Dual-energy CT for distinguishing between T3 and T4 stages of locally advanced nasopharyngeal carcinoma

Introduction

Nasopharyngeal carcinoma (NPC) is a malignant tumor originating from nasopharyngeal epithelial cells and is more common in middle-aged and elderly people than in young adults (1). In 2020, a total of 133,354 new cases and 80,008 deaths from NPC were reported worldwide, with a high incidence in Southern China (2,3). Treatment strategies for NPC includes radiation therapy such as intensity-modulated radiation therapy (IMRT), chemotherapy and surgery (4,5). Most cases are diagnosed at an advanced stage (stage III and IV), missing the early disease (stage I) which may have better prognosis (6).

Primary NPC with T3 or T4 stage is referred as locally advanced NPC (LA-NPC) (stage III and IVA: T_3-4N_0-3M₀ or T_1-2N_2-3M₀) regardless of the presence or absence of cervical lymph node metastasis according to the 8^th edition of the Union for International Cancer Control and the American Joint Committee on Cancer (UICC/AJCC) tumor-node-metastasis (TNM) staging system (7). A combined therapy of induction chemotherapy with concurrent radiotherapy is recommended for patients with LA-NPC (stage III and IVA). However, the efficacy of therapy and prognosis of each stage vary greatly (8), and the 5-year local recurrence-free survival of patients with T3–T4 stage NPC is decreased with the increasing T stage (9). A study by Sanford et al. showed that personalized target area delineation based on the invasion range (T stage) and invasion pathway of NPC could reduce unnecessary radiation damage and may improve treatment response in patients with NPC (10). Therefore, assessment of T stage for patients with LA-NPC can assist in treatment planning for modalities such as IMRT or current chemoradiation, which may improve prognosis.

Magnetic resonance imaging (MRI) has been a commonly used imaging tool for NPC staging (11), because of its high soft tissue resolution and high sensitivity for detecting early bone marrow infiltration of tumors in the skull base (T3 stage) (12). However, bone marrow changes seen on MRI may be non-specific as it could be challenging to differentiate between inflammation surrounding a tumor and bone marrow infiltration by a tumor, leading to false positives (13). The apparent diffusion coefficient (ADC) as a functional sequence parameter from MRI, may quantitatively reflect the intrinsic biological characteristics of tumors and help predict tumor staging and treatment response (14,15). However, ADC parameters are acquired from diffusion-weighted imaging and subsequent calculation, which could be challenging to acquire in busy clinical practice (16). Although traditional single energy computed tomography has high sensitivity in detecting sclerotic and osteolytic bone destruction, it is prone to misdiagnosis of early bone marrow invasion (17). Therefore, a new non-invasive method is needed to evaluate the T staging for patients with LA-NPC.

Dual-energy computed tomography (DECT) uses the “three substance decomposition algorithm” to quantify and distinguish substances with similar attenuation measurement values, generating specific material images, energy images, and multiple quantitative parameters (13,18). Previous studies have shown that DECT has great potential in assessing tumor invasion, lymph node metastasis, treatment efficacy, and prognosis of head and neck cancers (19-21). Several quantitative parameters of DECT have been used in differentiating between nasopharyngeal lymphoid tissue hyperplasia and T1 stage NPC lesions (22). Compared with traditional single energy CT and MRI, DECT has better diagnostic performance for skull base invasion (T3 stage) of NPC, and the DECT parameters such as the electron density related to water (Rho), effective atomic number (Zeff) value and normalized iodine content (NIC) values are helpful in detecting tumor infiltration into the skull base (13). DECT allows calculation of the relative electron density, i.e., Rho, which compares the electron density of a tissue to the electron density of water (23). DECT-derived Zeff parameter represents the average atomic number of composite materials, which is a parameter based on monochromatic data and mass attenuation coefficient, providing information on blood vessels, cell composition, and apoptosis/proliferation ratio (24). DECT-derived NIC parameter reflects the degree of tissue perfusion at a specific time point and can provide information on the microvascular environment of diseased tissue (25). However, it is not clear whether the quantitative parameters of DECT are useful for differentiating T3 from T4 stages for patients with LA-NPC prior to treatment.

Here, we conducted a prospective study on patients with LA-NPC, and assessed DECT parameters for distinguishing between T3 and T4 stages. In addition, we developed a combined model based on the independent variables from DECT parameters to predict T staging in patients with LA-NPC. We aimed to evaluate the quantitative parameters of DECT for detecting skull base invasion and for TNM staging. We hypothesized that DECT may perform better in T3/T4 staging than MRI. The results from this study should be able to enhance our understanding of LA-NPC and facilitate treatment planning. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-124/rc).

