[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
pISSN 1229-6929
eISSN 2005-8330
Open Access, Peer-reviewed
    Korean J Radiol. 2024 Dec;25(12):1070-1082. English.
    Published online Oct 29, 2024.
    https://doi.org/10.3348/kjr.2024.0335
    Copyright © 2024 The Korean Society of Radiology
    Original Article

    Determining Disease Activity and Glucocorticoid Response in Thyroid-Associated Ophthalmopathy: Preliminary Study Using Dynamic Contrast-Enhanced MRI

    Hao Hu,1,* Xiong-Ying Pu,1,* Jiang Zhou,1 Wen-Hao Jiang,1 Qian Wu,1 Jin-Ling Lu,1 Fei-Yun Wu,1 Huan-Huan Chen,2 and Xiao-Quan Xu1
      • 1Department of Radiology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China.
      • 2Department of Endocrinology, The First Affiliated Hospital of Nanjing Medical University, Nanjing, China.
    • Corresponding author: Xiao-Quan Xu, MD, PhD. Department of Radiology, The First Affiliated Hospital of Nanjing Medical University, No. 300, Guangzhou Rd, Gulou District, Nanjing, Jiangsu 210029, China. Email: xiaoquanxu_1987@163.com
      Corresponding author: Huan-Huan Chen, MD, PhD. Department of Endocrinology, The First Affiliated Hospital of Nanjing Medical University, No. 300, Guangzhou Rd, Gulou District, Nanjing, Jiangsu 210029, China. Email: drchenhuanhuan@njmu.edu.cn

      *These authors contributed equally to this work.
    Received February 20, 2024; Revised July 15, 2024; Accepted September 01, 2024.

    This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

    Abstract

    Objective

    To assess the role of dynamic contrast-enhanced (DCE)-MRI of the extraocular muscles (EOMs) for determining the activity of thyroid-associated ophthalmopathy (TAO) and treatment response to glucocorticoids (GCs).

    Materials and Methods

    We prospectively enrolled 65 patients with TAO (41 active, 82 eyes; 24 inactive, 48 eyes). Twenty-two active patients completed the GC treatment and follow-up assessment, including 15 patients (30 eyes) and 7 patients (14 eyes), defined as responsive and unresponsive, respectively. Model-free (time to peak [TTP], area under the curve [AUC], and Slopemax) and model-based (Ktrans, Kep, and Ve) parameters of EOMs in embedded simplified histogram analyses were calculated and compared between groups. Multivariable logistic regression analysis was used to identify the independent predictors. The area under the receiver operating characteristic curve (AUROC) was used to evaluate the diagnostic performance.

    Results

    Active patients exhibited significantly higher TTP at the 10th percentile (-10th), TTP-mean, and TTP at the 90th percentile (-90th); AUC-10th, AUC-mean, AUC-90th, and AUC-max; Ktrans-10th and Ktrans-mean; and Ve-10th, Ve-mean, Ve-90th, and Ve-max than inactive patients (P < 0.05). Responsive patients exhibited significantly lower TTP-min; higher Ktrans-mean and Ktrans-max; and higher Kep-10th, Kep-mean, and Kep-max than unresponsive patients (P < 0.05). TTP-mean and Ve-mean were independent variables for determining disease activity (P = 0.017 and 0.022, respectively). A combination of the two parameters could determine active TAO with moderate performance (AUROC = 0.687). TTP-min and Ktrans-mean were independent predictors of the response to GCs (P = 0.023 and 0.004, respectively), uniting which could determine the response to GCs with decent performance (AUROC = 0.821).

    Conclusion

    DCE-MRI-derived model-free and model-based parameters of EOMs can assist in the evaluation of TAO. In particular, TTP-mean and Ve-mean could be useful for determining the activity of TAO, whereas TTP-min and Ktrans-mean could be promising biomarkers for determining the response to GCs.

    Keywords
    Thyroid-associated ophthalmopathy; Magnetic resonance imaging; Dynamic contrast-enhanced; Disease activity; Treatment response

    INTRODUCTION

    Thyroid-associated ophthalmopathy (TAO) is a common inflammatory eye disease associated with thyroid dysfunction in adults [1]. The natural course of this disease includes an initial active phase and a subsequent inactive phase. The former features orbital tissue inflammation; therefore, it could be treated with systemic glucocorticoid (GC) therapy. However, the latter is characterized by orbital tissue fibrosis; therefore, a follow-up or symptomatic surgical treatment is usually suggested [1, 2]. Hence, the first crucial task during the clinical evaluation of TAO is to accurately determine disease activity. The clinical activity score (CAS) is widely used to assess disease activity [3]. However, the subjectivity potentially limits its accuracy, which is sometimes challenging for junior physicians [3, 4]. Once active TAO is defined, the next important task is to precisely determine the treatment response to GCs, considering that 20%–40% of patients are unresponsive to GCs [5]. For this special group of refractory patients, alternative immunosuppressive or radiation therapy can be selected to achieve treatment benefits and avoid the systemic side effects of GCs [6]. However, efficient clinical approaches for assessing the responsiveness of GCs before treatment are lacking [5, 7].

