Abstract
Purpose: To see if current guidelines for the early diagnosis of sarcomas can be improved.
Method: Data on 1100 patients referred to our unit with a lump suspicious of sarcoma was analyzed to try and identify clinical features more common in malignant than benign lumps. The following five items were analysed: size, history of increasing size, presence of pain, depth, age. For each of these items sensitivity, specificity, accuracy and weights of evidence were collected. ROC curves were used to identify the most sensitive cut off for continuous data.
Results: The best cut off predicting malignancy for size was 8cm and for age 53 years
The weights of evidence (WE) are logs of the likelihood ratios and can be added and a probability then calculated. e.g. a 36 yr old with a 10cm, deep, painless lump that is increasing in size scores −0.39 + 0.4 + 0.4 – 0.11 + 0.58 = 0.88. This equates to a risk of the lump being malignant of 70%.
Conclusion: This analysis shows that increase in size is the strongest predictor of malignancy/benignancy followed by age > 53 and size > 8cm. This data can help formulate strategies for earlier detection of soft tissue sarcomas.
The abstracts were prepared by Mr Roger Tillman. Correspondence should be addressed to BOOS at the Royal College of Surgeons, 35–43 Lincoln’s Inn Fields, London WC2A 3PN
