Skip to main content
Open Access Publications from the University of California

Dermatology Online Journal

Dermatology Online Journal bannerUC Davis

Cutaneous Epithelioid Angiomatous Nodule: Different views or interpretations in the analysis of ten new cases

Main Content

Cutaneous Epithelioid Angiomatous Nodule: Different views or interpretations in the analysis of ten new cases
W I Al-Daraji1, RJ Prescott2, A Abdellaoui3, MM Khan1, K Kulkarni1, MM Youssef4, BG Zelger5, B Zelger6
Dermatology Online Journal 15 (3): 2

1. Department of Histopathology, Nottingham, United Kingdom.
2. Blackburn Royal Infirmary, Blackburn, United Kingdom
3. Radiology Department, Derriford Hospital, Plymouth, United Kingdom
4. Department of Histopathology, Michigan
5. Department of Pathology
6. Departments of Dermatology and Venerology, Innsbruck, Austria


Classification schemes proposed for vascular lesions are the subjects of significant controversy. Cutaneous epithelioid angiomatous nodule (CEAN) was described in 2004, but there is no agreement as to whether this is a distinct entity or a type of either epithelioid hemangioma or angiolymphoid hyperplasia with eosinophilia. We present a typical case of CEAN and discuss nine other cases from our institution. We then provide two opposing viewpoints concerning its classification.


The nomenclature of vascular lesions is at best confusing and often incoherent, because the pathogenesis of many lesions is poorly understood or not understood at all. Thus, classification schemes have been proposed, but none is entirely satisfactory at the present [1].

Recently, Brenn and Fletcher (2004) described a new distinct clinicopathologic entity that they called cutaneous epithelioid angiomatous nodule (CEAN), resulting in more confusion among experts in the field [2]. In 2008, Sanguenza et al. [3] described an additional ten cases of CEAN and proposed that they should be classified as epithelioid hemangioma (EH) or angiolymphoid hyperplasia with eosinophilia (ALHE). In addition, three single case reports were also described [4, 5, 6]. Interestingly, one of these cases demonstrated the absence of microsatellite instability and another case was in an unusual location on the external ear [4, 5].

The main aim of this paper is to discuss opposing opinions regarding CEAN and whether it is a distinct clinicopathologic entity or a variant of EH or pyogenic granuloma (PyG) [3]. The two opposing opinions are expressed between the first author (WAD) and the senior author (BZ) according to the analysis of ten new cases, a review of the literature, and personal experiences.

Cutaneous epithelioid angiomatous nodule is a benign lesion that usually presents as a reddish to bluish papule or nodule. In the 28 cases described to date, no recurrence or metastasis has been documented. Treatment was usually simple surgical excision [2, 3].

The lesion frequently consists of a solid, often well-circumscribed, proliferation of large epithelioid endothelial cells with frequent intracytoplasmic vacuoles in the superficial dermis. The endothelial lined channels can only be identified focally and these cells are positive for endothelial markers such as CD31, CD34, and von Willebrand factor. Mitotic activity is variable (up to 5/10 HPF). However, atypical mitoses are absent [2, 3].

Materials and methods

Cases were obtained from archival files of all authors' institutions. Years 1998-2008 were screened for the diagnosis of cases that fit with CEAN current histological criteria and revealed ten cases; nine further cases were excluded because they did not exactly fit the current criteria available or showed collision features with PyG and EH. All cases presented as a single lesion. Table 1 summarizes the clinical data. In each case, a slide was stained with hematoxylin and eosin. Careful follow-up was obtained from clinical charts and treating physicians; the results were tabulated and analyzed. Standard immunohistochemical methods were used to stain sections from each case (when sufficient material was available, n=7) for CD31, CD34, S-100 protein, HHV-8 and cytokeratins (Cam 5.2 and MNF-116). Special stains for microorganisms included Steiner (n=2) and Warthin-Starry (n=2). Clinical photographs were available for two patients.

All results were used to create a database, on which statistical analysis by means of an SPSS (Statistical Package for Social Sciences) rel.8 System was performed. P<0.05 was considered statistically significant.


