Asthmatic Granulomatosis

A Novel Disease with Asthmatic and Granulomatous Features

Sally E. Wenzel1, Catherine A. Vitari1, Manisha Shende2, Diane C. Strollo3, Allyson Larkin4, and
Samuel A. Yousem5
1
University of Pittsburgh Asthma Institute at University of Pittsburgh Medical Center/University of Pittsburgh School of Medicine, Pulmonary Allergy
and Critical Care Medicine Division, Department of Medicine, 2Department of Surgery, 3Department of Radiology, 4Department of Pediatrics, and
5
Department of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania

Rationale: Severe asthma represents 5–10% of all asthma, yet

remains problematic and poorly understood. Although it is increas- AT A GLANCE COMMENTARY
ingly recognized as consisting of numerous heterogenous pheno-
types, their immunopathology, particularly in the distal airways Scientific Knowledge on the Subject
and interstitium, remains poorly described. Diseases with reversible airflow limitation are traditionally
Objectives: To identify the pathobiology of atypical difficult asthma. grouped together under the term “asthma.” However, many
Methods: We report 10 from a total of 19 patients (17 women and 2 of these patients can have atypical features, including
men) meeting asthma and severe asthma definitions, requiring daily later age at onset and low diffusing capacities. Very little
systemic corticosteroid (CS) use, with inconsistent abnormalities on
is understood regarding the pathobiology of these atypical
chest computed tomography scans, who underwent video-assisted
patients.
thoracoscopic biopsies for further diagnosis and management.
Measurements and Main Results: The pathology of 10 of the 19 cases
revealed small airway changes consistent with asthma (eosinophilia, What This Study Adds to the Field
goblet cell hyperplasia), but with the unexpected finding of intersti-
tial nonnecrotizing granulomas. These patients had no evidence for We present video-assisted thoracoscopic biopsy findings
hypersensitivity pneumonitis, but 70% of cases had a personal or from 10 patients, previously diagnosed with severe asthma
family history of autoimmune-like disease. The 10 cases were treated and meeting criteria for asthma. Pathobiologically, these
with azathioprine, mycophenolic acid, methotrexate, or infliximab. patients have evidence for asthmatic small airway inflam-
Nine of 10 showed decreased CS requirements and improved or mation and infrequent nonnecrotizing granulomas with in-
maintained FEV1 despite lower CS doses. Of the remaining nine terstitial inflammation. This distinct pathobiology in addition
patients, six manifested asthmatic small airway disease, alone or in to their response to cytotoxic agents suggests that these
combination with alveolar septal mononuclear cells, but no granu- patients represent a newly described disease, which we term
lomas, whereas three manifested other pathologic findings (aspira- asthmatic granulomatosis.
tion, pneumonia, or thromboemboli).
Conclusions: These data suggest that a subset of severe “asthma”
manifests a granulomatous pathology, which we term “asthmatic
granulomatosis.” Although identification of this disease currently
requires a thorascopic biopsy, alternative approaches to therapy or reversible airflow limitation, requirements met by a broad
lead to improvement in outcomes. range of patients (1). Cohort studies in asthma and severe asthma
are now uniformly identifying different subtypes or phenotypes
Keywords: asthma; hypersensitivity pneumonitis; granulomas; eosino-
that thus far differ by degree of airflow limitation, inflammatory
phils; autoimmunity
or allergic processes, and age at onset (2–4). These phenotypes are
perhaps best differentiated in patients identified with “severe
The term “asthma” has been loosely defined as the presence of ap- asthma,” a grouping of patients again defined clinically and
propriate symptoms in the setting of airway hyperresponsiveness physiologically (4–6). The pathology behind these different phe-
notypes is poorly understood.
Pathologic studies of severe asthma are generally limited to
small bronchoscopically obtained partial-thickness large airway
(Received in original form March 15, 2012; accepted in final form June 24, 2012)
biopsies, which do not include smaller airways or alveolated pa-
Supported by the Dellenback Research Fund. renchyma. Although transbronchial and distal lung biopsies have
Author Contributions: S.E.W. saw patients, entered and analyzed data, and wrote been obtained they are not without risk, and the amount of small
the manuscript. C.A.V. collected data, served as asthma nurse educator address- airway tissue obtained is limited (7, 8). Most descriptions of
ing adherence and environmental factors, and reviewed the manuscript. M.S.
performed video-assisted thoracoscopic biopsies on most patients and reviewed
distal lung pathology in asthma are from autopsies of fatal
the manuscript. D.C.S. interpreted most of the chest CTs, overread the CTs for asthma exacerbations, with additional rare descriptions of pa-
the manuscript, analyzed CT data, and reviewed the manuscript. A.L. saw thology in lobes resected for tumors, where clinical and physi-
patients and reviewed the manuscript. S.A.Y. identified abnormal pathology, ologic details are lacking (9–12). This contrasts with studies of
did special stains, and wrote and reviewed the manuscript. interstitial lung disease where large-volume lung biopsies, often
Correspondence and requests for reprints should be addressed to Sally E. Wenzel, driven by radiologic abnormalities on computed tomography
M.D., NW 931 Montefiore, 3459 Fifth Avenue, Pittsburgh, PA 15213. E-mail: (CT) scans, have served to identify pathologic subtypes. Thus,
[email protected] for nearly all presentations of severe forms of asthma, limited
This article has an online supplement, which is accessible from this issue’s table of pathologic data are available that could inform phenotypes,
contents at www.atsjournals.org mechanisms, and therapies.
Am J Respir Crit Care Med Vol 186, Iss. 6, pp 501–507, Sep 15, 2012 The University of Pittsburgh and University of Pittsburgh
Copyright ª 2012 by the American Thoracic Society
Originally Published in Press as DOI: 10.1164/rccm.201203-0476OC on July 5, 2012 Medical Center operate a Difficult Asthma Clinic that receives
Internet address: www.atsjournals.org referrals of patients with difficult asthma locally, regionally, and
502 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 186 2012

