Skip to main content
. Author manuscript; available in PMC: 2009 Dec 1.
Published in final edited form as: Ann Surg Oncol. 2008 Oct 1;15(12):3550–3560. doi: 10.1245/s10434-008-0163-0

Table 4.

Multivariate Model of Death from Locally Recurrent Disease for Patients Age ≥ 16 with Low Grade Soft Tissue Sarcoma Resected with Curative Intent

Variable* Hazard Ratio for Death
from Soft Tissue Sarcoma
(95% Confidence Interval)
Primary Site**
   Extremity 1.00 (referent)
   Trunk 5.98 (1.01 – 35.48)
   Other 19.80 (4.26 – 92.10)
   Visceral 16.09 (2.86 – 90.62)
   Retroperitoneal 59.10 (13.22 – 264.32)
 Primary Tumor Size
   ≤ 5 cm 1.00 (referent)
   > 5 cm ≤ 10 cm 6.78 (1.98 – 23.14)
   > 10 cm 10.23 (3.06 – 34.27)
 Histology
   Other¶¶ 1.00 (referent)
   Malignant Fibrous 1.05 (0.32 – 3.48)
   Histiocytoma (MFH)
   Leiomyosarcoma 2.14 (0.73 – 6.27)
   Liposarcoma 1.87 (1.08 – 3.25)
 Margin Status
   R0 1.00 (referent)
   R1 2.26 (1.40 – 3.68)
   R2 5.86 (3.42 – 10.04)
*

Variables significant at the 0.10 level on univariate analysis were included in the multivariate model. Gender, depth, and age were not statistically significant variables on multivariate analysis. Treatment-related variables were not included in the multivariate model since their association with the outcome of interest could not be assumed to be independent.

**

Although there were significant differences in DOLR by primary tumor site, multivariate analysis demonstrated liposarcoma histology to be a more reliable predictor of DOLR than site, likely because site and histology are tightly linked covariables.

Includes thoracic, head and neck, and skin.

Includes gastrointestinal, gynecologic, and genitourinary.

¶¶

Includes all histologies except MFH, leiomyosarcoma, and liposarcoma.