Eosinophilic angiocentric fibrosis is a rare indolent lesion of the head and neck region that has characteristic histologic findings of onionskin fibrosis and prominent eosinophils. Its pathogenesis has been poorly understood and has been most commonly attributed to hypersensitivity or previous trauma. Recently, the lesion has been included in the spectrum of immunoglobulin G4 (IgG4)–related disease. However, few of the existing cases of eosinophilic angiocentric fibrosis have been evaluated for IgG4+ and IgG+ plasma cells. Therefore, we provide an update on the clinical and histologic features of eosinophilic angiocentric fibrosis to increase awareness of the entity and encourage its further characterization as an IgG4-related disease.

Eosinophilic angiocentric fibrosis (EAF) is a rare lesion of the sinonasal and upper respiratory tract, most commonly presenting with prolonged obstructive symptoms. Eosinophilic angiocentric fibrosis was first described by Holmes and Panje1  in 1983 as “intranasal granuloma faciale” because of the histologic appearance of concentric fibrosis surrounding small-caliber arterioles, with numerous eosinophils scattered throughout the lesion. In 1985, Roberts and McCann2  described 3 similar cases and coined the histologically descriptive name. To date, there have been about 54 cases reported in the literature. In 2004, a case of EAF was presented as a resident review in the Archives of Pathology & Laboratory Medicine.3  At that time, the etiology of EAF was unclear and was most frequently attributed to hypersensitivity, prior trauma, or surgical manipulation. Since then, there has been varied discussion within the literature regarding its etiology; in 2011, EAF was first included in the spectrum of immunoglobulin G4 (IgG4)–related diseases.4  However, controversy remains as to whether EAF is a neoplasm or a fibroinflammatory lesion mimicking a neoplasm. Eosinophilic angiocentric fibrosis is an intriguing and clinically important entity of which pathologists and clinicians alike should be aware. Here, we report the clinicopathologic findings of EAF, with an update on its course, histologic features, etiology, and management.

Eosinophilic angiocentric fibrosis presents most commonly as a lesion in the nasal cavity in patients with a prolonged history of nasal obstruction or deformity. A history of allergies, nasal trauma, or surgery is common. Rarely, EAF may present in the larynx and lower respiratory tracts or orbit, causing dyspnea and epiphora, respectively.58  Men and women are equally affected, and the age at diagnosis ranges from 16 to 79 years.9  The lesion has an indolent course, and patients commonly present several years after symptom onset—up to 20 years.4  Rarely, EAF may present with a prior or concurrent history of granuloma faciale.10  There is a clear absence, however, of systemic diseases such as granulomatosis with polyangiitis, Churg-Strauss syndrome, sarcoidosis, Sjögren syndrome, and infectious granulomatosis.11 

Eosinophilic angiocentric fibrosis most commonly occupies the anterior nasal cavity and involves the nasal septum and lateral nasal wall.12,13  The lesion may extend into the adjacent sinuses and elicit reactive changes in adjacent bone, with rare bone destruction.12,14  Grossly, the lesion is a tan-white to pink, fleshy submucosal mass of variable size and extent. Mucosal ulceration or plaque formation may be present.5,1517 

Eosinophilic angiocentric fibrosis is a progressive fibrosing lesion demonstrating a spectrum of histologic findings dependent on the stage of disease.2,15,16  Key morphologic findings are summarized in Table 1. The early stage comprises a patchy eosinophilic vasculitis involving the small vessels of the submucosa. The eosinophils cluster and migrate through the vessel wall and show evidence of degranulation. There are also variable numbers of plasma cells and lymphocytes, both critical in the diagnosis of IgG4-related disease (Figure 1).

Table 1

Morphologic Features of Eosinophilic Angiocentric Fibrosisa

Morphologic Features of Eosinophilic Angiocentric Fibrosisa
Morphologic Features of Eosinophilic Angiocentric Fibrosisa
Figure 1

Vasculitis with lymphoplasmacytic infiltrate and prominent eosinophils (hematoxylin-eosin, original magnification ×400).

Figure 2Pseudogranulomatous appearance of fibrosis, lymphoid follicle (hematoxylin-eosin, original magnification ×200).

Figure 3Onionskin-like lamellar fibrosis with scattered eosinophils (hematoxylin-eosin, original magnification ×200).

Figure 4Lack of necrosis or true granulomas (hematoxylin-eosin, original magnification ×100).