nationally. Despite patients receiving an extensive evaluation phase (FEF25–75% predicted) who did not improve or worsened over time
and optimization of treatment, many of these patients remain were recommended for VATS biopsies. Most of the surgeries were per-
poorly controlled. We suspected that larger-volume lung tissue formed by a single surgeon (M.S.). Because these patients had no local-
biopsies obtained through video-assisted thorascopic surgery ized or parenchymal radiologic abnormalities by chest CT, multiple
random biopsies were taken from each lobe of a single lung. Double-
(VATS) could identify novel histopathologies with distinct
lumen endotracheal intubation with single-lung ventilation was used for
immunoinflammatory elements to suggest new approaches to all patients. Care was taken to ensure that generous wedge resections
therapy. We report 10 adult patients, all previously diagnosed were done to include deep tissue. Tissue samples were sent for culture
with severe asthma, whose lung tissue revealed small airway (all negative for bacteria, fungus, and mycobacteria) and pathologic anal-
disease in association with poorly formed nonnecrotizing inter- ysis. Initially, patients gave consent using general lung tissue repository
stitial granulomas. In addition to their clinical, inflammatory, and consent. However, since 2010, consent was obtained using an airway
physiologic characteristics, we report their improvement with disease–specific consent to collect lung tissue for future research proto-
alternative nonsteroidal antiinflammatory drugs. Some of these cols and to allow use of clinical, physiologic, radiologic, and immunoin-
results have previously been reported in abstracts (13, 14). flammatory data. The protocols and consents were approved by the
University of Pittsburgh Institutional Review Board.

METHODS
Pathologic Analysis
Evaluation
All biopsies were entirely processed in formalin; embedded in paraffin;
Patients were referred for evaluation of severe corticosteroid (CS) re- and stained with hematoxylin and eosin, Verhoeff-van Gieson, acid-fast,
fractory asthma to the Difficult Asthma Clinic at the Comprehensive and Gomori methenamine silver stains. Immunohistochemistry for
Lung Center of the University of Pittsburgh Medical Center. Asthma L26/CD20, CD3, CD4, CD8, Beta F1, TIA, CD79, CD138, S100 protein,
was defined using American Thoracic Society criteria and severe asth- Langerin, and mast cell tryptase were performed as previously described
ma defined by the American Thoracic Society workshop definition, in- (15). The ratio of CD4 to CD8 cells was calculated by manually count-
cluding use of high-dose inhaled (.880 mg fluticasone or equivalent) ing the number of T cells in three identical fields adjacent to the small
or systemic CSs with at least two of seven minor criteria (1, 5). airways (16).
All received a standardized evaluation, with extensive history and
physical examination; pulmonary function testing with bronchodilator
reversibility to 720 mg of inhaled albuterol using an Aerochamber Follow-up
(Monaghan, Plattsburgh, NY); diffusing capacity of carbon monoxide Seventeen of the 19 patients were offered treatment with alternative
(DLCO); chest CT and in some cases high-resolution CT (HRCT); antiinflammatory or cytotoxic medications: azathioprine (Imuran),
IgE and specific IgE/allergen skin testing; complete blood count; and methotrexate, or anti–tumor necrosis factor-a (Remicade). Treatment
evaluation of CS side effects. Most patients had measurement of with these antiinflammatory medications was instituted for 3 months
exhaled nitric oxide (FENO), C-reactive protein, antinuclear antigen, and continued for 12–48 months if response was seen. Treatment was
rheumatoid factor, screening precipitins for hypersensitivity pneumo- stopped if ineffective or poorly tolerated. A patient was considered
nitis (HSP), quantitative immunoglobulins, and sinus CT scans. They a “responder” to alternative therapy if he or she (1) was able to reduce
were asked questions on age at asthma onset, use and adherence to his or her prednisone dose to 5 mg or less, (2) maintained or improved
CSs, existence of other chronic diseases including sinusitis, environ- their FEV1 % predicted despite lower oral CS doses, and (3) had one
mental exposures, and family history of allergic and autoimmune dis- or fewer oral CS-requiring exacerbations during the year following.
eases. All met with a certified asthma educator. Patients were assessed Patients were monitored monthly with complete blood counts, liver
for 3–24 months to optimize asthma therapy and to better understand function, and pulmonary function tests.
their disease, its variability, and need for systemic CSs. All CT scans
were overread for this study by a chest radiologist (D.C.S.). Approxi-
mately 170 unique patients with severe asthma have been followed in
RESULTS
the Difficult Asthma Clinic from 2007 to 2011. Demographics and Clinical Characteristics
Demographics and clinical characteristics are presented in Table 1.
Thoracoscopic Biopsy Ten of the 19 patients with a clinical diagnosis of severe asthma
Nineteen patients with severe asthma with CT scans without parenchymal who underwent VATS biopsy between 2007 and 2011 had a com-
abnormalities and atypical asthma presentations (low DLCO), adult onset, bination of asthmatic small airway changes with nonnecrotizing
persistent eosinophilia, or very low forced expiratory flow midexpiratory interstitial and airway-centric granulomas on pathology. All