    Orbital MRI is increasingly used to evaluate TAO. Many studies have confirmed the role of multiparametric MRI of the extraocular muscles (EOMs) in determining the activity of TAO, including T2-based, diffusion-based, and magnetization transfer imaging [8, 9, 10]. Other researchers have also reported that MRI can detect the heterogeneity of EOMs and subsequently help in determining the treatment response to GCs [11, 12]. However, previous imaging-related studies have mainly focused on the pathological projection of orbital tissue-associated inflammation and fibrosis, and further physiological alterations within EOMs are rarely mentioned. Because both tissue inflammation and fibrosis are accompanied by changes in microenvironment perfusion, such as varied microvascular permeability and extracellular space [13, 14], the perfusion characteristics of EOMs in TAO could also be disturbed. Uncovering the inherent perfusion pattern of TAO can facilitate a deeper understanding of the physiological mechanism of the disease and provide novel potential for monitoring the disease course and improving evaluation efficiency.

    After a bolus injection of a gadolinium-based contrast agent and continuous scanning, dynamic contrast-enhanced (DCE)-MRI can provide quantitative information about the microcirculatory perfusion and permeability of various tissues [15]. Only one previous study used DCE-MRI to evaluate perfusion changes in the EOMs of patients with TAO [16]. The researchers reported that the time-to-peak (TTP) enhancement, enhancement ratio, and washout ratio of the EOMs differed significantly between the active and inactive groups, and all these parameters were significantly correlated with CAS [16]. However, they analyzed DCE-MRI using a model-free approach, and the major shortcoming of the model-free calculation was that it was not necessarily relevant to the physical essence [17]. In addition, the internal relationship between the perfusion characteristics of EOMs and the treatment response to GCs has not been explored.

    Therefore, this study aimed to assess the role of DCE-MRI-derived model-free and model-based parameters in discriminating active from inactive TAO phases and responsive patients with TAO after GCs treatment from non-responsive.

    MATERIALS AND METHODS

    Patients

    This prospective study was approved by the Institutional Review Board of The First Affiliated Hospital of Nanjing Medical University (IRB No. 2021-SRFA-024), and all patients who underwent pretreatment orbital MRI assessment provided signed informed consent. Between May 2022 and September 2023, 75 consecutive patients with TAO were recruited from the Department of Endocrinology at our hospital. The clinical diagnosis of TAO was based on the Bartley criteria [18]. The inclusion criteria were as follows: 1) patients’ age ≥18 years, 2) image quality adequate for further analysis, 3) bilateral eyes involved, 4) no history of surgical decompression, and 5) no other orbital pathologies. Finally, a total of 65 patients (29 male and 36 female; mean age, 48.7 ± 15.3 years) with 130 involved eyes were enrolled in this study.

    Clinical Evaluation

    Disease activity was assessed using a modified 7-point CAS [3] and recorded for each unit of eye. If CAS was ≥3, the eye was considered active; otherwise, it was considered inactive (CAS <3). Thereafter, 41 patients with TAO with 82 active eyes and 24 patients with TAO with 48 inactive eyes were included. Disease severity was assessed based on the European Group on Graves’ orbitopathy classification and was classified as mild, moderate-to-severe, or sight-threatening [6]. In addition, the history of smoking, serum levels of free triiodothyronine (FT3), free thyroxine (FT4), thyroid-stimulating hormone (TSH), and thyrotropin receptor antibody (TRAb) for each patient, and the duration of TAO symptoms, visual acuity (VA), visual field index (VFI), proptosis, and intraocular pressure (IOP) for each eye were also recorded.

