The Changes in Ocular Biomechanical Response Parameters and Intraocular Pressure After Surgical Treatment for Thyroid Eye Disease

Abstract

Purpose: To investigate changes in ocular biomechanical response parameters and intraocular pressure (IOP) in patients with thyroid eye disease (TED) undergoing orbital decompression or anterior blepharotomy.

Methods: Eighty-three eyes from 46 patients receiving orbital decompression (the orbital decompression group) and 45 eyes from 28 patients receiving anterior blepharotomy (the anterior blepharotomy group) were retrospectively enrolled from a tertiary center. Corvis ST tonometry was used to assess ocular biomechanical response and biomechanically corrected IOP (bIOP) pre- and postoperatively. Non-contact tonometry (IOP-NCT) was also performed.

Results: In the anterior blepharotomy group, the margin reflex distance decreased (P < 0001). The highest concavity radius (P = 0.026) and whole eye movement (P = 0.003) increased. Neither IOP-NCT nor bIOP had a significant change. In the orbital decompression group, the extent of exophthalmos decreased (P < 0.001). The A2 length (P = 0.009) decreased. The bIOP did not show a significant change (16.4 ± 2.7 vs. 16.7 ± 4.5; P = 0.415), but the IOP-NCT decreased significantly (17.5 ± 3.3 vs. 16.0 ± 3.3; P < 0.001). Higher baseline IOP-NCT (β = −0.40, P < 0.001) and greater reduction in stiffness parameter A1 (SP-A1; β = 0.05, P = 0.002) were associated with more significant IOP-NCT reduction after the orbital decompression.

Conclusions: Ocular biomechanical response parameters may change after TED surgery, potentially affecting IOP measurements, particularly in patients receiving orbital decompression.

Thyroid eye disease (TED) is an autoimmune disorder characterized by inflammation and remodeling of the orbital and periorbital tissues,^1,2 leading to various ocular changes such as progressive exophthalmos due to enlargement of extraocular muscles (EOMs) and expansion of orbital fat volume, as well as upper lid retraction caused by enlargement and fibrosis of the levator palpebrae superioris and Müller's muscle.^2,3 These changes can result in vision-threatening complications such as dysthyroid optic neuropathy and exposure keratopathy.^2,4,5 Orbital decompression surgery is a well-established procedure to correct disfiguring proptosis and related complications.⁶ Various types of non-surgical treatment modalities and lid surgeries,^2,7 such as anterior blepharotomy,^3,7 could be performed to correct upper eyelid retraction. 

Patients with TED often have altered ocular biomechanics and the resulting biomechanical response to an air puff applied to the cornea^8–15 due to the deposition of glycosaminoglycans, orbital tissue edema, and fibrotic changes of EOMs, which can decrease orbital tissue compliance and corneal hysteresis.^9,16 The inflammatory cytokine in the active stage may also lead to the changes in corneal microstructures, as well as viscoelastic changes in corneal hysteresis.^8,10,11 These changes in ocular biomechanics can affect the accuracy of intraocular pressure (IOP) measurement,^10,17 which is particularly important given the higher prevalence of ocular hypertension and glaucoma in patients with TED.¹⁸ Although intravenous steroids have been shown to alter ocular biomechanics in patients with active Graves’ orbitopathy,¹⁹ insufficient data are available on how ocular biomechanics change following surgical management of TED. 

The Corvis ST tonometer (Oculus Optikgeräte, Wetzlar, Germany), which utilizes Scheimpflug technology to visualize the dynamic corneal response to an air pulse, can provide valuable insights into ocular biomechanics. Biomechanically corrected IOP (bIOP) is a novel parameter that incorporates corneal structure and dynamic response to present an IOP measurement minimally affected by ocular biomechanics.²⁰ This study aimed to investigate the changes in ocular biomechanical response parameters and bIOP in patients with TED receiving orbital decompression surgery or anterior blepharotomy using the Corvis ST tonometer. 

