Introduction

Graves ophthalmopathy (GO) is the most common extrathyroidal manifestation of Graves disease (GD) [1, 2]. Clinical management of GO is the most important challenge for clinicians and early diagnosis and correct staging of GO and its inflammatory activity is important for timing and choice of treatment [3]. There are clinical studies that suggest a direct LG involvement leading to a decrease in tear secretion with an LG volume significantly higher in patients with GO than in healthy controls demonstrated at CT and MRI [4,5,6]. To date, no data are available about the difference in LG involvement between patients with different levels of GO activity that is commonly reported in terms of clinical activity score (CAS). Since different treatment strategies are required for patients with active and inactive GO [7], the aim of this study was to evaluate a simple quantitative MRI-derived parameter able to evaluate LG involvement (in terms of lacrimal gland herniation, LGH) in relation to GO activity.

Materials and methods

We enrolled thirty-two consecutive Caucasian patients, 10 men and 22 women, median age 49.5 years (IR 30–68), affected by GD with a diagnosis of GO. Patients were than grouped according to CAS score in group A (16 with inactive GO, CAS < 3) and group B (16 with active GO, CAS ≥ 3). Patients gave their informed consent after approval by the local Ethical Committee. All subjects underwent orbital MRI scan. A 1.5-T MRI scanner (Signa HDxt; GE Medical Systems) was used for orbital evaluation with an 8-channel phased-array head coil. For an in-depth description of inclusion and exclusion criteria, thyroid and ophthalmologic evaluation, MRI protocol, and the statistical analysis methods, see supplementary material.

LGH was evaluated on axial 3-mm-thick FSE-T2w sequences by drawing a line between the right and left ventral zygomatic borders (interzygomatic line). This line is commonly drawn at the level of the lens to evaluate the amount of proptosis (from there, a perpendicular line is taken to the apex of the globe, depicting the measurement of proptosis considering a Hertel-index of ≥ 22 mm pathological [8]). We also drew the interzygomatic line at the level of the maximum depicted LGH on post-gadolinium axial 3-mm-thick T1-w FSE FAT-SAT (or 3D-EFGRE’s MPRs) sequences and, taking a perpendicular line, we measured the amount of lacrimal gland parenchyma protruding anteriorly.

Results

The ophthalmological clinical parameters are shown in Table 1 together with MRI-derived ones. The clinical, biochemical, and ultrasound parameters of 32 patients, grouped according to CAS, are shown in Table 2 (see supplementary material). The two groups were matched for gender, age, and body mass index. No difference was found between the thyroid biochemical and ultrasound parameters except for TRAb levels that were slightly higher, although not significantly, in group B and A (see supplementary material). No differences between the two groups of patients were found for all the evaluated parameters, except for diplopia, which was significantly more prevalent in patients with active than inactive GO (81 vs. 31%; p = 0.003).

Table 1 Ophthalmological and MRI-derived parameters of 32 GO patients grouped according to CAS
Full size table

LGH was significantly higher for both right (10.1 (7.3–17) vs. 7 (0–13.4) mm; p = 0.004) and left (8.5 (6.6–13) vs. 5.8 (0–12) mm; p = 0.026) eyes in group B than in those of A (Fig. 1 and Table 1).

Fig. 1
figure 1

Axial 3-mm-thick FSE T2-w FAT-SAT (a, c) and FSE T1-w FAT-SAT after i.v. administration of gadolinium (b, d). a, b Patient from group A (inactive GO) shows no proptosis but some slight hyperintense signal of ocular muscles if compared with facial muscles (a); lacrimal glands (arrows in b) look almost unremarkable in their size and location. c, d Patient from group B (active GO) shows the typical GO-related MRI findings of active disease: enlargement, bright signal, and enhancement of ocular muscles (c, d), augmentation and slight bright signal from intraconal fat tissue (c), severe proptosis of both eyes (c), and bilateral severe herniation of lacrimal glands (16 and 12.9 mm on the right and on the left side respectively) that, after i.v. administration of gadolinium, also shows typical bright enhancement (d)

Full size image

Similarly, MRI-derived proptosis (Hertel index) was significantly higher for both right (23.3 (20–27.7) vs. 20 (15.1–26.2); p = 0.021) and left (23.7 (18–26.9) vs. 19.8 (13.3–26.2) mm; p = 0.036) eyes in group B than those in A. A linear correlation was found between TRAb levels and LGH measurement (Rho 0.462, p = 0.009) in all patients.

Discussion

In this study, we demonstrated significantly higher LGH, evaluated by the use of orbital MRI examination, in patients with active over those with inactive GO. This parameter appears useful for the differentiation between inactive and active GO and was correlated with TRAb levels.

Early diagnosis of GO and correct grading of the inflammation activity have primary importance in GO management [3]. The CAS, proposed by Mourits et al [9], is a validated scoring system for the identification of the active phase of the autoimmune process with a high predictive value for the outcome of immunosuppressive treatment in GO patients [10] and a lower predictive value for response to radiotherapy [3, 9]. In this study, we demonstrate that MRI-derived LGH index may be a good and simple parameter to differentiate inactive from active GO, which could be useful to improve the diagnostic ability of CAS in some clinical cases. Since CAS is operator-dependent and offers an imperfect differentiation of the various degrees of inflammation activity, integration with other imaging-based examinations may be required. Previous studies performed on patients with GO found a greater volume of LG, evaluated by CT, over those of healthy controls, and more recently the importance of LG volume calculation was suggested for the diagnosis of GO [5, 6]. To date, a simple LG quantitative parameter that could correlate with the GO staging has not yet been found. Our study demonstrates that LG involvement could also be assessed by means of herniation using a simple quantitative measurement, although with the limits related to the small sample, the lack of evaluation in healthy controls, and the decision to use MRI to take an osseous anatomical landmark. In addition, LGH could also be related to the growth of retro-orbital structures, as demonstrated by the significant higher proptosis evaluated by MRI for both eyes in patients with active GO over those in patients with inactive GO in our cohort. However, LG involvement may also have an independent pathogenic role in GO as suggested by the evidence of TSH-R in LG tissues [4]. These data are in agreement with our study, which demonstrates a linear correlation between LGH and TRAb levels. Finally, if the patient’s clinical history is not clear, or in the unlikely case of an isolated LG involvement without all the typical imaging findings of GO (i.e., myositis, cellulitis), then it will also be necessary to exclude another cause (i.e., dacryoadenitis, tumors).

In conclusion, we present a simple, quick, and easy method to perform in daily practice for the measurement of lacrimal gland herniation, which is closely correlated to GO disease activity. Studies of a larger group of patients with different degrees of GO activity may be needed to confirm these preliminary data.