Clinical findings

Figure 1
Figure 1. Clinical photograph of cutaneous epithelioid angiomatous nodule from the lower lip of a 33-year-old male (patient 4, see acknowledgement)

The patients included three females and seven males with a gender predilection to males (0.428:1), age 10 to 82 (median 46.1 years, mode 45.5 years). Clinically, the lesions showed a reddish or bluish papule to nodule of short duration, less than 13 months (mean in this series is 4.3 months but only four cases had this information available). One case presented as an ulcer on the lower lip (Fig. 1). The lesion sizes ranged between 3 mm and 6 mm (median=4.3 mm, mode=4 mm) in maximum dimension. The anatomic sites at which CEAN presented were the trunk (abdomen (n=1), back and breast (n=3), vulva (n=1), extremities (n=2) and head (n=3). There were no other known concurrent or prior vascular lesions, and the clinical history was negative for infectious diseases (n=5) including HIV (n=2). In addition, no other medical problems or underlying medical diseases were reported (n=6). X-ray studies were done on three patients and showed no evidence of underlying bone involvement.

In a follow-up period between four months and 84 months (mean=29.1, and mode=23), there has been no recurrence or progression. A summary of the follow-up data is found in Table 1.

Pathologic features

Figure 2aFigure 2b
Figure 2a. Scanning and Figure 2b. High power magnification of cutaneous epithelioid angiomatous nodule (from WA) showing a well-circumscribed nodule with sheets of epithelioid cells with hypochromatic nuclei, ill-defined cytoplasm, intracytoplasmic vacuoles. and some moderate vascularity.

Figure 2cFigure 3
Figure 2c. Immunohistochemistry for CD31.
Figure 3. Cutaneous epithelioid angiomatous nodule (from BZ) with characteristic silhouette of well defined dermal nodule and sheets of epithelioid endothelial cells in scant stroma.

All tumors were located in the superficial dermis without extension into deep dermis or subcutis (n=9) and submucosa (n=1). They all had a well circumscribed, unilobular, nodular appearance (Fig. 2a). The lesions were composed mainly of solid and sheet-like proliferation of large polygonal epithelioid cells with frequent intracytoplasmic vacuoles in the superficial dermis (Figs. 2b & 3). The epithelioid cells contained both clear or abundant eosinophilic cytoplasm and vesicular nuclei with prominent nucleoli. Mitotic figures were quite low (2/10 high power field) compared to what Brenn found (5/10 high power field) [2], but similar to Brenn and Sanguenza [3]; no atypical mitoses or nuclear atypia were noted. Dispersed within the lobules of all 10 cases were open channels displaying a single cell-lined endothelium. In all cases, a constant finding was the presence of variable numbers of inflammatory cells consisting mainly of lymphocytes and notably eosinophils (average 4-6/10 high power field). Hemosiderin as well as numerous RBCs were present within and outside the vessels. The dermis surrounding the lesions contained dilated vessels; mild fibrosis was present in four cases [2, 3].

By immunohistochemistry, the epithelioid cells stained positively for at least CD31 or CD34 (n=7/7). Immunohistochemistry for other markers mentioned above were negative as were special stains for microorganisms, when sufficient material was available [2, 3].


Cutaneous epithelioid angiomatous nodule is a recently described entity. It is an acquired, probably reactive process due to its short duration of development. It is characterized by vascular features with some overlapping findings with other reactive vascular proliferations, leading to some confusion regarding the nature of CEAN. Herein, we express two different opinions as to whether CEAN is a distinct pathologic entity or not, followed by concluding remarks; we leave the discussion open for the readers to consider whether they may favor one opinion or the other.

WAD Opinion

In the first 18 cases published, CEAN has been separated as a distinct entity, followed by another series of ten cases published in 2008 by Sangueza et al. [3], in which the authors proposed classification of this neoplasm as a variant of EH. All the above demonstrated the confusion that this new entity has caused among experts in the field.

However, in my experience CEAN is a distinct clinicopathologic entity for many reasons and for several lines of evidence listed below.