Figure 1

Vasculitis with lymphoplasmacytic infiltrate and prominent eosinophils (hematoxylin-eosin, original magnification ×400).

Figure 2Pseudogranulomatous appearance of fibrosis, lymphoid follicle (hematoxylin-eosin, original magnification ×200).

Figure 3Onionskin-like lamellar fibrosis with scattered eosinophils (hematoxylin-eosin, original magnification ×200).

Figure 4Lack of necrosis or true granulomas (hematoxylin-eosin, original magnification ×100).

Close modal

As the lesion matures, foci of early fibrosis arise. Spindle-shaped fibroblasts produce a pseudogranulomatous appearance, and adjacent reactive lymphoid follicles may be present. There is no true granulomatous reaction or cytologic atypia (Figure 2).

The late stage is characterized by dense fibrous thickening of the subepithelial stroma, with progressive loss of plasma cells and lymphocytes.16  The eosinophils remain prominent. The concentric lamellar deposition of collagen creates the characteristic onionskin appearance. There is consistently a lack of necrosis, mitotic activity, or true granulomas (Figures 3 and 4).5,7,16,17 

Although EAF may be recognized without the use of ancillary studies, IgG and IgG4 immunohistochemistry should be performed to evaluate for IgG4-related disease. In addition to morphologic features, such as storiform-type fibrosis, the histologic criteria for IgG4-related disease includes an increased number of IgG4+ plasma cells (more than 50 IgG4+ plasma cells per high-power field is considered highly specific) and a ratio of IgG4+:IgG+ plasma cells greater than 40%.18  These histologic findings, in addition to the presence of a tumefactive lesion, involvement of multiple classic organs (synchronously or metachronously), subacute onset, elevated serum IgG4, elevated plasmablast levels, and swift response to immunosuppressive therapy, strongly favor the diagnosis of IgG4-related disease.19  Of note, other inflammatory conditions, such as primary sclerosing cholangitis and rheumatoid arthritis, and malignancies, including lymphoma, are associated with tumefactive lesions and elevated serum IgG4 levels.18  Therefore, the presence of systemic diseases mimicking the above clinical presentation should be ruled out in order to diagnose IgG4-related disease.

Additional immunohistochemical and flow cytometric studies, although not necessary for diagnosis, elucidate components of the lesion. The inflammatory infiltrate is composed of both CD20+ B lymphocytes and CD5+ T lymphocytes.20  Flow cytometry demonstrates no unusual T-cell population, and the plasma cells are polyclonal.15,20,21  The remaining infiltrate is composed of neutrophils, histiocytes, and eosinophils.14,15,21  The stromal cells are positive for vimentin, and few to none stain for smooth muscle actin or HHF35.15,16,21  The lesional cells are negative for S100 and CD35.15,21  Fungal, acid-fast, and bacterial stains have invariably negative results.5,22 

The differential diagnosis of EAF is broad and includes reactive conditions (granulomatosis with polyangiitis, Churg-Strauss syndrome, Kimura disease, angiolymphoid hyperplasia with eosinophilia, granuloma faciale, erythema elevatum diutinum), and neoplastic lesions (angiofibroma, schwannoma, fibroma, fibromatosis). Granulomatosis with polyangiitis, previously Wegener granulomatosis, and Churg-Strauss syndrome are systemic vasculitides whose clinical and histologic features aid in distinguishing them from EAF. Microscopically, the lack of giant cells, granulomas, and necrosis excludes granulomatosis with polyangiitis and Churg-Strauss syndrome in the diagnosis of EAF.11,1517,23  Kimura disease is a rare entity that occurs predominantly in the Asian population. It has a mixed inflammatory infiltrate with prominent fibrosis that may resemble EAF. However, its dense lymphoid aggregates with prominent germinal centers and lack of whorling fibrosis distinguish it from EAF.1517  Inflammatory myofibroblastic tumor is composed of spindled myofibroblasts in a mixed inflammatory background, which includes eosinophils but lacks the concentric whorling pattern characteristic of EAF.15,24  Angiolymphoid hyperplasia with eosinophilia, also known as epithelioid hemangioma, is a rare idiopathic condition with proliferating blood vessels and enlarged vesicular endothelial cells. Eosinophils may be numerous in this lesion, but its distinct endothelial cell morphology and minimal fibrosis distinguish it from EAF.15,17,23  Juvenile nasopharyngeal angiofibroma affects predominantly adolescent males and is characterized by an irregular collagenous background with haphazardly arranged vessels. Inflammation is minimal, however, and the onionskin pattern of collagen deposition seen in EAF is not present.16  Erythema elevatum diutinum has perivascular inflammatory infiltrates and fibrinoid degeneration. It is predominantly a necrotizing vasculitis, often demonstrating marked vascular hyalinization, both of which are lacking in EAF.16,17 

