ABSTRACT
Malignant peripheral nerve sheath tumors (MPNSTs) are aggressive soft-tissue tumors prevalent in neurofibromatosis type 1 (NF1) patients, posing a significant risk of metastasis and recurrence. Current magnetic resonance imaging (MRI) imaging lacks decisiveness in distinguishing benign peripheral nerve sheath tumors (BPNSTs) and MPNSTs, necessitating invasive biopsies. This study aims to develop a radiomics model using quantitative imaging features and machine learning to distinguish MPNSTs from BPNSTs. Clinical data and MRIs from MPNST and BPNST patients (2000-2019) were collected at a tertiary sarcoma referral center. Lesions were manually and semi-automatically segmented on MRI scans, and radiomics features were extracted using the Workflow for Optimal Radiomics Classification (WORC) algorithm, employing automated machine learning. The evaluation was conducted using a 100× random-split cross-validation. A total of 35 MPNSTs and 74 BPNSTs were included. The T1-weighted (T1w) MRI radiomics model outperformed others with an area under the curve (AUC) of 0.71. The incorporation of additional MRI scans did not enhance performance. Combining T1w MRI with clinical features achieved an AUC of 0.74. Experienced radiologists achieved AUCs of 0.75 and 0.66, respectively. Radiomics based on T1w MRI scans and clinical features show some ability to distinguish MPNSTs from BPNSTs, potentially aiding in the management of these tumors.
ABSTRACT
BACKGROUND: Malignant peripheral nerve sheath tumours (MPNSTs) have high local recurrence (LR) rates. Literature varies on LR risk factors and treatment. This study aimed to elucidate treatment options and risk factors for first and second LRs (LR1 and LR2) in a large multicentre cohort. METHOD: Surgically treated primary MPNSTs between 1988 and 2019 in the MONACO multicentre cohort were included. Cox regression analysed LR1 and LR2 risk factors and overall survival (OS) after LR1. Treatment of LR1 and LR2 was evaluated. RESULTS: Among 507 patients, 28% developed LR1. Median follow-up was 66.9 months, and for survivors 111.1 months. Independent LR1 risk factors included high-grade tumours (HR 2.63; 95% c.i. 1.15 to 5.99), microscopically positive margins (HR 2.19; 95% c.i. 1.51 to 3.16) and large tumour size (HR 2.14; 95% c.i. 1.21 to 3.78). Perioperative radiotherapy (HR 0.62; 95% c.i. 0.43 to 0.89) reduced the risk. LR1 patients had poorer OS. Synchronous metastasis worsened OS (HR 1.79; 95% c.i. 1.02 to 3.14) post-LR1, while surgically treated LR was associated with better OS (HR 0.38; 95% c.i. 0.22 to 0.64) compared to non-surgical cases. Two-year survival after surgical treatment was 71% (95% c.i. 63 to 82%) versus 28% (95% c.i. 18 to 44%) for non-surgical LR1 patients. Most LR1 (75.4%) and LR2 (73.7%) patients received curative-intent treatment, often surgery alone (64.9% versus 47.4%). Radiotherapy combined with surgery was given to 11.3% of LR1 and 7.9% of LR2 patients. CONCLUSION: Large, high-grade MPNSTs with R1 resections are at higher LR1 risk, potentially reduced by radiotherapy. Surgically treated recurrences may provide improved survival in highly selected cases.