TABLE 1. DEMOGRAPHICS AND LUNG FUNCTION BY INDIVIDUAL PATIENT

Patient Age Age at Onset FEV1 % FEV/FVC FEF25–75% % Bronchodilator Response* DLCO

#1 49 29 63 60 28 15/9 78
#2 60 5 56 82 53 13/61 56
#3 59 23 84 74 52 22/50† 68
#4 62 50 57 63 27 12/19 57
#5 36 30 39 42 10 12/16 108
#6 52 45 73 78 46 5/4 75
#7 30 5 86 72 58 1/26‡ 62
#8 55 5 69 69 36 20/35 67
#9 63 60 97 62 32 24/22† 95
#10 56 28 45 52 16 19/73 96
Mean 6 SD or SEM 52 6 11 28 6 20 66 6 6 65 6 4 36 6 5 12 6 3 and 26 6 9 76 6 6

Definition of abbreviations: DLCO ¼ diffusing capacity of carbon monoxide; FEF25–75% ¼ forced expiratory flow, midexpir-
atory phase.
* % BD response for FEV1/FEF25–75.
y
Positive methacholine challenge PC20 ¼ 3.6 mg/ml (#3), 0.4 mg/ml (#9).
z
Patient with greater than 12% reversibility to prednisone.
Wenzel, Vitari, Shende, et al.: Asthmatic Granulomatosis: A New Disease? 503

TABLE 2. IMMUNOINFLAMMATORY AND RADIOLOGIC FEATURES

Patient Blood (eos/ml) IgE (IU/ml) Atopy (Y/N) FENO (ppb) CRP Sinus CT Chest CT

#1 780 5 No 10 0.7 ND Normal

#2 100 15 No 19 Normal Normal
#3 1,800 72 Yes 113 0.4 Polyps/sinusitis Bronchial wall thickening
#4 200 7 Yes 57 0.34 Pansinusitis Normal
#5 200 16 Yes 10 0.39 Mild sinusitis Air trapping
#6 100 280 Yes 22 3.5 Sinusitis Mosaicism
#7 200 19 No 88 0.25 Mild sinusitis Normal
#8 350 630 Yes 39 0.83 Mild thickening Normal
#9 1,390 99 No 58 5.6 Sinusitis/polyps Air trapping
#10 300 166 Yes 14.3 Mild thickening Bronchial dilatation
Median (25th–75th percentiles) 325 (200–760) 54 (15–252) 6/4 42 (15–79) 0.8 (0.4–4)

Definition of abbreviations: CRP ¼ C-reactive protein; CT ¼ computed tomography; FENO ¼ exhaled nitric oxide.

patients had been diagnosed with severe asthma by referring (n ¼ 1) during inspiration. In some cases, 1.25-mm thin HRCT
physicians, were oral CS dependent (5–60 mg/day), and had images (n ¼ 4), 0.625-mm source images (n ¼ 10), or chest CT
no evidence for other respiratory disease. One patient was on performed during end-expiration (n ¼ 1) were available for
omalizumab, none on methotrexate. All patients were white, 9 review. One or more comparison chest CTs, to include expira-
of 10 were female, and 8 of 10 were middle-aged. The cases tory CT (n ¼ 1), were available in six cases. Five patients had
were referred from western Pennsylvania, upstate New York, a normal CT. The other five CTs revealed nonspecific findings
and Maryland. Most were diagnosed with asthma as adults and of bronchiolar dilation (n ¼ 5); bronchiolar wall thickening
all had a history of sinus disease (see Figure E1 in the online (n ¼ 4); patchy predominantly subsegmental air trapping
supplement). The patients were all currently symptomatic with (n ¼ 2); subsegmental atelectasis (n ¼ 2); tree-in-bud opacities
cough, wheeze, chest tightness, and shortness of breath. There (n ¼ 1); or small mucosal nodules in the central airways (n ¼ 1)
was no history of untreated and symptomatic gastroesophageal (see Figure E2). Focal or extensive bronchiectasis, thoracic
reflux or aspiration, whereas 1 of 10 reported aspirin-exacerbated lymph node enlargement, and consolidation were absent.
respiratory disease. All were treated with traditional asthma Peripheral blood eosinophilia was present in most and high
medications, including inhaled CSs, long-acting b-agonists and, FENO despite systemic CSs. There were no consistent or sugges-
leukotriene modifiers. Nocturnal awakenings were not com- tive environmental exposures, the IgE was not consistently
mon. Three or more exacerbations in the previous year were elevated, and precipitating antibody screens for HSP were neg-
common and often triggered by worsening sinus disease. All ative. Although IgG levels were occasionally low (likely on the
smoked less than 10 pack-years and none had smoked in the basis of chronic systemic CS use), pre- and posttetanus and
last year. Although wheezing was common, no crackles were pneumococcal pneumonia vaccination titer responses were nor-
heard on examination. mal. Atopy was present in 6 of 10. Nine of 10 had significant
sinus disease by CT scan, with 5 of 10 having a history of sinus
Physiologic, Radiologic, and Immunologic Findings surgery. Of the 10 patients, 60% had a family history of auto-
immune disease: rheumatoid arthritis (two); Crohn disease
Physiologic, radiologic, and immunologic findings are presented (one); mixed connective tissue disease (one); CRST syndrome
in Table 2. All 10 cases had evidence of airflow limitation, re- (one); and dermatomyositis (one). Two patients had a personal
versibility, and in two cases a provocative concentration of
methacholine less than 8 mg/ml despite high-dose inhaled and
systemic CSs. The FEF25–75% predicted was disproportionately
low compared with the FEV1 % predicted and FEV1/FVC.
Unlike traditional asthma, the DLCO was modestly reduced in
most cases (76 6 6 % predicted).
All patients had a chest CT performed within 10 months of
surgery. Images were reconstructed at 5 (n ¼ 9) or 2.5 mm