The treatment approach for EAF most commonly includes surgery (either debulking or complete excision), systemic and intralesional corticosteroids, or a combination of these modalities. Patient outcome is the most favorable in those who undergo complete surgical resection with or without medical therapy,9  and is expectedly least favorable with the sole use of corticosteroids or incomplete excision.1,2,15,16,20,22,25  Rituximab has rapidly improved serologic tests and tumefactive lesions in IgG4-related disease and has shown improvement in a few cases of EAF.4  Because of the high recurrence risk of EAF lesions, long-term follow up is recommended; however, to date, there has been no known malignant transformation.

The pathogenesis of EAF has been unclear and broadly attributed to trauma, surgical manipulation, and hypersensitivity based on a high tumor recurrence rate and a consistent histologic abundance of eosinophils.3  More recently, EAF has been noted to share both clinical and histologic characteristics of the nonneoplastic immune-mediated spectrum of IgG4-related disease, which involves multiple classic organs.4,19  The diagnostic criteria are complex and involve both clinical and histologic findings, which have been summarized in Table 2. Clinically, the tumefactive growth pattern and indolent clinical presentation seen with EAF is characteristic. Histologically, the storiform-type fibrosis and lymphoplasmacytic infiltrate fulfill 2 of 3 proposed major morphologic criteria for IgG4-related disease.18  The third criterion, obliterative phlebitis, is not seen in EAF. Additional “minor” features that are seen in IgG4-related disease are nonobliterative phlebitis and prominent eosinophils, the latter of which EAF demonstrates.

Table 2. 

Characteristic Features of Immunoglobulin G4 (IgG4)–Related Disease

Characteristic Features of Immunoglobulin G4 (IgG4)–Related Disease
Characteristic Features of Immunoglobulin G4 (IgG4)–Related Disease

In addition to these morphologic features, elevated levels of IgG4+ plasma cells or an elevated IgG4:IgG ratio greater than 40% are required for definitive histologic diagnosis of IgG4-related disease.18  Although cases of EAF occurring prior to the description of IgG4-related disease were not stained with IgG4 or IgG, in 2011 Deshpande et al4  found that 4 of 5 cases of EAF fulfilled histologic criteria, including elevated IgG4+ plasma cells and an IgG4:IgG ratio of greater than 40%. The single case without increased IgG4 plasma cells was attributed to tissue evaluation at a late stage of disease. Among subsequently reported cases, only 1 had IgG4 and IgG stains performed: There were 10 IgG4+ plasma cells per high-power field, an IgG4:IgG ratio of 10%, and no clinical signs suggestive of IgG4-related disease.10  Therefore, in this case, a definitive diagnosis of IgG4-related disease could not be made. With the variability in staining of IgG4 plasma cells and a paucity of reported cases fulfilling IgG4-related disease criteria, it is important for pathologists to be aware of this entity and continue its characterization with IgG4 and IgG.

The pathogenesis of IgG4-related disease is incompletely understood, but there is growing evidence that the collaboration of clonal cytotoxic T cells and activated B-cell subsets plays a central role in the secretion of key profibrotic cytokines.26  The precise role of eosinophils in EAF, otherwise present in a minority of IgG4-related disease, is also unclear. However, tissue eosinophilia, classically due to immune hypersensitivity and recruitment by interleukin 5, is known to produce cytotoxic inflammatory mediators, several of which stimulate tissue fibroblasts.27  It is notable that eosinophils are capable of surviving in tissue for several weeks, depending on the presence of cytokines.27  This likely reflects the persistence of eosinophils within EAF throughout its evolution as well as its indolent clinical course. These findings, in addition to the lack of reported malignant transformation, support that EAF is either nonneoplastic or is a very indolent neoplasm. Further studies are needed to completely elucidate the initiating events in EAF pathogenesis. Of note, EAF shares many features with granuloma faciale, a rare, benign cutaneous disorder with an unclear etiology. When EAF was first described in 1985, it was suggested to be a mucosal variant of granuloma faciale2  because both entities have an eosinophilic infiltrate, vasculitis, and fibrosis.28  Furthermore, approximately 11% of EAF cases have a concurrent diagnosis of granuloma faciale,9  whether synchronous or metachronous in occurrence. The 2 entities therefore have been suggested to be different stages of the same lesion.1,2,10,14,22,25,28  Workup for IgG4-related disease in granuloma faciale has demonstrated mixed results.29,30 

