Available online at www.annclinlabsci.org

An Unusual Variant of Hepatic Inflammatory Angiomyolipoma: Report of a Rare Entity Hongxia Sun1 and Mary Yang1,2

1Department of Pathology and Laboratory Medicine, University of Texas Health Science Center - Medical School at Houston, Houston, TX, and 2Department of Pathology, LBJ General Hospital, Houston, TX, USA

Abstract. Hepatic angiomyolipoma is a rare mesenchymal tumor with a broad spectrum of histological ap-pearance. A 44-year-old female presented with abdominal discomfort. MRI revealed a 3.6 cm well-defined hepatic lesion. Ultrasound-guided core biopsy with imprint smears was performed. The smears showed histiocytes with a few epithelioid cells, mimicking an inflammatory lesion. The biopsy showed proliferation of epithelioid cells with aggregates of foamy histiocytes and scattered adipocyts, resembling hepatocellular neoplasm. The diagnosis of angiomyolipoma was clinched by the immunoprofile of the epithelioid cells, which were diffusely positive for HMB45, Mart-1 and focally positive for smooth muscle actin. To our best knowledge, this is the first reported case of biopsy diagnosed hepatic inflammatory angiomyolipoma with an unusual inflammatory infiltrate composed mainly of aggregates of foamy histiocytes.

Keywords: hepatic angiomyolipoma, inflammatory variant, hepatocellular neoplasm.

Introduction

Angiomyolipoma (AML) is a mesenchymal tumor composed of a mixture of various proportions of three heterogeneous tissue elements: blood vessels, myoid cells and adipocytes. AML arises most com-monly in the kidney either sporadically or in asso-ciation with tuberous sclerosis. Its occurrence is relatively uncommon in the liver. Hepatic AML was first described by Ishak in 1976 [1]. To date, about 300 cases have been reported in the English literature [2]. AML has a broad spectrum of histo-logical appearance and has been subclassified into multiple variants according to its predominant composed elements and its growth pattern [3,4]. Hepatic AML is essentially a benign lesion with oc-casional rupture and hemorrhage [5]. Malignant AML is extremely rare [6,7]. Preoperative imaging diagnosis of AML is challenging due to this wide range of variation in the proportions of the various components of the tumor. Imaging study of AML can closely mimic hepatocellular carcinoma, par-ticularly in cases of AML with minimal fat compo-nent [8]. Thus, imaging-guided cytology and his-tology evaluations are essential for establishing the

diagnosis for proper management. The diagnosis of classical AML is usually straightforward when the three different typical elements are apparent on the cytology smears or on the biopsy sections. However, unusual variants may be challenging. Here, we re-ported a case of biopsy diagnosed hepatic inflam-matory AML with striking aggregates of foamy his-tiocytes with discussion on the histological diagnostic challenges.

Case Report

A 44 year old female patient presented with history of hepatitis C and liver mass detected 2 years ago without work-up. The patient did not have any features of tuber-ous sclerosis complex. She complained of abdominal dis-comfort and fullness without fever, headache, or abdom-inal pain. Liver function tests were within normal range and serum alpha fetoprotein was 2.6 ng/ml. Magnetic resonance imaging of abdomen revealed a 3.6x3.1x3 cm well-defined hepatic lesion in the right lobe of the liver with no evidence of cirrhosis. Radiological differential diagnoses were hepatic adenoma vs. well-differentiated hepatocellular carcinoma.

Ultrasound-guided core biopsy was performed. Multiple tan-white cylindrical soft tissue cores were obtained from the patient. Two air-dried imprint smears were prepared from the core biopsy and were stained with DiffQuick for immediate assessment. The core biopsies were

0091-7370/16/0100-078. © 2016 by the Association of Clinical Scientists, Inc.

Address correspondence to Mary Yang, MD, Department of Pathology, LBJ General Hospital, 5656 Kelley Street, 3PA 93002A, Houston, Texas, 77026, USA; phone: 713 566 5710; fax: 713 440 1394; e mail: mary.y.yang@uth.tmc.edu