Methods

We conducted a retrospective analysis of patients with TED who underwent surgery at National Taiwan University Hospital between July 2020 and July 2022. Patients with a history of refractive or intraocular surgery were excluded. Diagnosis of TED was based on the consensus of the European Group on Graves’ Orbitopathy.²¹ Our previous work described the treatment algorithm and methods for TED.^3,22 In brief, patients with disfiguring exophthalmos underwent fat-removal (Hertel exophthalmometer ≧ 22 mm) or bone-removal (Hertel exophthalmometer ≧ 24 mm) orbital decompression surgery based on the Hertel exophthalmometry results; patients with upper lid retraction underwent anterior blepharotomy³ and were assigned to the orbital decompression and anterior blepharotomy groups, respectively. Patients with active TED were treated with systemic steroids or other immunosuppressants before surgery. Except for the vision-threatening condition, surgery was performed after the patient had achieved stable thyroid function and inactive TED, defined as having a seven-item clinical activity score of less than three.²³ For patients with dysthyroid optic neuropathy, bone-removal orbital decompression surgery was performed. Dysthyroid optic neuropathy was diagnosed if the patient suffered from decreased visual acuity (logarithm of the minimal angle of resolution > 0.2) and had the following signs: relative afferent pupillary defect, and/or impaired color vision, and/or visual field defect, and/or reduced amplitude and prolonged latency of visual evoked potentials test, and apical crowding on orbital imaging using computerized tomography.^4,22 

All patients underwent a comprehensive ophthalmic examination, including autorefraction (Topcon, Inc., Tokyo, Japan), slit-lamp biomicroscopy, Hertel exophthalmometry, margin to reflex distance (MRD) measurement, Oculus Corvis ST evaluation, orbital computed tomography, and thyroid function blood tests (thyroid-stimulating hormone and free T4). The Corvis ST was used to assess ocular biomechanical response parameters, including A1 and A2 lengths, A1 and A2 velocities, peak distance, highest concavity (HC) radius, HC deformation amplitude, HC deflection amplitude, deformation amplitude ratio at 2 mm, integrated inverse radius, stiffness parameter (stiffness parameter A1 [SP-A1]),²⁴ Corvis biomechanical index (CBI), maximum whole eye movement (WEM), central corneal thickness, and bIOP.²⁵ Preoperative Corvis ST measurements were performed within 1 month before surgery, and postoperative evaluations were conducted between 3 and 6 months postoperatively. IOP was also measured using non-contact tonometry (IOP-NCT; CT-80 Computerized Tonometer; Topcon) on the same day as the Corvis ST measurements. The study adhered to the tenets of the Declaration of Helsinki, and Institutional Review Board/Ethics Committee approval was obtained. Informed consent was waived due to the retrospective nature of the study. 

Statistical Analysis

In this study, all statistical analyses were conducted using R 4.2.0 (R Foundation for Statistical Computing, Vienna, Austria). Descriptive statistics were reported as mean and standard deviation for parametric data, and percentages were reported for categorical data. To account for correlated data from both eyes of a patient, a generalized linear model and generalized estimating equation were used to compare clinical characteristics and corneal biomechanical response parameters before and after surgeries. We adjusted for potential confounding variables such as age, sex, and bIOP. Because bIOP was incorporated into the formula for calculating SP-A1 and the CBI, we adjusted the IOP-NCT as a confounding variable instead for these two parameters.^24,25 We identified ocular biomechanical parameters that demonstrated statistically significant changes following orbital decompression surgery and anterior blepharotomy, respectively. To determine factors associated with changes in corneal biomechanical response parameters, we performed univariable regression analysis and included variables with P < 0.1 in the multivariable model. Adjusted P values were calculated for multiple comparisons using the Holm adjustment.²⁶ P < 0.05 was considered statistically significant. 

Results

We conducted a retrospective study involving 83 eyes from 46 patients who underwent orbital decompression surgery (the orbital decompression group) and 45 eyes from 28 patients who underwent anterior blepharotomy (the anterior blepharotomy group) for inactive TED. Table 1 presents the clinical characteristics of the patients. Among the eyes in the orbital decompression group, 20 (24.1%) underwent inferomedial bone-removal orbital decompression surgery, and the remaining eyes underwent fat-removal orbital decompression surgery, with an average removed fat volume²⁷ of 4.2 ± 1.1 mL. 

Table 1.

 

Clinical Characteristics of Patients With TED Receiving Orbital Decompression Surgery and Anterior Blepharotomy

 

Table 2 presents a comparison of the preoperative and postoperative clinical characteristics and ocular biomechanical response parameters of patients in the anterior blepharotomy group. Notably, the MRD1 significantly decreased from 7.4 ± 1.3 mm to 5.0 ± 1.2 mm (P < 0.001), indicating the effectiveness of the surgery in correcting upper lid retraction. Additionally, the HC radius increased from 6.41 ± 0.81 mm to 6.75 ± 0.75 mm (P = 0.026, adjusted P = 0.312), suggesting increased corneal stiffness. The WEM also significantly increased from 0.20 ± 0.05 mm to 0.24 ± 0.06 mm (P = 0.003, adjusted P = 0.039), indicating enhanced globe movement in response to an air pulse. However, no statistically significant changes were observed in the other parameters of Corvis ST. 

