# Predictive Factors for Response to Percutaneous Bleomycin in Lymphatic–Venous Malformations of the Head and Neck

## TL;DR

This study identifies imaging features that predict successful treatment of head and neck lymphatic–venous malformations using bleomycin.

## Contribution

The study introduces predictive imaging parameters for response to bleomycin in lymphatic–venous malformations.

## Key findings

- Cavitary morphology was the strongest predictor of treatment response.
- Pure lymphatic malformations and macrocystic components were associated with better outcomes.
- Neck location and deep-seated lesions also predicted favorable responses.

## Abstract

Background/Objectives: This study aims to identify baseline imaging parameters, across various imaging modalities, that can predict the response to bleomycin sclerotherapy in patients with head and neck lymphatic–venous malformations (LVMs). Methods: A retrospective analysis of 80 patients (85 lesions) treated at a tertiary care center between January 2018 and December 2022 was conducted. Imaging modalities, including CT, MRI, ultrasonography, and dynamic digital radiographic images, were reviewed for lesion characteristics. Factors including lesion type, volume, morphology, location, and contrast opacification patterns were analyzed for their association with treatment response, defined as a >50% reduction in lesion size and symptom improvement. Univariable and multivariable logistic regression analyses were performed. Results: Of 85 lesions, 45 (52.9%) responded to treatment. Univariable analysis showed that pure lymphatic malformations (OR = 6.12, p = 0.004), macrocystic components (OR = 10, p = 0.016), cavitary morphology on dynamic digital radiographic images (OR = 8.90, p < 0.001), neck location (OR = 4, p = 0.03), and deep-seated lesions (OR = 3.69, p = 0.03) were significantly associated with better outcomes. Multivariable analysis identified cavitary morphology as the strongest predictor (p = 0.04). A combination of cavitary morphology, macrocystic components, and pure LM type yielded the highest predictive accuracy (AUC = 0.80, p = 0.03). Conclusions: The presence of lymphatic channels or large cystic venous spaces—such as macrocystic features on imaging or cavitary morphology—along with neck or deep-seated lesion location, predicts a favorable response to bleomycin sclerotherapy in head and neck LVMs.