Although CEAN is a reactive process such as PyG, as we all agree, PyG usually affects face and mouth and they are usually post traumatic. None of the CEAN cases was preceded by trauma and most of the cases in our series and those of Brenn and Fletcher [2] affected the trunk and extremities. In our series only one single case presented clinically as herpes labialis, which failed to respond to treatment. In addition, none of the cases in our series extended into the deep dermis, subcutis, or intravascular space (which may be seen in PyG). In addition, most cases described were solitary except for a few and none had a history of recurrence, dissemination, or satellitosis after trauma or removal. Histologically, PyG features are: lobular proliferation of capillaries consisting of cytologically bland endothelial cells, surrounded by pericytes. This raises the question: Why do bland endothelial cells have the appearances of "ugly" epithelioid cells with relatively "scary" looking appearances in CEAN?

If it were a variant of EH as suggested by both Sangueza and BZ, undoubtedly, CEAN would show considerable overlap with EH. However, the latter consists of prominent capillary proliferation with at least partially epithelioid endothelial morphology, but with bland morphology without mitotic figures [7-14]. Clinically EHs form a cutaneous or subcutaneous entity that often appears in the head and neck. This predisposition for head and neck was only noted in the majority of cases described by Sangueza et al. (2008), but not in Brenn's cases and the current series in which trunk and extremities were mostly involved. EH is more common (compared to CEAN) in middle aged and older women. It can be multiple, involving arterial walls, and is usually characterized by a more prominent vasoformative component (not seen in CEAN); it is rarely solidly cellular as is CEAN. In contrast to the presence of inflammatory cells (including eosinophils) in EH, no lymphoid follicles or germinal centers were present in any CEAN described here. Furthermore, the overlying epidermis is usually hyperplastic, hyperkeratotic, and acanthotic; the stroma (when encountered) in CEAN is usually in a perilesional distribution rather than within the lesion itself as in EH. Very recently in an update by Goh and Calonje [15], CEAN was considered as a distinct entity. However, they did not state why this was the case as detailed here.

In conclusion, in my experience CEAN represents a distinct group of both clinically and morphologically distinct vascular lesions in the skin that defies classification according to current criteria and should be separated from PyG, EH, and other reactive vascular entities.

BZ Opinion

The initial observation and description of Brenn and Fletcher is correct [2]; all three cases from Innsbruck (and others from slide seminars and dermatopathology clubs as well as personal communications and discussions) fall into the setting of CEAN presented in the original presentation. My main point regards the teleological character of this lesion: what type of disease is CEAN? Again I agree with Brenn and Fletcher: a reactive process. In this sense CEAN is related to pyogenic granuloma (PyG) and epithelioid hemangioma (EH), the latter also known under a variety of different names such as histiocytoid hemangioma, angiolymphoid hyperplasia with eosinophilia (ALHE), and many others [12, 13, 14, 16]. Pyogenic granuloma is reactive and occurs in easily traumatized areas, particularly around the mouth or acral, and frequently in children or young adults. But it can occur at any age, anywhere on the body, and not only superficially but in deep locations such as in subcutaneous or soft tissue or intravascularly (PyG). Moreover, agminated variants may show a number of lesions spread into a localized area, frequently the trunk or proximal extremities; sometimes they may manifest with numerous lesions ("recurrences") when a "primary" has been resected or sometimes without previous surgical intervention. Histology shows granulation tissue with endothelial cells and pericytes in a loose, more or less myxoid, later fibrosclerotic, stroma with a lobulated appearance and a variable admixture of inflammatory cells including neutrophils, eosinophils, lymphocytes, and macrophages.

Epithelioid hemangioma (or ALHE) is a similar reactive vascular process, mostly on the head of young women around the ear [12, 13, 14, 16]. But again, this lesion has been described in all ages, in many other locations all over the body. The histomorphological characteristics are epithelioid endothelial cells with intracellular or subendothelial blebs, i.e. lumen formations, which are highly characteristic. Lesions are vaguely circumscribed and are intermingled by mostly loose stroma with a frequently prominent admixture of lymphocytes and eosinophils.