TABLE 3. HISTOPATHOLOGIC FEATURES OF TEN CASES OF

ASTHMATIC GRANULOMATOSIS
Small airways
Submucosal inflammation 10/10
Eosinophils 7/10
Neutrophils 2/10
Lymphocytic transmigration 8/10
Submucosal fibrosis 6/10
Mucus plugging 8/10
Peribronchiolar fibrosis 6/10
Bronchiolectasis 6/10
Muscular hypertrophy 10/10 Figure 1. Asthmatic granulomatosis. The small airways are occluded by
Interstitial mononuclear pneumonia large mucus plugs containing eosinophils with the submucosa ex-
Granulomas 10/10 panded by a chronic inflammatory cell infiltrate. A patchy interstitial
Airspace organization 5/10
mononuclear infiltrate with poorly formed granulomas is also noted
Lymphoid aggregates 4/10
(at right) (hematoxylin and eosin, 3100).
504 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 186 2012

Figure 2. Asthmatic granulomatosis. Mucus plugs within the airways Figure 4. Asthmatic granulomatosis. A granuloma is seen within the
contain degenerating eosinophils, whereas the bronchiolar epithe- interstitium (hematoxylin and eosin, 3400).
lium shows goblet cell metaplasia, basement membrane thickening,
and a mixed submucosal infiltrate with eosinophils (hematoxylin and
eosin, 3200). all cases but was pronounced in five (Figure 2). A mild to mod-
erate submucosal chronic inflammatory cell infiltrate was noted
in all cases, with epithelial transmigration in eight. Eosinophils
history of autoimmune disease (psoriasis and inactive Crohn
were easily identified in 70% of biopsies (Figures 2 and 3) and
disease by recent colonic biopsy).
plasmacytoid lymphocytes and plasma cells were present in
aggregates in 60%. Patchy submucosal and peribronchiolar fi-
Pathologic Analysis brosis was noted in most cases and muscular hypertrophy of the
Multiple unilateral wedge biopsies from the right (three) or left bronchiolar muscularis was ubiquitous, but dramatic in eight
(two) lung were taken from all patients. The histopathology instances. No fibrous obliteration or significant compromise of
of the 10 cases is detailed in Table 3. Histologic changes con- airway lumens was observed.
sisted of an asthmatic bronchiolitis accompanied by a patchy A patchy interstitial mononuclear infiltrate associated with
interstitial infiltrate of alveolar septal mononuclear cells with the airway changes was localized primarily to the perivenular
poorly formed nonnecrotizing granulomas in all 10 (Figure 1). regions and alveolar septa, accompanied by airspace fibrin
Terminal bronchioles displayed papillary epithelial hyperplasia, and granulation tissue in 50%. Ill-defined, nonnecrotizing gran-
basal cell hyperplasia, and goblet cell metaplasia in all cases ulomas comprised of epithelioid histiocyte aggregates with or
(Figure 2). Mucous plugging was noted in 8 of the 10 cases without giant cells were noted in all 10 cases, but were relatively
(80%) as was a mild bronchiolectasia in 60%. Subepithelial base- infrequent (Figures 3, 4, and 5). Interstitial lymphoid aggregates
ment membrane thickening, consistent with asthma, was seen in were seen in four cases.

Figure 3. Asthmatic granulomatosis. Small airway changes were asso- Figure 5. Asthmatic granulomatosis. A patchy mononuclear infiltrate
ciated with a patchy alveolar septal mononuclear infiltrate with airspace with interstitial and bronchiolocentric poorly formed granulomas con-
granulation tissue and interstitial nonnecrotizing granulomas and loose taining giant cells, epithelioid histiocytes, and occasional asteroid bodies
aggregates of giant cells and epithelioid histiocytes (hematoxylin and (inset) is common (hematoxylin and eosin, 3200; inset, hematoxylin and
eosin, 3200). eosin, 3400).
Wenzel, Vitari, Shende, et al.: Asthmatic Granulomatosis: A New Disease? 505

