Liposarcoma is a nonrhabdomyosarcoma soft tissue sarcoma (NRSTS) arising from primitive mesenchymal cells which undergo lipomatous differentiation. While liposarcomas comprise 20% to 30% of adult NRSTS tumors and account for approximately 2000 new diagnoses in the United States annually, they are much less common in the pediatric population, comprising fewer than 3% of reported cases of NRSTS [
With Institutional Review Board approval, a retrospective review was conducted to identify patients ≤22 years of age with the pathological diagnosis of liposarcoma who were treated from February 1960 to May 2011. All pathology was re-reviewed by a single pathologist (CRA) with expertise in soft tissue sarcomas. Histologic subtypes were classified according to the World Health Organization classification of tumors of soft tissue and bone. Strict criteria were used to define the histologic subtypes of liposarcomas as well differentiated, dedifferentiated, myxoid/round cell, or pleomorphic, in accordance with previously published recommendations [
Patient medical records were surveyed for sex, age, race, and the nature and duration of presenting symptoms. The primary tumor was classified as central if its location was in the head/neck, thorax (including chest wall), or abdomen/pelvic region. Tumors of either the upper or lower extremities, including both the inguinal and gluteal regions, were considered peripheral. Surgical resection, tumor size, margin status, and forms of adjuvant therapy were also reviewed. Tumor depth was defined as superficial or deep based upon the tumor’s position relative to the investing fascia of the extremity, chest, or abdominal wall. All intra-abdominal or intrathoracic tumors were considered deep. The overall stage of disease was determined based upon the tumor-node-metastasis (TNM) classification and was consistent with the American Joint Committee on Cancer (AJCC) staging system for soft tissue sarcoma [
Statistical analyses were performed using SPSS software version 19.0 for Windows (SPSS Inc., Chicago, IL). Descriptive statistics were used for all patient characteristics, and overall survival (OS) and progression-free survival (PFS) were evaluated using the Kaplan-Meier method. OS calculations were based on the time from the date of diagnosis to either the date of confirmed death or the date of last recorded contact with the patient. PFS was calculated based on the time from diagnosis to the date of disease progression or relapse. The log-rank test was implemented to compare survival between groups, and the chi-square test was used to compare categorical data. Statistical significance was defined as
Thirty-four patients with a pathological diagnosis of liposarcoma were identified. Nineteen patients were male (56%) and 15 were female (44%). The median age at diagnosis was 18.1 years (range, 0.3–22.8 years). The age distribution was as follows: 1 patient aged 0–5 years, 1 patient aged 6–10 years, 7 patients ages 11–15 years, 21 patients ages 16–20 years, and 4 patients ages 21-22 years. Patients presented after a mean of 3 months (range, 0–60 months) of symptoms. Patients most commonly presented with an enlarging mass (
A majority of patients (
Twenty-nine patients (85%) had low-grade tumors, among which myxoid histology was the most common subtype (25 of 29; 86%). The remaining low-grade tumors were well differentiated (
All patients underwent surgical resection of the primary tumor (Table
Treatment and outcomes for 3 different histologic tumor subtypes.
Myxoid ( |
Well differentiated ( |
Pleomorphic ( |
|
|
---|---|---|---|---|
Treatment | ||||
Surgery | 16 | 2 | 1 | .08 |
Surgery + XRT | 6 | 1 | 0 | |
Surgery + XRT + CT | 2 | 1 | 3 | |
Surgery + CT | 1 | 0 | 1 | |
| ||||
Median PFS (years) | 0.3 | 3.1 | 0.5 | .34 |
| ||||
Mean radiation dose, Gy (range) | 55 (42–67) | 35 (24–45) | 64 (50–84) | .38 |
Median followup (years) | 7.3 | 5.4 | 1.9 | .01 |
Number of recurrences | 7 | 1 | 3 | .31 |
| ||||
Median followup (years) for patients with recurrence | 8.38 | 00 | 2.39 | <.001 |
XRT: radiation therapy; CT: chemotherapy; PFS: progression-free survival.