    Among the 41 patients with active TAO, 27 received a standardized 12-week intravenous GC administration with a cumulative dose of 4.5 g methylprednisolone [6, 11, 12], and the remaining 14 were not administered GC due to preference towards other treatments or contraindications. Twenty-two of the 27 patients with active TAO completed the GC treatment, underwent consistent follow-up assessments, and completed the determination of treatment response at the endpoint of the current collection. All clinical assessments were performed cooperatively by the same endocrinologist (with 21 years of experience in endocrinology) and an ophthalmologist (with 32 years of experience in ophthalmology) before and three months after the end of GC treatment. Each eye was grouped into either “responsive” or “unresponsive” according to the therapeutic outcome. The responsive group was defined as 1) CAS reduction ≥2 points and CAS <3 and 2) at least one of the following improvements, without worsening of any other factors (decrease of proptosis ≥2 mm; decrease of lid width ≥2 mm; decrease in the Gorman score; and improvement of VA ≥1 Snellen line). The unresponsive group was defined as CAS reduction <2 points or staying active (CAS ≥3) [12, 19]. Finally, 15 patients with TAO with 30 responsive eyes and 7 patients with 14 unresponsive eyes were included. A flowchart of patient enrollment is presented in Figure 1.

    Fig. 1
    Flowchart of patient enrollment. TAO = thyroid-associated ophthalmopathy, GC = glucocorticoid

    Image Acquisition

    MRI scans were performed within one week before treatment on a 3T unit (Magnetom Skyra; Siemens Healthcare, Erlangen, Germany) with a 20-channel head and neck coil. Routine imaging protocols included axial T1-weighted imaging (repetition time [TR]/echo time [TE], 635/6.7 ms) and axial, coronal, and sagittal T2-weighted imaging with fat suppression (TR/TE, 4000/75–117 ms).

    Coronal DCE-MRI was performed using a three-dimensional volumetric interpolated breath-hold examination sequence. Before dynamic acquisition, an unenhanced T1 map based on three flip angles of 5°, 10°, and 15° was obtained using the same sequence. Gadolinium-diethylene triamine pentaacetic acid (Magnevist; Bayer Schering Pharma AG, Berlin, Germany) was administered by intravenous bolus injection using a power injector at a rate of 3 mL/s at a dose of 0.1 mmol/kg, followed by a 10-mL bolus of saline administered at the same injection rate. DCE-MRI acquisition included 50 phases at a temporal resolution of 6.36 s, and the total acquisition time was 5 min 18 s. The other imaging parameters were as follows: TR/TE, 3.89/1.31 ms; flip angle, 15°; number of excitations, 1; field of view, 203.3 × 230 mm2; matrix, 143 × 224; slice thickness, 4.0 mm; slice number, 24.

    Image Analysis

    All quantitative measurements were performed for each eye unit. Two radiologists (with 12 and four years of experience in head and neck radiology) who were blinded to the clinical information and study design independently assessed the data. The measurement results of the two radiologists were used to evaluate interobserver reproducibility. The measurement results of the senior radiologist (with 12 years of experience in head and neck radiology) were used for further statistical analyses.

    First, imaging data from DCE-MRI were imported into a dedicated post-processing software (Omni-Kinetics; GE Healthcare, Shanghai, China) to perform pixel-by-pixel pharmacokinetic calculations for analysis. DCE-MRI quantitative tracer kinetic modeling was based on a two-compartment modified Tofts model [20]. The vascular input function was extracted by manually setting a small circular region of interest (ROI) on the superior sagittal sinus, located proximal to the orbit [21]. Once the vascular input function was defined, voxel-wise perfusion maps, including both model-free and model-based parameters, were automatically generated. The model-free parameters included time–intensity curve–based TTP, area under the curve (AUC), and maximum ascending slope (Slopemax). The model-based parameters included the volume transfer constant between the plasma and extracellular extravascular space (EES) (Ktrans), the rate constant from the EES to blood plasma (Kep), and the volume fraction of the EES, which equals the ratio Ktrans/Kep (Ve) [15, 22].

    Polygonal ROIs were then manually outlined on the inferior, medial, and lateral EOMs and the superior rectus–levator complex to locate the maximum cross-section behind the eyeball representing the site of the muscle bellies. The surrounding fatty tissue and borders of the muscles were carefully avoided. Once the ROIs were placed, a ‘merge’ function in the software was used to unite all the ROIs. Subsequently, simplified histogram parameters of the ROI union, including the minimum, 10th percentile (-10th), mean, 90th percentile (-90th), and maximum values of the model-free and model-based parameters, were automatically obtained.

