Deep juvenile xanthogranuloma invading the left tensor fasciae latae muscle: a case report and a literature review

Abstract

Juvenile xanthogranuloma (JXG) is a rare benign non-Langerhans cell histiocytosis that usually occurs in cutaneous lesions on the head, neck, or upper trunk of neonates and young children. Intramuscular JXG, which invades muscle tissue, accounts for only 0.6% of all JXGs and mostly occurs in the skeletal muscles of the extremities or trunk. A 5-month-old girl was referred to our hospital. At the age of 3 months, she presented with a slow-growing lump on her left thigh. Magnetic resonance imaging (MRI) showed a 22 × 19 × 18 mm oval mass in her left thigh. First, needle biopsy results suggested deep JXG or myeloid sarcoma. Therefore, marginal resection was performed. Intraoperatively, the tumor adhered to the left tensor fasciae latae muscle and was resected together. Histopathological examination revealed a diffuse monotonous sheet-like proliferation of mononuclear histiocyte-like cells with pale, eosinophilic, foamy cytoplasm with a background of muscle and fatty tissue. Minimal mitotic figures and no nuclear atypia or multinucleated giant cells were observed. Immunohistochemical analysis was positive for CD68 (KP-1) and CD163; weakly positive for lysozyme; and negative for CD1a, S100, myeloperoxidase, and CD34. No blast proliferation was observed in the bone marrow. The patient was diagnosed with deep JXG and scheduled for periodic physical examination and MRI. Despite positive margins, the patient fared well without local recurrence 48 months after tumor removal. Understanding the unique pathology of deep JXG and detailed histological evaluation are important for decision-making.

INTRODUCTION

Juvenile xanthogranuloma (JXG) is a benign non-Langerhans cell histiocytosis.¹ When the lesion is located in deep subcutaneous and intramuscular tissues, it is called deep JXG.² In particular, intramuscular JXG is very rare.¹ The characteristic pathologic findings of deep JXG may require differentiation from malignancy.³ We present a case of deep JXG localized subcutaneously in the left thigh and infiltrating the left tensor fasciae latae muscle, complicating the differential diagnosis between intramuscular JXG and myeloid sarcoma.

CASE REPORT

The patient was a healthy girl with no specific history other than left congenital cataract. There was no family history of note. At 3 months of age, the patient presented with a slow-growing lump on her left thigh. Computed tomography (CT) and magnetic resonance imaging (MRI) revealed a subcutaneous mass in the left thigh, and she was referred to our hospital at 5 months of age. Based on the result of needle biopsy, deep JXG or myeloid sarcoma was suspected, and the patient was admitted for surgical resection and systemic examination.

On physical examination, a 20 × 20 mm elastic firm and mobile mass on the left thigh was palpated. No other skin lesions, fever, lymphadenopathy, or other findings were observed. Blood count, biochemical, and coagulation parameters were all within normal limits, and no blasts were observed in a peripheral blood smear. MRI revealed a well-defined oval tumor in the left thigh, partly contiguous with the left tensor fasciae latae muscle (Fig. 1A–B). As the contrast effect was seen around the edge of the mass in the contrast-enhanced MRI, and the boundary was unclear, malignant lymphoma, rhabdomyosarcoma, and fibrosarcoma were listed as possible differential diagnoses. No metastatic lesions were observed.

Fig. 1.

Contrast-enhanced magnetic resonance imaging of a tumor. A is a contrast-enhanced T1 weighted image (CE-T1WI), and B is a T2 weighted image, respectively. The tumor (yellow oval) shows lower intensity in contrast-enhanced CE-T1WI and mildly higher intensity in T2WI compared with the muscle. Contrast-enhanced T1WI shows a heterogeneous contrast effect at the margins and inside the tumor, and part of the tumor is contiguous with the left tensor fasciae latae muscle (red arrow). The size of the tumor is 22 × 19 × 18 mm.

Needle biopsy was performed at 5 months of age, and histopathological examination revealed a dense proliferation of cells with oval nuclei and pale eosinophilic cytoplasm. Immunohistochemical analysis showed cells positive for CD68 and CD163; weakly positive for lysozyme; and negative for CD1a, S100, myeloperoxidase (MPO), and CD34 (Fig. 2A–C). Although a pathologically deep JXG was suggested, myeloid sarcoma was also considered due to the monotonous proliferation of cells positive for the myeloid cell surface marker CD68 but negative for MPO and CD34.

Fig. 2.

Gross image and histology of the mass. A, B, and C are histological images of a needle-biopsied tumor. D is a gross image and E and F are histological images of the resected tumor. (A) High-powered view of the tumor. Dense proliferation of cells with oval nuclei and pale eosinophilic cytoplasm with a background of abundant eosinophilic infiltrate (hematoxylin and eosin stain). Immunohistochemical analysis is positive for (B) CD68 and (C) CD163. (D) The tumor is 23 mm, yellow to ocherous, substantial, and covered with a thin membrane. (E) Diffuse monotonous sheet-like proliferation of mononuclear histiocyte-like cells with pale eosinophilic, foamy cytoplasm and a background of muscle and fatty tissue, some of which were infiltrative. There is eosinophil and lymphocyte infiltration, minimal mitotic figures (<1 per 10 high power field), and no nuclear atypia or multinucleated giant cells (hematoxylin and eosin stain). (F) High-powered view of the tumor (hematoxylin and eosin stain). Bars; (A–C, F) 20 μm, (E) 100 μm.

