Quantitative MRI-based Radiomic Analysis of Percutaneous Sclerotherapy Outcomes in Peripheral Extremity Venous Malformations

Wednesday, March 8, 2023

3:00 PM – 3:09 PM

Location: Phoenix Convention Center, 225AB

Presenting Author(s)

Adham Khalil, MD

Postdoctoral Research Fellow
Johns Hopkins University, Department of Radiology

Disclosure(s): No relevant disclosure to display

Author/Co-author(s)

Amanda R. Laguna, BS

Medical Student
Warren Alpert Medical School of Brown University
TM

Tej I. Mehta, MD

Interventional Radiology Resident
Johns Hopkins Hospital
PG

Prateek C. Gowda, BS

Medical Student
Johns Hopkins University
Anna J. Gong, BA

Medical Student
Johns Hopkins University
Tushar Garg, MD (he/him/his)

Postdoctoral Research Fellow
Johns Hopkins University School of Medicine
Robert M. Weinstein, BS

Medical Student
The Johns Hopkins University
Natalie Y. Ring, MD

Resident
Johns Hopkins University
Ryan W. England, MD

Resident
Johns Hopkins University
David Gullotti, MD

Resident
Johns Hopkins University
CJ

Craig K. Jones, PhD

Assistant Professor
Johns Hopkins University
CW

Clifford Weiss, MD, FSIR, FCIRSE

Professor of Radiology and Biomedical Engineering
Johns Hopkins Univeristy School of Medicine

Disclosure(s):

Adham Khalil, MD: No relevant disclosure to display

Amanda R. Laguna, BS: No financial relationships to disclose

Tej I. Mehta, MD: No financial relationships to disclose

Prateek C. Gowda, BS: No financial relationships to disclose

Ryan W. England, MD: No financial relationships to disclose

Clifford Weiss, MD, FSIR, FCIRSE: Boston Scientific: Consultant (Ongoing), Intellectual Property Rights (Ongoing), Research Grant Recipient (Ongoing); Guerbet: Research Grant Recipient (Ongoing); Medtronic: Consultant (Ongoing), Research Grant Recipient (Ongoing); Siemens Healthineers: Consultant (Ongoing), Research Grant Recipient (Ongoing)

Purpose: To study the changes in MRI radiomic features occurring in peripheral venous malformations (VMs) after treatment using percutaneous sclerotherapy.

Materials and Methods: Our vascular anomalies database was searched for patients with peripheral extremity VMs who were treated by image-guided percutaneous sclerotherapy and had completed their treatment plan, and clinical and imaging follow-up between 2005-2022. VMs were manually segmented on pre- and post-treatment T2-weighted (T2-WI) MRI using 3D Slicer software to assess the changes in lesion volume and signal intensity (SI). To account for signal parameter differences between pre- and post-treatment T2-WI, the post-treatment T2-WI MRI signal was re-scaled to the pre-treatment T2-WI using a simple histogram matching algorithm, which allowed for precise and true calculation of SI change after sclerotherapy. Therapeutic response was categorized as 0 = worse or unchanged, and 1 = improvement based on clinical evolution after treatment. Clinical outcome assessment was also scored on a 7-point scale, ranging from -3 (worst) to +3 (maximum improvement), based on patient’s perception of symptom improvement. Spearman's rank correlation coefficient (ρ) and Paired t-test (t) were used for statistical analysis.

Results: Eighty-one patients (mean age: 20±14 years; 47 females) with peripheral upper (23 lesions) and lower (58) extremity VMs underwent 125 percutaneous sclerotherapy treatments (range: 1-6). Different sclerosants were used: alcohol (52 sessions), bleomycin (38), and sotradecol (35). Most patients (77) reported clinical improvement following sclerotherapy, including mild (8 patients), moderate (22), and significant (47) improvement. The mean change in lesion volume was -7.9±24.6 cm3 (P=.005) and in mean SI was -123.1±162.9 (P< .001). Overall, there was a significant correlation between the change in lesion volume and treatment response (ρ=-.3, P=.004). On subgroup analysis, SI change correlated with clinical outcomes of patients only receiving one treatment (n=51; ρ=-.3, P=.01), and lesions treated with bleomycin (n=22; ρ=-.4, P=.04). While lesion volume change correlated with clinical outcomes of pediatric patients (n=50; ρ=-.3, P=.03), lesions treated with sotradecol (n=17; ρ=-.5, P=.02), and lesions located in the foot (n=10; ρ=-.6, P=.04).

Conclusion:
MRI radiomic features (such as changes in lesion volume and signal intensity) correlate with the clinical outcomes for peripheral venous malformations treated using percutaneous sclerotherapy.