Methods

Participants

A total of 210 consecutive patients with newly diagnosed histopathologically confirmed NPC (all stages) between October 2021 and December 2023 were prospectively enrolled into this study. 136 out of the 210 patients had LA-NPC, and underwent both MRI and DECT scans (with a maximum of 7 days between the two examinations) 1–3 days prior to treatment. One hundred and thirteen out of the 210 patients had their DECT quantitative parameter data available for final analysis. Details of the enrollment process including exclusion criteria and the reasons for exclusion are presented in Figure 1.

Figure 1 Enrollment flowchart of the study cohort. DECT, dual-energy computed tomography; MRI, magnetic resonance imaging; NPC, nasopharyngeal carcinoma.

The TNM staging for the study patients with NPC was determined by the clinical team through comprehensive assessment. This included various imaging examinations (pre-treatment simulated CT for localization, MRI, ultrasound, whole body bone scan, and positron emission tomography/computed tomography scan, etc.), clinical symptom evaluation, laboratory tests, and pathological biopsy results. Clinical team used the 8th edition of the UICC/AJCC TNM staging system to determine the clinical TNM staging (7). A total of 136 LA-NPC patients (T3 stage: 60 cases, T4 stage: 76 cases) underwent MRI and DECT examinations before treatment in this study.

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the First Affiliated Hospital of Guangxi Medical University (No. 2022-KT hengxiang-026) and informed consent was obtained from all study participants.

MRI acquisition

MRI examinations were performed on a 1.5T MR imaging scanner (Signa 1.5T, GE Healthcare). Routine standardized neck MR scanning was performed, including T1-weighted images (T1WI), T1-weighted with gadolinium contrast agent (T1WI + C), T2-weighted images (T2WI), and T2-weighted with fluid-attenuated inversion recovery (T2-FLAIR). The parameters for T1WI were as follows: repetition time (TR) =600 ms, echo time (TE) =15 ms, matrix =220×182, and field-of-view (FOV) =220×220 mm. The parameters for T2WI-FS were the following: TR =1,140 ms, TE =60 ms, matrix =180×150, and FOV =200×200 mm. The parameters for T1WI+C were as followings: TR =372 ms, TE =12 ms, matrix =220×174, FOV =220×220 mm, slice thickness =5 mm, and interlayer spacing =1 mm. Gadopentetate dimeglumine (Gd-DTPA) was injected at a dosage of 0.2 mmol/kg and an injection rate of 2.5 mL/s, with a high-pressure injector (26).

DECT image acquisition

All examinations were performed on a third-generation dual-source CT system (SOMATOM Force; Siemens Healthineers, Erlangen, Germany) was used to acquire the DECT images. The scan range extended from the skull base to the subclavian margin. The DECT scan parameters in the dual-source, dual-energy mode were as follows: tube A 100 kVp, tube B Sn150 kVp, 128×0.6 mm detector collimation; 0.5-sec tube rotation time; 0.6 pitch; 1.0-mm-thick sections; 1.0-mm-thick section increments. Automated tube-current modulation was performed in all studies (CAREdose4D; Siemens Healthineers). For contrast-enhanced DECT scans, non-ionic iodinated contrast medium (350 mg I/mL) at a rate of 3.0 mL/s through a pump injector was given intravenously through the antecubital vein of the patients, with the dose calculated based on the height and weight of the patient. DECT images of arterial and venous phases were acquired after 30- and 55-second of delay, respectively.

DECT image analysis

All DECT datasets were reconstructed using a strength level of 3 and the Qr40 kernel with advanced modeled iterative reconstruction (ADMIRE; Siemens Healthineers). Linear-blended images with a standard blending ratio of 0.6 (termed Mix0.6) were generated simultaneously. All images were then transferred to a dedicated dual-energy post-processing workstation (syngo.via VB20, Siemens Healthineers, Forchheim, Germany). Virtual non-contrast (VNC) images and iodine maps were generated using the VNC algorithm for three-component decomposition (iodine, soft tissue, and fat). Electron density relative to water and effective atomic number (Rho/Zeff) images, as well as virtual monoenergetic imaging (VMI) series, were derived using the Rho/Zeff and Monoenergetic+ algorithms on the same platform. To ensure representative tumor sampling while avoiding small vessels and necrotic regions, three consecutive axial slices displaying the largest cross-sectional area of the lesion were selected (27). The iodine concentration (IC) in the left carotid artery was measured on the iodine map and used as a reference. The normalized iodine concentration (NIC) was then calculated as follows: NIC =IC_lesion/IC_{left carotid artery} (28). CT values from the Mix0.6 images were also recorded. CT attenuation values at different keV levels were extracted from the spectral curve, and the spectral slope (λ_Hu) was computed using the formula: λ_Hu = (CT_{40 keV} − CT_{110 keV})/(110 − 40). Rho and Zeff values were obtained directly from the corresponding Rho/Zeff images generated by the Rho/Zeff algorithm.

