Volume 62, Issue 1 pp. 73-86
Research Article

The Potential of Model-Free Parameters Derived From IVIM in Evaluating Pathological Indicators and Long-Term Survival in NPC

Fan Yang MD

Fan Yang MD

Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

Department of Radiology, Division of Neuroimaging and Neurointervention, Stanford University, Stanford University Medical Center, Stanford, California, USA

Search for more papers by this author
Haoran Wei MD

Haoran Wei MD

Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

Search for more papers by this author
Xiaolu Li MD, PhD

Xiaolu Li MD, PhD

Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

Search for more papers by this author
Xiaoduo Yu MD, PhD

Xiaoduo Yu MD, PhD

Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

Search for more papers by this author
Lin Li MD, PhD

Lin Li MD, PhD

Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

Search for more papers by this author
Yanfeng Zhao MD, PhD

Yanfeng Zhao MD, PhD

Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

Search for more papers by this author
Lizhi Xie PhD

Lizhi Xie PhD

MR Research China, GE Healthcare, Beijing, China

Search for more papers by this author
Yueluan Jiang PhD

Yueluan Jiang PhD

MR Research Collaboration, Siemens Healthineers, Beijing, China

Search for more papers by this author
Meng Lin MD, PhD

Corresponding Author

Meng Lin MD, PhD

Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

Address reprint requests to: M.L., #17 Panjiayuan Nanli, Chaoyang District, Beijing 100021, China. E-mail: [email protected], or H.Z., #17 Panjiayuan Nanli, Chaoyang District, Beijing 100021, China. E-mail: [email protected]

Search for more papers by this author
Hongmei Zhang MD, PhD

Corresponding Author

Hongmei Zhang MD, PhD

Department of Radiology, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China

Address reprint requests to: M.L., #17 Panjiayuan Nanli, Chaoyang District, Beijing 100021, China. E-mail: [email protected], or H.Z., #17 Panjiayuan Nanli, Chaoyang District, Beijing 100021, China. E-mail: [email protected]

Search for more papers by this author
First published: 10 March 2025
Citations: 2

Fan Yang and Haoran Wei contributed equally to this work.

Abstract

Background

IVIM is a useful quantitative tool for predicting prognosis, but it is labor-intensive. Simplified b-value settings and post-processing could be more practicable for clinical applications.

Purpose

To assess the value of model-free parameters (virtual MR elastography [vMRE] and signature-index [S-index]) derived from IVIM in evaluating pathological indicators and long-term survival in nasopharyngeal carcinoma (NPC), and to compare those with model-based parameters.

Study Type

Prospective.

Population

One hundred six patients (median: 49.5 years, 19.8% females).

Field Strength/Sequence

3-T, IVIM, T1- and T2-weighted imaging, contrast-enhanced T1-weighted imaging.

Assessment

The volumes of primary tumors were separately delineated. vMRE and S-index were calculated based on two b-values (200 and 1500 s/mm2), and correlations with pathological indicators (histology, epidermal growth factor receptor [EGFR] level) were analyzed. Six types of model-based parameters were constructed using all b-values. The value in predicting overall survival (OS) and progression-free survival (PFS) was calculated. The vMRE/S-index-Clinic model and tumor-node-metastasis (TNM) staging model were constructed.

Statistical Test

Spearman and intraclass correlation analysis (ICC), univariate and multivariate Cox analyses with seven clinical factors. Significance level: P-value < 0.05.

Results

The median follow-up duration was 63.0 (interquartile range: 55.9, 100.3) months. All parameters showed excellent agreement (all ICC ≥ 0.790). The vMRE/S-index was significantly correlated with EGFR level and histology. Moreover, the vMRE/S-index was an independent predictor of PFS (hazard ratio [HR]: 0.639 and 2.365, respectively) and OS (HR: 0.690 and 2.116, respectively). The performance of the vMRE/S-index (C-index: 0.585 ~ 0.631) was comparable to that of the model-based parameters (C-index: 0.551 ~ 0.630). The vMRE-Clinic model (C-index: 0.700 and 0.702 for PFS and OS, respectively) and the S-index-Clinic model (C-index: 0.703 and 0.709 for PFS and OS, respectively) demonstrated superior performance compared to TNM staging model.

Conclusion

Model-free parameters from IVIM can predict long-term survival in NPC.

Plain Language Summary

This study employed two simplified model-free parameters: the virtual MR Elastography (vMRE) to measure tissue stiffness and the signature index (S-index) to reveal the intricate tissues microstructures. These two parameters showed the comparable value in EGFR expression level and histology diagnosis as well as long-term survival prediction in nasopharyngeal carcinoma. Patients with high vMRE and low S-index value have favorable treatment outcome. Taking advantage of shorter MRI scan time and post-processing difficulty, the two model-free parameters showed good prospects in clinical application.

Evidence Level

2

Technical Efficacy

Stage 2

The full text of this article hosted at iucr.org is unavailable due to technical difficulties.