Therefore, marginal resection was performed in the same month. Intraoperatively, the tumor adhered to the left tensor fasciae latae muscle, partially within the left vastus lateralis muscle, and was resected together. Grossly, the tumor was 23 mm, yellow to ocherous, substantial, and covered by a thin membrane. Histopathological examination revealed a diffuse, monotonous, sheet-like proliferation of mononuclear histiocyte-like cells with pale eosinophilic, foamy cytoplasm and a background of muscle and fatty tissue, some of which were infiltrative. Eosinophil and lymphocyte infiltration, minimal mitotic figures (<1 per 10 high-power fields), and no nuclear atypia or multinucleated giant cells were observed (Fig. 2D–F). Immunohistochemical analysis showed cells positive for CD68 (KP-1) and CD163, weakly positive for lysozyme, and negative for CD1a, S100, MPO, and CD34, similar to the needle biopsy results. The Ki-67 labeling index of the excised tissue was less than 10%, and malignancy, including histiocytosarcoma, was unlikely. No blast proliferation was observed in the bone marrow, and bone marrow flow cytometry indicated normal patterns. Immunohistochemical analysis of bone marrow revealed no proliferation of MPO- or CD34-positive cells, and we ruled out the possibility of myeloid sarcoma. The patient was diagnosed with deep JXG and scheduled for periodic physical examination and MRI. Despite having positive margins, she fared well, without local recurrence 48 months after surgery.

DISCUSSION

JXG is an uncommon benign non-Langerhans cell histiocytosis that usually occurs in the head, neck, or upper trunk in neonates and young children. Lesions most frequently present as solitary cutaneous nodules.¹ Deep subcutaneous and intramuscular localization of this tumor is particularly rare and is called deep JXG.² Among these, intramuscular JXG, invading muscle tissue, accounts for only 0.6% of all JXG cases and mostly occurs in the skeletal muscles of the extremities or trunk.^1,2

Deep JXG and intramuscular JXG were first reported by Janney et al. in 1991.⁴ The prognosis is good for deep JXG, including intramuscular JXG, while systemic JXG with multiple lesions in the liver, central nervous system, and lungs is sometimes lethal.¹ Dehner et al. reported 26 cases of deep JXG, including intramuscular JXG, most of which underwent excision, and no local recurrences or JXG development elsewhere were reported in these children. Due to the self-limiting nature of deep JXG, once the pathological diagnosis of deep JXG has been established with confidence, re-excision of a positive marginal tumor will not be necessary unless clinical circumstances dictate otherwise.^1,3,4

Including the present case, 15 cases of intramuscular JXG have been reported (Table 1).^1–12 The median age was 4 months, and most cases occurred in infancy. However, two cases occurred in adulthood. According to these reports, there was no sexual bias. The paraspinal muscles were the most common sites of onset, followed by the upper extremities and muscles around the head and neck. Surgical resection was performed in 14 cases, and no recurrence was observed in any case. At least five of these described cases had positive margins; one of them had local recurrence after 5 weeks and another after 2 months, and they underwent re-excision. Although the re-excisions were incomplete and positive for margins, no subsequent recurrences were observed.^7,12 One case of intramuscular JXG of the paraspinal muscle resolved spontaneously after only surgical biopsy.⁶ Only one patient developed intramuscular JXG of the right forearm extensor muscle 4 years after the occurrence of cutaneous JXG in the nose, while the other patients had no complications of cutaneous or other organ involvement.

Table 1. Clinical features of previously reported cases of intramuscular JXG.

