Back to 2026 Abstracts
MRI-Based Diagnostic Model Using Tumor Features for Peripheral Nerve Sheath Tumors: Enhancing Differentiation and Guiding Clinical Management
Christianne Y.M.N. Jansma, BSc
1, Walter Taal, MD, PhD
2, David Hanff, MD
3, Galied Muradin, MD
4, Dirk J Grunhagen, MD, PhD
5, David van Klaveren, PhD
4, Henk Coert, MD
6, Cornelis Verhoef, MD, PhD
2; Enrico Martin, MD, PhD
7(1)University Medical Center Utrecht, Utrecht, Utrecht, Netherlands, (2)Erasmus Medical Center, Rotterdam, Zuid-Holland, Netherlands, (3)Erasmus MC, Rotterdam, Zuid-Holland, Netherlands, (4)Erasmus MC Cancer Institute, Rotterdam, Zuid-Holland, Netherlands, (5)Erasmus Medical Center, Rotterdam, Netherlands, (6)University Medical Center Utrecht, Utrecht, Netherlands, (7)UMC Utrecht, Utrecht, Utrecht, Netherlands
Background: Peripheral nerve sheath tumors (PNSTs) are classified as either benign (BPNSTs) or malignant (MPNSTs). Accurate differentiation is crucial, especially in patients with neurofibromatosis type 1, who frequently develop multiple benign tumors but have increased risk of malignant transformation. Conventional MRI is the gold standard for soft tissue tumor evaluation, but its ability to reliably distinguish MPNSTs from BPNSTs remains limited. Because of this diagnostic challenge, patients often undergo invasive biopsies, which can lead to complications such as pain and nerve damage. Furthermore, repeated imaging is frequently required, contributing to patient burden. Currently, no standardized MRI-based diagnostic model exists, often leading to reliance on invasive biopsies. This study aims to develop an MRI-based predictive model and diagnostic scoring system to differentiate MPNSTs from BPNSTs, reducing the need for invasive biopsies and improving clinical decision-making.
Material and Methods: A retrospective analysis was conducted on 101 histopathologically (72 BPNST, 29 MPNST) confirmed PNSTs from 79 patients treated at Erasmus Medical Center between 2000 and 2019. Key MRI-features, including tumor size, shape, margin definition, necrosis, and signal characteristics, were analyzed. Logistic regression was used to construct a predictive model, and a diagnostic scoring system was developed based on significant MRI-features.
Results: MPNSTs were significantly associated with irregular shape (OR = 7.98, p = 0.001), ill-defined margins (OR = 7.59, p = 0.036), and necrosis (OR = 7.66, p = 0.004), whereas the presence of a reticular shape was negatively correlated with malignancy. The predictive model achieved an area under the curve (AUC) of 0.84, with a specificity of 98.6% when three or more key MRI features were present. The diagnostic scoring system provided a structured approach to assess malignancy probability, aiding further diagnostic and treatment decision-making.
Conclusion: This MRI-based predictive model provides a method for differentiating MPNSTs from BPNSTs with high accuracy. Implementing this model in clinical practice may help reduce unnecessary biopsies and guide optimal patient management. Further validation is required to confirm its clinical utility.
Back to 2026 Abstracts
