Atypical Lipomatous Masses of the Extremities

Outcome of Surgical Treatment

Abstract

Atypical lipomatous tumors occur predominantly in middle-aged patients and often present as painless, slow-growing masses in the extremities. The clinical outcomes of surgically treated superficial or deep atypical lipomas of the extremities were reviewed. Thirty-one patients were included in this study. There were 16 men and 15 women, with an average age of 57 years (range, 32–87 years). The mean followup was 7 years (range, 1–28.8 years). Twenty-five tumors occurred in the lower extremity and six in the upper extremity. Sixteen patients (52%) had a recurrence at an average of 4.7 years after resection. Twelve (39%) patients required additional surgical procedures to treat their tumor. Occurrence of a deep lesion and positive margins at the time of the initial surgery correlated closely with the rate of recurrence and need for additional surgical treatment. Dedifferentiation to high-grade liposarcoma developed in four (13%) patients. Atypical lipomas have a high propensity for local recurrence and a potential for malignant dedifferentiation. Patients with atypical lipomas require careful evaluation, complete surgical excision when possible, and close clinical followup extending beyond 5 years.

Atypical lipomatous tumors occur predominantly in middle-aged patients and often present as painless, slow growing masses in the extremities. Because of the site-dependent differences in behavior of these masses, the nomenclature of differentiated lipomatous tumors has been the subject of debate. Despite identical histologic features among all well-differentiated liposarcomas, those outside the retroperitoneum typically follow a more benign course. ⁴ The designation of atypical lipoma for previously termed well-differentiated liposarcomas outside the retroperitoneal space was introduced with the intention of portraying a better prognosis. ^5,6 Although this terminology has been embraced by some authors, others continue to classify retroperitoneal and extremity tumors as well-differentiated liposarcomas. ¹¹

On magnetic resonance imaging (MRI), atypical lipomas have a homogenous signal and are differentiated from simple lipomas by their size (atypical lipomas tend to be larger) and increased number of fibrous septations. ^2,3 Grossly, atypical lipomas are smooth, lobulated, encapsulated lesions. Histologically, they are characterized by mature appearing adipocytes of varying sizes (as seen in simple lipomas) and atypical cells with hyperchromatic nuclei and scattered lipoblasts. ^6,11 High-grade liposarcoma, however, is characterized by undifferentiated fusiform or stellate cells, small lipoblasts in a myxoid matrix, and primitive round cells with more hyperchromatic nuclei. ¹⁰

Atypical lipomatous masses arise in a wide variety of locations, including subcutaneous, intramuscular, and central body sites. The most common sites for superficial masses are the posterior neck and upper back, whereas deep lipomatous tumors are common in the thigh, arm, and buttocks. ⁶ The differences in nomenclature and clinical behavior have generated confusion about the best way to clinically treat these tumors. Although treatment usually is accomplished by complete excision, previous series have reported a high rate of recurrence and instances of dedifferentiation into high-grade liposarcoma. ^{1,4–6,8–11}

Several studies have attempted to better define the histologic features of atypical lipomatous tumors ^1,5,6,11 but no reports exist in the orthopaedic literature documenting the clinical course, identifying risk factors for recurrence, or examining treatment options for these common masses. The purpose of this study was to describe the history and clinical outcome of 31 patients with superficial or deep atypical lipomas of the extremities who were surgically treated.

MATERIALS AND METHODS

The records of 31 patients treated at the authors’ institution for atypical lipomas between 1976 and 1999 were reviewed. Criteria for inclusion were a diagnosis of superficial or deep atypical lipomatous tumor determined by pathologic examination, and a minimum followup of 1 year from the time of original diagnosis. Slides of pathologic specimens were reviewed by an experienced pathologist. The description of mature adipocytes of varying sizes, atypical cells with hyperchromatic nuclei, and lipoblasts was necessary to meet the definition of an atypical lipoma. Patients with histologic dedifferentiation at original presentation, defined as zones of increased atypia, mitotic figures, or high-grade sarcoma within the lipomatous mass, were excluded from the study. The patients’ age at presentation, gender, anatomic site, presenting symptom, and treatment methods were identified.

The tumor originally was diagnosed by clinical examination, MRI, or computed tomography (CT). Patients were followed up with biannual serial examinations and, in patients with residual masses or recurrences, with yearly MRI. Two patients moved during the study and were not available for the most recent followup. In these two patients, information was obtained through telephone interviews and referring physicians.

