Abstract
PurposeTo explore the value of MRI-based radiomics features in predicting risk in disease progression for nasopharyngeal carcinoma (NPC).Methods199 patients confirmed with NPC were retrospectively included and then divided into training and validation set using a hold-out validation (159: 40). Discriminative radiomic features were selected with a Wilcoxon signed-rank test from tumors and normal masticatory muscles of 37 NPC patients. LASSO Cox regression and Pearson correlation analysis were applied to further confirm the differential expression of the radiomic features in the training set. Using the multiple Cox regression model, we built a radiomic feature-based classifier, Rad-Score. The prognostic and predictive performance of Rad-Score was validated in the validation cohort and illustrated in all included 199 patients.ResultsWe identified 1832 differentially expressed radiomic features between tumors and normal tissue. Rad-Score was built based on one radiomic feature: CET1-w_wavelet.LLH_GLDM_Dependence-Entropy. Rad-Score showed a satisfactory performance to predict disease progression in NPC with an area under the curve (AUC) of 0.604, 0.732, 0.626 in the training, validation, and the combined cohort (all 199 patients included) respectively. Rad-Score improved risk stratification, and disease progression-free survival was significantly different between these groups in every cohort of patients (p = 0.044 or p < 0.01). Combining radiomics and clinical features, higher AUC was achieved of the prediction of 3-year disease progression-free survival (PFS) (AUC, 0.78) and 5-year disease PFS (AUC, 0.73), although there was no statistical difference.ConclusionThe radiomics classifier, Rad-Score, was proven useful for pretreatment prognosis prediction and showed potential in risk stratification for NPC.
Highlights
Nasopharyngeal carcinoma (NPC) is the most common tumor occurring in the nasopharynx
Among all 2818 radiomic features extracted from T2-weighted fat suppression images (T2WI/FS) and CET1WI images, we identified 1846 radiomics features including features differentially expressed between tumors and normal tissue and shape-related features
We developed and validated a novel prognostic tool, Rad-Score, based on one radiomic feature to improve the prediction of disease progression for patients with NPC after pathological complete remission
Summary
Nasopharyngeal carcinoma (NPC) is the most common tumor occurring in the nasopharynx. With superb soft-tissue contrast and no radiation, MRI has been integrated into the whole workflow of NPC management, including lesion detection and diagnosis, TNM staging, radiotherapy guiding, treatment response evaluation, etc. With the advent of a new age of big data mining, radiomics emerges as a brand-new way to non-invasively and quantitatively translate lesion spatial heterogeneity into high-dimensional image features that used to be nearly impossible to be quantified by our bare eyes and could provide additional valuable evidence to improve decision-making in cancer management [4]. Several studies have attempted to investigate the potential clinical relevance of radiomics features and their value in predicting treatment response, clinical staging, survival and prognosis, prediction in NPC [5, 6].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
