Main

Leiomyosarcoma, which is a soft tissue sarcoma demonstrating smooth muscle differentiation, is a tumor prone to local and distant relapse. At present, histological grade, tumor size, and anatomical location are the only pathological factors that are independently predictive of patient outcome.1, 2 However, significant heterogeneity exists within these risk groups, and patients with tumors deemed low to intermediate risk have reported 5-year metastasis rates of approximately 8 and 32%, respectively.1, 3 Further research into prognostic markers and alternative treatment options is greatly required.

Recently, the downregulation of integrin alpha 7 (ITGA7) gene expression in leiomyosarcoma has been correlated with an increased risk of metastasis and decreased disease-free survival.4, 5 Integrins are critical mediators of the interactions between cells and the extracellular matrix, and as such are important regulators of cellular differentiation and migration. ITGA7 is a member of the α-integrin family, which when heterodimerized with integrin β-1, is a integral membrane receptor for laminin, a major component of the extracellular basement membrane.6 Other experimental data have implicated the oncoprotein c-Myc as a direct repressor of ITGA7 gene transcription.7

The c-Myc gene was originally recognized as a transforming sequence in the avian myelocytomatosis retrovirus and is a member of the Myc gene family that also includes N-Myc and L-Myc. The c-Myc protein acts as a transcription factor involved in the regulation of numerous critical cell processes, including cell-cycle regulation, cellular differentiation, apoptosis, adhesion, and migration.8, 9 The currently recognized mechanisms of its oncogenic activation include gene amplification, chromosomal translocation, and insertional mutagenesis.8 c-Myc expression has been correlated with poor prognosis in many solid tumors, including osteosarcoma and synovial sarcoma.10, 11, 12 Overexpression of c-Myc has also been described in high-grade liposarcoma, skeletal chondrosarcoma, and uterine leiomyosarcoma.13, 14, 15, 16 However, the c-Myc oncoprotein has not been studied extensively in other sarcomas, particularly leiomyosarcoma occurring in the somatic soft tissues.17, 18, 19

In light of recent data showing that c-Myc downregulates ITGA7 expression, we evaluated whether immunohistochemical staining for c-Myc might serve as a surrogate prognostic marker for decreased ITGA7 expression in leiomyosarcoma. We hypothesized that the overexpression of c-Myc would be associated with decreased overall and metastasis-free survival.

Materials and methods

Study Cases

This study was approved by the Institutional Review Board at Vanderbilt University; the requirement for informed consent was waived by this committee. The surgical pathology archives at the Vanderbilt University Medical Center (Nashville, TN, USA) were searched for all primary surgical resections of leiomyosarcoma performed between 1994 and 2006. Cases occurring in the subcutaneous soft tissues or involving viscera, including the female genital tract were specifically excluded. In all, 29 cases were retrieved. One low-grade tumor sample recorded as a pelvic mass from a 61-year-old female was excluded because it strongly and diffusely expressed estrogen and progesterone receptors and was considered a smooth muscle neoplasm of gynecological type.

Original hematoxylin and eosin-stained slides were reviewed and blocks were selected for immunohistochemistry. All cases showed classic histopathological features suggestive of leiomyosarcoma (elongate cells with blunt-ended nuclei and relatively abundant eosinophilic cytoplasm forming interweaving fascicles), at least focally. Overall, 17 of the 20 cases tested were positive for desmin expression. The three desmin-negative tumors showed strong and diffuse staining for α-smooth muscle actin. Histological grade was assessed using the French Federation of Cancer Centers Sarcoma Group criteria.20 Clinicopathological parameters, including patient age, sex, tumor location, size (greatest linear dimension), and American Joint Committee on Cancer TNM stage (6th edn), were recorded through a retrospective chart review.21

c-Myc Immunohistochemistry

Formalin-fixed, paraffin-embedded tissue sections were pretreated in citrate buffer in a steamer for 25 min and subsequently incubated with a 1:200 dilution of monoclonal anti-c-Myc antibody (clone 9E10, Sigma-Aldrich, St Louis, MO, USA) for 1 h at room temperature. Specific antibody binding was detected using an EnVision+ kit according to the manufacturer's recommendations (Dako, Carpinteria, CA, USA). A representative section of leiomyosarcoma incubated with diluent alone instead of primary antibody was run in parallel as a negative control. In accordance with previous studies, tumors with specific staining in >5% of neoplastic nuclei were scored as positive for c-Myc overexpression.10 Slides were independently reviewed by two of the authors (ACT and JMMC) blinded to patient outcome; there were no discrepancies in the recorded c-Myc scores.

