Impact of Axial Orbital Depth on Extraocular Muscle Diameter Changes Following Retrobulbar Glucocorticoid Injection in Graves Ophthalmopathy: A Retrospective Cohort Study

Abstract

Objective:

The objective of this study was to explore the impact of the bony orbit on the alteration of extraocular muscle diameters (EMDs) following retrobulbar injection of glucocorticoids (GCs) in patients with Graves ophthalmopathy (GO), which is a common therapeutic approach.

Methods:

Retrospective collection and analysis of data were conducted on GO patients who underwent retrobulbar GC injection. Measurements of bony orbital dimensions before and after injection were acquired through orbital computed tomography. Univariate and multifactorial logistic regression analyses were employed to identify bony orbital parameters as potential risk factors contributing to post-treatment EMD changes in GO patients.

Results:

A total of 60 GO patients with 102 eyes were included, with a male-to-female ratio of 1.07:1.00 and a mean age of 49.28 ± 9.40 years. After retrobulbar GC injection, EMDs decreased in size in 36 eyes, while remaining unchanged or increased in 66 eyes. An axial orbital depth threshold effect on EMD alteration was observed. When axial orbital depth was below 41.28 mm, there was no significant correlation between EMD changes and axial orbital depth (OR: 1.09, 95% CI: 0.75–1.57, P = 0.6644). However, a protective effect was evident when axial orbital depth exceeded 41.28 mm (OR: 0.41, 95% CI: 0.18–0.90, P = 0.0260). For every 1-unit increment in axial orbital depth, the risk of non-decrease in EMD decreased by 59% (95% CI: 0.18–0.99, P = 0.00260).

Conclusion:

The alteration of EMDs following retrobulbar GC injection in GO patients is associated with axial orbital depth.

Graves ophthalmopathy (GO) stands as a chronic autoimmune disorder, ranking as the foremost among adult orbital ailments and representing a key cause of both unilateral and bilateral ocular protrusion.¹ This condition chiefly encompasses the extraocular muscles (EOMs), orbital fat, and lacrimal glands. A notable pathological change associated with GO is the enlargement of EOM volume.^2,3

Glucocorticosteroids [glucocorticoid (GCs)] serve as the primary therapeutic agents for GO, and retrobulbar injection treatment with GCs has garnered increased attention due to its association with fewer systemic side effects.^4–7 Ebner et al documented a reduction in the volume of initially enlarged EOMs subsequent to retrobulbar GCs injection,⁴ whereas Winder and colleagues demonstrated that variations in orbital anatomical structure can lead to differences in drug dispersion within the orbit following retrobulbar injections.^8–10 As a result, we postulated the potential influence of disparities in bony orbital characteristics on the outcomes of altered extraocular muscle diameters (EMDs) after retrobulbar GC injection in individuals with GO.

In this study, we conducted measurements using 2 computed tomography (CT) scans, both taken before and after retrobulbar injections in patients diagnosed with GO. The objective was to ascertain data related to the bone orbit as well as alterations in EMDs. Moreover, we aimed to explore potential correlations between bony orbital parameters and changes in EMDs subsequent to retrobulbar injections. In addition, we investigated risk factors contributing to EMD alteration. By pursuing this comprehensive approach, individuals with GO can formulate proactive treatment strategies and enhance their clinical decision-making process.

METHODS

Study Design and Data Collection

This is a retrospective cohort study. In an unselective and continuous manner, we gathered data from 60 patients (102 eyes) who underwent retrobulbar GC injections between January 1, 2015 and June 1, 2022, at the inpatient ophthalmology departments of Union Hospital, Tong Medical College, Huazhong University of Science and Technology, China.

The retrobulbar injection consisted of a combination of triamcinolone acetonide (20 mg) and dexamethasone (2.5 mg). Administered by the same physician, the corticosteroid mixture was introduced gradually using a disposable syringe (22–27 gauge, 3 cm in length). The needle was inserted in a vertical orientation into the lower outer quadrant of the orbit. Once a depth of around 1.5 cm was achieved, the insertion was altered to direct the needle medially and cephalad, reaching an additional 1.5 cm toward the apex of the orbit. Intraocular pressure (IOP) measurements were taken before the injection and on the subsequent day following the injection. A baseline orbital CT scan was conducted before the initial injection, followed by subsequent injections at 1-month intervals, totaling 3 injections. Upon completion of the 3 injections, a follow-up orbital CT scan was performed.