Table 2.

 

Changes in Corneal Biomechanical Response Parameters and Clinical Characteristics of Patients With TED Receiving Anterior Blepharotomy and Orbital Decompression Adjusted With Sex, Age, and bIOP

 

Table 2 presents a comparison of preoperative and postoperative clinical characteristics and corneal biomechanical parameters of patients in the orbital decompression group. The results showed a significant decrease in the extent of exophthalmos from 22.3 ± 2.2 mm to 17.7 ± 1.1 mm (P < 0.001). Additionally, the A2 length decreased from 1.89 ± 0.39 mm to 1.76 ± 0.34 mm (P = 0.009, adjusted P = 0.117). This change suggested that the corneal resistance to deformation decreased after orbital decompression surgery. The changes in other parameters of Corvis ST were not statistically significant. We also compared the changes in corneal biomechanical response parameters between patients receiving fat-removal and bone-removal orbital decompression surgery. There were no significant differences in the observed changes within these parameters between the two subgroups of patients. 

Table 3 presents the clinical factors associated with changes in WEM and HC radius in the anterior blepharotomy group. Univariable regression analysis showed that an increase in WEM was positively associated with the extent of proptosis (P = 0.036) and negatively associated with the spherical equivalent (P = 0.005). Patients with previous muscle surgeries for diplopia correction had less change in WEM after anterior blepharotomy (P = 0.029). In the multivariable regression analysis, more myopic eyes tended to have a greater increment in WEM after anterior blepharotomy (β = −0.005, P = 0.029). Univariable regression analysis also showed that eyes with higher baseline bIOP had a greater increment in HC radius (P = 0.001) postoperatively, which remained statistically significant in the multivariable model (β = 0.035, P < 0.001). 

Table 3.

 

Factors Associated With Changes in Corneal Biomechanical Response Parameters After Anterior Blepharotomy in Patients With TED

 

Table 4 presents the clinical characteristics associated with changes in A2 length in the orbital decompression group. In the univariate regression analysis, more hyperopic eyes had a greater decrease in A2 length after receiving orbital decompression surgery (P = 0.017). Eyes with lower bIOP postoperatively (P < 0.001) and requiring additional muscle surgery (P = 0.003) had a greater decrease in A2 length. In the multivariate regression analysis, changes in bIOP (β = 0.039, P < 0.001) and spherical equivalent (β = −0.027, P = 0.012) remained associated with changes in A2 length. 

Table 4.

 

Factors Associated With Changes in Corneal Biomechanical Response Parameters After Orbital Decompression in Patients With TED

 

The Figure illustrates the changes in bIOP and IOP-NCT before and after surgery in the orbital decompression group. There was no significant change in bIOP between the preoperative and postoperative measurements (16.4 ± 2.7 mmHg vs. 16.7 ± 4.5 mmHg; P = 0.415). However, a significant decrease in IOP-NCT was observed after surgery (17.5 ± 3.3 mmHg vs. 16.0 ± 3.3 mmHg; P < 0.001). In contrast, there was no significant change in either bIOP or IOP-NCT in the anterior blepharotomy group (bIOP: 15.9 ± 3.8 mmHg vs. 15.3 ± 1.8 mmHg, P = 0.261; IOP-NCT: 16.2 ± 3.5 mmHg vs. 16.7 ± 2.7 mmHg, P = 0.406). 

Figure.

 

Preoperative and postoperative IOP measured using the Corvis ST (bIOP-Corvis) and non-contact tonometry (IOP-NCT) in the orbital decompression group. (Left) The biomechanically corrected IOP (bIOP-Corvis) did not show significant change after the orbital decompression surgery. (Right) The IOP-NCT decreased significantly after the orbital decompression surgery.

Table 5 presents the clinical characteristics that are associated with changes in IOP-NCT after orbital decompression surgery. In the univariable regression analysis, eyes with a greater reduction in IOP-NCT after the surgery had a higher baseline IOP-NCT (β = −0.38, P < 0.001), and a greater reduction in SP-A1 after the surgery (β = 0.04, P = 0.014). In the multivariable regression analysis, eyes with a greater reduction in IOP-NCT after the surgery remained to have a higher baseline IOP-NCT (β = −0.40, P < 0.001) and a greater reduction in SP-A1 (β = 0.05, P = 0.002). 