    Statistical Analysis

    All statistical analyses were performed using the SPSS software (version 23.0; IBM Corp., Armonk, NY, USA) and MedCalc (version 20.022; MedCalc Software Inc., Ostend, Belgium). Continuous variables were expressed as mean ± standard deviation or median with interquartile range (25%, 75%), based on whether they were normally distributed. Independent-samples t-test or the Mann–Whitney U test was used to compare continuous variables between the active and inactive TAO groups and between the responsive and unresponsive TAO groups. Categorical variables were compared using the chi-squared test or Fisher’s exact test. Univariable and multivariable logistic regression analyses using generalized estimating equations were performed to identify independent imaging predictors for determining disease activity and treatment response to GCs. The imaging variables with a P-value of <0.1 in the univariable analysis were adopted into the multivariable analysis. Odds ratios and 95% confidence intervals were determined. The identified significant imaging predictors were combined using binary logistic regression. Receiver operating characteristic (ROC) curve analysis and area under the ROC curve (AUROC) were used to evaluate the performance of the imaging parameters and their combinations. The sensitivity and specificity were calculated using a threshold criterion determined as the value would maximize the Youden index (Youden index = sensitivity + specificity − 1). Spearman’s rank correlation analyses were used to evaluate the relationship between imaging parameters and duration of TAO. Intraclass correlation coefficients (ICCs) were used to assess the interobserver reproducibility of model-free and model-based measurements. It was defined as follows: <0.40, poor reproducibility; 0.40–0.60, moderate; 0.61–0.80, good; and ≥0.81, excellent [9]. A two-sided P-value of <0.05 was considered statistically significant.

    RESULTS

    Patients’ Characteristics

    Table 1 summarizes the detailed clinical and demographic characteristics of the study cohort. Duration of TAO, CAS, FT4, VA, VFI, and proptosis differed significantly between the active and inactive TAO groups (all P < 0.05). No significant differences were observed in patients’ age, sex, smoking history, FT3, TSH, TRAb, or IOP between the two groups (all P > 0.05). In addition, only the duration of TAO differed significantly between the responsive and unresponsive groups (P = 0.036).

    Table 1
    Clinical and demographic data of study patients

    Interobserver Reproducibility in DCE-MRI-Derived Parameters of EOMs

    Good to excellent interobserver reproducibility was acquired for measurements of model-free and model-based parameters (ICCs: TTP-min, 0.846; TTP-10th, 0.875; TTP-mean, 0.967; TTP-90th, 0.825; TTP-max, 0.737; AUC-min, 0.861; AUC-10th, 0.885; AUC-mean, 0.932; AUC-90th, 0.935; AUC-max, 0.962; Slopemax-min, 0.800; Slopemax-10th, 0.784; Slopemax-mean, 0.935; Slopemax-90th, 0.890; Slopemax-max, 0.726; Ktrans-min, 0.778; Ktrans-10th, 0.820; Ktrans-mean, 0.782; Ktrans-90th, 0.741; Ktrans-max, 0.796; Kep-min, 0.746; Kep-10th, 0.706, Kep-mean, 0.838; Kep-90th, 0.767; Kep-max, 0.857; Ve-min, 0.770; Ve-10th, 0.723; Ve-mean, 0.912; Ve-90th, 0.870; Ve-max, 0.680).

    DCE-MRI-Derived Parameters of EOMs Between Active and Inactive, as Well as Between Responsive and Unresponsive TAO Groups

    Patients with active TAO exhibited significantly higher TTP-10th, TTP-mean, and TTP-90th; AUC-10th, AUC-mean, AUC-90th, and AUC-max; Ktrans-10th and Ktrans-mean; and Ve-10th, Ve-mean, Ve-90th, and Ve-max than inactive patients (all P < 0.05). Responsive patients with TAO showed significantly lower TTP-min, higher Ktrans-mean and Ktrans-max, and higher Kep-10th, Kep-mean, and Kep-max values than unresponsive patients (all P < 0.05). Detailed comparisons of the model-free and model-based parameters are presented in Table 2 and Supplementary Figure 1.

    Table 2
    Comparisons of model-free and model-based parameters between active and inactive, and responsive and unresponsive groups (per-eye results)

    Independent Imaging Predictors for Determining Disease Activity and Response to GCs

    Multivariable logistic regression analysis indicated that TTP-mean and Ve-mean were independent predictors of the disease activity of TAO (odds ratio = 0.440 and 38.361, respectively; P = 0.017 and 0.022, respectively), whereas TTP-min and Ktrans-mean were independent predictors of the treatment response to GCs (odds ratio = 0.450 and 141.022, respectively; P = 0.023 and 0.004, respectively). Detailed univariable and multivariable logistic regression analysis results are presented in Tables 3, 4, and Supplementary Tables 1, 2.