The above tumor measurement was independently completed by a head and neck radiologist (Yan Wen, MD, with more than 15 years of experience in head and neck imaging). In addition, CT images of 20 patients were randomly selected and re-measured by the same head and neck radiologist within 2 weeks and the intra-observer intraclass correlation (ICC) values ranged from 0.769 to 0.910.

Assessment of DECT and MRI images

The DECT and MRI images presented in random orders were independently reviewed by two experienced radiologists specialized in head and neck imaging (reader 1: Z.Z., and reader 2: Y.Q., with 20 and 25 years of experience in head and neck imaging, respectively) and the results were recorded. The final imaging result was determined by consensus between the two head and neck radiologists. When there was a disagreement between the two head and neck radiologists, a final decision was made by a senior radiologist (L.L., with 30 years of experience in head and neck imaging). The Cohen’s kappa coefficient values between reader 1 and reader 2 for diagnosing the T3 and T4 stages of LA-NPC ranged from 0.813 to 0.931. All head and neck radiologists were unaware of the clinical information of the patients and the results from other medical imaging examinations.

Statistical analysis

Statistical analysis was performed with IBM SPSS Statistics (versions 25.0, IBM) and MedCalc Statistics (versions 20.0, IBM). First, we performed the Kolmogorov-Smirnov test to assess the normality of all continuous parameters. For data that does not follow a normal distribution, we used the Mann-Whitney U test. Otherwise, a two-independent-sample t-test was used to test for intergroup differences. We used the variance inflation factor (VIF) to measure the degree of multicollinearity between variables, where VIF >10 indicated a high correlation between variables (29). Subsequently, non-multicollinearity indicators were included in the multivariate binary logistic regression, and independent factors for distinguishing between T3 and T4 stages were obtained. After that, all independent factors were used to create a logistic regression model. Receiver operating characteristic curve (ROC) analysis was performed to obtain the area under curve (AUC) values to assess the prediction efficacy of independent variables and the combined model. The comparison of ROC curves was performed using Delong’s test. A significant difference was defined as P<0.05.

Results

Performance of qualitative diagnosis based on DECT and MRI morphology

There were 136 patients with LA-NPC (male: 104, female: 32; mean age: 47.57±11.58 years) who underwent both MRI and DECT examinations before treatment. Based on the qualitative diagnosis of morphology using MRI and DECT, there were 124 (124/136=91.18%) patients correctly diagnosed by DECT and 118 (118/136=86.76%) patients correctly diagnosed by MRI (Table 1).

The representative DECT and MRI images of a patient with T3-stage NPC are presented in Figure 2, showing skull base involvement. In the VMI, the bone density on the left sphenoid bone was significantly higher than that on the right side, indicating that the area may have been invaded by the tumor. The deposition of iodine contrast agent on the left sphenoid bone invaded by the tumor was visually displayed as pseudo-color images (red arrows) on the iodine maps and on the Rho/Zeff images of the DECT. However, no gross changes were observed in the bone marrow signal in the left sphenoid bone on the MRI images. The representative MRI images of a T4-stage NPC patient (Figure 3) showed subtle perineural spread of the tumor in the left foramen rotundum (red arrows), which also extended to the left cavernous sinus and the second branch of the trigeminal nerve (V2) that traveled through the foramen rotundum. Nevertheless, the DECT images in Figure 3 clearly showed the involvement of the tumor lesion in the left foramen rotundum and the left cranial nerve V2.

Figure 2 Representative imaging findings of DECT (A-C) and MRI (D-F) for a 39-year-old male patient with a T3 nasopharyngeal carcinoma. (A), (B) and (C) presenting stronger color gradient and bone density in the left sphenoid bone as compared to the contralateral right side (the red arrows) on DECT; (D), (E) and (F) showing no abnormal signal intensity in the left sphenoid bone (the red arrows) on MRI. (A) DECT_electron density relative to water and effective atomic number images; (B) DECT_iodine maps; (C) DECT_the virtual monochromatic images at 150 keV; (D) MRI_T1-weighted images; (E) MRI_T1-weighted axial contrast enhanced images; (F) MRI_T1-weighted coronal contrast enhanced images. DECT, dual-energy computed tomography; MRI, magnetic resonance imaging.

Figure 3 Representative imaging findings of DECT (A-C) and MRI (D-F) for a 61-year-old male patient with T4 stage of nasopharyngeal carcinoma. (A), (B) and (C) presenting tumor involvement of left foramen rotundum (the red arrows) with bone sclerosis and more contrast enchantment than that of the contralateral soft tissue on DECT. (D), (E) and (F) showing subtle contrast enhancement through left foramen rotundum (the red arrows) on MRI. (A) DECT_electron density relative to water and effective atomic number images; (B) DECT_iodine maps; (C) DECT_the virtual monochromatic images at 40 keV; (D) MRI_T2-weighted images; (E) pre-contrast MRI_T1-weighted axial images; (F) MRI_T1-weighted axial contrast enhanced images. DECT, dual-energy computed tomography; MRI, magnetic resonance imaging.