Annals of Clinical & Laboratory Science, vol. 46, no. 1, 201678

preserved in formalin. H&E sections were prepared. Acid fast stain (AFB), Gomori’s methenamine silver stain (GMS) and reticulin stain were performed.Immunohistochemistry was performed for smooth mus-cle actin (SMA) (Ventana, Tucson, AZ,mouse monoclo-nal, 1A4, 1:200 dilution), CD68 (Ventana, Tucson, AZ, mouse monoclonal, KP-1, 1:200 dilution), hepatocyte specific antigen (HSA) (Ventana, Tucson, AZ, mouse monoclonal, 0CH1E5, 1:200 dilustion), S100 (Ventana, Tucson, AZ, rabbit polyclonal, 1:200 dilution), vimen-tin (Ventana, Tucson, AZ, mouse monoclonal, V9, 1:200 dilution), CD34 (Ventana, Tuscon, AZ, mouse monoclonal, QBEnd/10, 1:200 dilution), HMB45 (Ventana, Tucson, AZ, mouse monoclonal, HMB45, 1:200 dilution), Mart-1 (Ventana, Tucson, AZ, mouse monoclonal, A103, 1:200 dilution), Cam5.2 (Ventana, Tucson, AZ, mouse monoclonal, Cam 5.2, 1:200 dilu-tion), and AE1/AE3 (Ventana, Tucson, AZ, mouse monoclonal, AE1/AE3/PCK26, 1:200 dilution).

Results

The air-dried DiffQuick smears showed mostly foamy histiocytes and rare single or small aggregates of epithelioid cells with ample granular cytoplasm (Figure 1). The epithelioid cells showed oval to elongated nuclei with fine chromatin and minimal atypia. Definite polygonal hepatocytes with central

round nuclei, adipocytes or thick-walled blood vessels were not pres-ent. A possible inflammatory or granuloma-tous lesion was suspected on immediate assessment.

Microscopic H&E sections showed strips of tissue composed of sheets of epi-thelioid cells with eosino-philic granular cytoplasm. Fat vacuoles/fat cells were pres-ent among the

epithelioid cells. Prominent aggregates of cells with foamy cytoplasm, minimal lymphocytic infiltrate, and focal thick-walled blood vessels were present (Figure 2A and B). No microorganisms were re-vealed on AFB and GMS stains. Reticulin stain showed inconspicuous reticulin fibers. The epithe-loid cells were diffusely positive for HMB45, Mart-1 (Figure 2C and D), vimentin, focally positive for SMA (Figure 2E) and negative for HSA, S100, Cam 5.2, AE1/AE3, and desmin. The foamy cells were positive for CD68 (Figure 2F), vimentin and negative for AE1/AE3, Cam 5.2, HSA, HMB45, Mart-1, SMA, desmin, confirming the cells to be foamy histiocytes. CD34 highlighted focal thick-walled blood vessels and focal areas suggestive of thick trabeculae with sinusoidal pattern. A diagno-sis of angiomyolipoma was rendered.

Discussion

According to the policies of our ethical committee, this study was exempted from review by our institu-tional review board. AML is currently considered to be a member of perivascular epithelioid cell tumor (PEComa) family in the WHO classification. This group of tumor is characterized histologically by

Figure 1. Cytological features. Foamy histiocytes and epithelioid cell with oval to elongated nuclei, and minimal atypia on imprint (A: Diff-Quick x20; B: Diff-Quick x20; C: Diff-Quick x40; D: Diff-Quick x40).

Hepatic inflammatory angiomyolipoma 79

perivascular epithelioid cells with clear or granular cytoplasm and immunohistochemically by HMB-45 positivity. Most cases are also positive for Melan-A, but HMB-45 seems to be more sensitive [9]. Pleuripotent primitive mesenchymal cells were speculated to be the progenitor cells [10]. With the presence of melanosome proteins in the tumor cells, a neural crest lineage origin has also been proposed [11]. By far, the precursor cell of PEComa is uncertain.

AML arises classi-cally in the kidney in association with tuberous sclerosis. The occurrence of tuberous sclerosis is relatively un-common in the liver. Similar to its renal counterpart, hepatic AML is composed of vari-able proportions of blood vessels, myoid cells and adipocytes. Nonomura et al classified hepatic AML into 10 his-tological types ac-cording to the rel-ative proportion of its three com-ponents: mixed (classic), lipoma-tous, angioma-tous, myomatous, angiomyomatous, myoangiomatous, myolipomatous, lipomyomatous, lipoangiomatous angiolipomatous [3]. Other report-ed unusual pat-terns include tra-becular, pelioid, inflammatory, on-cocytic, pleomor-phic etc. [4]. Morphology of the myoid element

can show a spectrum of appearance from spindle to intermediate to epithelioid. The high variation in the morphology of tumor occasionally causes diagnostic confusion. Hepatic AML of epithelioid variant and trabecular pattern can closely mimic hepatocellular carcinoma [12,13], while the in-flammatory variant with prominent inflammatory cells can resemble inflammatory pseudo-tumor [14].

