Introduction The aim of the investigation was to assess the significance of the extent of surgical margin on the chance of death, metastasis and local recurrence in soft tissue sarcoma.

Methods The review consisted of 324 patients. Surgical margin data was unavailable for 21, and of the remaining 303 patients, 10 patients had no residual tumour, margins were not defined for 24 patients and nine patients had radical resections. Wide margins were achieved for the remaining 260 patients. Fifty-four percent had surgical margins of under five millimetres. Cox Proportional Hazards Regression modelling was used to consider the impact of surgical margin with an overall survival, disease-free survival and metastasis-free survival. Results were expressed as survival rate ratios and graphics represented as model based survival curves. All associations that were statistically significant as well as any associations for which the rate ratios were 2.0 or greater were reported. Follow-up ranged from 53 days to 187 months, with a median of 40 months.

Results Overall survival time for the 279 patients with complete information was 124 months. There was a significant association between overall survival and extent of the surgical margin (chi-squared test statistics = 14.7, 8df, p = 0.043). There was a significantly higher death rate in patients who had a wide contaminated margin or a radical resection indicating likely poorer prognostic groups. There was however no difference between any margin from one to 20 mm. With respect to disease relapse, there were 27 local recurrences among 279 patients, and for statistical reasons, 24 local recurrences among 213 patients were reviewed. There was a significant association between the extent of surgical margin and disease-free survival (chi-squared test statistics = 9.5, 4df, p = 0.051). With relation to metastasis, 68 of 257 patients were reviewed, there being no statistical association between metastasis-free survival and margin extent.

Conclusions There is significant statistical evidence to suggest overall and disease-free survival increase with increasing width of surgical margin. The evidence is not convincing in our assessment of metastasis-free survival

In relation to the conduct of this study, one or more of the authors is in receipt of a research grant from a non-commercial source.