Marshall-Taylor and Fanburg-Smith [9] found HFLL/HFT in 10 (0.2%) of the benign lipomatous lesions in the soft tissue registry of the Armed Forces Institute of Pathology. Including our report, there have been 31 documented cases of HFLL/HFT [1, 2, 5, 7–11]. However, the clinicopathologic characteristics of these uncommon and unusual lesions are variably reported.

Physical examination revealed a soft, mobile, and tender mass on the dorsal aspect of the right foot. There was no associated edema, erythema, or increased warmth. Strength, ROM, neurologic examination, and vascular examination were unremarkable.

MRI revealed a 7.0- × 5.0- × 2.0-cm mass in the subcutaneous tissue of the dorsal right foot (Fig. 1) with ill-defined margins. Predominant signal intensity followed fat in all sequences: bright on T1 (Fig. 1A) and intermediate on fat-suppressed T2 (Fig. 1B–C). Internal stranding/septations of low signal intensity were seen. There was no enhancement with contrast. The differential diagnosis at this time included simple lipoma, lipoma variant, and atypical lipoma/well-differentiated liposarcoma.

Given the clinical history and radiographic findings, a benign process was strongly suspected. Surgical removal of the mass therefore was recommended and subsequently performed. Intraoperatively, because the lesion lacked a pseudocapsule and was adherent to the overlying skin, an intralesional excision was required instead of a marginal excision. Small portions of the remaining lesion required piecemeal removal with a rongeur. Grossly, it was a fatty tumor with areas of hemosiderin deposition.

Histologic analysis revealed mature fat and spindle cells, with strands of dissecting fibrous tissue, and abundant macrophage-engulfed hemosiderin (Fig. 2A). Scattered chronic inflammatory cells and rare giant cells were present. The spindle cells were consistent with fibroblasts or fibrohistiocytic cells; no atypia was identified (Fig. 2B). We interpreted these findings as diagnostic for HFLL/HFT.

Marshall-Taylor and Fanburg-Smith [9] reported a history of acute or repetitive local trauma in seven of eight patients in their seminal report. This history, their usual superficial location, and several histologic features, including lack of encapsulation, homogeneously sized adipocytes, and inflammatory infiltrates, prompted them to postulate these lesions are attributable to reactive processes. Such an etiology is possible in our patient given that she had HFLL/HFT develop at a prior surgical site. Kazakov et al. [7] similarly supported this reactive hypothesis by proposing a clinical correlation with tissue injury from venous stasis. Newer studies have not identified a history of trauma or vascular disease as frequently [1, 2, 10], but our review revealed 13 of 28 (46%) reported patients had a history of previous trauma or substantial vasculopathy [1, 2, 5, 7–11]. The consistency of the lesion’s histologic appearance, immunohistochemistry, and cytogenetic analyses revealing clonal translocations between chromosomes 1 and 10 (t(1;10)(p31;q25) and t(1;10)(p22;q24)), also prompted some authors to suggest HFLL/HFT is neoplastic in origin [1, 3, 5, 11].

Our review suggests HFLL/HFT predominately occurs during the fifth and sixth decades of life (mean age, 49.5 years; range, 0.67–74 years) [1, 2, 5, 7–11]. There is a strong female predilection of 22:9 [1, 2, 5, 7–11]. The lesion typically presents as a painful, slow-growing mass in subcutaneous tissue [1, 2, 5, 7–11]. It most frequently occurs in the foot/ankle (84%). Other sites include the hand, calf, thigh, and cheek [1, 2, 5, 7–11]. The average lesion size is 7.7 cm (range, 0.1–19 cm) [1, 2, 5, 7–11].

Our case is in agreement with the few previous reports of HFLL/HFT imaging studies describing fatty tumors with ill-defined borders. West et al. [10] and de Vreeze et al. [2] presented cases that were isointense to fat on T1- and T2-weighted sequences. These authors also noted scattered reticular structures of intermediate to low signal intensity, presumably correlating to the fibrous components of the lesion. These basic features also were seen in our case but are not diagnostic of one specific tumor. Given the pathologic presence of hemosiderin deposition in these lesions, a T2* or other susceptibility-weighted imaging sequence is expected to reveal scattered dot-like dark deposits. This feature could be helpful in radiographically distinguishing HFLL/HFT from other fatty lesions. However, de Vreeze et al. [2] did not recognize such a finding in their dynamic gradient echo sequences and no other T2* or susceptibility-weighted imaging has been published for HFLL/HFT. MRI appearance after contrast is unclear, as one reported case was enhanced whereas ours was not [2]. West et al. [10] also reported radiographs of HFLL/HFT with heterogeneous thickening of subcutaneous fat and dystrophic calcification in the lesion area.

