Abstract
Current guidelines suggest that most lumps less than 5cm in size will not be malignant. We reviewed our experiences of small lumps at the Royal Orthopaedic Hospital, Birmingham, in order to try and identify factors associated with malignancy. All referrals to our clinic were included and data on size, symptoms, depth and diagnosis were analysed. We identified 325 patients, with 57 (17.5%) malignant and 268 (82.5%) benign lumps. Size >3cm was the most accurate predictor of malignancy (35.8% malignant vs 9.9% <3cm). 23.6% of lumps reported as increasing size were found to be malignant, compared to 8.6% of those that were not. Pain was found to be a poor predictor of malignancy, with only 15.5% of painful lumps found to be malignant and 18.9% of painless lumps found to be malignant. ‘Size >3cm’ has a sensitivity of 73.6% and a specificity of 56.3%, ‘Depth’ has a sensitivity of 73.6% and a specificity of 30.2%, ‘Increasing in Size’ a sensitivity of 71.4% and a specificity of 56.7% and ‘Pain’ a sensitivity of 33.3% and a specificity of 57.1%. If a lump has none of the above features then one can be 89.5% certain that it is not malignant. If it has 1 feature there is a 16.1% chance of its becoming malignant, 2 features a 58% chance, 3 features an 87% chance and if a lump has all 4 features there is a 93.5% chance of its being found to be malignant.
The main limitation with this study is that the data set is skewed because all lumps referred to the Oncology Department are by their nature suspicious. This study identifies those factors which suggest malignancy in small lumps and provides the basis for carrying out a large, community-based study on all lumps, to identify features associated with malignancy.
