From February 2000 and January 2009, a total of 45 patients with localized STS were treated with neoadjuvant external beam RT followed by resection with curative intent. These patients were prospectively entered and tracked in a computerized cancer center database. Fourteen patients with low-grade primary tumors and six patients who received combination preoperative chemoradiotherapy were excluded from this analysis. The remaining 25 patients formed the basis of this study.
Preoperative RT was administered in 2-Gy fractions over 25 sessions for a total dose of 50 Gy. Surgical resection was performed approximately 4 to 6 weeks after the completion of RT to allow for acute toxicities to resolve.
After approval for this study by the Institutional Review Board, clinical, pathologic, and treatment data were reviewed and analyzed with respect to percentage pathologic necrosis in the tumor specimen at the time of surgical resection as well as Response Evaluation Criteria in Solid Tumors (RECIST) response on pre- and post-RT cross-sectional imaging. Local recurrence-free survival, distant recurrence-free survival, disease-specific survival, and overall survival were also calculated.
Hematoxylin and eosin–stained slides were reviewed in a blinded fashion by a single STS pathologist (D.B.). The percentage of histologically intact tumor and the percentage of necrotic tumor were scored per slide, and the mean percentage of tumor necrosis was calculated for the entire specimen, excluding nonneoplastic tissue. The median number of stained slides examined per patient was 10 (range, 8–18).
Cross-sectional images obtained before and after RT were reviewed by a single musculoskeletal radiologist (W.M.) in a blinded fashion. Among extremity STS, magnetic resonance imaging was used exclusively in all 18 cases. Among retroperitoneal STS, magnetic resonance imaging was used in five of seven cases, computed tomographic scan in one of seven cases, and both modalities in one of seven cases. Tumors were evaluated in three dimensions, and maximal tumor diameter was ascertained. Percentage change in maximal radiographic tumor dimension and RECIST response were calculated (http://ctep.cancer.gov/protocolDevelopment/docs/recist_guideline.pdf
Data were collected on age, sex, tumor location, histologic type, primary tumor size, maximal tumor diameter, histologic grade, tumor depth, margin status, extent of resection, presence of myxoid histology, and disease recurrence. Major wound complications were classified according to standardized definitions.
Histologic grade was classified by a three-tiered system (grade I through III). Age was determined from the date of diagnosis of the primary tumor. Depth was categorized as either superficial or deep to the investing fascia. By convention, size of the primary tumor was divided into three groups: ≤5 cm, >5 cm but ≤10 cm, and >10 cm. Sites included extremity (upper at or distal to the shoulder/axilla, and lower at or distal to the buttock/groin) and retroperitoneal.
Histologic diagnosis was assigned by the published criteria of the World Health Organization Classification of Tumors of Soft Tissue and Bone. Margin status was determined either clinically (R2 for gross residual tumor left behind) or as part of the histopathologic assessment (R1 for microscopically positive margins, and R0 for microscopically negative margins).
The date of recurrent disease was defined either by biopsy or by the radiographic detection of suspicious lesions when no biopsy was performed. Peritoneal recurrences of intra-abdominal and RPS were considered local recurrences, whereas liver metastases were considered distant recurrences. Follow-up was counted from the date of diagnosis until the date of death or date of last follow-up. Freedom from local recurrence was counted from the date of resection. Patients who were free from recurrence or death were censored according to the date of their last follow-up.
Summary statistics were reported as mean ± standard deviation with median (range) where appropriate. Statistical analyses were performed by SAS version 9.1 (SAS Institute, Cary, NC). Nonparametric numerical variables were compared by the Mann–Whitney U
-test for two groups or the Kruskal–Wallis test for three groups. The Kaplan–Meier approach was used to estimate survival curves, and survival differences were analyzed by the log rank test.8
Significance was set at P