    Table 3
    Results of univariable and multivariable logistic regression analysis identifying independent imaging predictors for disease activity (i.e., active vs. inactive)

    Table 4
    Results of univariable and multivariable logistic regression analysis identifying independent imaging predictors for treatment response (i.e., responsive vs. unresponsive)

    Diagnostic Performance of Significant Imaging Parameters

    To determine the disease activity of TAO, ROC analysis revealed that the integration of TTP-mean and Ve-mean performed the best (AUROC = 0.687), with a sensitivity of 68.3% and a specificity of 64.6%, followed by Ve-mean alone (AUROC = 0.681) and then TTP-mean alone (AUROC = 0.613) (Table 5, Supplementary Fig. 2A). Representative patients with active and inactive TAO are shown in Figure 2.

    Fig. 2
    Representative images of a 56-year-old female with active TAO (A) and a 65-year-old male with inactive TAO (B). The TTP-mean and Ve-mean were 4.398/4.530 and 0.877/0.784 on the left/right orbit for the active patient (A), respectively, whereas 3.605/3.502 and 0.679/0.600 on the left/right orbit for the inactive patient (B), respectively. TAO = thyroid-associated ophthalmopathy, TTP = time to peak, T2WI = T2-weighted imaging, ROI = region of interest, DCE = dynamic contrast-enhanced, TIC = time-intensity curve

    Table 5
    Performance of models using DCE-MRI derived parameters in determining disease activity and GC response

    To determine the treatment response to GCs, combining TTP-min and Ktrans-mean exhibited good performance, with an AUROC of 0.821, sensitivity of 76.7%, and specificity of 78.6%, followed by TTP-min alone (AUROC = 0.765) and Ktrans-mean alone (AUROC = 0.736) (Table 5, Supplementary Fig. 2B). Representative cases of responsive and unresponsive patients with TAO are shown in Figure 3.

    Fig. 3
    Representative images of a 29-year-old male with responsive TAO (A) and a 70-year-old male with unresponsive TAO (B). The TTP-min and Ktrans-mean were 1.695/1.801 and 0.600/0.767 on the left/right orbit for the responsive patient (A), respectively, whereas 2.860/3.284 and 0.371/0.351 on the left/right orbit for the unresponsive patient (B), respectively. TAO = thyroid-associated ophthalmopathy, TTP = time to peak, T2WI = T2-weighted imaging, ROI = region of interest, DCE = dynamic contrast-enhanced, TIC = time-intensity curve

    Relationship Between Imaging Parameters and Duration of TAO

    TTP-min was positively correlated with TAO duration (r = 0.177, P = 0.045). Besides, Ktrans-10th, Ktrans-90th, Kep-max, Kep-10th, Kep-90th, TTP-max, and TTP-mean exhibited a close to significant correlation with TAO duration (P = 0.057, 0.051, 0.055, 0.052, 0.056, 0.082, and 0.087, respectively).

    DISCUSSION

    Our study has two major findings. First, DCE-MRI-derived model-free and model-based parameters differed significantly between the active and inactive TAO groups. The TTP-mean and Ve-mean were independent predictors of disease activity. Integrating TTP-mean and Ve-mean can achieve the highest efficacy in determining disease activity. Second, DCE-MRI-derived parameters differed significantly between the responsive and unresponsive TAO groups. The TTP-min and Ktrans-mean were independent predictors of treatment response, and their combination exhibited good performance in assessing the treatment response to GCs.

    TTP is considered a useful model-free parameter reflecting tissue vascularity and is closely related to microvessel count [23]. According to previous studies, tissue inflammation is usually accompanied by hypervascularization, whereas fibrosis can be followed by hypovascular changes [24, 25]. Considering the different pathological properties of TAO [1, 2], we speculated that during the active phase, patients may develop increased vascularization within the EOMs. The abundant number of microvessels may lead to an extended peak time for contrast agent accumulation, thereby increasing the TTP. In addition, Jiang et al. [16] found that the manually calculated time–intensity curve–based enhancement ratio of EOMs was higher in patients with active TAO than in inactive patients. Another study confirmed an increase in the signal intensity ratio of EOMs on contrast-enhanced T1-weighted imaging in an active cohort [8]. The AUC reflects the total amount of gadolinium accumulated in the EOMs within a specific time interval [26]. As a result, a longer TTP with a more pronounced enhancement level was more likely to result in an increased AUC during the active period.

