Abstract
Nasopharyngeal Carcinoma (NC) refers to the malignant tumor that occurs at the top and side walls of the nasopharyngeal cavity. The NC incidence rate always dominates the first among the malignant tumors of the ear, nose and throat, and mainly occurs in Asia. NC cases are mainly concentrated in southern provinces in China, with about 4 million existing NC. With the pollution of environment and pickled diet, and the increase of life pressure, the domestic NC incidence rate has reached 4.5-6.5/100000 and is increasing year by year. It was reported that the known main causes of NC include hereditary factor, genetic mutations, and EB virus infection, common clinical symptoms of NC include nasal congestion, bloody mucus, etc. About 90% of NC is highly sensitive to radiotherapy which is regard as the preferred treatment method; However, for NC with lower differentiation, larger volume, and recurrence after treatment, surgical resection and local protons and heavy ions therapy are also indispensable means. According to reports, the subtle heterogeneity and diversity exists in some NC, with about 80% of NC undergone radiotherapy and about 25% experienced recurrence and death within five years after radiotherapy in China. Therefore, screening the NC population with suspected recurrence after concurrent chemoradiotherapy may improve survival rates in current clinical decision-making.
Keywords: magnetic resonance imaging, delta radiomics, nasopharyngeal carcinoma, nomogram
Plain language summary
NC is one of the prevalent malignancies of the head and neck region with poor prognosis. The aim of this study is to establish a predictive model for assessing NC prognosis using clinical and MR radiomics data.
NC, also known as nasopharyngeal carcinoma, is a type of cancer that develops in the upper and lateral regions of the nasopharyngeal cavity.1-4 The NC incidence rate always dominates the first among the malignant tumors of the ear, nose and throat, and mainly occurs in Asia. NC cases are mainly concentrated in southern provinces in China, with about 4 million existing NC. With the pollution of environment and pickled diet, and the increase of life pressure, the domestic NC incidence rate has reached 4.5-6.5/100000 and is increasing year by year. It was reported that the known main causes of NC include hereditary factor, genetic mutations, and EB virus infection, frequent clinical signs of NC comprise stuffy nose, bloody nasal discharge, sore throat, etc. Approximately 90% of patients with NC exhibit remarkable responsiveness to radiation therapy which is regard as the preferred treatment method; However, for NC with lower differentiation, larger volume, and recurrence after treatment, surgical resection and local protons and heavy ions therapy are also indispensable means. According to reports, the subtle heterogeneity and diversity exists in some NC, with about 80% of NC undergone radiotherapy and about 25% experienced recurrence and death within five years after radiotherapy in China. Therefore, screening the NC population with suspected recurrence after concurrent chemoradiotherapy may improve survival rates in current clinical decision-making.1-4 At present, AJCC guidelines recommend MR as the preferred screening method before NC surgery and after chemoradiotherapy. However, there may be minor defects or missed diagnosis in the interpretation of MR by human naked eyes, and it may also miss out on suspicious subtle signs of MR that cannot be distinguished by the eyes, such as slight thickening and uneven signal of nasopharyngeal soft tissue,etc. As a result, our team has introduced a relatively new method called Delta radiomics in predicting the recurrence or metastasis of NC, leading to a substantial enhancement in performance of the prediction model and NC poor prognosis screening through time-region-related radiomics texture difference analysis.5-7 We have reviewed the previous references in PUBMED over past 30 years, there have been no topic reports on related research content. In this study, we divided 215 enrolled NC patients as a recurrence or metastasis group (n = 79) and a normal NC cases group (n = 136) via inclusion, exclusion criteria and 5-10 years follow-up, the clinical and enhanced MR data of the above NC cases were collected for retrospective study. And then all NCs were divided into training set and testing set by a time cutoff and 7/3 ratio. Then, three prediction models were established in the training set by incorporating NC clinicopathological data and Delta enhanced MR Radiomics scores. The combined model demonstrated superior predictive ability [AUC:0.880, 95%CI(.818-.927)] compared to both the clinical data model [AUC:0.660, 95%CI(.579-.735);P < .05] and the radiomics model [AUC:0.842, 95%CI(.774-.896); P = .08]. The combined model had higher clinical net benefits was also confirmed by decision curve, and next, these same findings were confirmed in the validation set{combined model [AUC:0.848, 95%CI(.735-.926)] excelled over the clinical data model [AUC:0.702, 95%CI(.574-.811),P = .02] and radiomics model [AUC:0.786, 95%CI(.664-.878)],P = .09}. Finally, the combined nomogram constructed using the combined model also received better clinical use, providing reliable information support for NC clinical treatment decision making and improving NC survival rates (Figures 1-6 & Tables 1-3).8-11
Figure 2.
The simplified schematic diagram of NC cases enrollment and grouping in this study.
Figure 3.
The flow-process diagram of delta enhanced MR radiomic analysis of NC lesion in this study. The depiction on the left illustrated the NC lesion’s outline, the delineation of the volume of interest, and the three-dimensional reconstruction; the right figure represented the extraction of NC radiomic texture features and generation of radscore using the method of Least absolute shrinkage and selection operator (LASSO) with 10-fold cross-validation. Notes: Delta radiomic features = enhanced MRI radiomic feature value (after concurrent chemoradiotherapy - before concurrent chemoradiotherapy), and then generate delta radscore.