In my experience, CEAN seems to combine features of PyG and EH, i.e., the lobular growth pattern and the epithelioid character of cells, which in exaggerated expression of this process become so prominent that the tumor forms sheets of moderately atypical cells. When this occurs, it may cause diagnostic difficulties in its differentiation from angiosarcoma, as correctly stated by Brenn and Fletcher in their initial report. Yet, the scanning magnification usually is helpful; it shows the well-structured lobular character, the collarette phenomenon of the covering epidermis, and the absence of any architectural criteria of irregularity. These findings combine with a clinical appearance that is unremarkable, usually vascular, and often thought to represent PyG.

Figure 4aFigure 4b
Figure 4a. Characteristic lobulated silhouette of pyogenic granuloma that on high power magnification (Figure 4b) reveals areas with sheets of epithelioid cells similar to cutaneous epithelioid angiomatous nodule.

Moreover, when looking at this group (CEAN, PyG, EH) one will not infrequently find overlapping features, as in the nine cases excluded from our series. In seven instances from Innsbruck, features of CEAN were seen as part of vascular lesions that also showed features of PyG or EH (Figs. 4 & 5) and, thus, not included in this series. This collective consisted of four female and three male patients, between 15 and 61 years of age. Five tumors were found on the head or face, one on the breast, and another on the leg.

Figure 5aFigure 5b
Figure 5a. Eroded lesion of epithelioid hemangioma from the face that reveals a moderately well-circumscribed nodule that on high power magnification (Figure 5b) reveals characteristic hobnail appearance of peripheral vessels, but central features of sheets of epithelioid cells like cutaneous epithelioid angiomatous nodule.

This may to some degree explain why initially CEAN was diagnosed as EH (see also Brenn and Fletcher) and why the observer will see PyG with CEAN features, at least focally. Similarly, Sangueza et al. reported similarities of CEAN to EH. This supports the interpretation of CEAN as a reactive vascular process with variations. One could speculate about possible reasons for these variations including location, age, "immune status," and individual characteristics. Yet in reality, we do not know. In this sense CEAN, including PyG and EH, are similar to the phenomena of variation seen in dermatofibromas [17] and xanthogranulomas[18], which are inflammatory processes with predominantly fibrocytic and macrophage differentiation, respectively. Unusual presentations of dermatofibromas may be epithelioid, cellular, and granular; they may also exhibit clear cells. Unusual presentations of xanthogranuloma include variants with a confusing array of so-called "non-X histiocytic types" such as reticulohistiocytoma, multicentric reticulohistiocytosis, xanthoma dissenminatum, generalized eruptive histiocytomas, and many others. Moreover, some cases of dermatofibromas may be difficult to separate from xanthogranulomas: such as epithelioid cell histiocytoma from mononulear xanthogranuloma, or lipidized fibrous histiocytoma from papular xanthoma. In CEAN, PyG, and EH the endothelial cells and pericytes, which vary in their appearance and architecture, result in different variations of a reactive vascular process (according to the present literature and in my experience). There are even some similarities between reactive vascular, fibrocytic, and macrophage disorders e.g., sclerosing hemangiomas that are vascular rich as opposed to the classic dermatofibroma or fibrous histiocytoma [17].

In short, we split out entities such as CEAN, but should not forget the importance of lumping, i.e., to know what type of lesions we are discussing. CEAN is a reactive vascular process related to PyG and EH and the knowledge of this relationship is important for the correct management of patients.

Concluding remarks

The differential diagnoses of CEAN have been detailed in papers by Brenn and Fletcher (2004) [2] and Sanguenza (2008) [3]. It is not the aim of this paper to list the differential diagnoses, but rather to discuss with an open mind whether CEAN is a distinct entity and for the readers to draw to their own conclusions.

Acknowledgement: The authors are grateful to A von Stein, Dresden, Germany, for providing the clinical photograph of Figure 1 (patient 4) to these series.