Figure 2. Histological features and Immunohistochemistal staining. A-B: Section revealed epi-thelioid cells, fat vacuoles/fat cells and prominent aggregates of foamy cells (A. H&E x10; B. H&E x20). C-E: The epithelioid cells were positive for HMB45, Mart-1, and focally positive for SMA (C: HMB45 x20; D: Mart-1 x20; E: SMA x20). F: The foamy cells were positive for CD68 (CD68 x20).

Annals of Clinical & Laboratory Science, vol. 46, no. 1, 201680

Hepatic inflammatory angiomyolipoma

Most of the reported cases of hepatic AML were diagnosed on resected specimens. There were only a few reported cases with cytological findings. Xie et al. have a detailed summary of the reported cyto-morphology of fifteen hepatic AML, nine of which were classic AML and six of which were epithelioid AML [12]. Thirteen of the cases were findings of fine needle aspirations and two of them were im-prints of biopsy specimens. All of the reported cases showed hypercellular smears with epithelioid cells and variable spindle cells, adipocytes and thick-walled blood vessels. Lymphocytes and mast cells were reported in one case. In contrast to the previ-ously reported cytological findings, imprints of the present case showed mainly foamy histiocytes with only a few single and small aggregates of epithelioid cells. The predominance of histiocytes and lack of adipocytes or blood vessels on the smears had led to an initial impression of a possible inflammatory or granulomatous process.

On H&E core biopsy sections, the predominant epithelioid cells with eosinophilic granular cyto-plasm superficially mimicked hepatocytic cells while the intermingled fat cells resembled macrove-sicular fatty metamorphosis. With the lack of portal tracts or bile ducts, the features were suggestive of hepatocellular neoplasm with fatty metamorphosis. On top of these, the lack of reticulin fibers and fo-cal sinusoidal pattern highlighted by CD34 further led towards the impression of hepatocellular carci-noma. The aggregates of foamy histiocytes among the epithelioid cells also resembled hepatocytic cells with ballooning degeneration and raised the possi-bility of hepatocelluar carcinoma with foamy his-tiocyte-like appearance [15].

Although it is well known that fatty metamorphosis can occur in hepatoceullar adenoma or hepatocel-lular carcinoma, the larger size of the adipocytes in comparison with that of fat vacuoles in macrove-sicular fatty metamorphosis was an important clue to the true nature of the adipocytes. In addition, hepatocellular adenoma or hepatocellular carcino-ma predominantly shows vascular sinusoids histo-logically. The presence of thick-walled blood vessels would be unusual. The fine chromatin and minimal nuclei atypia of the epithelioid cells on the imprint smears also disproved a malignant diagnosis. A unique and striking feature of the present case was the prominent aggregation of cells with foamy

cytoplasm, which were confirmed to be histiocytes by positive CD68 staining. This feature led to the impression of possible inflammatory or granuloma-tous process. No microorganisms were revealed on AFB and GMS stains. The final diagnosis of inflam-matory angiomyolipoma was clinched by immu-nopositivity for HMB-45 and Mart-1 and focal expression of SMA of the epithelioid cells.

The inflammatory variant is the least common vari-ant of hepatic AML, in which the inflammatory component constituted more than 50% of the tu-mor. When the inflammatory component predom-inates, it can create diagnostic confusion with other hepatic neoplasms, such as inflammatory myofi-broblastic tumor, or follicular dendritic cell tumor. To date, only about nine cases of hepatic inflamma-tory AML have been reported in the English litera-ture [10,14,16-18], and all these cases were diag-nosed on tumor resection specimens, with one misdiagnosed as inflammatory pseudotumor on bi-opsy. On histological sections, all of the reported cases showed prominent mixed inflammatory infil-trate, including small lymphocytes, plasma cells, and histiocytes, throughout the lesions. Differ from the previous reported cases, the present case of he-patic inflammatory AML showed an unusual histo-logical feature with the inflammatory infiltrate composed mainly of aggregates of foamy histio-cytes, and minimal lymphocytic infiltrate. This morphologic feature has not been previously reported.

Cytological findings of inflammatory AML have not been reported in the literature for comparison. Our report illustrated the importance of keeping AML in mind for radiologically well-defined he-patic lesion even when the smears did not show the classical triphasic pattern. Diligent search for and recognizing epithelioid myoid cells on the smears and correlation with histological findings would be critical leading to the correct diagnosis.

In summary, we reported a biopsy diagnosed un-usual variant of hepatic inflammatory AML with the inflammatory infiltrate composed predomi-nantly of aggregates of foamy histiocytes, a variant which has not been previously reported. Considering the generally benign nature of hepatic AML and possible conservative management with close fol-low up for asymptomatic tumors less than 5 cm,

81

awareness of this rare entity and features differenti-ating it from inflammatory lesion and hepatocellu-lar neoplasm are essential for correct preoperative diagnosis and proper management.