The ROC curve in Figure 4 showed that the qualitative diagnostic value of MRI [area under the curve (AUC) value 0.88, 95% confidence interval (CI): 0.81–0.93] was lower than that of DECT (AUC value 0.92, 95% CI: 0.86–0.96) (P=0.19>0.05) for T3 and T4 in patients with LA-NPC. Moreover, the sensitivity and specificity of DECT and MRI were 85.53%, 98.33%, and 77.63%, 98.33%, respectively. Therefore, DECT may be a better qualitative diagnosis than MRI when morphological methods are used to differentiate between T3 and T4 staging of LA-NPC.

Figure 4 ROC curve and the area under the curve for diagnostic performance of MRI and DECT for T3 and T4 stages of LA-NPC. AUC, area under the curve; DECT, dual-energy computed tomography; LA-NPC, locally advanced nasopharyngeal carcinoma; MRI, magnetic resonance imaging; ROC, receiver operating characteristic.

Diagnostic performance of DECT-based quantitative parameters

Table 2 shows the clinical characteristics and quantitative parameters of DECT for the 113 patients with LA-NPC. There were 46 patients with T3 stage and 67 patients with T4 stage, and the patients with the T4 stage having a higher average value of the quantitative parameters than patients with T3 stage.

The results of univariate logistic regression analysis showed statistical significance (P<0.05) in the following DECT parameters: Rho and Mix-0.6 in the arterial phase, IC, NIC, Rho, Zeff, Mix-0.6, λ_HU, 70 keV, and 60 keV in the venous phase. The results of VIF evaluation showed that only A-Rho (VIF =1.05), V-NIC (VIF =2.64), and Mix-0.6 of arterial phase (VIF =1.29) had no multicollinearity. Subsequently, multivariate logistic regression analysis was applied to quantitative parameters with a VIF <10, and the findings indicated that V-NIC and A-Rho were independent factors for differentiating T3 and T4 stages of LA-NPC (Table 3).

Table 4 shows that the AUC values, sensitivity, and specificity of A-Rho and V-NIC being 0.81 (95% CI: 0.73–0.89), 89.55%, 60.87%, and 0.64 (95% CI: 0.53–0.74), 46.27%, and 82.61%, respectively. The combined model based on V-NIC and A-Rho had an AUC value of 0.84 (95% CI: 0.76–0.91), a sensitivity of 77.61%, and a specificity of 80.43% for predicting T3 and T4 stages. The comparison of ROC curves using Delong’s test as shown in Figure 5 showed that the AUC [0.84 (0.76–0.91)] of the combined model for detecting T3 and T4 staging of LA-NPC was higher than the A-Rho [0.81 (95% CI: 0.73–0.89); P=0.25>0.05] or the V-NIC alone [0.64 (95% CI: 0.53–0.74); P<0.001]. As a result, DECT-based quantitative parameters are useful for differentiating between LA-NPC T3 and T4 staging.

Figure 5 ROC curve of DECT quantitative parameters including models by the A-Rho or the V-NIC alone and the combined model. A-Rho, electron density relative to water of the arterial phase; DECT, dual-energy computed tomography; ROC, receiver operating characteristic; V-NIC, normalized iodine concentration of the venous phase.

Discussion

In this study, we found that the DECT-based quantitative parameters appeared to perform well for distinguishing T3 and T4 stages of LA-NPC, similar to the performance of MRI. The DECT quantitative parameters such as the A-Rho and V-NIC were independent factors for predicting T3 and T4 staging, and the combined model built with these two DECT quantitative parameters showed robust performance in predicting the T staging.

This study showed the qualitative diagnostic efficacy of MRI being lower than that of DECT for detecting the T3- and T4-stages of LA-NPC, which is generally in agreement with literature. It is challenging to assess NPC as the presence of metal implants such as dental fillers may cause alterations in the static magnetic field and may lead to various artifacts on MRI imaging (30). In addition, the nasopharynx is located at the base of the skull, and the surrounding bones and air could lead to an increase in magnetic susceptibility, which in turn affects the uniformity of the magnetic field and thus resulting in poor imaging quality on MRI (31). On the other hand, DECT modality may show better bony details and may largely avoid the pitfall of MRI in skull base imaging. A study by Stolzmann et al. found that using DECT technology for imaging after dental restoration reduced metal artifacts, and using high-energy VMI significantly reduced beam hardening artifacts caused by dental restoration (32). A study by Zhan et al. indicated that the Zeff and NIC values from DECT had higher sensitivity and accuracy, which was more effective than MRI in diagnosing early skull base bone invasion. They speculated that when nasopharyngeal tumors invaded the skull bone, there would be rapid growth of tumor cells and vascular networks in the affected bone, leading to a significant increase in the expression of microvascular density and thus higher DECT parameter values for the vasculature involved (13).