1. Mentzel T, Brown LF, Dvorak HF et al. The association between tumour progression and vascularity in myxofibrosarcoma and myxoid/round cell liposarcoma. Virchows Arch 2001;438;13-22. [PubMed]

2. Brenn T, Fletcher CD. Cutaneous epithelioid angiomatous nodule: a distinct lesion in the morphologic spectrum of epithelioid vascular tumors. Am J Dermatopathol 2004;26;14-21. [PubMed]

3. Sangueza OP, Walsh SN, Sheehan DJ, Orland AF, Llombart B, Requena L. Cutaneous epithelioid angiomatous nodule: a case series and proposed classification. Am J Dermatopathol 2008;30;16-20. [PubMed]

4. Alvarez-Arguelles-Cabrera H, Guimera-Martin-Neda F, Carrasco JL, Garc'a-Castro MC, Hernandez-Leon CN, D'Az-Flores L. Cutaneous epithelioid angiomatous nodule. J Eur Acad Dermatol Venereol 2008. [PubMed]

5. Kantrow S, Martin JD, Vnencak-Jones CL, Boyd AS. Cutaneous epithelioid angiomatous nodule: report of a case and absence of microsatellite instability. J Cutan Pathol 2007;34;515-516. [PubMed]

6. Zamecnik M. Relationship between cutaneous epithelioid angiomatous nodule and epithelioid hemangioma. Am J Dermatopathol 2004;26;351-352. [PubMed]

7. McMenamin ME, Fletcher CD. Malignant myopericytoma: expanding the spectrum of tumours with myopericytic differentiation. Histopathology 2002;41;450-460. [PubMed]

8. Mentzel T, Beham A, Calonje E, Katenkamp D, Fletcher CD. Epithelioid hemangioendothelioma of skin and soft tissues: clinicopathologic and immunohistochemical study of 30 cases. Am J Surg Pathol 1997;21;363-374. [PubMed]

9. Mentzel T, Calonje E, Fletcher CD. [Vascular tumors of the skin and soft tissue. Overview of newly characterized entities and variants]. Pathologe 1994;15;259-270. [PubMed]

10. Mentzel T, Calonje E, Nascimento AG, Fletcher CD. Infantile hemangiopericytoma versus infantile myofibromatosis. Study of a series suggesting a continuous spectrum of infantile myofibroblastic lesions. Am J Surg Pathol 1994;18;922-930. [PubMed]

11. Mentzel T, Mazzoleni G, Dei Tos AP, Fletcher CD. Kaposiform hemangioendothelioma in adults. Clinicopathologic and immunohistochemical analysis of three cases. Am J Clin Pathol 1997;108;450-455. [PubMed]

12. Olsen TG, Helwig EB. Angiolymphoid hyperplasia with eosinophilia. A clinicopathologic study of 116 patients. J Am Acad Dermatol 1985;12;781-796. [PubMed]

13. Quante M, Patel NK, Hill S et al. Epithelioid hemangioendothelioma presenting in the skin: a clinicopathologic study of eight cases. Am J Dermatopathol 1998;20;541-546. [PubMed]

14. Weiss SW, Ishak KG, Dail DH, Sweet DE, Enzinger FM. Epithelioid hemangioendothelioma and related lesions. Semin Diagn Pathol 1986;3;259-287. [PubMed]

15. Cutaneous vascular tumours: an update. Goh SG, Calonje E. Histopathology. 2008 May;52(6):661-73. Epub 2008 Feb 6. Review. [PubMed]

16. Wells GC, Whimster IW. Subcutaneous angiolymphoid hyperplasia with eosinophilia. Br J Dermatol 1969;81;1-14. [PubMed]

17. Zelger B, Zelger BG, Burgdorf WH. Dermatofibroma - a critical evaluation. Int J Surg Pathol 2004; 12; 333-344. [PubMed]

18. Burgdorf WHC, Zelger B. The non-Langerhans cell histiocytoses in childhood. Cutis 1996; 58; 201-207. [PubMed]

© 2009 Dermatology Online Journal