References

1. Ishak KG. Mesenchymal tumors of the liver. In hepatocellular Carcinoma. Okuda K, Peters RL. New York, John Wiley & Sons, 1976, 247-307.

2. Agaimy A, Vassos N, Croner RS, Strobel D, Lell M. Hepatic angiomyolipoma: a series of six cases with emphasis on patho-logical-radiological correlations and unusual variants diag-nosed by core needle biopsy. Int J Clin Exp Pathol 2012; 5:512-521.

3. Nonomura A, Enomoto Y, Takeda M, Takano M, Morita K, Kasai T. Angiomyoliopma of the liver: a reappraisal of morpho-logical features and delineation of new characteristic histologi-cal features from the clinicopathological findings of 55 tu-mours in 47 patients. Histopathology 2012;61:863-880.

4. Tsui WM, Colombari R, Portmann BC, Bonetti F, Thung SN, Ferrell LD, Nakanuma Y, Snover DC, Bioulac-Sage P, Dhillon AP. Hepatic angiomyolipoma: a clinicopathologic study of 30 cases and delineation of unusual morphologic variants. Am J Surg Pathol 1999;23:34-48.

5. Occhionorelli S, Dellachiesa L, Stano R, Cappellari L, Tartarini D, Sever S, Palini GM, Pansini GC, Vasquez G: Spontaneous rupture of a hepatic epithelioid angiomyolipoma: damage control surgery. A case report. G Chir 2013;34:320-322.

6. Nguyen TT, Gorman B, Shileds D, Goodman Z. Malignant hepatic angiomyolipoma: report of a case and review of the lit-erature. Am J Surg Pathol 2008;32:793-798.

7. Deng YF, Lin Q, Zhang SH, Ling YM, He JK, Chen XF. Malignant angiomyolipoma in the liver: A case report with pathological and molecular analysis. Pathol Res Pract 2008;204:911-918.

8. Lee SJ, Kim SY, Kim KW, Shin YM, Kim HJ, Lee JS, Kim SA. Hepatic angiomyolipoma with minimal fat, mimicking hepa-tocellular carcinoma. Clin Mol Hepatol 2012;18:330-335.

9. Makhlouf HR, Ishak KG, Shekar R, Sesterhenn IA, Young DY, Fanburg-Smith JC. Melanoma markers in angiomyolipoma of the liver and kidney: a comparative study. Arch Pathol Lab Med 2002;126:49-55.

10. Liu Y, Wang J, Lin XY, Xu HT, Qiu XS, Wang EH. Inflammatory angiomyolipoma of the liver: a rare hepatic tu-mor. Diagn Pathol 2012;7:122.

11. Folpe AL, Kwiatkowski DJ. Perivascular epithelioid cell neo-plasms: Pathology and pathogenesis. Hum Pathol 2010;41:1-15.

12. Xie LJ, Jessurun J, Manivel JC, Pambuccian SE. Hepatic epi-thelioid angiomyolipoma with trabecular growth pattern: a mimic of hepatocellular carcinoma on fine needle aspiration cytology. Diagn Cytopathol 2012;40:639-650.

13. Szekely E, Schaff Z, Madaras L, Kupcsulik P, Zsirka A. Trabecular angiomyolipoma mimicking hepatic cell carcino-ma. Pathol Oncol Res 2000;6:224-226.

14. Kojima M, Nakamupra S, Ohno Y, Sugihara S, Sakata N, Nasawa N. Heptatic angiomyolipoma resembling an inflam-matory pseudotumor of the liver. A case report. Patho Res Pract 2004;200:713-716.

15. Noro T, Gotohda N, Kojima M, Konishi M, Nakaghori T, Takahashi S, Hasebe T, Kinoshita T. Hepatocellular carcino-ma with foamy histiocyte-like appearance: a deceptively clear cell carcinoma appearing variant. Case Rep Gastroenterol 2010;4:286-292.

16. Shi H, Cao D, Wei L, Sun L, Guo A. Inflammatory angiomyo-lipomas of the liver: a clinicopathologic and immunohisto-chemical analysis of 5 cases. Ann Diagn Pathol 2010;14:240-246.

17. C R J, Menon DP Augustine J, Abdul Siyad AK. Epithelioid angiomyolipoma of liver with an inflammatory component: a case report. Case Reports Hepatol 2013;2013:738708

18. Agaimy A, Markl B. Inflammatory angiomyolipoma of the liver: an unusual case suggesting relationship to IgG4-related pseudotumor. Int J Clin Exp Pathol 2013;6:771-779.

Annals of Clinical & Laboratory Science, vol. 46, no. 1, 201682