Surgical margins after the first definitive procedure were positive in 12 (35%) patients and negative in 22 (65%). Four (33%) patients with positive margins, all with central tumors, had gross residual disease (R2 resection) at the completion of surgery. Overall, 75% (9/12) of patients with positive margins had central tumors.
The median followup for the entire cohort was 5.4 years (range, 0.3–30.3 years). There was no difference in survival between male and female patients. Tumor grade and histology correlated most significantly with survival. Patients with high-grade pleomorphic tumors had a median survival of 1.9 years (range, 0.3–30.3 years) versus 7.3 years (range, 1.5–28.6 years) in those with low-grade myxoid/well-differentiated tumors (
Univariate analysis evaluating 5-year survival.
|
Number of events | 5-Year survival | 95% CI |
|
|
---|---|---|---|---|---|
Overall survival | 34 (100%) | 8 | 78% | 63–93 | |
| |||||
Male | 19 (56%) | 4 | 76% | 55.2–96.8 | .71 |
Female | 15 (44%) | 4 | 79% | 57–100 | |
| |||||
High grade | 5 (15%) | 3 | 25% | 0–68.4 | .003 |
Low grade | 29 (85%) | 5 | 85% | 71.4–98.6 | |
| |||||
Well differentiated | 4 (11%) | 1 | 100% | 0 | .013 |
Myxoid | 25 (74%) | 4 | 83% | 67.2–98.8 | |
Pleomorphic | 5 (15%) | 3 | 25% | 0–68.4 | |
Dedifferentiated | 0 | 0% | |||
| |||||
Tumor ≤ 5 cm | 14 (41%) | 2 | 84% | 63.2–100 | .312 |
Tumor > 5 cm | 20 (59%) | 6 | 74% | 53.6–94.4 | |
| |||||
(+) Margins | 12 (35%) | 7 | 50% | 21.2–78.8 | .001 |
(−) Margins | 22 (65%) | 1 | 95% | 85.8–100 | |
| |||||
Peripheral | 22 (65%) | 0 | 100% | 0 | <.001 |
Central | 12 (35%) | 8 | 42% | 13.6–70.4 | |
| |||||
Superficial | 14 (41%) | 1 | 92% | 79.2–100 | .103 |
Deep | 20 (59%) | 7 | 69% | 47.6–90.4 | |
| |||||
Stage 1A | 11 (32%) | 1 | 91% | 73.5–100 | <.001 |
Stage 1B | 16 (47%) | 2 | 86% | 67.3–100 | |
Stage 2A | 2 (6%) | 0 | 100% | 0 | |
Stage 2B | 0 | 0 | 0% | 0 | |
Stage 3 | 3 (9%) | 3 | 33% | 0–87.7 | |
Stage 4 | 2 (6%) | 2 | NR | 0 |
Overall survival for different subtypes of liposarcoma.
Overall survival and tumor grade.
Overall survival and margin status.
There were 8 (24%) deaths in this series, all from progression of disease, and all in patients with centrally located lesions. In contrast, all patients with peripheral lesions are alive and well. The Kaplan-Meier analysis demonstrated the significant correlation of central lesion location with poorer survival (
Overall survival for peripheral versus central primary tumor location.
Two patients with high-grade, centrally located tumors presented with metastatic disease in the lung and/or bone. Three additional patients, all with central tumors, developed lung and/or bone metastases during the course of followup. Of the 5 patients with metastatic disease, 4 have died due to disease progression. The fifth was a foreign national lost to followup, but is presumed to have died of disease.
