Quantifying Head and Neck Lymphatic Malformation Volume and Composition During Treatment

Abstract

Objective: Head and neck lymphatic malformation (HNLM) treatment involves a combination of observation, surgery, sclerotherapy, and targeted medical therapy. Objective comparison of these differing treatments is unstandardized due to heterogeneity of HNLM location, size, and variable components. Our objective was to develop a protocol for standardized assessment of magnetic resonance imaging (MRI) image sets through novel semiautomated algorithms. We aimed to obtain reproducible multimodal tissue level data that can be analyzed for interval HNLM treatment response. Methods: Patients who were undergoing therapy for HNLM were queried from an institutional database between 2015 and 2020. MRI sequences were registered in 3D Slicer, and in MATLAB a multimodal tool was developed to quantitate HNLM volume and changes in composition. Volume measures were normalized to normal childhood growth using the nasion-basion distance. Reproducibility studies were conducted to evaluate interrater reliability. Results: HNLM undergoing excision (n = 3) had a 92.3%–96.7% decrease malformation volume. HNLM having medical treatment with sirolimus and aspirin had a 7.3%–36.4% decrease in normalized volume, the majority of which was due to a decrease in cystic fluid content (reduced by 27.0%–36.4%). HNLM treated with sclerotherapy had no normalized volume change following treatment. One HNLM first treated with sirolimus had a 27.8% decrease in normalized volume and then combined normalized volume reduction of 75.6% after resection. Conclusion: This proof-of-concept use of longitudinal HNLM MRI in pediatric patients undergoing treatment demonstrates that objective information can be obtained through this method. This information can be used to determine treatment efficacy and optimize lesion specific treatment strategies.