The inclusion criteria utilized in this study were as follows:

• Age between 18 and 80 years.

• Patients who received retrobulbar GC injections while being hospitalized in our ophthalmology department.

• Absence of systemic diseases other than thyroid conditions.

• Having undergone orbital CT within 1 week before retrobulbar GCs injection, followed by a subsequent orbital CT within 2 to 3 months post-injection.

• Patients who received low-dose (0.5 mg/kg) oral maintenance treatment with methylprednisolone tablets post-injection, with no additional treatment administered.

• Presence of one or more instances of rectus muscle enlargement.

• Before the injection, patients exhibited IOP ranging from 10 to 21 mm Hg in both eyes. Subsequently, post-injection IOP did not exceed 21 mm Hg.

Conversely, the exclusion criteria were as follows:

• Age below 18 years or above 80 years.

• History of ocular trauma or surgery.

• Use of GCs or radiation therapy within the preceding 6 months.

• Pregnancy in the case of female patients.

• Absence of data pertaining to general examinations and CT scans.

• Intraocular pressure exceeded 21 mm Hg before or after each injection.

Ethical Approval

This study received ethical approval from the Human Ethics Committee of Wuhan Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, and it was conducted in accordance with the principles outlined in the Declaration of Helsinki. The study has also been registered with the Chinese Clinical Trials Registry (www.chictr.org.cn, registration number: ChiCTR2200063429, date of registration: September 6, 2022). The authors had access to de-identified and anonymized patient data, which could not be used to identify individual participants during or after data collection. The study received ethical approval from the Human Ethics Committee of Wuhan Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, and the requirement for informed consent was waived due to the retrospective nature of the study and the thorough de-identification and anonymization procedures applied to patient data within the cohort.

Data Collection

Data pertaining to the health status and CT images of the enrolled patients were sourced from the electronic medical record system of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology. For CT data acquisition, we employed a picture archiving and communication system, with window position/window width configured to 60/300 Hounsfield units. Each measurement index was independently recorded 3 times and subsequently averaged. A comprehensive assessment of the variations in each EOM before and after treatment was subsequently executed. The data were compiled and organized within the EDC data collection and management system (study.empoweredc.com, Shanghai).

• Orbital distance: Measured on axial CT, the distance between the highest points of the zygomatic bones on both sides.

• Ocular prominence: Measured on axial CT, the vertical distance between the most protruding point of the eye and the interzygomaitic line.

• EOM assessment: Axial CT measurements reveal the horizontal diameter of the medial rectus and lateral rectus muscle belly at their widest point. Conversely, sagittal CT measurements involve determining the vertical diameter of the superior rectus muscle-levator complex and the widest point of the inferior rectus muscle belly. The superior rectus and superior levator muscles above are treated as a unified muscle group and measured jointly.

• Medial wall length: This measurement is taken from axial CT scans, representing the distance between the medial orbital rim and the medial edge of the optic foramen.

• Lateral wall length: Evaluated using axial CT scans, this metric corresponds to the distance from the lateral orbital rim to the lateral edge of the optic foramen.

• Orbital horizontal diameter: Determined through axial CT scans, this value represents the distance from the medial orbital rim to the lateral orbital rim.

• Orbital pituitary diameter: Measured through sagittal CT scans, this measurement captures the distance from the superior orbital rim to the inferior orbital rim.

• Axial orbital depth: Assessed using axial CT scans, this measurement quantifies the distance from the orbital apex to the interzygomatic line. The orbital apex is directed posteriorly and medially, serving as the location of the optic nerve canal leading into the skull.

• Sagittal orbital depth: This measurement is obtained from sagittal CT scans, representing the distance from the orbital apex to the diameter of the orbital pendulum.

• Orbital rim angle: Determined through axial CT scans, this angle is formed by the medial and lateral orbital walls.

