Low Signal Intensities of MRI T1 Mapping Predict Refractory Diplopia in Graves' Ophthalmopathy

Kazuhiko Matsuzawa; Shoichiro Izawa; Ayumi Kato; Kenji Fukaya; Kazuhisa Matsumoto; Tsuyoshi Okura; Dai Miyazaki; Masamichi Kurosaki; Shinya Fujii; Shin-ichi Taniguchi; Masahiko Kato; Kazuhiro Yamamoto


Clin Endocrinol. 2020;92(6):536-544. 

In This Article

Abstract and Introduction


Objective: In Graves' ophthalmopathy (GO), fibrosis in extraocular muscles (EOMs) may be related to intravenous glucocorticoid (ivGC)-resistant diplopia. Signal intensity (SI) of magnetic resonance imaging (MRI) T1 mapping can quantify properties of EOM components, including fibrosis. We investigated EOM features of GO patients with diplopia using T1 mapping SI and the predictive value of T1 mapping SI in the response of diplopia to ivGCs.

Design: We performed a cross-sectional study that included 13 active GO patients, 34 inactive GO patients with history of diplopia, including 20 with a history of diplopia disappearance, 14 GO patients with refractory diplopia and 35 control subjects. In nine active GO patients, the relationship between T1 mapping SI at pretreatment and at diplopia outcome after ivGC treatment was prospectively investigated.

Methods: T1 mapping SI of left and right inferior rectus and medial rectus muscles was measured in all participants.

Results: T1 mapping SI in inactive GO patients with refractory diplopia was significantly lower than that of other groups in all evaluated EOMs. Diagnostic accuracy for refractory diplopia by T1 mapping SI in GO patients with a history of diplopia disappearance was excellent (AUC 0.89) compared with other assessments. Furthermore, among nine active GO patients, pretreatment T1 mapping SI in four patients with ivGC-resistant diplopia tended to be low compared with the other five patients with improved diplopia.

Conclusions: Low intensity T1 mapping in EOMs is likely to be associated with refractory diplopia and may be useful in predicting the response of diplopia to ivGCs.


In Graves' ophthalmopathy (GO), clinical activity score (CAS) is the most widely used index of inflammatory activity and reactivity to intravenous glucocorticoids (ivGCs).[1–3] However, in a Korean study, over 40% of patients with GO who had progressive diplopia had a low CAS.[4] In Japanese patients with GO, symptoms that constitute CAS, such as eyelid oedema, are less frequent than in Caucasians.[5,6] The cause is not clear, but may result from differences in skin colour, orbital structure[7] and the curvature of the face. CAS symptoms may, therefore, be less conspicuous in races that have a similar facial appearance to that of the Japanese, such as Koreans. In these patients with GO, it is difficult to decide whether ivGCs are appropriate or not when only diplopia progresses without change in other CAS symptoms. Magnetic resonance imaging (MRI) is widely used to assess inflammation in EOMs and can predict a therapeutic response.[8–11] In Japanese patients with GO, even in those with a low CAS, MRI T2-weighted imaging (T2WI) high intensity is expected to reveal ivGC effects.[9] However, it is not clear whether assessment of the inflammatory activity of EOMs is sufficient to predict a therapeutic effect on diplopia. Although ivGCs have a high rate of adverse effects, including severe liver damage, cardiovascular and cerebrovascular complications, and severe infection,[12] diplopia is the most debilitating symptom in GO.[13] Therefore, a novel clinical index is needed that can select patients for whom ivGCs will be highly effective in treating diplopia. Diplopia in the active phase results from restriction of muscle movement because of inflammatory EOM enlargement, which is associated with oedema and an infiltration of inflammatory cells.[14] Diplopia in this phase is thought to be responsive to ivGCs. EOMs in GO with diplopia that is refractory to ivGCs are infiltrated with fibrosis;[14] therefore, a clinical index that represents tissue fibrosis may be useful to predict refractory diplopia.

MRI T1 mapping is used as a noninvasive technique for the evaluation of tissue properties. T1 mapping quantitatively evaluates the T1 relaxation time for each pixel and displays these data on a map. The T1 relaxation time is specific for each tissue composition; therefore, T1 mapping signal intensity is expected to reflect tissue properties. Originally, precontrast and postcontrast T1 mapping were mainly used to evaluate myocardial characteristics.[15] Now, however, MRI T1 mapping can evaluate the degree of fibrosis in various organs.[16–18] For example, MRI T1 mapping can show a more than 95% diagnostic specificity for four grades of liver fibrosis, as classified following liver biopsy.[17] In addition, pre-enhanced native T1 mapping can potentially assess tissue properties.[16,19,20]

Recent clinical trials have shown several nonsurgical agents to be useful in GO.[21] This study focused on predicting the effects of ivGCs on diplopia using MRI T1 mapping. However, new treatment strategies for GO are being developed; therefore, it is necessary to establish various indicators for selecting appropriate treatment. This study describes how MRI T1 mapping has the potential to become such an indicator for selecting treatment.

To address our hypothesis that noncontrast MRI T1 mapping can reflect the histological features of EOMs in patients with diplopia and can predict the response to ivGCs, T1 mapping signal intensity was assessed in patients with GO. In patients with active GO and diplopia, we also investigated the relationship between T1 mapping signal intensity before and after ivGC treatment.