The hemosiderotic fibrohistiocytic lipomatous lesion (HFLL) is first described by Marshall-Taylor in 2000 [2]. The incidence of HFLL is estimated to be less than 0.2% of all benign lipomatous lesions [2].

There is an ongoing debate about the resemblance of early pleomorphic hyalinizing angiectatic tumor (PHAT) and HFLL; some consider HFLL a precursor lesion of PHAT, implicating HFLL to be a neoplastic lesion [3, 4], others consider HFLL an individual more reactive lesion [5, 6].

Based on the cases described so far, HFLL is most common in middle aged females, however there is a wide age spectrum. HFLL is typically located on the distal extremities, particularly on the dorsal side of the foot and may be associated with venous stasis and trauma [3]. The median size at clinical presentation is 50 mm and ranges between 1 and 170 mm [2, 5]. Surgical removal is mainstay treatment for this lesion.

Local recurrences appear in approximately 50% of cases and become apparent within one year [2, 5]. Distant metastases have not been reported. Characteristic histopathological features are the spindled cells morphology and the presence of variably prominent hemosiderin pigment. The most common immunoprofile is diffuse staining of the spindled cell with CD34.

MR imaging, by Philips 3T Achieva and intravenous contrast series with Dotarem, showed a lipomatous lesion of the left thigh measuring 19 × 8 × 4 cm with irregular boundaries. The lesion showed multiple far reaching intramuscular and subfascial extensions. The assessment of internal structures showed a homogeneous, lobulated lesion. Figure 1 illustrates the high signal intensity of the lesion on T1- and T2 (STIR) weighted images with foci of hyperintensity on the fat-saturated (STIR) images. The signal intensity, particularly on T1 weighted images, was substantially lower than that of surrounding subcutaneous lipomatous tissue. Dynamic MR imaging was performed to characterize the enhancement pattern of the tumor, which showed homogeneous enhancement.

These combined imaging features were suggestive for a benign lesion or low grade sarcoma. However, we could not unequivocally define these MR images to a specific diagnosis. As intermediate or high grade sarcoma could not be ruled out, and these lesions in our institute are preferably treated by preoperative radiotherapy, a trucut biopsy was performed. A thoracic computed tomography scan was made which did not show distant metastasis.

Histopathological analysis did not allow a definitive diagnosis and suggested a not otherwise classifiable benign or low-grade lipomatous lesion; an intermediate or high-grade liposarcoma was unlikely.

Based on these findings, a surgical resection was planned.

Immunohistochemical analysis showed a notable staining primarily of the spindled cells and multinucleated cells for CD34. S-100 was positive in mature fat cells. CD31, PPARγ, and P53 were negative.

This case, unlike most described HFLL lesions, lacked a history of prior trauma, presented at an unusual location and was large. However, the macroscopic aspect, morphological and immunohistochemical features were consistent with HFLL. The specific histological HFLL features could be correlated to the MR characteristics. The fatty component of the tumor was macroscopically darker brown than normal fat tissue and correlated with lower signal intensity of the HFLL compared to the surrounding subcutaneous fat. The absence of a fibrous capsule could be recognized with MR findings as well as the wide intramuscular extension. The internal tumor structures seen on MR imaging were a reflection of the characteristic fibrous bands in the overall histological architecture. The spindled cell CD34 positive areas were corresponding to the enhanced parts of the tumor on the MR images. The abundant homogenous adipose areas were reflecting on MR imaging in the lobulated areas. The scarce specific iron deposition seen at histopathological review however was not recognized on MR images or dynamic gradient echo sequence, since the signal intensity on MR was not decreased by iron deposition of HFLL. Furthermore, the typical inflammatory parts of HFLL could not be correlated with characteristic MR features.