Upon analyzing 2 CT images of the GO patient, taken before and after the injection, 2 individuals will assess whether alterations have occurred in the EMDs of the lesion (increased, decreased, or relatively unchanged). In the event of discordance between the assessments made by the initial 2 individuals, a third party will be engaged to resolve the disagreement

Statistical Analysis

Quantitative data were presented as means ± SD (X ± SD) and median (Q1–Q3). For qualitative data, a frequency (%) representation was employed. All measurement data underwent a Kolmogorov-Smirnov normality test. When the data adhered to a normal distribution, the t test was utilized to establish the presence of significant differences between the two groups. Alternatively, in instances where the data did not exhibit normality, the Wilcoxon rank-sum test was employed. Qualitative data were assessed using the χ² test to ascertain the significance of disparities between groups.

A univariate analysis was performed to determine whether bony orbital parameters were associated with decreased EMDs. Considering the possibility of a nonlinear relationship between bony orbital parameters and decreased EMDs, a generalized additive model was used to identify the nonlinearity. In the presence of a nonlinear relationship between them, a dichotomous linear regression model was employed to estimate thresholds for decreased EMDs.

We performed 2 multifactorial analyses, adjusting for the age, sex, and Clinical Activity Score (CAS) score groupings, to demonstrate the independent effect of axial orbital depth on non-decreased EMDs before and after the inflection point.

Statistical analysis was performed with R version 4.0.3 (https://www.R-project.org) and EasyR (https://www.easyr.cc Solutions, Inc., Shanghai). A 2-tailed P value of <0.05 was considered statistically significant in all analyses.

RESULTS

Baseline Characteristics

Initially, a total of 130 patients were enrolled in this study. However, subsequent refinement was carried out according to exclusion criteria: 4 juveniles, 10 patients with a history of ocular trauma and ocular surgery, 30 patients who had undergone GCs and radiation therapy within a 6-month timeframe, and 26 patients with incomplete CT data. Consequently, the final analysis comprised 60 patients, encompassing 102 eyes.

The assessment of bony orbital parameters in GO patients was carried out by the same investigator using the picture archiving and communication system system. A meticulous analysis was performed to ascertain the disparities between bony orbital parameters measured before and after retrobulbar injection. Significantly, no statistically significant differences were observed between the two sets of measurements (Supplemental Digital Content, Table 1, http://links.lww.com/SCS/H504). Thus, bony orbital parameters before injection were employed for the study.

Based on the alterations in EMDs following retrobulbar injection of GCs, the 102 eyes were categorized into a decreased group (EMDs decreased) and a non-decreased group (EMDs unchanged or increased). The decreased group encompassed 36 eyes, while the non-decreased group comprised 66 eyes.

The baseline characteristics of the patients are outlined in Supplemental Table 2 (Supplemental Digital Content, Table 2, http://links.lww.com/SCS/H504). The study included 31 (51.67%) male and 29 (48.33%) female participants. The mean age of the enrolled patients was 49.28 ± 9.40 years. When comparing the decreased group to the non-decreased group, it was observed that the former exhibited a larger diameter of the medial rectus muscle and a higher proportion of CAS ≥3 (P = 0.048). Importantly, no statistically significant differences were observed in the distribution of bony orbital parameters (all P values >0.05).

The demographic features of the patients are shown in Supplemental Table 1 (Supplemental Digital Content, Table 1, http://links.lww.com/SCS/H504). A total of 303 patients with GO were included. The median age was 47 (37–54) years. There were 146 (48.2%) males and 157 (51.8%) females. In total, 28 (9.2%) experienced eyelid redness, 164 (54.1%) had conjunctival redness, 159 (52.5%) had eyelid edema, 63 (20.8%) had chemosis of the conjunctiva, and 45 (14.9%) had inflammation of the caruncle and/or plica (Supplemental Digital Content, Table 1, http://links.lww.com/SCS/H504).

Univariate Analysis of Extraocular Muscle Diameters Change After Retrobulbar Injection of Glucocorticoids

We conducted a univariate logistic regression analysis based on the collected data (as presented in Supplemental Digital Content, Table 3, http://links.lww.com/SCS/H504). Our analysis indicated that the risk of non-decreasing EMDs following GC injection was 0.57 times lower in GO patients with CAS ≥3 compared with those with CAS <3, with a 95% CI of (0.18, 1.00), P = 0.00495. Moreover, an enlarged medial rectus muscle emerged as a protective factor against non-decreasing EMDs following retrobulbar GC injection. Specifically, the risk of non-decreasing EMDs was reduced by 18% with each 1-unit increase in medial rectus muscle diameter, 95% CI (0.69, 0.98), P = 0.00275. No significant correlation was identified between bony orbital parameters and changes in EMDs through linear correlation analysis (P > 0.05).