There were 15 women and 16 men, ranging in age from 32 to 87 years (average, 57 years). Average duration of followup was 7.1 years (range, 1–28.8 years). All 31 patients were treated operatively by the senior author (RDL). Nine patients had an initial marginal excision at another institution and were referred to the senior author after tumor recurrence.

Most commonly, the tumor presented as a solitary, asymptomatic mass. Twelve patients, however, had pain and one patient with a pelvic tumor adherent to the sciatic nerve had paresthesias. In addition, one patient with a large intrapelvic tumor was referred after detection of a mass during a rectal examination. The duration of symptoms before presentation ranged from 1 month to 7 years (mean, 8.5 months). All lesions were associated with a history of gradual increase in size.

Twenty-three patients had an MRI before surgery (Fig 1). In all cases, MRI revealed a heterogeneous fatty mass. Tumors had high signals on T1-weighted images, with hypointense septumlike areas on T1-and T2-weighted images. Of the 23 masses seen on MRI scans, 17 were deep and five were subcutaneous. Two patients with masses deep in the pelvic cavity and buttock region, and one patient with a shoulder mass, had a CT scan before their surgical procedure (Fig 2). Two patients who were referred from an outside institution each had their diagnosis confirmed by needle biopsy. The remainder of the patients had their diagnoses confirmed at the time of surgery by excisional biopsy.

Fig 1.: A T1-weighted MRI scan shows an atypical lipoma arising in the buttock (Patient 22). The mass consists of a largely fatty matrix with fibrous septae of low signal intensity.

Fig 2.: An atypical lipoma of the thigh can be seen on this CT scan. Numerous fibrous septae are evident (Patient 19).

Twenty-six tumors occurred in the lower extremity and six in the upper extremity. The most common location was the thigh (n = 19), followed by the buttock (n = 4), shoulder (n = 4), and knee (n = 2). One tumor each occurred in the calf, the forearm, and the hand.

Statistical analysis was done using chi square analysis for nonparametric data (JMP statistics program, SAS Corporation, Cary, NC).

RESULTS

The initial treatment consisted of surgery in all 31 patients. Two patients (6%) with masses in close proximity to neurovascular structures were treated with partial excision and the remainder with marginal excision. Because atypical lipomas are usually low-grade malignancies, wide resections, with high associated morbidity, were not done at the time of initial surgery. At the first operative procedure, 22 patients had documented positive margins and nine had negative margins. The nine patients who initially were treated at an outside institution all had positive margins at the time of the first resection.

The atypical lipomatous tumors were relatively large, ranging from 4 × 2 × 1.4 cm to 40 × 22 × 9 cm. Grossly, most tumors were yellow-to-tan, firm, rubbery masses composed of fat lobules separated by fibrous septae. Seven tumors were entirely encapsulated. The microscopic pathologic features showed atypical lipomatous masses without dedifferentiation in all cases. The specimens primarily were made of adult fat with enlarged, hyperchromatic atypical cells, often located in areas of loosely arranged fibrous tissue (Fig 3).

Fig 3.: A photomicrograph of an atypical lipoma of the thigh shows narrow bands of loose fibrous tissue separating variably sized adipocytes and occasional lipoblasts (Stain, hematoxylin and eosin; magnification, ×150).

Sixteen patients (52%) had a recurrence at an average of 4.7 years after the initial resection (range, 2.3–10 years). Two of these patients (including the patient with recurrence at 2.3 months) were first treated with partial excision because of the location of the tumor and close proximity of neurovascular structures. All patients with recurrences had enlarging masses with multiseptated fatty lesions seen on MRI scans obtained at followup. Eleven recurrences occurred in the thigh or buttock, one in the knee, one in the forearm, and one in the shoulder. All but one recurrence occurred in patients with deep masses (94%). Of the 16 patients with recurrences, 15 (94%) had positive margins and one patient (6%) had negative margins at the time of original surgery (p < 0.05). Size and age at presentation were not statistically significant predictors of recurrence (p > 0.05).