Immunohistochemical analysis revealed most inflammatory Although azathioprine is the first drug used (and most success-
cells to be Beta F1 CD3 positive T cells with a CD4/CD8 ratio ful), it is not always well tolerated, because three discontinued
of 4.2 (range, 2.8–5.1), with most intraepithelial lymphocytes for nausea and one for liver function test abnormalities. One
a Beta F1 CD3/CD4 positive, TIA-negative phenotype. L26/ patient has not been able to tolerate any alternative therapy to
CD20 stains highlighted lymphoid aggregates and scattered sub- date and therefore response cannot be evaluated. The nine res-
mucosal B cells, whereas CD79A and CD138 studies showed ponders have had improvements in FEV1, FVC, and FEF25–75,
plasmacytoid lymphocytes and plasma cells in aggregates in the in their level of symptoms and activity despite lower (or stable)
submucosa of those cases with the most prominent inflammatory CS doses. Despite frequent CS burst pre-VATS, no responding
infiltrates. S100 protein and Langerin stains displayed rare intra- patient has required more than one CS burst per year after
epithelial and submucosal Langerhans cells, whereas mast cell response to therapy. In three patients treated for more than
tryptase staining showed nonaggregated and scattered mast cells 1 year, attempts to taper azathioprine have been unsuccessful,
throughout the interstitium. leading to worsening clinical symptoms, worsening airway ob-
There was no evidence for microorganisms by silver and acid- struction, and increased CS doses. When the azathioprine dose
fast stains, vasculitis, necrosis, fibroblastic foci, honeycomb change, was increased back to 250–300 mg per day, the symptomatology
or aspirated material. and lung function returned to the prereduction baseline.

Additional Surgical Cases Responses to Therapy in the Nine Additional Cases

The remaining nine cases (see Tables E1 and E2 for clinical and Table E3 shows responses to therapy in the nine additional cases.
immunoinflammatory details) included one aspiration (food One patient (aspiration) was lost to follow-up. In the four
particles in the airways); one eosinophilic airway inflammation patients with small airway disease only, three of four could be
with superimposed acute pneumonia; and one case with marked characterized as “responders” to azathioprine, whereas one
thromboembollic disease in the presence of asthmatic and re- would not (patient #2). Both patients with airway disease and
spiratory bronchiolitis and interstitial granulomas (15 pack-year mononuclear septal infiltrates without granulomas have had
smoking history). Two patients manifested eosinophilic small reductions in oral CS requirements, but mixed improvement
airway disease and alveolar septal mononuclear cell infiltrates in FEV1 % predicted (one azathioprine, one myophenolic acid
without granulomas, whereas four manifested only small airway [after LFT increases with azathioprine]). The case with throm-
disease consisting of goblet cell hyperplasia and mixed submu- boembolic disease has been anticoagulated and is under evalu-
cosal eosinophilic/mononuclear inflammation. In these six ation by rheumatology for recent positive antiphospholipid IgM
patients, small airway inflammatory changes were similar to antibody. The case with eosinophilic inflammation and pneu-
the 10 cases. In several cases, low DLCO, low FEF25–75% pre- monia has not responded to any therapy, including anti–IL-5
dicted, high blood eosinophils, and high FENOs were observed. (mepolizumab).

Asthma Control Subjects DISCUSSION

Seventeen deidentified autopsies of adult patients (nine women Except for fatal asthma, the distal lung pathology of severe re-
and eight men; median age, 58; range, 26–76) admitted with versible obstructive airway disease (asthma) and its phenotypes
a known diagnosis of inactive chronic but well-controlled asthma remains poorly described. We present thoracoscopic biopsy data
who died of nonpulmonary causes were reviewed as a control from 19 adults, all diagnosed with “severe asthma,” 10 of which
group. Although changes of asthmatic bronchitis and bronchio- reveal granulomas in association with more traditional features
litis were seen, no granulomas were identified. A patchy incon- of asthmatic or eosinophilic airway inflammation. Importantly,
spicuous interstitial infiltrate of mononuclear cells was seen in preliminary follow-up studies suggest that nonsteroidal cyto-
35% of patients. toxic or antiinflammatory therapies improve the course of disease,
supporting a continuation of this invasive diagnostic approach in
selected cases. We believe that these data support the identifica-
Response to Therapy tion of a newly described clinical and pathologic entity, which we
All 10 cases were treated with oral CSs ranging from 5 to 60 mg/ propose to name “asthmatic granulomatosis.”
day and had been CS dependent for as long as 20 years. Of the 10, All 19 cases described here met physiologic and clinical cri-
six patients on azathioprine, one on methotrexate (intramuscu- teria for asthma: symptoms; reversible airflow limitation (to b2
lar), one on infliximab, and one on mycophenolic acid have suc- agonists or CSs); or airway hyperresponsiveness in the absence
cessfully reduced their systemic CS dose to 5 mg per day (Table 4). of parenchymal abnormalities on HRCT. Interestingly, the one

TABLE 4. RESPONSE TO NONSTEROIDAL CYTOTOXIC/ANTIINFLAMMATORY THERAPY

Patient Prednisone Pre (mg/d) Prednisone Post (mg/d) FEV1% Pre-Rx FEV1% Post-Rx Therapy

#1 (R) 10 5 63 71 AZ
#2 (R) 50 5 58 65 AZ
#3 (R) 20 5 84 91 AZ
#4 (R) 10 5 57 65 AZ
#5 (R) 5 5 39 44 AZ failure (nausea) now infliximab
#6 (R) 30 5 73 71 AZ
#7 (PR) 40 5 86 71 AZ failure (nausea) now MTX IM
#8 (R) 5 5 69 80 AZ
#9 (NE) 13 13 97 112 AZ, MP, MTX failures (nausea, LFTs)
#10 (R) 20 5 45 54 AZ failure (nausea)/now MP acid

