Lung, liver and lymph node metastases in follow-up MSCT: comprehensive volumetric assessment of lesion size changes

Abstract

Purpose: To investigate measurement accuracy in terms of precision and inter-rater variability in the simultaneous volumetric assessment of lung, liver and lymph node metastasis size change over time in comparison to RECIST 1.1.

Materials and methods: Three independent readers evaluated multislice CT data from clinical follow-up studies (chest/abdomen) in 50 patients with metastases. A total of 117 lung, 77 liver and 97 lymph node metastases were assessed manually (RECIST 1.1) and by volumetry with semi-automated software. The quality of segmentation and need for manual adjustments were recorded. Volumes were converted to effective diameters to allow comparison to RECIST. For statistical assessment of precision and interobserver agreement, the Wilcoxon-signed rank test and Bland-Altman plots were utilized.

Results: The quality of segmentation after manual correction was acceptable to excellent in 95 % of lesions and manual corrections were applied in 21 - 36 % of all lesions, most predominantly in lymph nodes. Mean precision was 2.6 - 6.3 % (manual) with 0.2 - 1.5 % (effective) relative measurement deviation (p <.001). Inter-reader median variation coefficients ranged from 9.4 - 12.8 % (manual) and 2.9 - 8.2 % (volumetric) for different lesion types (p < .001). The limits of agreement were ± 9.8 to ± 11.2 % for volumetric assessment.

Conclusion: Superior precision and inter-rater variability of volumetric over manual measurement of lesion change over time was demonstrated in a whole body setting.