Nonrhabdomyosarcoma soft tissue sarcomas, though rare, comprise a heterogeneous group of tumors that respond differently to chemo- and radiotherapy; have vastly different local, locoregional, and distant metastatic potential; and are associated with disparate overall survival rates. Their rarity and heterogeneity have made it difficult to gather prospective data on sufficient numbers of patients to be able to predict disease progression and to generate optimized paradigms of care. Because liposarcoma is among the most common types of NRSTS in adults, several large retrospective series analyzed during the past decade (which in aggregate account for more than 1500 cases) have provided a foundation for understanding the presentation and progression of adult liposarcoma and its treatment response [
In contradistinction to NRSTS in adults, liposarcoma in the pediatric population is very rare. Accordingly, there are very few published studies of any size that attempt to correlate pediatric disease presentation and tumor biology with outcome or to differentiate the presentation and progression in children from that of adults. Moreover, the largest of these studies have been multi-institutional and have spanned the course of decades, further complicating the synthesis and analysis of data. We present here the largest series of patients with pediatric liposarcoma. Our results suggest that the presentation of pediatric liposarcoma is different from its adult counterpart and that the two most significant factors that affect overall survival are the location of the primary tumor and the completeness of surgical excision. As observed in other studies, we found that pediatric patients are more likely to present with histologically low-grade tumors (29 of 34; 85%) than adults and, within this category, are more likely to present with myxoid histology (25 of 29; 86%) than adults, in whom well-differentiated histology predominates [
Surgical excision was the mainstay of treatment for all patients in this cohort. Only two cases of recurrent disease were associated with primary tumors in the periphery, both in the lower extremity and both with initially negative margins, a rate of local recurrence (2 of 22; 9%) similar to that noted in adult studies of peripheral tumors with complete excision [
Central tumors have a much poorer prognosis than peripheral lesions, with a 5-year survival of only 43%. This poorer prognosis is likely due to a combination of later presentation and more difficulty in achieving a complete resection. Central lesions may also have more aggressive tumor biology, as observed in this study, in which 4 of the 5 pleomorphic tumors were centrally located. Three of 4 patients with central pleomorphic tumors died of disease; 2 of these experienced disease recurrence less than a year after primary resection. Among the 8 patients who died of disease, all had centrally located tumors, and all but one had positive surgical margins after the first definitive resection. Only 2 patients with centrally located tumors did not receive adjuvant therapy. Of these, one had a high-grade pleomorphic tumor of the chest wall/shoulder girdle with negative margins after surgery. The other patient had a low-grade tumor in the retroperitoneum. Both are alive with no evidence of disease.
The role of chemotherapy is poorly defined in pediatric patients with LS. In our series, chemotherapy was administered primarily to patients with central tumors. The only patient with a peripheral lesion who received chemotherapy had a high-grade pleomorphic tumor of the right buttock that exhibited rapid enlargement during the 2 months prior to excision. Diagnosis of liposarcoma was commonly delayed in patients with central tumors, in whom larger tumor size and high-grade histology were also associated with a higher rate of positive surgical margins. In this population of patients with advanced disease, adjuvant therapy was offered as a salvage technique in the hope of extending survival, and thus it was no surprise that adjuvant therapy was associated with decreased survival when compared to those patients who did not receive additional therapies. However, size limitations in this study precluded meaningful statistical analysis of the contribution of adjuvant therapy to progression-free or overall survival. Some adult studies [
Although this study is the largest single-institution report of pediatric liposarcomas, the strength of our analysis is limited by its retrospective nature, relatively small sample size, and 50-year study period (during which time significant changes in the diagnosis and treatment of soft tissue sarcomas have occurred). We are optimistic that the NRSTS database currently being compiled by the Children’s Oncology Group will provide more comprehensive insight into the presentation and progression of pediatric liposarcoma, as well as its response to treatment.
In conclusion, pediatric liposarcoma appears to have a presentation and clinical course that is distinct from its adult counterpart. A majority of tumors are peripherally located with myxoid histology; patients with these tumors have an excellent overall prognosis following surgical excision with negative margins as single-modality treatment. High-grade tumors are rare, but are associated with a more aggressive course and poorer overall survival. Optimized treatment algorithms with neoadjuvant and/or adjuvant therapies have yet to be defined. Patients with centrally located tumors have a poor prognosis, due in part to a greater percentage of high-grade histologies, and the difficulty in achieving negative resection margins in such lesions. These are the patients who would most benefit from multicenter, randomized trials evaluating the role of neoadjuvant and adjuvant therapies.