We observed the DECT parameter V-NIC being useful for differentiating between T3 and T4 stages of LA-NPC. Iodine concentration (IC) is influenced by the quantity of blood vessels and has been thought to be a direct reaction to blood flow (33). A prior study of patients with gastric cancer at the T4 stage (serosa invasion) showed a significantly higher IC in the peri-gastric fat tissue when compared to patients at the T1–T3 stage (intact serosa). Tumor cell invasion or malignant tumor cell leakage into the peri-gastric fat tissue may cause vascular proliferative reaction, which in turn increases blood flow perfusion in the area invaded by tumor (34). A study of NPC showed the T1-stage NPC lesions having a higher NIC than benign nasopharyngeal hyperplasia (20). This could be partially explained by the following speculations. There may be more accumulation of iodine contrast agent in the NPC tumors than non-tumorous hyperplasia due to the tumor neovascularity and incomplete vascular endothelium of the tumor (35). In addition, higher T-stage tumor may have more tumor perfusion and thus higher NIC value from DECT. Therefore, DECT parameters for iodine contrast may help to determine the degree of tumor vascular enhancement and identify benign versus malignant tumors (36), as well as differentiating varying T-stage tumors (20,34). However, the AUC value of V-NIC at 0.64 for distinguishing between T3 and T4 stage of LA-NPC was relatively low, which may be due to the modest sample size in our study.

Our study indicated that the A-Rho parameter from DECT was an independent predictor for distinguishing between T3 and T4 stages and the A-Rho in T4 stage NPC lesions was higher than that of T3 stage NPC, which is consistent with literature. Saito et al. used DECT technology to convert CT values into relative electron density values of substances, and presented a single linear relationship between CT values and the Rho values (37). Our observation of the A-Rho value in T4 stage NPC lesions being higher than that of T3 stage was consistent with the Cheng et al.’s study on T staging of esophageal cancer. They found that the Rho in advanced esophageal cancer (T3–T4 stage) was significantly higher than that in early esophageal cancer (T1–T2 stage), which may be related to the increased content of connective tissue with higher soft tissue density in T4 stage of the primary tumor (38). Additional studies have shown that the expression of connective tissue growth factor is closely related to tumor development. When the expression level of connective tissue growth factor increases, the risk of tumor progression to higher clinical stages also increases accordingly (39). However, this finding was divergent from the study by Yu et al., who observed that the level of connective tissue growth factor in T3–T4 stage NPC was lower than that in T1–T2 stage. As the size and local spread of NPC increased (T stage increased), the expression level of connective tissue growth factor decreased (40). It has been speculated that the interactions between proteins containing the connective tissue growth factor domain had different effects on tumor development, leading to the different functions of connective tissue growth factor in tumors (41). Our study showed higher A-Rho value in T4 than in T3, which might be partly due to our cohort focusing towards higher T stages with more T4 than T3 stage cases. More studies are needed to assess the differences of the DECT parameters among the various T stages of NPC.

In this study, the combined model based on the V-NIC and the A-Rho parameters from DECT performed better than either model alone with V-NIC or with A-Rho. This may be related to the pathways of iodine contrast agent uptake by tumor lesions at different phases during the DECT scan. During CT enhanced scanning, the arterial phase enhancement of primary tumor lesions is mainly influenced by the density of neovascularization within the tumor and the degree of microvascular disorder and tortuosity, while the venous phase enhancement mainly depends on the accumulation of iodine contrast agent into the tumor interstitial space and in the extracellular space outside the blood vessels (42,43). In this study, the combined model included indicators for both arterial and venous phases, thereby more accurately depicting the distribution of iodine contrast agent within the tumor. Therefore, the diagnostic performance of the combined model based on quantitative parameters of both arterial and venous phases was enhanced.

There are several limitations to this study. First, this was a single center study with a modest sample size, which may limit general applications of the study results. An independent external cohort is needed to validate the results of this study. Second, this study focused on advanced stages of NPC and was deficient in lower T stage analysis. A comprehensive assessment of all stages of NPC would have provided more information on the DECT performance in predicting NPC staging. Third, this study did not take into consideration of clinical demographic data such as risk factors including family history of NPC, smoking and alcohol abuse, history of being infected with Epstein-Barr virus or conventional imaging data such as the primary NPC tumor size, and nodal status in the final analysis. These additional data could have improved the prediction model performance. Fourth, the diagnostic model was developed and tested in the same internal cohort, without external validation or cross-validation, which limited the model’s generalizability. Multi-center prospective studies with sufficient sample size and statistical power are needed to validate the model performance for differentiating T3 and T4 stage of LA-NPC. Fifth, the early acquisition of the venous phase DECT may result in reduced quantification of iodine uptake by tumors. Lastly, we had no follow-up data on this cohort, and no DECT post-treatment imaging to assess the longitudinal changes of the DECT parameters for patients with NPC. Prospective longitudinal studies are needed to improve the prediction model performance with machine-learning algorithm.