Definition of abbreviations: AZ ¼ azathioprine; IM ¼ intramuscular; LFT ¼ liver function tests; MP ¼ mycophenolate; MTX ¼ methotrexate; NE ¼
not evaluable because not on any drug for a least 1 month because of side effects; PR ¼ partial clinical responder; R ¼ clinical responder.
506 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 186 2012

patient who did not meet these criteria before diagnosis (patient Stains for microorganisms and all cultures were negative. This
#6) developed bronchodilator responsiveness after successful prominence of plasmacytoid and plasma cells might suggest a link
treatment with azathioprine. All 19 also met diagnostic criteria to autoimmunity, as does the presence of the nonnecrotizing gran-
for severe asthma, being on daily systemic CSs; high-dose in- ulomas themselves. An autoimmune link is further supported by
haled CS, usually with a long-term controller; having daily sym- the family or personal history of autoimmunity in 70%.
ptoms; and averaging three CS bursts in the previous year. Most Although patient #9 had inactive Crohn disease, she reported
patients reported either adult-onset disease or worsening in asthma since childhood, had a high IgE level, and had been
adulthood. VATS biopsies were performed on these patients treated with anti-IgE therapy with some improvement in symp-
to determine the presence of a traditional asthma masquerader, toms, also supporting an overlap in immune processes. Interest-
such as a limited Churg-Strauss syndrome or constrictive bron- ingly, genetic factors, including such genes as SLC22A5 and
chiolitis, and if not present, to better understand the pathobiol- RORA, which are shared by asthma and Crohn disease, could
ogy of their severe disease such that alternative therapies might predispose to granulomatous inflammation, perhaps through
be suggested. The biopsies were all done using a deep cone modulation of responses to microbial exposures (26).
approach to obtain the airways tissue vital to revealing the de- Four of nine female patients described onset or worsening of
scribed pathology. There have been no complications and all their disease during pregnancy, whereas the one man in the data-
have gone home within 3 days of the procedure. set is currently treated for low testosterone levels. The Severe
Ten of the 19 cases manifested a combination of three path- Asthma Research Program identified a cluster of women with
ologic findings not previously reported together (asthmatic air- late-onset disease and an association with hormonal influences
way inflammation, alveolar septal mononuclear inflammation, (4). Eosinophilic inflammation, as measured in sputum, was
and poorly demarcated nonnecrotizing granulomas). Seven of prominent, suggesting that at least some of the women de-
10 cases either had an elevated FENO (.30 ppb) or peripher- scribed here may have overlap with the predominantly female
al blood eosinophilia (.300 per milliliter) despite continuous cluster in the Severe Asthma Research Program. Women in
systemic CSs, consistent with an “asthmatic” Th2-like inflam- general are predisposed to autoimmune disease, supporting
matory process (17–19). Although the accompanying granulo- a complex role for hormones in autoimmunity, and perhaps in
matous process may be suggestive of HSP, these 10 cases have this newly described asthmatic granulomatosis (27). Although
asthmatic small airway changes including goblet cell metaplasia, asthma is not thought of as an autoimmune disease (and there
thick basement membranes, tissue eosinophilia, and marked were no increases in known autoantibodies), aspirin-exacerbated
muscular wall hypertrophy with mucus plugs that are not part respiratory disease and, more recently, nasal polyps have been
of the spectrum of infection, HSP, aspiration, or autoimmune associated with autoantibody formation (28, 29). Indeed, several
disease. The lack of interstitial or nodular changes on HRCT of the patients described here reported a history of nasal polyps
scan further argues against an HSP diagnosis (20). Crohn dis- in addition to severe chronic sinusitis.
ease of the lung is usually seen with active ileocolonic inflam- These 10 patients were accumulated over a 4-year period of
mation, whereas patient #9 was asymptomatic with negative time (2007–2011). Although VATS biopsies are not routinely
colonic biopsies. Other autoimmune diseases, such as Sjögren performed in difficult or severe asthma, this novel finding of in-
and rheumatoid arthritis, and sarcoidosis may have granulomas terstitial granulomas, associated with a clinical response to non-
but lack the asthmatic reactive airway pathology described here. steroidal medications, suggests this procedure may be justified
No vasculitis was noted in any case excluding granulomatosis in a select group of atypical patients, especially given the risks
with polyangitis (Wegener’s) and Churg-Strauss syndrome. of these immunosuppressive agents. Over the 4 years of the
Thus, although we acknowledge the histologic (and clinical) accumulation of these cases, approximately 10% of the 170
differential diagnoses that need to be considered, it seems that patients with severe asthma seen in the Difficult Asthma Clinic
the clinicopathologic profile in these 10 cases is best encom- underwent VATS biopsy, 50% of whom manifested the unique
passed in the term “asthmatic granulomatosis,” reflecting the combination of pathologic findings reported here. Of the
asthmatic small airway changes and the granulomatous disease remaining nine biopsies, seven manifested one or more of these
processes. findings, suggesting that the 10 cases may represent the far end
These three inflammatory elements were present despite sys- of a spectrum of CS-refractory asthma. Although this pathology
temic CS use, confirming a profound CS resistance, findings and disease are not as common as traditional asthma, over the
previously noted in endobronchial biopsies and sputum from same period of time, in the same clinic, two cases of allergic
subgroups of patients with severe asthma (3, 6). Although the bronchopulmonary aspergillosis and only one case of Churg-
mechanisms for the CS-resistant eosinophilic, likely Th2, pro- Strauss syndrome have been identified, suggesting this new pa-
cess remain poorly understood, studies of severe eosinophilic thology or disease is more common than these typical asthma
subjects with asthma support a role for the Th2 cytokine masqueraders.
IL-5 (21, 22). However, the presence of granulomas suggests Of the 10 cases reported here, nine have responded from
the immune process reported here is more complex than Th2 a moderate to a high degree to cytotoxic and antiinflammatory
alone including elements of innate, Th1, or Th17 immunity, all therapies including azathioprine, methotrexate, anti–tumor ne-
of which have been identified with granulomatous inflammation crosis factor-a, and mycophenolic acid. The remaining one has
(23–25). These processes may be accompanied by IL-8 had difficulty tolerating these therapies. Importantly, three res-
or interferon-g, which have been identified with granulomatous ponding patients have attempted to taper their azathioprine, but
inflammation in endobronchial bioposies and sputum from sub- had to return to full or partial dosing because of deterioration in
groups of patients with severe asthma such as those with eosin- symptoms, increasing steroid requirements, and decline in lung
ophilic pneumonia, and which could contribute to the asthmatic function. In the 1980s, studies inconsistently suggested efficacy
granulomatous histopathology observed here. of methotrexate in some severe CS-dependent people with
Initial characterization of the inflammatory mononuclear cell asthma (30). No pathology was obtained, but it is conceivable
bronchiolitis in these patients identifies a CD31 lymphocyte infil- the therapeutic responses observed here may be similar to those
tration, predominantly of CD4 cells. Plasmacytoid lymphocytes noted to methotrexate in the previous reports. Of the six
and plasma cells were present in submucosal lymphoid aggregates patients who did not manifest all three of the pathologic findings
in the regions and patients with the most prominent inflammation. reported here, a more mixed response to immunosuppressive
Wenzel, Vitari, Shende, et al.: Asthmatic Granulomatosis: A New Disease? 507