Eosinophilic angiocentric fibrosis is a rare, indolent tumefactive lesion presenting in the sinonasal cavity, orbit, and the upper and rarely lower respiratory tract that is now included in the spectrum of autoimmune-mediated IgG4-related disease. As a clinicopathologic diagnosis, it is important for pathologists to be aware of EAF and to perform the indicated histologic studies in order to direct clinical workup of IgG4-related disease if not already suspected. Furthermore, because most reported cases occurred prior to the classification of IgG4 disease, the continued immunophenotypic characterization of EAF as IgG4-related disease is critical. The best clinical management of EAF involves complete excision with close follow-up because of its high risk of recurrence. Rituximab has been used, with a favorable response.

The photographs included in this article are from a case of eosinophilic angiocentric fibrosis diagnosed at West Virginia University. We thank Dana Gray for her assistance editing the figures.

1
Holmes
DK,
Panje
WR.
Intranasal granuloma faciale
.
Am J Otolaryngol
.
1983
;
4
(
3
):
184
186
.
2
Roberts
PF,
McCann
BG.
Eosinophilic angiocentric fibrosis of the upper respiratory tract: a mucosal variant of granuloma faciale?: a report of three cases
.
Histopathology
.
1985
;
9
(
11
):
1217
1225
.
3
Onder
S,
Sungur
A.
Eosinophilic angiocentric fibrosis: an unusual entity of the sinonasal tract
.
Arch Pathol Lab Med
.
2004
;
128
(
1
):
90
91
.
4
Deshpande
V,
Khosroshahi
A,
Nielsen
GP,
Hamilos
DL,
Stone
JH.
Eosinophilic angiocentric fibrosis is a form of IgG4-related systemic disease
.
Am J Surg Pathol
.
2011
;
35
(
5
):
701
706
.
5
Fageeh
NA,
Mai
KT,
Odell
PF.
Eosinophilic angiocentric fibrosis of the subglottic region of the larynx and upper trachea
.
J Otolaryngol
.
1996
;
25
(
4
):
276
278
.
6
Valenzuela
AA,
Whitehead
KJ,
Brown
I,
Sullivan
TJ.
Eosinophilic angiocentric fibrosis: an unusual entity producing complete lacrimal duct obstruction
.
Orbit
.
2006
;
25
(
2
):
159
161
.
7
Leibovitch
I,
James
CL,
Wormald
PJ,
Selva
D.
Orbital eosinophilic angiocentric fibrosis case report and review of the literature
.
Ophthalmology
.
2006
;
113
(
1
):
148
152
.
8
Kiratli
H,
Onder
S,
Yildiz
S,
Ozseker
H.
Eosinophilic angiocentric fibrosis of the orbit
.
Clin Exp Ophthalmol
.
2008
;
36
(
3
):
274
276
.
9
Fang
CH,
Mady
LJ,
Mirani
NM,
Baredes
S,
Eloy
JA.
Sinonasal eosinophilic angiocentric fibrosis: a systematic review
.
Int Forum Allergy Rhinol
.
2014
;
4
(
9
):
745
752
.
10
Chen
VH,
Grossniklaus
HE,
DelGaudio
JM,
Kim
HJ.
A concomitant case of orbital granuloma faciale and eosinophilic angiocentric fibrosis
.
Ophthal Plast Reconstr Surg
.
2017
;
33
(
2
):
e47
e49
.
11
Magro
CM,
Dyrsen
M.
Angiocentric lesions of the head and neck
.
Head Neck Pathol
.
2008
;
2
(
2
):
116
130
.
12
Yang
BT,
Wang
YZ,
Wang
XY,
Wang
ZC.
Nasal cavity eosinophilic angiocentric fibrosis: CT and MR imaging findings
.
AJNR Am J Neuroradiol
.
2011
;
32
(
11
):
2149
2153
.
13
Sunde
J,
Alexander
KA,
Reddy
VV,