Statistical Analysis

The principal study end points were overall survival and metastasis-free survival. Relationships between c-Myc expression and other clinicopathological and prognostic variables were also explored. Overall survival was measured as the time elapsed from the date of primary surgical resection to the date of death from any cause and was censored only for patients known to be alive at last contact. Metastasis-free survival was calculated from the date of surgical resection to the date on which radiographic evidence of metastatic disease was first documented and was censored for death and patients without radiographic evidence of metastasis at the last follow-up. Dates of death, last clinical follow-up, or documentation of distant metastasis were obtained by chart review of electronic medical records. Date of patient death was confirmed by searching the Social Security Death Index.22 There were no instances of missing data in this study cohort.

Clinicopathological and prognostic parameters were correlated with c-Myc status using standard univariate statistical methods (GraphPad Prism v5.01, La Jolla, CA, USA). For ordinal and non-Gaussian continuous data (TNM stage, histological grade, and tumor size), the Mann–Whitney test was used. Categorical data were evaluated by Fisher's exact test using a web-based calculator.23 Unless otherwise stated, two-tailed P-values ≤0.05 were considered statistically significant.

Kaplan–Meier survival curves for the different subgroups of prognostic variables were compared using the log-rank (Mantel–Haenszel) test. Survival probabilities for continuous variables were evaluated using both univariate regression analysis and the log-rank test for trend after stratification. Tumor size was stratified into tertiles (<4.99, 5–9.99, and >10.0 cm) and patient age was approximately stratified into quartiles (<39, 40–59, 60–79, and ≥80 years). The effect of c-Myc status at the time of primary surgical resection of leiomyosarcoma was assessed by multivariable survival analysis according to the Cox proportional hazards regression model, adjusting for potential confounding prognostic factors, such as histological grade, tumor size, and TNM stage. Other clinicopathological parameters associated with the value of P <0.20 in univariate analysis were also included in the regression models. Multivariable survival analysis was carried out using the survival package installed on R software (v2.8.1).24, 25 For multivariate analysis of the effect of AJCC TNM stage on metastasis-free survival, stage IV cases were reclassified as if metastases were not present at the time of diagnosis. No interactions between c-Myc expression and any of the covariables studied were detected when interaction terms were included in the model.

Results

The cohort consisted of 14 females and 14 males. Neoadjuvant radiation therapy was administered to four patients; one patient received preoperative combination radiation and chemotherapy. The median follow-up for the entire cohort was 23 months (range: 6–94 months) during which time 12 patients (43%) died and 12 patients (43%) developed metastatic disease; metastases were detected during initial staging evaluations in 3 other patients. The median overall survival for the entire cohort was 35 months, and the median metastasis-free survival was 23 months. Nuclear immunoreactivity for c-Myc was detected in 15 of 28 tumors (54%) (Figure 1). The expression of c-Myc did not correlate with prognostic variables, such as histological grade, tumor size, anatomical location, or TNM stage (Table 1). In addition, no significant associations between c-Myc expression and patient age, gender, or exposure to neoadjuvant radiation therapy were observed.

Figure 1
figure 1

Immunohistochemical stains for c-Myc in soft tissue leiomyosarcoma (a, H&E) representative of positive (b) and negative (c) cases (original magnification, Ă— 200).

Full size image
Table 1 Comparison of clinicopathological features of c-Myc-negative and c-Myc-positive soft tissue leiomyosarcoma
Full size table

Log-rank tests performed on Kaplan–Meier survival curves disclosed significantly decreased metastasis-free survival for patients with c-Myc-positive leiomyosarcoma compared with c-Myc-negative tumors (median, 9 months vs. >94 months; P=0.014) (Table 2 and Figure 2a). Increased risk of metastasis was also noted for patients with high-grade (grade 3) leiomyosarcoma and for those patients who received preoperative radiation therapy. Positive c-Myc status was the only variable that reached statistical significance for decreased metastasis-free survival in subsequent Cox proportional hazards regression analysis (Table 2).

Table 2 Univariate and multivariate metastasis-free survival analysis
Full size table
Figure 2
figure 2

Metastasis-free (a) and overall (b) Kaplan–Meier survival curves of patients with c-Myc-positive (—) and c-Myc-negative (- - - - ) soft tissue leiomyosarcoma.

Full size image

Overall survival intervals were also significantly shorter in patients with c-Myc-positive leiomyosarcoma (median, 23 months vs. >94 months; P=0.017) (Table 3 and Figure 2b). None of the other parameters that were studied showed a significant effect on overall survival in univariate analysis, although a trend was noted for histological grade. Despite a relatively wide 95% confidence interval, c-Myc status was the only factor independently associated with increased risk of death in multivariate regression analysis (Table 3).