    Ve represents the volume of EES [13, 22]. Previous diffusion-based MRI studies have reported that patients with active TAO have less restricted diffusivity of EOMs than inactive patients [8, 9, 27]. These studies explained that inflammatory changes in the active phase could lead to interstitial edema with increased interstitial volume, whereas fibrotic infiltration in the inactive phase may result in reduced interstitial volume [8, 9, 27]. Consistent with these findings, in the present study, the Ve values in patients with active TAO were higher than those in inactive patients. In addition, active patients had higher Ktrans-10th and Ktrans-mean in contrast to the inactive patients. Ktrans and Kep are well-behaved perfusion parameters that reflect vascular permeability within tissues. As it is known, increased permeability is a hallmark of active inflammation [13, 28]. Previously, in a study by Lee et al. [28] on Crohn’s disease, they detected higher Ktrans in the active phase than in the inactive phase, which corresponded to synchronous histopathology. Hence, the currently observed higher Ktrans values in the active TAO group may also indicate more impaired vascular permeability within the EOMs than that in the inactive cohort.

    The advantage of model-based perfusion parameters was further highlighted when determining treatment response to GCs. Ktrans-mean, Ktrans-max, Kep-10th, Kep-mean, and Kep-max were significantly higher in responsive TAO patients than in unresponsive patients with TAO, whereas only TTP-min showed differences between the groups with regard to model-free parameters. Previous studies revealed that responsive patients with TAO have higher high-percentile T2 signal intensity binding with more prominent inflammatory changes [11]. In contrast, unresponsive patients with TAO with a longer disease duration have a lower minimum T2 signal intensity associated with fibrosis, leading to a distinct treatment response [12, 29]. Therefore, regarding the aforementioned relationship between vascular permeability and inflammation, the responsive group was expected to exhibit increased Ktrans and Kep values, indicating a more significantly altered microvascular perfusion of EOMs. In addition, interstitial fibrosis was deemed to have an impact on TTP, prolonging it by limiting the extravasation of contrast agents into the EES [23, 30]. Hence, unresponsive patients with TAO with more prominent fibrosis changes would show a longer TTP. The positive correlation between TTP values and TAO duration may further support our conjecture.

    Through further multivariable logistic regression analysis, one model-free parameter and one model-based parameter, namely TTP-mean and Ve-mean, were identified as independent predictors of disease activity. Similarly, TTP-min and Ktrans-mean were independent predictors of treatment response. Integrating TTP-mean and Ve-mean showed moderate diagnostic performance for active TAO group, whereas combining TTP-min and Ktrans-mean enabled appropriate diagnostic performance for responsive TAO group. The initial exploration and novel findings indicated potential vascular remodeling and permeability alterations related to the activity of TAO, which could be a reliable basis for the noninvasive assessment of pathophysiological changes in TAO. Notably, another strength of the current study is the model-based quantitative evaluation of DCE-MRI perfusion parameters in the TAO, which provides more robust metrics that directly reflect microcirculation physiology than model-free measurement. Subtle perfusion differences between responsive and unresponsive patients with TAO during the parallel-active phase could be detected based on the cooperation of model-based and model-free parameters.

    This study has several limitations. First, the sample size was relatively small, especially for the responsive and unresponsive groups. Future studies with larger sample sizes are warranted. Second, only DCE-MRI and its derived simplified histogram metrics were analyzed. Future studies integrating multimodal MRI, radiomics, and machine learning may improve determination efficiency. Third, histopathological control was still not achieved owing to the difficulty in acquiring histological specimens of EOMs. Future studies identifying correlations between imaging indicators and histopathology are needed.

    In conclusion, our preliminary study indicated that DCE-MRI-derived perfusion parameters could help determine the activity of TAO and the treatment response to GCs. Our study establishes an important foundation for the use of DCE-MRI to evaluate patients with TAO in clinical practice.

    Supplement

    The Supplement is available with this article at https://doi.org/10.3348/kjr.2024.0335.

    Supplement

    Click here to view.(368K, pdf)

    Notes

    Conflicts of Interest:The authors have no potential conflicts of interest to disclose.

    Author Contributions:

    • Conceptualization: Hao Hu, Xiao-Quan Xu, Fei-Yun Wu.

    • Data curation: Huan-Huan Chen, Wen-Hao Jiang, Qian Wu, Jin-Ling Lu.

    • Funding acquisition: Hao Hu, Xiao-Quan Xu, Fei-Yun Wu.

    • Investigation: all authors.

    • Methodology: Xiong-Ying Pu, Jiang Zhou, Qian Wu, Jin-Ling Lu.

    • Supervision: Xiao-Quan Xu, Huan-Huan Chen.

    • Validation: all authors.

    • Visualization: Xiong-Ying Pu, Jiang Zhou, Wen-Hao Jiang.

    • Writing–original draft: Hao Hu, Xiong-Ying Pu, Jiang Zhou.

    • Writing–review & editing: all authors.