Woodworth
BA.
Intranasal eosinophilic angiocentric fibrosis: a case report and review
.
Head Neck Pathol
.
2010
;
4
(
3
):
246
248
.
14
Narayan
J,
Douglas-Jones
AG.
Eosinophilic angiocentric fibrosis and granuloma faciale: analysis of cellular infiltrate and review of literature
.
Ann Otol Rhinol Laryngol
.
2005
;
114
(
1, pt 1
):
35
42
.
15
Altemani
AM,
Pilch
BZ,
Sakano
E,
Altemani
JM.
Eosinophilic angiocentric fibrosis of the nasal cavity
.
Mod Pathol
.
1997
;
10
(
4
):
391
393
.
16
Thompson
LD,
Heffner
DK.
Sinonasal tract eosinophilic angiocentric fibrosis: a report of three cases
.
Am J Clin Pathol
.
2001
;
115
(
2
):
243
248
.
17
Kosarac
O,
Luna
MA,
Ro
JY,
Ayala
AG.
Eosinophilic angiocentric fibrosis of the sinonasal tract
.
Ann Diagn Pathol
.
2008
;
12
(
4
):
267
270
.
18
Deshpande
V,
Zen
Y,
Chan
JK,
et al.
Consensus statement on the pathology of IgG4-related disease
.
Mod Pathol
.
2012
;
25
(
9
):
1181
1192
.
19
Deshpande
V
.
IgG4 related disease of the head and neck
.
Head Neck Pathol
.
2015
;
9
(
1
):
24
31
.
20
Jain
R,
Robblee
JV,
O'Sullivan-Mejia
E,
et al.
Sinonasal eosinophilic angiocentric fibrosis: a report of four cases and review of literature
.
Head Neck Pathol
.
2008
;
2
(
4
):
309
315
.
21
Tabaee
A,
Zadeh
MH,
Proytcheva
M,
LaBruna
A.
Eosinophilic angiocentric fibrosis
.
J Laryngol Otol
.
2003
;
117
(
5
):
410
413
.
22
Burns
BV,
Roberts
PF,
De Carpentier
J,
Zarod
AP.
Eosinophilic angiocentric fibrosis affecting the nasal cavity: a mucosal variant of the skin lesion granuloma faciale
.
J Laryngol Otol
.
2001
;
115
(
3
):
223
226
.
23
Li
Y,
Liu
H,
Han
D,
Zang
H,
Wang
T,
Hu
B.
Eosinophilic angiocentric fibrosis of the nasal septum
.
Case Rep Otolaryngol
.
2013
;
2013
:
267285
.
24
Werner
RL,
Castle
JT.
Inflammatory myofibroblastic tumor of the nasal cavity
.
Head Neck Pathol
.
2016
;
10
(
3
):
336
339
.
25
Yung
A,
Wachsmuth
R,
Ramnath
R,
Merchant
W,
Myatt
AE,
Sheehan-Dare
R.
Eosinophilic angiocentric fibrosis–a rare mucosal variant of granuloma faciale which may present to the dermatologist
.
Br J Dermatol
.
2005
;
152
(
3
):
574
576
.
26
Mattoo
H,
Stone
JH,
Pillai
S.
Clonally expanded cytotoxic CD4+ T cells and the pathogenesis of IgG4-related disease
.
Autoimmunity
.
2017
;
50
(
1
):
19
24
.
27
Rothenberg
ME.
Eosinophilia
.
N Engl J Med
.
1998
;
338
(
22
):
1592
1600
.
28
Stelini
RF,
Moyses
MD,
Cintra
ML,
et al.
Granuloma faciale and eosinophilic angiocentric fibrosis: similar entities in different anatomic sites
.
Appl Immunohistochem Mol Morphol
.
2017
;
25
(
3
):
213
220
.
29
Oliveira
CC,
Ianhez
PE,
Marques
SA,
Marques
ME.
Granuloma faciale: clinical, morphological and immunohistochemical aspects in a series of 10 patients
.
An Bras Dermatol
.
2016
;
91
(
6
):
803
807
.
30
Cesinaro
AM,
Lonardi
S,
Facchetti
F.
Granuloma faciale: a cutaneous lesion sharing features with IgG4-associated sclerosing diseases
.
Am J Surg Pathol
.
2013
;
37
(
1
):
66
73
.

Author notes

The authors have no relevant financial interest in the products or companies described in this article.