Abstract
The use of peritumoral oedema on magnetic resonance (MR) imaging to predict soft tissue tumour grade is controversial. The clinical significance of oedema visualised on MR scans is poorly defined in the literature. We undertook this study to ascertain a diagnostic relationship between peritumoral oedema surrounding soft tissue sarcomas and the histological grade of the tumour.
One hundred and ten consecutive soft tissue tumours were extracted from the New Zealand Bone and Soft Tissue Tumour Registry. Key inclusion criteria were tumours deep to fascia, measuring more than 5cm in any dimension. Both benign and malignant sarcomas were included. MR scans and histology were reviewed, separately and in random order by a single author. Histology was graded as benign, low or high grade (based on the American Joint Committee on Cancer grading system).
Peritumoral oedema was defined as the increased signal intensity, on T2 or STIR images, immediately surrounding a discrete lesion. It was measured on two or more planes with the largest value used in diagnostic calculations. Oedema greater than or equal to 20mm was defined as a positive test result. Twenty five random scans were double read to ensure inter-observer reliability
Data was obtained for 83 tumours, 36 benign and 47 malignant (34 high grade and 13 low grade). The tumours in all groups were matched for size. The mean peritumoral oedema was 10.5mm for benign tumours, 20.6mm for low grade sarcomas (p<0.1), 28.1mm for high grade tumours (p<0.01) and 26.1mm if all malignant tumours were included as a single group (p<0.01). Using peritumoral oedema as a diagnostic test for tumour grade resulted in a specificity of 72%. The highest diagnostic ability was found when comparing benign to high grade tumours which yielded sensitivity of 59% and a positive likelihood ratio of 2.1. This data suggests a high false negative rate and that the test adds little to the diagnostic process.
To our knowledge this is the first study which assesses the diagnostic accuracy of peritumoral oedema to predict the histological grade of soft tissue sarcomas. Our results show a statistically significant difference, in surrounding peritumoral oedema, exists when comparing benign to high grade sarcomas and to all malignant tumours. This relationship is not apparent for low grade tumours. As a diagnostic test, using only peritumoral oedema to predict histological grade is unreliable.