    Funding Statement:This work was supported by National Natural Science Foundation of China (NSFC) (81801659 to Hao Hu), Jiangsu Province Hospital (The First Affiliated Hospital of Nanjing Medical University) Clinical Capacity Enhancement Project (JSPH-MC-2021-8 to Xiao-Quan Xu) and Jiangsu Province Capability Improvement Project through Science, Technology and Education (JSDW202243 to Fei-Yun Wu).

    Availability of Data and Material

    The data that support the findings of this study are available from the corresponding author upon reasonable request.

    References

      1. Neag EJ, Smith TJ. 2021 update on thyroid-associated ophthalmopathy. J Endocrinol Invest 2022;45:235–259.
      1. Taylor PN, Zhang L, Lee RWJ, Muller I, Ezra DG, Dayan CM, et al. New insights into the pathogenesis and nonsurgical management of Graves orbitopathy. Nat Rev Endocrinol 2020;16:104–116.
      1. Mourits MP, Prummel MF, Wiersinga WM, Koornneef L. Clinical activity score as a guide in the management of patients with Graves’ ophthalmopathy. Clin Endocrinol (Oxf) 1997;47:9–14.
      1. Tortora F, Cirillo M, Ferrara M, Belfiore MP, Carella C, Caranci F, et al. Disease activity in Graves’ ophthalmopathy: diagnosis with orbital MR imaging and correlation with clinical score. Neuroradiol J 2013;26:555–564.
      1. Vannucchi G, Covelli D, Campi I, Origo D, Currò N, Cirello V, et al. The therapeutic outcome to intravenous steroid therapy for active Graves’ orbitopathy is influenced by the time of response but not polymorphisms of the glucocorticoid receptor. Eur J Endocrinol 2013;170:55–61.
      1. Bartalena L, Kahaly GJ, Baldeschi L, Dayan CM, Eckstein A, Marcocci C, et al. The 2021 European Group on Graves’ orbitopathy (EUGOGO) clinical practice guidelines for the medical management of Graves’ orbitopathy. Eur J Endocrinol 2021;185:G43–G67.
      1. Hart RH, Kendall-Taylor P, Crombie A, Perros P. Early response to intravenous glucocorticoids for severe thyroid-associated ophthalmopathy predicts treatment outcome. J Ocul Pharmacol Ther 2005;21:328–336.
      1. Politi LS, Godi C, Cammarata G, Ambrosi A, Iadanza A, Lanzi R, et al. Magnetic resonance imaging with diffusion-weighted imaging in the evaluation of thyroid-associated orbitopathy: getting below the tip of the iceberg. Eur Radiol 2014;24:1118–1126.
      1. Hu H, Chen L, Zhou J, Chen W, Chen HH, Zhang JL, et al. Multiparametric magnetic resonance imaging for differentiating active from inactive thyroid-associated ophthalmopathy: added value from magnetization transfer imaging. Eur J Radiol 2022;151:110295
      1. Li Z, Luo Y, Feng X, Zhang Q, Zhong Q, Weng C, et al. Application of multiparameter quantitative magnetic resonance imaging in the evaluation of Graves’ ophthalmopathy. J Magn Reson Imaging 2023;58:1279–1289.
      1. Liu P, Luo B, Chen L, Wang QX, Yuan G, Jiang GH, et al. Baseline volumetric T2 relaxation time histogram analysis: can it be used to predict the response to intravenous methylprednisolone therapy in patients with thyroid-associated ophthalmopathy? Front Endocrinol (Lausanne) 2021;12:61453
      1. Hu H, Chen L, Zhang JL, Chen W, Chen HH, Liu H, et al. T2-weighted MR imaging-derived radiomics for pretreatment determination of therapeutic response to glucocorticoid in patients with thyroid-associated ophthalmopathy: comparison with semiquantitative evaluation. J Magn Reson Imaging 2022;56:862–872.
      1. de Vries BA, van der Heijden RA, Poot DHJ, van Middelkoop M, Meuffels DE, Krestin GP, et al. Quantitative DCE-MRI demonstrates increased blood perfusion in Hoffa’s fat pad signal abnormalities in knee osteoarthritis, but not in patellofemoral pain. Eur Radiol 2020;30:3401–3408.
      1. Zou L, Jiang J, Zhang H, Zhong W, Xiao M, Xin S, et al. Comparing and combining MRE, T1ρ, SWI, IVIM, and DCE-MRI for the staging of liver fibrosis in rabbits: assessment of a predictive model based on multiparametric MRI. Magn Reson Med 2022;87:2424–2435.