Conclusions

This study showed that DECT performs better diagnostically than MRI in differentiating between T3 and T4 staging in LA-NPC based on both quantitative and qualitative comparative analysis, which may serve as a potentially useful imaging tool for personalized treatment planning of patients with NPC.

Acknowledgments

We would like to thank staff members in the Departments of Radiology, Radiotherapy and Otolaryngology at all participating hospitals for their efforts in collecting the information used in this study and for their helpful discussion and assistance in the data analysis and manuscript preparation. We acknowledge the technical support of Dr. Shushen Lin from Siemens Healthineers. We thank Yan Wen, Fuling Huang, and Liling Long, from the First Affiliated Hospital of Guangxi Medical University, for their helpful suggestions for collecting the information.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-124/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-124/dss

Funding: This work was supported by the National Science Foundation of China (grant No. 82260344), Ruiying Fund of Beijing Medical Award Foundation (grant No. YXJL-2022-0105-0151), Medical Excellence Award from the Creative Research Development Grant at the First Affiliated Hospital of Guangxi Medical University, China.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-124/coif). L.C. is a current employee of Siemens Healthineers Ltd. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the First Affiliated Hospital of Guangxi Medical University (No. 2022-KT hengxiang-026) and informed consent was obtained from all study participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bai R, Sun J, Xu Y, Sun Z, Zhao X. Incidence and mortality trends of nasopharynx cancer from 1990 to 2019 in China: an age-period-cohort analysis. BMC Public Health 2022;22:1351. [Crossref] [PubMed]
2. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
3. Guo X, Cui J, Yuan X, Gao Z, Yu G, Wu H, Kou C. Long-term trends of nasopharyngeal carcinoma mortality in China from 2006 to 2020 by region and sex: an age-period-cohort analysis. BMC Public Health 2023;23:2057. [Crossref] [PubMed]
4. Chen D, Cai SB, Soon YY, Cheo T, Vellayappan B, Tan CW, Ho F. Dosimetric comparison between Intensity Modulated Radiation Therapy (IMRT) vs dual arc Volumetric Arc Therapy (VMAT) for nasopharyngeal cancer (NPC): Systematic review and meta-analysis. J Med Imaging Radiat Sci 2023;54:167-77. [Crossref] [PubMed]
5. Huang TL, Chien CY, Tsai WL, Liao KC, Chou SY, Lin HC, Dean Luo S, Lee TF, Lee CH, Fang FM. Long-term late toxicities and quality of life for survivors of nasopharyngeal carcinoma treated with intensity-modulated radiotherapy versus non-intensity-modulated radiotherapy. Head Neck 2016;38:E1026-32. [Crossref] [PubMed]
6. Zhang L, Chen QY, Liu H, Tang LQ, Mai HQ. Emerging treatment options for nasopharyngeal carcinoma. Drug Des Devel Ther 2013;7:37-52. [Crossref] [PubMed]
7. Pan XX, Tong LH, Chen YF, Li FL, Tang WB, Liu YJ, Yang W. A simplified T classification based on the 8th edition of the UICC/AJCC staging system for nasopharyngeal carcinoma. Cancer Manag Res 2019;11:3163-9.
8. Hui EP, Ma BBY, Chan ATC. The emerging data on choice of optimal therapy for locally advanced nasopharyngeal carcinoma. Curr Opin Oncol 2020;32:187-95. [Crossref] [PubMed]
9. Su SF, Han F, Zhao C, Huang Y, Chen CY, Xiao WW, Li JX, Lu TX. Treatment outcomes for different subgroups of nasopharyngeal carcinoma patients treated with intensity-modulated radiation therapy. Chin J Cancer 2011;30:565-73. [Crossref] [PubMed]
10. Sanford NN, Lau J, Lam MB, Juliano AF, Adams JA, Goldberg SI, Lu HM, Lu YC, Liebsch NJ, Curtin HD, Chan AW. Individualization of Clinical Target Volume Delineation Based on Stepwise Spread of Nasopharyngeal Carcinoma: Outcome of More Than a Decade of Clinical Experience. Int J Radiat Oncol Biol Phys 2019;103:654-68. [Crossref] [PubMed]