therapy has been observed. However, four of the six have also 12. Sur S, Crotty TB, Kephart GM, Hyma BA, Colby TV, Reed CE, Hunt
been able to markedly decrease their oral CS requirements. LW, Gleich GJ. Sudden-onset fatal asthma. A distinct entity with few
This further supports the concept that the range of pathologies eosinophils and relatively more neutrophils in the airway submucosa?
Am Rev Respir Dis 1993;148:713–719.
reported here may represent varying spectrums of a similar dis-
13. Larkin ASYS, Wenzel SE. Clinical characteristics of a newly described
ease. Placebo-controlled trials are necessary to truly determine airway disease presenting with lymphocytic bronchiolitis and granulomas:
selective responsiveness of these patients. severe adult onset asthma misdiagnosed. J Allergy Clin Immunol 2011;
In conclusion, we report the unexpected finding of nonnecro- 127:AB223.
tizing granulomas in the airways of 10 patients with clinically and 14. Larkin ASYS, Wenzel SE. Small airway obstruction, eosinophilia and
physiologically defined severe asthma. We believe these patients airway granulomas: a new phenotype of severe asthma? Am J Respir
have a newly identified disease, which we name asthmatic gran- Crit Care Med 2011;183:A3723.
ulomatosis. This disease, which overlaps features of asthma, 15. Yousem SA, Martin T, Paradis IL, Keenan R, Griffith BP. Can immu-
autoimmunity, and granulomatous disease, seemingly responds nohistological analysis of transbronchial biopsy specimens predict
better to cytotoxic and antiinflammatory agents than to high- responder status in early acute rejection of lung allografts? Hum
Pathol 1994;25:525–529.
dose CSs. We suggest VATS biopsies be considered in atypical
16. Edinger JT, Kant JA, Swerdlow SH. Cutaneous marginal zone lym-
CS-dependent patients with severe asthma, especially females phomas have distinctive features and include 2 subsets. Am J Surg
who are characterized by adult-onset disease, a family history Pathol 2010;34:1830–1841.
of autoimmunity, low DLCO, disproportionately low FEF25–75% 17. Wenzel S, Wilbraham D, Fuller R, Getz EB, Longphre M. Effect of an
predicted with persistent blood eosinophils (>200 per milliliter) interleukin-4 variant on late phase asthmatic response to allergen
and FENO greater than 30 ppb despite systemic CS use. Further challenge in asthmatic patients: results of two phase 2a studies. Lancet
studies are needed to characterize the immunologic pathways 2007;370:1422–1431.
involved and to develop biomarkers that would more easily iden- 18. Chibana K, Trudeau JB, Mustovich AT, Hu H, Zhao J, Balzar S, Chu
tify and then treat this newly described disease. HW, Wenzel SE. IL-13 induced increases in nitrite levels are primarily
driven by increases in inducible nitric oxide synthase as compared with
Author disclosures are available with the text of this article at www.atsjournals.org. effects on arginases in human primary bronchial epithelial cells. Clin
Exp Allergy 2008;38:936–946.
Acknowledgment: The authors acknowledge the outside review of their patho-
logic findings by Dr. Thomas V. Colby (Mayo Clinic) and Dr. Jody Wright (Univer- 19. Corren J, Lemanske RF, Hanania NA, Korenblat PE, Parsey MV, Arron
sity of British Columbia). JR, Harris JM, Scheerens H, Wu LC, Su Z, et al. Lebrikizumab
treatment in adults with asthma. N Engl J Med 2011;365:1088–1098.
20. Bogaert P, Tournoy KG, Naessens T, Grooten J. Where asthma and
hypersensitivity pneumonitis meet and differ: noneosinophilic severe
References asthma. Am J Pathol 2009;174:3–13.
1. Standards for the diagnosis and care of patients with chronic obstructive 21. Nair P, Pizzichini MM, Kjarsgaard M, Inman MD, Efthimiadis A,
pulmonary disease (COPD) and asthma. This official statement of Pizzichini E, Hargreave FE, O’Byrne PM. Mepolizumab for