Table 5.

 

Factors Associated With Changes in IOP Measured Using NCT After Patients Received Orbital Decompression

 

Discussion

The aim of this study was to investigate the changes in IOP and ocular biomechanical response parameters in patients with inactive TED who underwent orbital decompression surgery and anterior blepharotomy. Our results demonstrated a significant reduction in IOP-NCT after orbital decompression surgery, whereas bIOP, as measured by the Corvis ST, did not exhibit a significant change. The discrepancy between these two measurements was associated with a decrease in corneal resistance after the surgery, as indicated by SP-A1. Moreover, anterior blepharotomy was observed to increase orbital tissue compliance, as demonstrated by the increase in WEM, and increase corneal resistance to deformation, as indicated by the increase in HC radius, particularly in patients with myopia and higher bIOP. 

TED is an autoimmune disease that targets the orbital fibroblasts, leading to subsequent orbital changes.²⁸ The overproduction of adipose tissue and extracellular matrix and the infiltration of lymphocytes exacerbate local inflammation and cause orbital tissue edema and expansion.^29,30 The enlargement and fibrotic changes in EOM result from the deposition of glycosaminoglycans and inflammatory cells.³⁰ Consequently, changes in orbital tissue compliance and corneal biomechanics may occur in patients with TED.^8–16 Higher orbital tissue tension and lower orbital compartment compliance have been demonstrated using direct orbital manometry.³¹ Congested orbital tissue reduces the ease of retropulsion of the eyeball posteriorly.³² The Corvis ST, an ultra-high-speed Scheimpflug camera, can measure orbital tissue compliance by visualizing corneal response to an air puff. The WEM was defined by the maximum globe displacement in an anteroposterior direction.^9,12 In patients with TED, reduced WEM and orbital tissue compliance were observed compared to healthy individuals.^9,12,33 These findings were found to be associated with the increased cross-sectional area of the EOM as visualized by orbital computed tomography.³³ Moreover, in patients with active TED, intravenous corticosteroid injection was found to increase WEM due to the reduction of orbital congestion.¹⁹ The preoperative WEM values reported in our study for the orbital decompression and anterior blepharotomy groups were consistent with previously reported values in the literature for patients with TED.^9,33 Specifically, we found preoperative WEM values of 0.23 and 0.20, respectively. In the anterior blepharotomy group, we observed a significant increase in WEM following surgery, likely due to the release of the retracted upper eyelid and fibrotic levator palpebrae muscle, which may have hindered eye movement.³² Myopic eyes were more susceptible to the effects of the fibrotic eyelid due to the weaker corneal biomechanics.³⁴ Therefore, a larger increase in WEM in myopic eyes following the anterior blepharotomy was noted. In contrast, the lack of a significant change in WEM in the orbital decompression group may be attributed to the fact that these patients had inactive TED and less congested orbital tissue at baseline. 

Previous studies have consistently reported altered corneal biomechanics in eyes with TED, as evidenced by lower corneal hysteresis and corneal resistance factor measurements using ocular response analyzer. This suggests lower damping ability when compared to normal eyes.^8,10,13–15 The cause of these changes may be related to the increased tear film osmolarity and inflammatory cytokines that are common in TED, leading to microstructural changes in the corneal epithelium and stroma.^35,36 Altered corneal biomechanics in TED eyes were also presented by Corvis ST.^9,11,12,19 In the present study, the HC radius increased in patients with TED following anterior blepharotomy, likely due to the relief of compression on the eyeball caused by the fibrotic eyelid. This effect was more pronounced in eyes with higher bIOP, which is associated with greater ocular rigidity. Conversely, we observed a significant decrease in A2 length following orbital decompression surgery in TED eyes, which was more significant in hyperopic eyes. The orbital fat was resected during fat-removal decompression, and the orbital volume was enlarged during the bone-removal decompression. These modifications to the orbital structure might alleviate the compression of periocular tissues on the globe, consequently influencing ocular biomechanics. 