Case no.	First author, year	Age	Sex	Complications	Site	Size, cm	Other lesions	Treatment	positive margins	Outcome
1	Dehner et al., 2014	1 mo	F	NA	retroauricular skeletal muscle	NA	none	resection	NA	NED
2	Koren et al., 2010	28 y	F	none	biceps brachii muscle	2.5	none	resection	negative	NED, 1 y
3	Nascimento et al., 1997	6 mo	M	none	Paraspinal muscles	3.5	none	resection	NA	NA
4	Nascimento et al., 1997	4 mo	M	none	rectus abdminus muscle	1.2	none	resection	NA	NED, 2 mo
5	Janner et al., 1991	8 mo	F	NA	Paraspinal muscles	1.5	none	resection	positive	NED, 7 y
6	White et al., 1991	3 mo	M	NA	Paraspinal muscles	4.0	none	resection	NA	NED, 11 mo
7	White et al., 1991	2 mo	M	NA	Paraspinal muscles	2.0	none	resection	NA	NED, 2 y
8	de Graaf et al., 1992	3.5 y	M	NA	Paraspinal muscles	5.0	none	none (only surgical biopsy)	positive	Spontaneous resolution NED, 2 y
9	Favara et al., 1995	1 mo	M	none	Occipital muscles	5.0	none *1	exisional biopsy	positive	Local recurrence in 5 w
								re-resection	positive	NED
10	Yamanaka et al., 1995	4 mo	F	NA	Psoas muscle	4.0	none	resection	NA	NED, 1 y
11	David et al., 2003	2 mo	M	none	triceps muscle	3.0	none	resection	negative	NED
12	Margulis et al., 2003	5 mo	F	none	mentalis muscle	3.0	none	resection	negative	NED, 4 y
13	Kuo et al., 2005	52 y	F	NA	right forearm extensor muscle	3.8	none *2	resection	NA	NED, 37 mo
14	Barocca et al., 2007	0 day	F	none	sternocleidomastoid muscle	5.3	none	resection	positive	Local recurrence in 2 mo
								re-resection	NA	NED, 17 mo
15	Present case	3 mo	F	left congenital cataract	left tensor fasciae latae muscle	2.3	none	resection	positive	NED

NA, not available; NED, no evidence of disease; w, weeks; mo, months; y, years; M, male; F, female

*1, Two lesions in the same location *2, Skin lesion on the nose 4 years prior to onset of intramuscular lesion.

Our patient had deep JXG invading the left tensor fasciae latae muscle, which was treated with marginal resection. Although margins were positive, we followed up with the patient for local recurrence by physical examination and MRI since deep JXG is a self-limiting disease and has a good prognosis. The first post-operative MRI was performed at 6 months and showed no recurrence. Thereafter, periodic follow-up examinations were performed at gradually increasing intervals. Thirty-one months after surgery, she was still in good health with no apparent JXG recurrence.

Histologically, a typical cutaneous JXG consists of mononuclear cells, foamy macrophages, multinucleated Touton-type cells, and spindle-shaped fibroblasts. Among these cells, a few eosinophils, lymphocytes, and sometimes plasmocytes are scattered.^2,3 Numerous mitoses are uncommon in JXG. Deep JXG differs from the cutaneous form histologically in its tendency to consist of homogeneous proliferation of histiocytes without xanthomatous cells and no Touton- or Langhans-type multinucleated giant cells. The monotonous proliferation of histiocytes in association with the lack of multinucleated giant cells, occasional mitotic activity, and infiltrative appearance can raise concerns about malignancy, such as childhood soft-tissue sarcomas (deep malignant fibrous histiocytoma, cellular subcutaneous neurogenic tumors including neurofibroma, and cellular schwannoma), and myeloproliferative disorders, such as granulocytic sarcomas.^2,3 In particular, the absence of multinucleated giant cells and the rarity of deep JXG itself complicate diagnosis.³ Koren et al. noted the absence of “true” nuclear atypia to differentiate deep JXG from malignancy by cytology.² Malignant tumors with such malignant histiocytoma-like morphology are usually of higher grade, and thus show well-visible and well-interpretable nuclear atypia and pleomorphism.²

Immunohistochemical study of JXG is generally positive for vimentin, CD68, and factor XIII-a; negative for CD1a; and partially positive for S100; with no major difference between deep JXG and cutaneous lesions.² In particular, phosphoglucomutase 1 (PG-M1), an anti-CD68 antibody, shows a stronger and more uniform staining pattern than KP-1 and is considered a reliable marker.³ Awareness of the possibility of this atypical presentation of JXG helps in making the correct histologic diagnosis, which can be supported by using appropriate immunomarkers of histiocytes, mainly PG-M1.³

In this case, histopathological examination revealed a diffuse monotonous proliferation of mononuclear histiocyte-like cells with pale eosinophilic, foamy cytoplasm and a background of muscle and fatty tissue, without multinucleated giant cells, nuclear atypia, or mitotic figures. Immunohistochemical analysis was positive for CD68 (KP-1) and CD 163; weakly positive for lysozyme; and negative for CD1a, S100, MPO, and CD34. These findings are consistent with those described above, and the patient was diagnosed with deep JXG. The pathological diagnosis was consistent with that of the National Center for Child Health and Development. Initially, myeloid sarcoma was considered because of the monotonous proliferation of cells positive for the myeloid cell surface marker CD68. Indeed, flow cytometry analysis of the excised tumor was positive for CD34 and MPO. However, it did not detect any monoclonal populations. We therefore checked the pathological tissue again and confirmed that there was no proliferation of myeloblasts, as both MPO and CD34 immunostaining of the tumor cells were negative. Thus, myeloid sarcoma was ruled out.

In conclusion, we report a rare case of deep JXG that invaded the left tensor fasciae latae muscle. Deep JXG is a benign, self-limiting disease that can be treated with conservative surgery. The characteristic pathological findings of deep JXG require differentiation from a malignancy. Understanding the unique pathology of deep JXG and detailed histological evaluation are important for diagnosis and decision-making.