the American Thoracic Society was adopted by the ATS Board of prednisone-dependent asthma with sputum eosinophilia. N Engl J
Directors, November 1986. Am Rev Respir Dis 1987;136:225–244. Med 2009;360:985–993.
2. Miranda C, Busacker A, Balzar S, Trudeau J, Wenzel SE. Distinguishing 22. Haldar P, Brightling CE, Hargadon B, Gupta S, Monteiro W, Sousa A,
severe asthma phenotypes: role of age at onset and eosinophilic in- Marshall RP, Bradding P, Green RH, Wardlaw AJ, et al. Mepolizu-
flammation. J Allergy Clin Immunol 2004;113:101–108. mab and exacerbations of refractory eosinophilic asthma. N Engl J
3. Haldar P, Pavord ID, Shaw DE, Berry MA, Thomas M, Brightling CE, Med 2009;360:973–984.
Wardlaw AJ, Green RH. Cluster analysis and clinical asthma phe- 23. Simonian PL, Roark CL, Wehrmann F, Lanham AK, Diaz del Valle F,
notypes. Am J Respir Crit Care Med 2008;178:218–224. Born WK, O’Brien RL, Fontenot AP. Th17-polarized immune re-
4. Moore WC, Meyers DA, Wenzel SE, Teague WG, Li H, Li X, D’Agostino sponse in a murine model of hypersensitivity pneumonitis and lung
R Jr, Castro M, Curran-Everett D, Fitzpatrick AM, et al. Identification fibrosis. J Immunol 2009;182:657–665.
of asthma phenotypes using cluster analysis in the Severe Asthma 24. Barrera L, Mendoza F, Zuniga J, Estrada A, Zamora AC, Melendro EI,
Research Program. Am J Respir Crit Care Med 2010;181:315–323. Ramirez R, Pardo A, Selman M. Functional diversity of T-cell sub-
5. Proceedings of the ATS workshop on refractory asthma: current un- populations in subacute and chronic hypersensitivity pneumonitis.
derstanding, recommendations, and unanswered questions. American Am J Respir Crit Care Med 2008;177:44–55.
Thoracic Society. Am J Respir Crit Care Med 2000;162:2341–2351. 25. Burke ML, Jones MK, Gobert GN, Li YS, Ellis MK, McManus DP.
6. Wenzel SE, Schwartz LB, Langmack EL, Halliday JL, Trudeau JB, Immunopathogenesis of human schistosomiasis. Parasite Immunol
Gibbs RL, Chu HW. Evidence that severe asthma can be divided 2009;31:163–176.
pathologically into two inflammatory subtypes with distinct physi- 26. Moffatt MF, Gut IG, Demenais F, Strachan DP, Bouzigon E, Heath S,
ologic and clinical characteristics. Am J Respir Crit Care Med 1999; von Mutius E, Farrall M, Lathrop M, Cookson WO, et al. A large-
160:1001–1008. scale, consortium-based genomewide association study of asthma.
7. Balzar S, Chu HW, Strand M, Wenzel S. Relationship of small airway N Engl J Med 2010;363:1211–1221.
chymase-positive mast cells and lung function in severe asthma. Am J 27. Rubtsov AV, Rubtsova K, Kappler JW, Marrack P. Genetic and hor-
Respir Crit Care Med 2005;171:431–439. monal factors in female-biased autoimmunity. Autoimmun Rev 2010;
8. Balzar S, Wenzel SE, Chu HW. Transbronchial biopsy as a tool to 9:494–498.
evaluate small airways in asthma. Eur Respir J 2002;20:254–259. 28. Szczeklik A, Nizankowska E, Serafin A, Dyczek A, Duplaga M, Musial J.
9. Carroll N, Carello S, Cooke C, James A. Airway structure and inflam- Autoimmune phenomena in bronchial asthma with special reference
matory cells in fatal attacks of asthma. Eur Respir J 1996;9:709–715. to aspirin intolerance. Am J Respir Crit Care Med 1995;152:1753–1756.
10. Hamid Q, Song Y, Kotsimbos TC, Minshall E, Bai TR, Hegele RG, 29. Tan BK, Li QZ, Suh L, Kato A, Conley DB, Chandra RK, Zhou J,
Hogg JC. Inflammation of small airways in asthma. J Allergy Clin Norton J, Carter R, Hinchcliff M, et al. Evidence for intranasal an-
Immunol 1997;100:44–51. tinuclear autoantibodies in patients with chronic rhinosinusitis with
11. Saetta M, Di Stefano A, Rosina C, Thiene G, Fabbri LM. Quantitative nasal polyps. J Allergy Clin Immunol2011;128:1198–206.
structural analysis of peripheral airways and arteries in sudden fatal 30. Davies H, Olson L, Gibson P. Methotrexate as a steroid sparing agent
asthma. Am Rev Respir Dis 1991;143:138–143. for asthma in adults. Cochrane Database Syst Rev 2000;CD000391.