11. Gorolay VV, Niles NN, Huo YR, Ahmadi N, Hanneman K, Thompson E, Chan MV. MRI detection of suspected nasopharyngeal carcinoma: a systematic review and meta-analysis. Neuroradiology 2022;64:1471-81. [Crossref] [PubMed]
12. Parlak S, Yazici G, Dolgun A, Ozgen B. The evolution of bone marrow signal changes at the skull base in nasopharyngeal carcinoma patients treated with radiation therapy. Radiol Med 2021;126:818-26. [Crossref] [PubMed]
13. Zhan Y, Wang P, Wang Y, Wang Y, Tang Z. Dual-energy CT for the detection of skull base invasion in nasopharyngeal carcinoma: comparison of simulated single-energy CT and MRI. Insights Imaging 2023;14:95. [Crossref] [PubMed]
14. Lee MK, Choi Y, Jung SL. Diffusion-weighted MRI for predicting treatment response in patients with nasopharyngeal carcinoma: a systematic review and meta-analysis. Sci Rep 2021;11:18986. [Crossref] [PubMed]
15. Huang TX, Lu N, Lian SS, Li H, Yin SH, Geng ZJ, Xie CM. The primary lesion apparent diffusion coefficient is a prognostic factor for locoregionally advanced nasopharyngeal carcinoma: a retrospective study. BMC Cancer 2019;19:470. [Crossref] [PubMed]
16. Qiang M, Li C, Sun Y, Sun Y, Ke L, Xie C, et al. A Prognostic Predictive System Based on Deep Learning for Locoregionally Advanced Nasopharyngeal Carcinoma. J Natl Cancer Inst 2021;113:606-15. [Crossref] [PubMed]
17. Chen J, Qiu Z, Jiang N, Xia Y, Li H, Chen S, Liu J, Xue Y. Detecting bone marrow infiltration in nonosteolytic multiple myeloma through separation of hydroxyapatite via the two-material decomposition technique in spectral computed tomography. Quant Imaging Med Surg 2024;14:2345-56. [Crossref] [PubMed]
18. McCollough CH, Leng S, Yu L, Fletcher JG. Dual- and Multi-Energy CT: Principles, Technical Approaches, and Clinical Applications. Radiology 2015;276:637-53. [Crossref] [PubMed]
19. Park SI, Guenette JP, Suh CH, Hanna GJ, Chung SR, Baek JH, Lee JH, Choi YJ. The diagnostic performance of CT and MRI for detecting extranodal extension in patients with head and neck squamous cell carcinoma: a systematic review and diagnostic meta-analysis. Eur Radiol 2021;31:2048-61. [Crossref] [PubMed]
20. Shen H, Yuan X, Liu D, Tu C, Wang X, Liu R, Wang X, Lan X, Fu K, Zhang J. Multiparametric dual-energy CT to differentiate stage T1 nasopharyngeal carcinoma from benign hyperplasia. Quant Imaging Med Surg 2021;11:4004-15. [Crossref] [PubMed]
21. Shen H, Yuan X, Liu D, Huang Y, Wang Y, Jiang S, Zhang J. Multiparametric dual-energy CT for distinguishing nasopharyngeal carcinoma from nasopharyngeal lymphoma. Eur J Radiol 2021;136:109532. [Crossref] [PubMed]
22. Wang P, Xiao Z, Tang Z, Wang J. Dual-energy CT in the differentiation of stage T1 nasopharyngeal carcinoma and lymphoid hyperplasia. Eur J Radiol 2020;124:108824. [Crossref] [PubMed]
23. Möhler C, Wohlfahrt P, Richter C, Greilich S. On the equivalence of image-based dual-energy CT methods for the determination of electron density and effective atomic number in radiotherapy. Phys Imaging Radiat Oncol 2018;5:108-10. [Crossref] [PubMed]
24. Mileto A, Allen BC, Pietryga JA, Farjat AE, Zarzour JG, Bellini D, Ebner L, Morgan DE. Characterization of Incidental Renal Mass With Dual-Energy CT: Diagnostic Accuracy of Effective Atomic Number Maps for Discriminating Nonenhancing Cysts From Enhancing Masses. AJR Am J Roentgenol 2017;209:W221-30. [Crossref] [PubMed]
25. Liang H, Zhou Y, Zheng Q, Yan G, Liao H, Du S, Zhang X, Lv F, Zhang Z, Li YM. Dual-energy CT with virtual monoenergetic images and iodine maps improves tumor conspicuity in patients with pancreatic ductal adenocarcinoma. Insights Imaging 2022;13:153. [Crossref] [PubMed]
26. Chen Z, Wang Z, Liu S, Zhang S, Zhou Y, Zhang R, Yang W. 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. [Crossref] [PubMed]
27. Wang P, Tang Z, Xiao Z, Hong R, Wang R, Wang Y, Zhan Y. Dual-energy CT in differentiating benign sinonasal lesions from malignant ones: comparison with simulated single-energy CT, conventional MRI, and DWI. Eur Radiol 2022;32:1095-105. [Crossref] [PubMed]