Based on imaging alone, the primary differential diagnosis for HFLL/HFT includes simple lipomas, lipoma variants, and atypical lipomas/well-differentiated liposarcomas. All of these lesions are grossly fatty masses with significant overlap in their range of appearances [2, 4]. Simple lipomas typically have a characteristic appearance on MR images: a discrete, encapsulated, homogeneous fatty mass. However, some simple lipomas can contain muscle fibers, blood vessels, fibrous septa, and areas of necrosis or inflammation that can mimic the findings associated with lipoma variants, atypical lipomas/well-differentiated liposarcomas, and HFLL/HFT [4]. Imaging features characteristic of atypical lipomas/well-differentiated liposarcomas include: a fatty mass with thickened or nodular internal septa, nonadipose components, foci of high T2 signal, and areas of enhancement [2, 4, 6]. Lipoma variants have an even wider range of appearances, reflecting their pathologic diversity, but often will include thin internal septa and other nonadipose components [4]. High-grade liposarcomas can be excluded from the differential diagnosis of grossly fatty lesions as they generally contain only minimal amounts of macroscopic fat [4].

Grossly, HFLL/HFT is described as dark yellow or yellow-brown fatty tissue [1, 2, 5, 7–10]. Several reports also note areas of macroscopic hemorrhage [1, 7]. Histologically, the lesions are composed of three main elements in varying proportions: mature adipocytes, spindle cells, and hemosiderin pigment [9]. The predominant component in most specimens is mature adipocytes, in correlation with the MRI and gross appearance of most HFLLs/HFTs [9]. The adipocytes are arranged in lobules, with fibrous septae composed of spindle cells and inflammatory cells percolating through in a honeycomb-like pattern [9]. The fibrous septae also can be seen surrounding blood vessels and individual adipocytes [9]. The spindle cells are described as plump, slightly pleomorphic, and fibrohistiocytic [9]. Prominent hemosiderin deposition is present in all lesions, mainly in macrophages but also in the cytoplasm and extracellular matrix of some spindle cells [9].

HFLL/HFT can be confused histologically for several other processes. The literature provides cases of HFLL/HFT mistakenly diagnosed as spindle-cell lipomas, well-differentiated liposarcomas (WDLS), and other lipomatous tumors [1, 9]. Unlike the heterogeneous adipocytes of most lipomatous neoplasms though, the adipocytes of HFLL/HFT are homogeneously sized [9]. HFLL/HFT also lacks the prominent myxoid stroma seen in spindle-cell lipoma, and the lipoblasts and scattered pleomorphic hyperchromatic stromal cells seen in WDLS [1, 9]. Other lesions that HFLL/HFT can be misdiagnosed as include fibromatosis, fibrous histiocytoma, nodular fasciitis, dermatofibrosarcoma protuberans, and pleomorphic hyalinizing angiectatic tumor (PHAT) [1, 5, 9]. The presence and prominence of adipocytes in HFLL/HFT distinguish it from these fibroblastic and fibrohistiocytic lesions [9]. However, some authors still propose HFLL/HFT is an early form of PHAT [3].

Although the paucity of HFLL/HFT cases makes it difficult to establish consensus, the prognosis for these lesions generally is considered good owing to their slow growth and benign histologic appearance [9]. The lack of metastases in any of the reported patients further supports this benign presumption. Of the 27 lesions reported in the literature with followup information available, we found 10 (37%) cases of local recurrence and two (7%) cases of residual disease [1, 5, 7–10]. Recurrences, on average, occurred 28 months (range, 11–60 months) after treatment. The two cases of residual disease appear to have been attributable to partial excisions at the time of treatment.

Of the cases with treatment details and followup information available, four of 15 (27%) lesions treated by marginal/intralesional excision recurred [1, 9]. In comparison, none of three lesions treated by wide excision recurred, although their average followup was only 19 months (range, 9–36 months) [1]. This high recurrence rate may be related to the difficulty in determining intraoperatively that a resection is complete, secondary to the lack of anatomic tumor boundaries such as a pseudocapsule as occurred in our case. Considering the lack of morbidity from recurrent or residual disease and the lack of any known metastasis, we do not believe wide excisions or other aggressive treatments need to be pursued for HFLL/HFT, especially at the risk of major surgical complications. Instead, patients should be educated about the risk of recurrence and consideration given to yearly clinical followup.