Nonlinear Correlation Analysis of Bony Orbital Parameters and Changes in Extraocular Muscle Diameters After Retrobulbar Injection of Glucocorticoids

Through smoothed curve fitting and generalized summation models, we observed a nonlinear relationship between bony orbital parameters and EMDs change (Fig. 1). After adjusting for age, gender, CAS score grouping, and EMDs, it became evident that a threshold effect was present in the association between axial orbital depth and EMDs change. According to a 2-piecewise linear model and recursive algorithm, the inflection point for axial orbital depth was identified as 41.28 mm. In the segment to the right of the inflection point, each 1-unit increase in axial orbital depth yielded a 59% reduction in the risk of non-decreasing EMDs, with a 95% CI of (0.18, 0.99), P = 0.0260 (as indicated in Supplemental Digital Content, Table 4, http://links.lww.com/SCS/H504).

FIGURE 1.

(A–H) Nonlinear relationship between bony orbital parameters and EMD change. EMD indicates extraocular muscle diameter.

DISCUSSION

In this retrospective cohort study, we examined the relationship between bony orbital parameters and changes in EMDs following retrobulbar GC injection in GO patients. Data from 60 patients, encompassing a total of 102 eyes, were analyzed. A notable correlation emerged between axial orbital depth and the likelihood of EMDs not decreasing. Specifically, the impact of axial orbital depth on EMD change lacked significance when the axial orbital depth was below 41.27 mm. However, once the axial orbital depth surpasses 41.27 mm, a pattern emerges: greater axial orbital depth correlates with a reduced risk of EMDs not decreasing.

The study’s findings demonstrated that among GO patients who underwent retrobulbar injection of GCs, a greater number of eyes experienced either a decrease or no change in EMDs (88 eyes), compared with those that exhibited an increase (14 eyes). This outcome aligns with the observations made by Ebner et al in their research. Retrobulbar GC injection has proven to be effective in addressing GO.^4,7,11 Upon comparing the baseline characteristics of the EMDs decreased group (36 cases, 35.29%) with the EMDs non-decreased group (66 cases, 64.71%), the analysis revealed that the risk of EMDs not decreasing post GCs injection was 2.33 times higher in patients with CAS <3 than in those with CAS ≥3, with a 95% CI of (0.18, 1.00), P = 0.0495. This finding underscores CAS <3 as a risk factor. Notably, a higher CAS score indicates a more active orbital inflammation in GO patients, which may correspondingly lead to a more favorable response to retrobulbar GC injection.¹² The results of our study further strengthen the notion that GO patients with a CAS score of ≥3 are more likely to experience decreased EMDs following retrobulbar GCs injection.

Although the CAS is widely utilized for assessing GO,^13,14 its subjectivity presents limitations. As a qualitative tool, it does not directly unveil orbital lesions. To address this, CT measurement of EMDs provides enhanced accuracy. Simple coronal slice measurements enable an estimate of rectus muscle volume, where the largest transverse diameter of the lateral rectus is equally applicable.¹⁵ Hence, we adopted the extraocular muscle belly diameter as a representative of muscle volume. Among the 102 eyes studied, bilateral EOM involvement frequency followed a descending sequence of the inferior rectus, memedial rectus, superior rectus muscle-levator complex, and lateral rectus muscles—consistent with the findings of Bahn.¹⁶

In our investigation, we noted a noteworthy correlation: as the diameter of the medial rectus muscle increases, the likelihood of non-decreasing EMDs following retrobulbar GCs injection decreases. Intriguingly, findings from studies by Weis et al¹⁷ and Berger et al¹⁸ indicated that individuals with larger memedial rectus muscles in GO face an elevated risk of dysthyroid optic neuropathy (DON). This suggests that an enlarged memedial rectus muscle is associated with a heightened DON risk. our results imply that a larger medial rectus muscle might confer greater benefits following retrobulbar GC injection. This intriguing connection raises the possibility that retrobulbar GC injection could serve as an effective preventive treatment for DON in patients with enlarged medial rectus muscles.