Twelve (39%) patients had a second resection at an average of 5.7 years after initial treatment (range, 2.7 months–277 years). Conservative treatment was continued in one patient with a mass adherent to the sciatic nerve and three patients declined additional intervention secondary to poor general health. At surgery, 10 patients had recurrent atypical lipomas with varying degrees of fibrosis and atypia. The pathologic examination in the remaining two patients revealed foci of dedifferentiation to high-grade liposarcoma.

Seven patients (23%) required additional surgery for a second recurrence at an average of 4.5 years after the first recurrence. Six patients had three or more recurrences: one patient with a slow-growing mass was followed up with close observation and four patients had at least one additional excision. All patients who had additional surgical treatment had positive margins at the time of their previous procedure.

In addition to the two previously mentioned patients, dedifferentiation to high-grade liposarcoma developed in two patients at the time of the fourth recurrence (Fig 4). Dedifferentiation was diagnosed when zones of increased atypia or mitotic figures were seen within the previously diagnosed atypical lipomatous mass (Fig 5). In all patients, dedifferentiation was diagnosed in a small focal area within the lipomatous mass. The average time from presentation to dedifferentiation for all four patients was 10.4 years (4.9, 5.5, 6.6, and 24.6 years). No patient had evidence of tumor metastases at the latest followup (an average of 3 years after dedifferentiation).

Fig 4A–B.: (A) T1 and (B) T2-weighted images of a recurrent atypical lipoma in the thigh (Patient 2) show a superior focus suspicious for dedifferentiation (white arrow). The dense area at the inferior pole of the lesion is an old hematoma.

Fig 5.: The dedifferentiated component of the lesion from Figure 4 shows a pattern of pleomorphic, hyperchromatic spindle cells (Stain, hematoxylin and eosin; magnification, ×150).

Patients with small areas of dedifferentiation within the atypical lipomatous mass were treated with wide resection and radiation therapy. Two patients also chose to receive adjunctive chemotherapy. All patients were followed up with MRI scans of the affected area and CT scans of the chest. Three patients with dedifferentiation were free of tumor at the latest followup. Postoperative complications included lower extremity phlebitis in one patient and a superficial wound hematoma in another.

At latest followup all but eight patients were free of tumor. No amputations were done, none of the tumors metastasized, and no deaths secondary to the tumor occurred. The findings are shown in Table 1.

TABLE 1: Data on Patients With Atypical Lipomas of the Extremities

TABLE 1: Continued.

DISCUSSION

Lipomatous masses are among the most common soft tissue tumors in adults. Patients usually present with a painless, slowly enlarging mass, which can be accurately delineated with CT or MRI. When a high fat content is seen in a tumor, the differential diagnosis includes lipoma, atypical lipomas, and liposarcoma. Lipomas usually are recognized because of their uniform fat density (usually over 95%) (Fig 6), whereas intermediate-to high-grade liposarcomas and myxoid liposarcomas are characterized by the absence of fat density on MRI. ³

Fig 6.: A T1-weighted MRI scan shows a simple lipoma of the anterior thigh (black arrow). No fibrous septations such as those that characterize atypical lipomatous lesions are seen.

Preoperative diagnosis of atypical lipomas, however, is often problematic, as these lesions are difficult to distinguish from low-grade liposarcoma. Computed tomography and MRI of atypical lipomas reveal a largely fatty matrix with areas of low signal intensity similar to those that characterize soft tissue sarcomas. ^2,3 Hosono et al ⁷ showed that on MRI scans, lipomas often have thin areas of linear enhancement, whereas thicker septae characterize atypical lipomas and liposarcomas. A review of MRI scans of atypical lipomas and low-to intermediate-grade liposarcomas showed the fat content to be less than 75% in all cases, making the separation between these entities difficult. ³ The distinction between atypical lipoma and low-to intermediate-grade liposarcoma cannot be made with imaging alone, making a diagnosis by histologic examination necessary. Unless directed at a specific area of suspicion, needle biopsy is often unreliable in low-grade fatty tumors, and surgical excision of the entire mass is the preferred treatment.

Atypical lipomatous masses have long been recognized as tumors of low-grade malignancy. Studies have suggested varying rates of recurrence for subcutaneous lesions. Although none of the subcutaneous tumors reported by Evans et al ⁶ recurred, Allen et al ¹ documented a 20% recurrence rate and Kindblom et al ⁸ reported a 50% recurrence rate among atypical superficial masses. The current study revealed one recurrence among five superficial masses, equivalent to a 20% rate of recurrence for that location.