Patients with TED commonly experience elevated IOP,³⁷ which is caused by various factors, including enlargement of EOMs, expansion of fat volume, increased episcleral venous pressure, and extracellular matrix deposition in the trabecular meshwork.^38,39 Several studies have reported a higher prevalence of ocular hypertension in patients with TED compared to the general population.^18,38 On the other hand, the measurement of IOP could be falsely elevated due to positioning. IOP measurement in a slight downgaze position using a Tono-Pen (Reichert Technologies, Buffalo, NY, USA) could be considered to confirm IOP elevation before further management.⁴⁰ Orbital decompression has been shown to significantly reduce IOP, with the degree of reduction varying depending on factors such as baseline IOP.^39–42 Patients who have higher baseline IOP or those who have undergone orbital decompression surgery due to dysthyroid optic neuropathy experience a greater reduction in IOP.³⁹ A study reported that patients with baseline IOP higher or lower than 21 mmHg had average reductions in IOP of 6.2 mmHg and 1.2 mmHg, respectively.³⁹ However, a study with a baseline IOP lower than 21 mmHg reported no significant reduction in IOP after intraconal fat removal surgery.⁴³ Similarly, another study on patients who received two-wall decompression did not show a significant reduction in IOP postoperatively.⁴⁴ In the present study, we observed a significant reduction in IOP-NCT (1.5 mmHg). This reduction is lower compared to previous studies, but it is comparable to patients with TED with a lower baseline IOP. It is worth noting that our patients had inactive TED, and the relatively less congested orbital content may have contributed to the lower reduction in IOP. In contrast, the bIOP measured using the Corvis ST did not show a significant change postoperatively in the present study. Interestingly, eyes with a greater reduction in IOP-NCT were associated with a more significant decrease in SP-A1. The lower SP-A1 in the post-decompression eyes may indicate weaker corneal biomechanics, which could have influenced the IOP measurement. Therefore, the significant reduction of IOP-NCT observed in our cohort might be partially attributed to the changes in the corneal biomechanics. The altered corneal biomechanics in TED eyes may affect the accuracy of IOP measurement.^10,45 To address this issue, the use of cornea-compensated IOP for the Ocular Response Analyzer and bIOP for the Corvis ST, which take into account corneal structure and response,²⁰ may be considered for more accurate IOP measurement in patients with TED, particularly those with glaucoma who require precise IOP control. 

The present study had several limitations that should be acknowledged. First, the patients enrolled in this study had inactive TED, which may have led to less significant changes in ocular biomechanical response parameters and IOP compared to patients with active TED. In the active phase, the increased inflammatory cytokine may affect the ocular biomechanical response parameters and result in a more pronounced increase in IOP.⁴⁶ Therefore, the subgroup analysis is necessary to clarify the characteristics in patients with active and inactive TED.⁴⁶ Future studies could investigate the corneal biomechanical response parameters in patients with active TED. Second, there was only one preoperative and one postoperative measurement with the Corvis ST, which may limit the reliability of the findings. Although the repeatability and reproducibility of the Corvis ST are good in healthy subjects,⁴⁷ it is unclear if these findings hold true in eyes with TED. Given the retrospective nature of this study, there were variations in the timing of the Corvis ST measurements among the patients. Additional prospective studies with the Corvis ST performed at multiple scheduled time points could provide more insight into the changes in ocular biomechanical response parameters after the operation and the evolving nature of these parameters over time. Third, we aimed to identify corneal biomechanical response parameters that exhibited changes after the surgery and conducted comparisons across several parameters. We recognized that performing multiple comparisons raises the risk of type I error and, as a result, we reported the adjusted P values and annotated the parameters that retained their significance after adjustment for the readers to interpret. Nonetheless, employing conservative adjustments can also increase the risk of type II error.^48,49 Therefore, we engaged in a discussion based on the results obtained without adjustment. Finally, the absence of a healthy control group in this study is another limitation. However, previous studies have extensively investigated the comparison of ocular biomechanical response parameters between patients with TED and healthy controls. Therefore, this study focused on the changes in ocular biomechanical response parameters after surgical treatment of TED. 

In conclusion, our study highlights that ocular biomechanics may be altered in patients with TED following orbital decompression surgery and anterior blepharotomy. We observed that the anterior blepharotomy increased orbital tissue compliance and corneal stiffness, whereas the ocular biomechanics weakened following orbital decompression surgery. Furthermore, we found that these changes in ocular biomechanical response parameters were associated with the refractive status of patients. Our results suggest that the IOP reduction recorded by non-contact tonometry after the orbital decompression surgery may be overestimated due to the weaker ocular biomechanics postoperatively. Hence, bIOP obtained using the Corvis ST, which takes into account corneal structure and response, may be a more appropriate method for measuring IOP in patients with TED. 

Acknowledgments

Disclosure: Y. Hsia, None; Y.-H. Wei, None; S.-L. Liao, None