28. Zhao W, Shen S, Ke T, Jiang J, Wang Y, Xie X, Hu X, Tang X, Han D, Chen J. Clinical value of dual-energy CT for predicting occult metastasis in central neck lymph nodes of papillary thyroid carcinoma. Eur Radiol 2024;34:16-25. [Crossref] [PubMed]
29. Katsuyama E, Miyawaki Y, Sada KE, Asano Y, Hayashi K, Yamamura Y, Hiramatsu-Asano S, Morishita M, Ohashi K, Watanabe H, Katsuyama T, Narazaki M, Matsumoto Y, Wada J. Association of explanatory histological findings and urinary protein and serum creatinine levels at renal biopsy in lupus nephritis: a cross-sectional study. BMC Nephrol 2020;21:208. [Crossref] [PubMed]
30. Jungmann PM, Agten CA, Pfirrmann CW, Sutter R. Advances in MRI around metal. J Magn Reson Imaging 2017;46:972-91. [Crossref] [PubMed]
31. Wong KL, Cheng KH, Lam SK, Liu C, Cai J. Review of functional magnetic resonance imaging in the assessment of nasopharyngeal carcinoma treatment response. Precision Radiation Oncology 2022;6:177-85.
32. Stolzmann P, Winklhofer S, Schwendener N, Alkadhi H, Thali MJ, Ruder TD. Monoenergetic computed tomography reconstructions reduce beam hardening artifacts from dental restorations. Forensic Sci Med Pathol 2013;9:327-32. [Crossref] [PubMed]
33. Zou Y, Zheng M, Qi Z, Guo Y, Ji X, Huang L, Gong Y, Lu X, Ma G, Xia S. Dual-energy computed tomography could reliably differentiate metastatic from non-metastatic lymph nodes of less than 0.5 cm in patients with papillary thyroid carcinoma. Quant Imaging Med Surg 2021;11:1354-67. [Crossref] [PubMed]
34. Küpeli A, Bulut E, Cansu A, Güner A, Soytürk M, Danışan G. Contribution of DECT in detecting serosal invasion of gastric cancer. Turk J Med Sci 2019;49:782-8. [Crossref] [PubMed]
35. Song C, Cheng P, Cheng J, Zhang Y, Sun M, Xie S, Zhang X. Differential diagnosis of nasopharyngeal carcinoma and nasopharyngeal lymphoma based on DCE-MRI and RESOLVE-DWI. Eur Radiol 2020;30:110-8. [Crossref] [PubMed]
36. Kato T, Uehara K, Ishigaki S, Nihashi T, Arimoto A, Nakamura H, Kamiya T, Oshiro T, Ebata T, Nagino M. Clinical significance of dual-energy CT-derived iodine quantification in the diagnosis of metastatic LN in colorectal cancer. Eur J Surg Oncol 2015;41:1464-70. [Crossref] [PubMed]
37. Saito M. Potential of dual-energy subtraction for converting CT numbers to electron density based on a single linear relationship. Med Phys 2012;39:2021-30. [Crossref] [PubMed]
38. Cheng F, Liu Y, Du L, Wang L, Li L, Shi J, Wang X, Zhang J. Evaluation of optimal monoenergetic images acquired by dual-energy CT in the diagnosis of T staging of thoracic esophageal cancer. Insights Imaging 2023;14:33. [Crossref] [PubMed]
39. Zhao F, Li C, Wu Y, Xia J, Zeng M, Li T, Xie K. Connective Tissue Growth Factor in Digestive System Cancers: A Review and Meta-Analysis. Biomed Res Int 2020;2020:8489093. [Crossref] [PubMed]
40. Yu X, Zhen Y, Yang H, Wang H, Zhou Y, Wang E, et al. Loss of connective tissue growth factor as an unfavorable prognosis factor activates miR-18b by PI3K/AKT/C-Jun and C-Myc and promotes cell growth in nasopharyngeal carcinoma. Cell Death Dis 2013;4:e634. [Crossref] [PubMed]
41. Chen Z, Zhang N, Chu HY, Yu Y, Zhang ZK, Zhang G, Zhang BT. Connective Tissue Growth Factor: From Molecular Understandings to Drug Discovery. Front Cell Dev Biol 2020;8:593269. [Crossref] [PubMed]
42. Li Q, Li X, Li XY, Huo JW, Lv FJ, Luo TY. Spectral CT in Lung Cancer: Usefulness of Iodine Concentration for Evaluation of Tumor Angiogenesis and Prognosis. AJR Am J Roentgenol 2020;215:595-602. [Crossref] [PubMed]
43. Yang X, Hu H, Zhang F, Li D, Yang Z, Shi G, Lu G, Jiang Y, Yang L, Wang Y, Duan X, Shen J. Preoperative Prediction of the Aggressiveness of Oral Tongue Squamous Cell Carcinoma with Quantitative Parameters from Dual-Energy Computed Tomography. Front Oncol 2022;12:904471. [Crossref] [PubMed]