical improvement during her hospitalization, Dr Mehta, why do you think the patient’s fevers and tracheitis, now known to be extraintestinal manifestations of her IBD, ultimately improved? Could it have been the antiinflammatory properties of the antibiotics, or something else?

I do not think it was the antibiotics. Although it is difficult to pin her improvement on steroids, it could have been the topical ophthalmic steroids she was receiving in the background for her uveitis that helped her. As we saw early in her course, her symptoms transiently improved when she received short steroid courses. Although much less than oral or intravenous administration, ocular steroids have long been known to have some degree of systemic absorption and resulting systemic side effects.^11–13  Review of the current literature noted a few pediatric case reports in which the use of ocular steroids resulted in the disruption of the hypopituitary–pituitary–adrenal axis, leading to Cushing’s syndrome with prolonged use or adrenal crisis if steroids were abruptly discontinued.^14–19  Given this, it could be possible our patient’s every 1- to 2-hour ocular steroid drops for her uveitis may have led to enough systemic absorption to provide some improvement in her symptoms. Years ago, we had a patient with diffuse lymphadenopathy and uveitis with a lot of fevers, ultimately diagnosed with sarcoidosis, who received hourly steroid topical drops for uveitis that improved the fevers completely, presumably from systemic steroid absorption. Our patient also began topical drops for uveitis, which could have made it seem like the antibiotics were leading to improvement.

Dr Kelly, were there any other possible contributors to her clinical improvement while hospitalized, given the absence of treatment of IBD?

As it concerns antimicrobial agents, the data on antibiotics for luminal Crohn disease is inconsistent. A 2012 meta-analysis of randomized controlled trials of antibiotic use found modest effect over placebo for induction of remission, as well as maintenance of remission in Crohn disease, with some limitations, which included heterogeneity of results and diverse antibiotics used.²⁰  It is unclear if the efficacy of antimicrobial treatment is related to undetected bacterial pathogens, bacterial overgrowth, unsuspected microperforation, or, perhaps, modulation of the microbiome.^20,21  Although the patient was intubated from HD 5 to HD 12, she was nothing per orem with IV fluid and was intermittently on elemental formula via a nasogastric tube. Retrospectively, it is possible that this relative bowel rest and intermittent elemental diet, proven as a valid therapy for IBD, could have contributed to her improvement in addition to systemically absorbed steroids, as noted by Dr Mehta. A course of exclusive enteral nutrition with elemental formula over 6 to 8 weeks has been shown to suppress inflammation and promote mucosal healing.²²  Per a meta-analysis of randomized trials conducted by Narula et al in 2018, exclusive enteral nutrition appears to be useful for inducing remission: 83% of children treated with exclusive enteral nutrition achieved remission compared with 61% of those treated with glucocorticoids, although that difference was not statistically significant (relative risk 1.35, 95% confidence interval 0.92–1.97).²³ 

In retrospect, the patient’s case can be explained by an overarching diagnosis of IBD, specifically Crohn disease, which explains her prolonged fevers, elevated inflammatory markers and fecal calprotectin, bilateral anterior uveitis, and her tracheitis. If given more weight, her history of aphthous ulcers, intermittent nonbloody diarrhea (originally attributed to multiple antibiotic courses), and Crohn disease in her paternal grandfather noted on admission could have led to an earlier diagnosis. However, this patient otherwise presented with none of the other typical signs or symptoms of IBD, including abdominal pain, arthritis, rashes, significant weight loss, or ocular complaints. Moreover, the patient’s chief complaint was her persistent, worsening cough with stridor, which clouded the clinical picture. With that said, the response of her cough and stridor to brief courses of steroids before hospitalization could have served as a sign that this was not a disease process of bacterial etiology.

IBD has long been known for its associated extraintestinal manifestations (EIMs); in fact, some authors argue that the prevalence of EIMs should lead to a shift in the view of IBD away from primarily a GI disease to a more systemic disorder. It is estimated that between 6% and 47% of adult patients with IBD and ∼25% to 29% of pediatric patients with IBD can have EIMs, with the reported prevalence of EIMs being greater in patients with Crohn disease than those with ulcerative colitis.^4,24,25  In Crohn disease, bronchopulmonary manifestations were previously estimated to be as low as 0.2%²⁶ ; however, this estimate may be low as pulmonary involvement is typically asymptomatic and commonly not considered as part of Crohn disease routine care. When they do occur, respiratory manifestations most often arise after the onset of bowel disease and may be aggravated during gastrointestinal flares.²⁷ 

IBD-related respiratory involvement typically affects the airway, specifically the intermediate-sized central bronchi. Less commonly, respiratory involvement may present as parenchymal lung disease, serositis, or pulmonary vascular disease. Airway inflammation may manifest as bronchiectasis, chronic bronchitis, suppurative large airway disease, or acute bronchitis. Small airway involvement is rarely encountered clinically. A disease of the more proximal airway, namely tracheal disease, including subglottic stenosis or diffuse tracheitis, as was seen in this case, is exceedingly rare.²⁸  After the first report in 1976 of IBD-related pulmonary manifestations in patients, several cases and case series in adult populations have been published. However, the association is less well-described in children. A literature search revealed only 17 cases of symptomatic pulmonary disease in children with IBD, as of 2012. Of these 17 cases, only 8 had not been previously diagnosed with Crohn disease; of these 8 cases, 7 had lung lesions and only 1 had airway involvement.^27,29  Thus, our authors have described one of the few known pediatric cases of airway involvement serving as the presenting manifestation of previously undiagnosed Crohn disease.

At this time, the pathogenesis of respiratory tract involvement in Crohn disease remains poorly understood. Many theories exist, most of which posit that pulmonary damage occurs secondarily to circulating immune complexes and inflammatory mediators. It has been hypothesized that lymphocytes sensitized to antigens at 1 mucosal site, namely the gastrointestinal tract, may circulate to different mucosal sites and lead to inflammatory cascades, potentially explaining IBD-related airway disease and recruitment of alveolar lymphocytes that cause interstitial lung disease. Also, both the colonic and respiratory epithelia share a common embryonic origin from the primitive gut and contain a submucosal lymphoid tissue, which plays a crucial role in the mucosal defense mechanism. This similarity in the mucosal immune system may explain the same pathologic changes observed in both organ systems.²⁷ 

Because of the lack of reported cases of laryngotracheal involvement and IBD, there is no consensus treatment and ongoing management approach. Although many adults respond rapidly to oral or IV corticosteroids, others have responded to inhaled corticosteroids and epinephrine. In refractory cases, procedures like a laser beam, balloon dilatation, stent placement, or tracheostomy may be necessary to maintain the patient’s airway.³⁰ 

The knowledge of IBD-related purulent aseptic tracheitis (a noncaseating granulomatous tracheitis) can contribute to pediatric care by helping to avoid unnecessary combinations of antibiotics and delayed correct treatment. Proper treatment of respiratory manifestations of IBD will lead to better short and long-term outcomes; if left untreated, this condition may progress to irreversible airway fibrosis.³¹ 

Ultimately, this case represents an exceedingly rare initial presentation of Crohn disease with respiratory tract manifestations serving as the presenting symptom.

ENT

otorhinolaryngology

OSH

outside hospital

CT

computed tomography

IV

intravenous

ESR

erythrocyte sedimentation rate

CRP

C-reactive protein

JIA

juvenile idiopathic arthritis

HD

hospital day

ANCA

anti-neutrophilic cytoplasmic autoantibody

IBD

inflammatory bowel disease

GI

gastrointestinal

EIM

extraintestinal manifestation