Figure 4.
The Delong test curve confirmed that the combined model (C) outperformed both the clinical data model (A) and the radiomics model (B) in terms of predictive performance in both the training (left) and test sets (right). This suggests that integrating all available data sources can lead to more accurate predictions compared to using only clinical or radiomics data alone.
Figure 5.
The decision curve confirmed that the combined model (C) demonstrated high predictive accuracy and clinical net benefit in both the training (on the left) and test sets (on the right).
Table 2.
The Table Outcomes From Radiomic Univariate and Multivariate Logistic Analysis Influencing the Recurrence or Metastasis of NC(Radiomics Model), *P < .05.
| Radiomics Model | Univariate Analysis | Multivariate Analysis | ||
|---|---|---|---|---|
| Factors | P | Hazard Ratio | P | Hazard Ratio |
| Delta radscore | <.05* | 2.85 (1.99-4.06) | <.05* | 2.85 (1.99-4.06) |
Figure 1.
Our team searched 100 highly cited references on NC from 1994 to 2024. From the major content of the search, NC has always been a hot topic of research, but most of it focuses on the molecular mechanisms of tumor metastasis and subsequent treatment management of NC, while there is little research on multimodal radiomics for predicting the recurrence or metastasis of NC. The novelty of this paper is to build an efficient prediction model using the magnetic resonance imaging delta radiomics differences before and after chemoradiotherapy by incorporating clinicopathological data, which has rarely been extensively explored in prior research.
Figure 6.
The novel nomogram developed using the screened independent risk factors received better clinical evaluation. We selected a case and scored it according to the nomogram, and compared it with the actual results, explaining the better use and value of the nomogram (Patient 52: Female (nomogram score 13,blue arrow), Delta-Radscore 6 (nomogram score 75,green arrow), a total of 88 nomogram score (red arrow). The nomogram indicates a 90% risk of recurrence, and the recurrence was confirmed by follow up) (left.nomogram, right. calibration curve).
Table 1.
The Table Outcomes From Clinical Univariate and Multivariate Logistic Analysis Influencing the Recurrence or Metastasis of NC(Clinical Data Model), *P < .05.
| Clinical Data Model | Univariate Analysis | Multivariate Analysis | ||
|---|---|---|---|---|
| Factors | P | Hazard Ratio | P | Hazard Ratio |
| Age | .13 | 1.04 (.98-1.09) | ||
| Gender | .03* | 2.20 (1.09-4.41) | .03* | 2.34 (1.10-4.96) |
| BMI | .87 | 1.01 (.93-1.09) | ||
| Smoking history | .08 | 1.04 (.99-1.10) | ||
| Alcohol use | .21 | .97 (.93-1.02) | ||
| Hypertension history | .85 | 1.01 (.96-1.05) | ||
| Diabetes history | .45 | 1.02 (.96-1.09) | ||
| Tumor volume | .66 | .99 (.97-1.02) | ||
| Peripheral lymph node metastasis | .71 | .95 (.72-1.25) | ||
| NC pathological differentiation | .03* | 1.91 (1.09-3.37) | ||
| Pathology type | .04* | 1.77 (1.01-3.13) | ||
| Clinical stage | .56 | 1.18 (.68-2.05) | ||
| MR enhanced pattern | .03* | 1.67 (1.04-2.70) | ||
| NLR | .04* | 1.06 (1.01-1.12) | ||
| WBC | .51 | .93 (.75-1.15) | ||
| PLT | .33 | 1.01 (.99-1.02) | ||
| PLR | .66 | 1.01 (.96-1.07) | ||
Table 3.
The Table Outcomes From Clinical-Radiomic Univariate and Multivariate Logistic Analysis Influencing the Recurrence or Metastasis of NC(Combined Model), *P < .05.
| Combined Model | Univariate Analysis | Multivariate Analysis | ||
|---|---|---|---|---|
| Factors | P | Hazard Ratio | P | Hazard Ratio |
| Gender | .03* | 2.20 (1.09-4.41) | .01* | 3.49 (1.33-9.21) |
| NLR | .04* | 1.06 (1.01-1.12) | ||
| Pathology type | .04* | 1.77 (1.01-3.13) | ||
| NC pathological differentiation | .03* | 1.91 (1.09-3.37) | ||
| MR enhanced pattern | .03* | 1.67 (1.04-2.70) | ||
| Delta radscore | <.05* | 2.85 (1.99-4.06) | <.05* | 3.19 (2.14-4.76) |
This is a single-center, non-prospective and small-sample study, and the results may be subject to selection bias. More clinical and imaging quantitative data are required to additionally verify the model’s efficacy. Furthermore, this paper did not include classification screening using multiple machine learning methods and semi-automatic lesion image delineation, which may also affect the study results.
In conclusion, the combined nomogram established derived from Delta MR radiomics shows significant clinical utility in the prediction of NC recurrence and metastasis, and it is of significance for subsequent clinical decision-making changes and improving survival.
