Prognostic Factors and Expression of MDM2 in Patients with Primary Extremity Liposarcoma

OBJECTIVE The objective of this study was to investigate MDM2 (murine double minute 2) protein expression and evaluate its relationship with some anatomical and pathological aspects, aiming also to identify prognostic factors concerning local recurrence-free survival, metastasis-free survival and overall survival in patients with primary liposarcomas of the extremities. MATERIALS AND METHODS Of 50 patients with primary liposarcomas of the extremities admitted to a Reference Service, between 1968 and 2004, 25 were enrolled in the study, following eligibility and exclusion criteria. RESULTS The adverse factors that influenced the risk for local recurrence in the univariant analysis included male sex (P = 0.023), pleomorphic histological subtype (P = 0.027), and high histological grade (P = 0.007). Concerning metastasis-free survival, age less than 50 years (P = 0.040), male sex (P = 0.040), pleomorphic subtype (P < 0.001), and high histological grade (P = 0.003) had a worse prognosis. Adverse factors for overall survival were age under 50 years (P = 0.040), male sex (P = 0.040), pleomorphic subtype (P < 0.001), and high histological grade (P = 0.003). CONCLUSIONS There was no correlation between immunohistochemically observed MDM2 protein expressions and the anatomical and pathological variables studied. The immunohistochemical expression of MDM2 protein was not considered to have a prognostic value for any of the surviving patients in this study (local recurrence-free survival, metastasis-free survival, or overall survival). The immunoexpression of MDM2 protein was a frequent event in the different subtypes of liposarcomas.


INTRODUCTION
Soft tissue sarcomas are rare neoplasias; however, they may be present across all age groups and all body sites where soft tissues exist. In the United States, soft tissue sarcomas (STSs) occur in 0.7% of cancer patients aged 16 years and over. 1 In Brazil, the actual incidence of these neoplasias is difficult to determine due to lack of appropriate records.
Although being rare, STSs have a relatively high mortality rate. This high mortality can be attributed in part to aggressive local invasion, but mostly to frequent metastases. 2 Liposarcomas can occur in any anatomical site; concerning the extremities, they occur more frequently in the thigh. [3][4][5][6][7] Adipose tissue tumors represent a group of injuries whose classification is an issue of continuous debate. 8,9 Liposarcomas have been shown to possess many histological patterns. 10 Many changes have occurred over the years concerning the histological classification of this type of soft tissue sarcoma. Currently, according to the Classification of Soft Tissue and Bone Tumors of the World Health Organization, it is accepted that this sarcoma presents as one of 5 histological subtypes: well-differentiated, dedifferentiated, myxoid/round-cell, pleomorphic, and mixed. 11 Most frequently, the proto-oncogene mdm2 is amplified in soft tissue sarcomas. Most of the known oncogenes are changed by amplification, resulting in overexpression and, consequently, high levels of their protein products.
The q 13-15 region of chromosome 12 is complex and contains different genes, such as mdm2, 12 that are amplified or reorganized in lipomatous tumors, which can be demonstrated by immunohistochemistry. 13,14 The number of publications on the molecular characteristics of STSs is constantly increasing, and research focuses on the search for additional prognostic factors, aiming to identify those related to risk of local recurrence, metastases, and death due to the disease, and to study their correlation with known clinical, anatomical, and pathological variables. 15 The determination of molecular variables related to anatomical and pathological aspects and to the prognosis could help identify subgroups of patients with better or worse prognoses. 15 Thus, with more prognostic factors identified, it would be possible to select the risk patients and work toward improved therapeutic outcomes.
The objective of this study, which is part of a general survey on sarcomas, 16 was to investigate mdm2 protein expression and evaluate its relationship with some anatomical and pathological aspects, aiming also to identify prognostic factors concerning local recurrence-free survival, metastasisfree survival, and overall survival in patients with primary liposarcomas of the extremities.

