Tumor habitat-based MRI features assessing early response in locally advanced nasopharyngeal carcinoma

Jinling Yuan ¹, Mengxing Wu ¹, Lei Qiu ¹

⁰Department of Radiation Oncology, The First Affiliated Hospital of Nanjing Medical University, Nanjing 210029, Jiangsu, China; The First School of Clinical Medicine, Nanjing Medical University, Nanjing 210029, Jiangsu, China.

Oral Oncology +

|

August 16, 2024

View abstract on PubMed

Summary

Predicting early treatment response in locally advanced nasopharyngeal carcinoma (LA-NPC) is crucial. A combined model using Intratumor Heterogeneity (ITH) score and clinical factors effectively predicts early response to chemoradiotherapy.

Area Of Science

Oncology
Radiology
Medical Imaging

Background

Early response to chemoradiotherapy in locally advanced nasopharyngeal carcinoma (LA-NPC) significantly impacts prognosis.
Predictive models are needed to assess treatment response early in LA-NPC patients.

Purpose Of The Study

To develop and validate a combined model for predicting early treatment response in LA-NPC.
To integrate sub-regional radiomics features from MRI with clinical factors and Intratumor Heterogeneity (ITH) score.

Main Methods

104 LA-NPC patients were divided into training and testing cohorts.
Radiomic features were extracted using K-means clustering and selected via LASSO regression.
Four models were compared: radiomics, clinical, ITH score-based, and a combined ITH score-clinical factor model.

Main Results

The combined model (ITH score + clinical factors) showed the highest predictive performance in the test cohort (AUC=0.838).
Models incorporating the ITH score demonstrated superior prognostic value in both training (AUC=0.888) and test (AUC=0.833) cohorts.

Conclusions

A combined model integrating the Intratumor Heterogeneity (ITH) score and clinical factors is superior for predicting early response to chemoradiotherapy in LA-NPC.
The ITH score is a valuable predictor of early treatment response in LA-NPC.

Keywords:

Early response Habitat analysis Locally advanced nasopharyngeal carcinoma Prognostic model Radiomics