Acknowledgments
The authors gratefully acknowledge J.H. An and Y.Z. An for assistance with translating references and Applied improved statistical technology.
Footnotes
Author Contributions: All authors conceived and drafted the manuscript, contributed to the literature review and are responsible for collecting and collating clinical data, revised the manuscript critically for important intellectual content. Dr An was responsible for the quality control of article statistics. Dr An, and Dr Ji approved the final version to be published and agreed to act as guarantors of the work. Dr Yu Shang, Junjie Liu, and Lina Song contributed equally to this work.
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The “323” Public Health Project of the Hubei health commission and the Xiangyang No.1 People’s Hospital (XYY2022-323). Graduate Education Research Project Fund of Hubei University of Medicine (YJ2022013), Xiangyang Science and Technology Plan Key Project Fund (2022YL34A). The Natural Science Foundation of Hubei Province [grant numbers 2022BCE021 and 2022CFD010]. Innovative Research Program of Xiangyang No.1 People’s Hospital (XYY2023QT04).
Ethical Statement
Ethical Approval
This study ethics has been approved by the Human Ethics Committee of Xiangyang No.1 People’s Hospital Ethics Committee (approval No.: XYYYE20230090). Our present medical research was conducted according to the principles expressed in the Declaration of Helsinki.
Consent to Participate
We also obtained, written consent from the patient’s family.
Consent for Publication
Written informed consent was obtained from the individuals or patient’s family.
ORCID iD
Data Availability Statement
All relevant data supporting the conclusions of this article are included within the article.*
References
- 1.Lin DF, Li HL, Liu T, et al. Radiomic signatures associated with tumor immune heterogeneity predict survival in locally recurrent nasopharyngeal carcinoma. J Natl Cancer Inst. 2024;19:djae081. doi: 10.1093/jnci/djae081 [DOI] [PubMed] [Google Scholar]
- 2.Chen Z, Wang Z, Liu S, et al. Nomograms based on multiparametric MRI radiomics integrated with clinical-radiological features for predicting the response to induction chemotherapy in nasopharyngeal carcinoma. Eur J Radiol. 2024;175:111438. doi: 10.1016/j.ejrad.2024.111438 [DOI] [PubMed] [Google Scholar]
- 3.Cao X, Wang X, Song J, Su Y, Wang L, Yin Y. Pretreatment multiparametric MRI radiomics-integrated clinical hematological biomarkers can predict early rapid metastasis in patients with nasopharyngeal carcinoma. BMC Cancer. 2024;24(1):435. doi: 10.1186/s12885-024-12209-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Wang SX, Yang Y, Xie H, et al. Radiomics-based nomogram guides adaptive de-intensification in locoregionally advanced nasopharyngeal carcinoma following induction chemotherapy. Eur Radiol. 2024;22:10678. doi: 10.1007/s00330-024-10678-8 [DOI] [PubMed] [Google Scholar]
- 5.Wang M, Xi Y, Wang L, Chen H, Jiang F, Ding Z. Predictive value of delta radiomics in xerostomia after chemoradiotherapy in patients with stage III-IV nasopharyngeal carcinoma. Radiat Oncol. 2024;19(1):26. doi: 10.1186/s13014-024-02417-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cheng J, Su W, Wang Y, et al. Magnetic resonance imaging based on radiomics for differentiating T1-category nasopharyngeal carcinoma from nasopharyngeal lymphoid hyperplasia: a multicenter study. Jpn J Radiol. 2024;27:01544. doi: 10.1007/s11604-024-01544-0 [DOI] [PubMed] [Google Scholar]
- 7.Khongwirotphan S, Oonsiri S, Kitpanit S, et al. Multimodality radiomics for tumor prognosis in nasopharyngeal carcinoma. PLoS One. 2024;19(2):e0298111. doi: 10.1371/journal.pone.0298111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zeng F, Ye Z, Zhou Q. CT-based peritumoral radiomics nomogram on prediction of response and survival to induction chemotherapy in locoregionally advanced nasopharyngeal carcinoma. J Cancer Res Clin Oncol. 2024;150(2):50. doi: 10.1007/s00432-023-05590-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Liu T, Dong D, Zhao X, et al. Radiomic signatures reveal multiscale intratumor heterogeneity associated with tissue tolerance and survival in re-irradiated nasopharyngeal carcinoma: a multicenter study. BMC Med. 2023;21(1):464. doi: 10.1186/s12916-023-03164-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wu Q, Chang Y, Yang C, et al. Adjuvant chemotherapy or no adjuvant chemotherapy? A prediction model for the risk stratification of recurrence or metastasis of nasopharyngeal carcinoma combining MRI radiomics with clinical factors. PLoS One. 2023;18(9):e0287031. doi: 10.1371/journal.pone.0287031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wang A, Xu H, Zhang C, Ren J, Liu J, Zhou P. Radiomic analysis of MRI for prediction of response to induction chemotherapy in nasopharyngeal carcinoma patients. Clin Radiol. 2023;78(9):e644-e653. doi: 10.1016/j.crad.2023.05.012 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All relevant data supporting the conclusions of this article are included within the article.*