MATERIALS AND METHODS
A review was performed of the medical records of patients who were anatomically and pathologically diagnosed with liposarcoma of the extremities and were admitted to the Institute of Orthopedics and Traumatology, Faculty of Medicine, University of São Paulo (IOT/FMUSP), SP, Brazil, between 1968 and 2004. Based on these medical records and after a detailed anatomical and pathological evaluation from the paraffin blocks of specimens obtained in the original biopsy, we selected 50 patients. Of these 50 patients, 5 were excluded for presenting local recurrence in the first evaluation, and another 3 because they underwent biopsy only, with no treatment. Another 17 cases were excluded because their follow-up periods were less than 24 months. The analysis of variables was then performed for 25 patients (and their medical records) who underwent treatment at IOT-HC/FMUSP. The anatomical and pathological evaluation was made from paraffin blocks containing tumor fragments that were obtained from the surgical specimens.
The following variables were studied:

Anatomical site
Fifteen (60%) patients had tumor formation in the thigh. For the other patients, the leg was affected in 4 (16%) cases and the forearm in 2 (8%) cases. The arm, axilla, scapula, and foot were also affected (1 of each). The right side of the body exceeded its contralateral side in the 1.5:1 ratio.

Clinical condition and delay in seeking medical advice
The unanimous clinical complaint was the insidious onset of usually painless local tumors. The secondary symptom was pain, occurring in only 3 (12%) cases.
The period between the onset of symptoms and the seeking of medical advice was 2.9 months, ranging between 2 and 120 months (mean, 36.0 ± 41.3 months; median, 24.0 months).

Remote metastasis
Metastasis was identified in 3 (12%) patients, 1 with high-grade myxoid/round-cell histology and 2 with pleomorphic histology. All 3 patients developed an extrapulmonary metastasis, usually in the lumbar spine. Only 1 among the pleomorphic subtypes presented additional lung metastasis. The patient with lung metastasis died after a less than 7-month follow-up period.

Follow-up period
The average follow-up period for our patients with liposarcoma was 68.3 months (standard deviation, 47.4 months; median, 54.0 months; range, 8-184 months) ( Table 1)

Tumor size and histology
All patients underwent surgical resection of the tumor. Variant myxoid/round-cell represents the most common histological type, affecting 13 (52%) of 25 cases. The case frequency by histological subtype of liposarcoma was distributed as follows: 9 (36%) well-differentiated, 13 (52%) myxoid/round-cell, and 3 (12%) pleomorphic. Using the adopted selection criteria, we did not find cases of dedifferentiated and mixed liposarcomas.

Surgical and adjuvant treatment
In all cases, the tumors were subjected to ample resection according to the Enneking classification. 17 Forty-eight percent of these patients underwent surgery only.
Thirteen (52%) patients received radiotherapy (RT) and/or chemotherapy (ChT) postoperatively. None of them received RT and/or ChT followed by surgery. Radiotherapy and/or ChT before and after surgery (both) were not instituted for any of the patients.

Local recurrence
Local recurrence was observed in 3 (12%) cases, 1 of them of the well-differentiated type and 2 of the myxoid/round-cell type (1 low-grade malignancy and 1 high-grade malignancy). Two low-grade cases presented 2 local recurrences each and had no adjuvant therapy. The patient with high-grade neoplasia developed only 1 episode of recurrence after receiving postoperative radiotherapy. None of the patients experiencing local recurrence presented remote metastasis.

Complications of treatment
Eleven (44%) of the 25 treated patients presented complications of 8 types, and some patients presented more than one type of associated form. The most frequent complications were 4 surface infections of the surgical wound, 4 peripheral nerve injuries, and 3 deep infections.

INCLUSION CRITERIA
All patients included in this study met the following eligibility criteria: (a) having undergone surgery at IOT for local treatment of the primary tumor; (b) having anatomical and pathological confirmation of liposarcoma (all slides were reviewed by the same pathologist who is an expert in  musculoskeletal tissues); and (c) having a tumor located in the extremities only, as described in the medical record.
To study the mdm2 nuclear expression marker, we selected the cases for which specimens had been preserved in paraffin blocks at the time of biopsy prior to any adjuvant therapy, and only primary tumors were included.