Table 3 Univariate and multivariate overall survival analysis
Full size table

Of the three patients with leiomyosarcoma of low histological grade (grade 1), two were positive for c-Myc expression. One of the c-Myc-positive cases presented with metastatic disease and died 13 months after surgical resection of the primary tumor; the other is alive without evidence of the disease 8 months after surgical resection. The one patient with a c-Myc-negative, low-grade leiomyosarcoma is an event-free survivor at 23 months of clinical follow-up. The three patients in whom metastatic disease was discovered during initial staging studies were all positive for c-Myc, and all three died within 23 months of surgical resection.

Discussion

We found that c-Myc expression significantly correlates with both metastasis-free and overall survival in patients with resected leiomyosarcoma. Multivariate survival analysis showed that c-Myc status was the only prognostic factor that independently predicted decreased survival intervals. Although histological grade and pathological stage are widely accepted prognostic factors for leiomyosarcoma, it remains very difficult to predict outcomes for individual patients stratified using these parameters alone.2 Recently, it has been shown that ITGA7 gene expression is downregulated 40-fold in leiomyosarcoma with high metastatic potential and is associated with decreased metastasis-free survival.4, 5 The lack of a commercially available antibody against ITGA7 for use on formalin-fixed, paraffin-embedded tissue prompted us to evaluate upstream regulators of ITGA7 as potential molecular prognostic markers for leiomyosarcoma.

c-Myc downregulates ITGA7 gene transcription by binding to a double E-box in the ITGA7 promoter region.7 Therefore, we examined whether the detection of nuclear c-Myc was prognostically significant in a series of leiomyosarcoma. Nuclear c-Myc expression did not correlate with other accepted prognostic markers for leiomyosarcoma. However, c-Myc expression was associated with decreased metastasis-free and overall survival, consistent with previous experimental data that showed poor outcomes in patients with leiomyosarcoma with decreased ITGA7 expression.

In C2C12 myoblasts, overexpression of ITGA7 is associated with increased binding of laminin and decreased binding of fibronectin.26, 27 Conversely, the downregulation of ITGA7 modulates interactions between integrins and the extracellular matrix such that fibronectin binding is favored.27 In mammary epithelial cells, fibronectin-dependent signaling has been shown to upregulate c-Myc expression through the activation of the MEK–ERK pathway.28 Although not yet established, it is possible that the c-Myc-mediated inhibition of ITGA7 increases fibronectin binding and represents a positive feedback loop that stimulates c-Myc overexpression in a growth factor-independent manner. The potential significance of this pathway is shown by evidence that the MEK–ERK inhibitor U0126 decreases proliferation of rhabdomyosarcoma cells concomitant with decreased intracellular levels of c-Myc.29, 30 Perhaps the c-Myc status of leiomyosarcoma might identify those tumors most susceptible to MEK–ERK inhibition.

The consequences of c-Myc overexpression are pleiotropic and include dysregulation of the cell cycle, suppression of apoptosis, and cellular transformation, in addition to inhibition of ITGA7 and subsequent alteration of cell adhesion and migratory properties.8, 9 Studies directly targeting c-Myc expression in transgenic mouse models have shown profound anti-neoplastic effects, including potentiation of chemotherapeutic agents, suggesting that that other therapeutic strategies may improve survival in patients with c-Myc-positive leiomyosarcoma.12, 31, 32, 33, 34 Strategies to inhibit transcription or translation of c-Myc mRNA that are currently under investigation or in development include the use of antisense oligonucleotides, triple helix-forming oligonucleotides, cationic porphyrins, and phosphorodiamidate morpholino oligomers (AVI-5126, ClinicalTrials.gov identifier: NCT00777842).35 Together with the results of this study, these advances should prompt further investigation into current therapeutic approaches to soft tissue leiomyosarcoma.36

In summary, our data suggest that immunohistochemical detection of c-Myc may be useful in predicting increased risk of metastasis and shorter overall survival in patients with leiomyosarcoma. Such ancillary information may assist in clinical decision making regarding the intensity and type of adjuvant therapy administered. Consideration of larger, multi-institutional studies is recommended to validate these data. Given the adverse prognosis that c-Myc overexpressing tumors confer on patients and other preclinical data collected thus far, clinical trials evaluating the role of anti-c-Myc or anti-MEK–ERK pharmacotherapeutic agents in treating sarcomas such as leiomyosarcoma should also be considered.