Results for deep lesions have been more consistent. Several series have shown that atypical lipomas in the deep soft tissues of the extremities and groin often recur but rarely metastasize. Evans ⁵ reported nine recurrences among 13 deep lesions (69%) and, in a review of well-differentiated liposarcomas of the extremities and retroperitoneum, Weiss and Rao ¹¹ found that 20 of 46 extremity tumors (43%) recurred. The results of the current study with 52% of tumors recurring and 39% of patients having additional operations, confirm that deep atypical lipomas have a high likelihood of recurrence. In addition, 94% of recurrences in the current series occurred in patients with deep masses. Although late first recurrences have been reported, ^1,5,9 most recurrences in this patient group occurred relatively early in the disease process, at an average of 4.7 years after the original diagnosis.

The type of surgical excision has a marked effect on the rate of recurrence. Lucas et al ⁹ reported a rate of recurrence of 11% in patients treated with a wide excision, but a 62% rate of recurrence in patients treated with marginal excision. In the current study, all but two masses were first treated with marginal excision. Of the 16 patients with recurrences in the current study, 15 (94%) had positive margins at the time of original surgery and only one recurrence was observed in a patient with documented negative surgical margins. Complete excision of the original lesion is thus a critical factor in preventing tumor recurrence.

Although size and age at presentation were not statistically significant predictors of recurrence, occurrence of a deep lesion and positive margins at the time of the initial excision were risk factors for recurrence. Although atypical lipomatous masses commonly are treated by excision, the surgeon must be aware of the high risk of recurrence that accompanies positive margins. When the size and location of the tumor allow this, negative margins significantly lessen the rate of recurrence and complete excision becomes the treatment of choice. Because most recurrences occur in the first 6 years after diagnosis, patients should be followed up biannually during this time and subsequently on an annual basis. Patients also should be instructed to watch for tumor recurrence and any return of symptoms.

In cases of recurrence, MRI is the preferred imaging method to define the lesion’s boundaries and monitor for dedifferentiation to high-grade sarcoma. When a recurrent mass exists, the surgeon has two choices: observation or repeat excision. Masses manifesting slow growth, few symptoms, and similar appearance on MRI scans to the initial tumor can be observed safely on a yearly basis. For more aggressive lesions, complete excision of the recurrence with negative margins should be done to minimize the risk of additional recurrence and to examine for the presence of dedifferentiated areas.

Dedifferentiation identifies tumors with more aggressive qualities in their potential to recur locally and to metastasize to distant sites. Although one report showed that dedifferentiation rarely occurred in lesions of the extremities, ⁴ another study showed that as much as 29% of groin and 6% of extremity lesions can have malignant dedifferentiation. ¹¹ Weiss and Rao ¹¹ defined dedifferentiation as a timedependent phenomenon. In their series, malignant dedifferentiation occurred in three of 46 lesions (7%) at an average of 9 years after the initial diagnosis. Patients in the current study followed a similar course. Four patients (13%), all with long-term followup, had areas of dedifferentiation to a high-grade sarcoma at an average of 10.4 years after diagnosis. Two of these patients had areas of dedifferentiation after multiple recurrences. The earliest case of dedifferentiation in this series occurred after 4.9 years, suggesting that greater than 5 years followup is necessary to monitor for the development of dedifferentiation.

Although metastatic lesions arising from dedifferentiated lipomatous masses have been described, ¹¹ none of the patients in the current study with areas of dedifferentiation had metastatic disease at latest followup. The potential for metastatic disease, as for most soft tissue sarcomas, is likely related to the amount of dedifferentiation present, and to the nature of adjunctive therapy administered. Masses with dedifferentiation mandate a wide resection, adjunctive therapy with radiation, and close monitoring for recurrence and metastatic disease. The current authors routinely observe these patients with biannual MRI scans of the affected area and CT scans of the chest.

The current results show that atypical lipomatous masses have a high propensity for local recurrence and a potential for malignant dedifferentiation. Patients with deep masses and those with positive surgical margins have a higher risk of recurrence and are more likely to require additional surgery. Patients with atypical lipomas require careful evaluation, complete surgical excision when possible, and close clinical followup extending beyond 5 years.

Acknowledgments

The authors thank Pedro K. Beredjiklian, MD, for his contribution.