EXCLUSION CRITERIA
Patients were excluded from the study if they (a) presented metastases and/or local recurrence during the first evaluation; (b) had postoperative follow-up periods of less than 2 years, except for those who died of cancer before that period; (c) underwent any treatment prior to enrollment at IOT.

ASSESSMENT OF IMMUNOHISTOCHEMICAL POSITIVITY
The mdm2 protein was classified as present or absent. For the immunohistochemical analysis, the tumors were classified as positive (positivity ≥ 10% of the sample) or negative (Figure 1).
Protein expression was assessed at 400x magnification, and the nuclei with unequivocally typical brown immunoexpression were considered positive. At least 400 nuclei of neoplastic cells were counted per case.

STATISTICAL ANALYSIS
The descriptive analysis of the sample was performed based on average, median, and percentage values. The simple frequencies of all variables studied were calculated. The accumulated survival probabilities were evaluated us-ing the Kaplan-Meier method, and the survival curves were compared using the log rank test.
Positive or negative indexes for mdm2 were evaluated as variables to determine their prognostic value and the value of the association with other (anatomical and pathological) variables. The associations were performed according to Fisher's exact test.
The rounding up of frequency and survival rates was used to simplify calculations. Statistical significance was defined for P values < 0.05.

RESULTS
The prognostic value of some important clinical, epidemiological, anatomical, pathologic, and immunohistochemical variables was also evaluated based on curves of local recurrence-free survival, metastasis-free survival, and overall survival.

Prognostic factors related to local recurrence-free survival
Male gender (P = 0.023) was considered an adverse factor; the pleomorphic subtype (P = 0.027) presented the highest index of local recurrence. High histological grade (P = 0.007) had the worst prognosis. The other factors investigated (age, site, side of body, delay in seeking medical advice, specimen size, or type if treatment performed) did not reach statistically significant levels. The presence or absence of mdm2 did not reach statistically significant levels and was not considered a prognostic factor for local recurrence-free survival.

Prognostic factors related to metastasis-free survival
Age less than 50 years (P = 0.040), male sex (P = 0.040), pleomorphic subtype (P < 0.001), and high histological grade (P = 0.003) were considered adverse factors for metastasis-free survival (Figure 2). The other factors studied did not reach statistically significant levels (site, side of body, delay in seeking medical advice, specimen size, and type of treatment performed). Immunohistochemical evidence of the presence or absence of mdm2 gene expression did not reach statistically significant levels and was not considered a prognostic factor for metastasis-free survival.

Prognostic factors related to overall survival
The overall survival curve for the 25 patients is shown in Figure 3. Age under 50 years (P = 0.036), male sex (P = 0.043), pleomorphic subtype (P = 0.001), and high histologi-   Site, side of body, delay in seeking medical advice, specimen size, and type of treatment performed did not reach statistically significant levels. Also, the presence or absence of mdm2 protein expression did not influence the overall survival.

Analysis of the correlation between mdm2 protein expression and anatomical and pathological variables
As shown in Table 3, there was no statistically significant association between the expression (or not) of mdm2 protein and the following anatomical and pathological variables: subtype, grade, and size of tumors.