      1. Hu H, Xu XQ, Liu H, Hong XN, Shi HB, Wu FY. Orbital benign and malignant lymphoproliferative disorders: differentiation using semi-quantitative and quantitative analysis of dynamic contrast-enhanced magnetic resonance imaging. Eur J Radiol 2017;88:88–94.
      1. Jiang H, Wang Z, Xian J, Li J, Chen Q, Ai L. Evaluation of rectus extraocular muscles using dynamic contrast-enhanced MR imaging in patients with Graves’ ophthalmopathy for assessment of disease activity. Acta Radiol 2012;53:87–94.
      1. Gaddikeri S, Gaddikeri RS, Tailor T, Anzai Y. Dynamic contrast-enhanced MR imaging in head and neck cancer: techniques and clinical applications. AJNR Am J Neuroradiol 2016;37:588–595.
      1. Bartley GB, Gorman CA. Diagnostic criteria for Graves’ ophthalmopathy. Am J Ophthalmol 1995;119:792–795.
      1. Xu L, Li L, Xie C, Guan M, Xue Y. Thickness of extraocular muscle and orbital fat in MRI predicts response to glucocorticoid therapy in Graves’ ophthalmopathy. Int J Endocrinol 2017;2017:3196059
      1. Tofts PS, Brix G, Buckley DL, Evelhoch JL, Henderson E, Knopp MV, et al. Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: standardized quantities and symbols. J Magn Reson Imaging 1999;10:223–232.
      1. Lewis D, Zhu X, Coope DJ, Zhao S, King AT, Cootes T, et al. Surrogate vascular input function measurements from the superior sagittal sinus are repeatable and provide tissue-validated kinetic parameters in brain DCE-MRI. Sci Rep 2022;12:8737
      1. Ang T, Juniat V, Patel S, Selva D. Evaluation of orbital lesions with DCE-MRI: a literature review. Orbit 2024;43:408–416.
      1. Yabuuchi H, Fukuya T, Tajima T, Hachitanda Y, Tomita K, Koga M. Salivary gland tumors: diagnostic value of gadolinium-enhanced dynamic MR imaging with histopathologic correlation. Radiology 2003;226:345–354.
      1. Korchi AM, Cengarle-Samak A, Okuno Y, Martel-Pelletier J, Pelletier JP, Boesen M, et al. Inflammation and hypervascularization in a large animal model of knee osteoarthritis: imaging with pathohistologic correlation. J Vasc Interv Radiol 2019;30:1116–1127.
      1. Wollina U, Verma SB, Ali FM, Patil K. Oral submucous fibrosis: an update. Clin Cosmet Investig Dermatol 2015;8:193–204.
      1. Chen BB, Hsu CY, Yu CW, Kao JH, Lee HS, Liang PC, et al. Hepatic necro-inflammation and elevated liver enzymes: evaluation with MRI perfusion imaging with gadoxetic acid in chronic hepatitis patients. Clin Radiol 2014;69:473–480.
      1. Chen HH, Hu H, Chen W, Cui D, Xu XQ, Wu FY, et al. Thyroid-associated orbitopathy: evaluating microstructural changes of extraocular muscles and optic nerves using readout-segmented echo-planar imaging-based diffusion tensor imaging. Korean J Radiol 2020;21:332–340.
      1. Lee S, Choi YH, Cho YJ, Cheon JE, Moon JS, Kang GH, et al. Quantitative evaluation of Crohn’s disease using dynamic contrast-enhanced MRI in children and young adults. Eur Radiol 2020;30:3168–3177.
      1. Matsuzawa K, Izawa S, Kato A, Fukaya K, Matsumoto K, Okura T, et al. Low signal intensities of MRI T1 mapping predict refractory diplopia in Graves’ ophthalmopathy. Clin Endocrinol (Oxf) 2020;92:536–544.
      1. Zhou L, Chen TW, Zhang XM, Yang Z, Tang HJ, Deng D, et al. Liver dynamic contrast-enhanced MRI for staging liver fibrosis in a piglet model. J Magn Reson Imaging 2014;39:872–878.
    Publication Types
    Original Article
    MeSH Terms
    Biomarkers
    Follow-Up Studies
    Glucocorticoids
    Glucocorticoids*
    Graves Ophthalmopathy*
    Humans
    Logistic Models
    Magnetic Resonance Imaging*
    Oculomotor Muscles
    Prospective Studies
    ROC Curve
    Troleandomycin
    Substance
    thiamine triphosphorate
    Metrics
    Share
    Figures
    slide
    slide
    slide

    1 / 3

    Tables
    slide
    slide
    slide
    slide
    slide

    1 / 5

    ORCID IDs
    Funding Information
    PERMALINK