In our current investigation, the orbital distance, measured on axial CT as the distance between the highest points of the zygomatic bones on both sides, exhibited a median value of 98.40 ± 3.67 mm. A study involving 109 individuals conducted by Furuta reported a mean orbital distance of 100.88 + 3.60 mm. In addition, our analysis revealed an axial CT orbital rim angle of 45.40 ± 2.56 degrees, whereas the study by Chan et al¹⁹ on GO patients in Singapore reported a rim angle of 42.4% + 2.5%. The variations in these values could potentially stem from ethnic distinctions, given that our measurements exclusively encompass Chinese individuals, whereas Furuta²⁰ focused on Japanese individuals, and Chan and colleagues on Singaporeans. Furthermore, the exploration by Barretto and Mathog²¹ underscores the influence of ethnic diversity on orbital anatomy.

Previous research has underscored associations between the orbital rim angle, lateral wall length, and EMDs with the emergence of DON.²² Furthermore, bony orbital parameters have been linked to alterations in prominence subsequent to orbital decompression surgery in individuals with GO.²³ Nevertheless, scarce investigations have delved into the correlation between bony orbital parameters and shifts in EMDs. Recognizing that orbital anatomical variations might influence drug dispersion, and consequently impact treatment outcomes, we explored this relationship. Our findings indicated that axial orbital depth demonstrated no substantial effect on outcomes when measuring <41.28 mm (P = 0.6644). However, a noteworthy protective effect was evident when axial orbital depth exceeded this threshold. For each incremental unit in axial orbital depth, the risk of EMDs not decreasing diminished by 59%, with a 95% CI (0.18, 0.99) and a P value of 0.00260. This could potentially stem from variances in drug dispersion and absorption within the orbit upon surpassing a particular depth, ultimately influencing EMDs.

We embarked on an investigation to discern risk factors associated with the non-decreasing of EMDs following GC injection in individuals with GO. Our findings suggest that GO patients harboring an enlarged medial rectus muscle, a CAS ≥3, and an axial orbital depth exceeding 41.28 mm could potentially exhibit a more favorable response to retrobulbar GCs injection treatment. This study equips clinicians with a straightforward and practical assessment tool, enhancing clinical decision-making capabilities. To enhance the precision of our findings, future endeavors should encompass a larger patient cohort drawn from a broader demographic. Consequently, a multicenter prospective randomized controlled clinical study with an expanded sample size emerges as a promising avenue for further exploration.

The study boasts several notable advantages: (1) Both orbital data were gleaned from precise CT measurements, ensuring the accuracy of EMD change assessments. (2) Diverging from earlier cross-sectional and case-control approaches, our study design embraces a cohort structure. However, it remains an observational study, rendering it susceptible to potential confounding factors. Notably, we executed rigorous adjustments for these variables and gauged result robustness through sensitivity analysis.

Limitations of the Study

This study has 3 primary limitations. First, its retrospective design may introduce data biases and limit causal inference. Second, the relatively small sample size (60 GO patients, 102 eyes) may affect statistical power and generalizability. Third, the study primarily focused on orbital bone dimensions, neglecting other potential factors like age, gender, disease severity, and treatment duration. These limitations should be considered when interpreting the study’s findings.

CONCLUSION

In this retrospective cohort study, we investigated the risk factors associated with non-decreasing EMDs after retrobulbar injection of GCs in patients with GO. Our findings suggest that individuals with GO who possess an enlarged medial rectus muscle, a CAS score ≥3, and an axial orbital depth >41.28 mm may exhibit a more favorable response to retrobulbar GCs injection. Interestingly, the influence of axial orbital depth on EMD change was insignificant when the depth was below 41.27 mm. However, beyond this threshold, deeper axial orbital depths were linked to a reduced risk of non-decreasing EMDs. Our study contributes a straightforward and practical assessment approach for clinicians. This empowers patients with GO to develop targeted treatment strategies and make informed clinical choices.

Supplementary Material

scs-36-1995-s001.docx ^{(24.1KB, docx)}