DISCUSSION
Although there are more than 30 histological types of soft tissue sarcomas, many of them are grouped for being similarly diagnosed, staged, and treated. 18 However, because these sarcomas include a large series of histological types, the clinical course of each type of sarcoma can be extremely different. 2 Soft tissue sarcomas account for less than 1% of all malignant tumors. 19 As shown in many publications, [19][20][21][22][23][24] of this percentage, between 10% and 20% are liposarcomas, which are considered the second most frequent type. 20,23 Concerning global survival, Reitan et al 5 asserted that younger patients presented better prognoses, which does not agree with our findings, where we found a significant difference regarding worse prognosis for patients under 50 years of age.
A few investigators have asserted that the most common form of liposarcoma is the well-differentiated sarcoma. 25 The most frequent histological type found in our study was the myxoid/round-cell type (52%), in agreement with other current studies. 23,26 One of the most important factors influencing the survival rate seems to be the histological type of the tumor. In our case series, we found that the histological subtype of liposarcoma is significantly related with survival; that is, patients with pleomorphic liposarcoma are at higher risk for death than those who have myxoid/round cell or welldifferentiated sarcomas. We also observed a significant risk of local recurrence and development of remote metastases and worse prognosis for patients with the pleomorphic histological type.
Tumor grade has been clearly recognized as a performance predictor, with high-grade tumors being associated with prognoses that are unfavorable for survival. 27 For liposarcomas of the extremities, this also appears to be the case. 2 The high average value of tumor size seems to be directly related to the long evolution time presented by these neoplasias. However, in this series, we did not find that the dimension of the tumor was a prognostic factor in any of the surviving patients studied.
As a group, liposarcomas present a postoperative local recurrence rate of approximately 50%. However, these indexes are lower if only injuries of the extremities are considered. 28.29 In our case series, of the 25 patients undergoing ample surgical resection, 3 presented local recurrence during the follow-up period: 1 of the well-differentiated type and 2 of the myxoid/round-cell type (1 low-grade and 1 high-grade). mdm2 gene amplification has been shown to occur at a high frequency in many studies on soft tissue sarcomas. The product of mdm2 interacts with the p53 protein and inhibits its ability to regulate genetic expression as a transcription factor. Thus, amplification (and subsequent overexpression) of mdm2 may have the same effect as that of mutations in p53. 30 The changes in the mdm2 gene are known to be a common mechanism in the genesis of liposarcomas. 31 The amplification of this gene and its mRNA overexpression can lead to overproduction of mdm2 protein. 32 The immunohistochemical analysis reveals nuclear location and overexpression of mdm2 in these tumors with amplification of the mdm2 gene. 33 In liposarcomas, amplification of the mdm2 gene is observed only in high-frequency well-differentiated tumors, not in other tumor subtypes. 32 Another author 34 mentions that amplification of the mdm2 gene is not also seen in myxoid liposarcomas, the largest subtype of sarcomas. On the other hand, Schneider-Stock et al 31 found amplification of the mdm2 gene in the myxoid and pleomorphic variants, although the well-differentiated liposarcoma is characterized by an already known high frequency.
Although immunohistochemistry can be used to demonstrate mdm2 overexpression, direct correlation between the gene amplification and protein overexpression is not the rule. 35 Nevertheless, mutations cause an irregular increase in stable proteins that can be detected by immunohistochemistry. 36 In our case series, immunohistochemistry showed mdm2 protein expression in 22 (88%) of the 25 cases. Of these 22 cases, 7 (32%) were of the well-differentiated subtype, 12 (54.5%) of the myxoid/round-cell type, and 3 (13.5%) of the pleomorphic type. Data from this study indicate a high frequency of the presence of mdm2 protein in the different histological subtypes of liposarcomas, suggesting preliminarily a high sensitivity of this marker in this pathology. In spite of this, currently, the diagnosis of these neoplasms is based on morphology, not immunocytochemical tests. High 7 1 (33%) 6 (27%) * P values using Fisher's exact test CONCLUSIONS 1) There was no correlation between the mdm2 protein expression as observed by immunohistochemistry and the anatomical and pathological variables studied; additionally, there was no relationship between those variables and the different prognostic factors in primary liposarcomas of the extremities. 2) Factors such as male sex, pleomorphic histological subtype, and high grade of malignancy are unfavorable for local recurrence-free survival. 3) Factors such as male sex, age under 50 years, pleomorphic subtype, and high histological grade are adverse for metastasis-free survival and overall survival. 4) mdm2 expression was a frequent event in the different subtypes of liposarcomas.