Optical Coherence Tomography Angiography in Patients of Thyroid-Associated Ophthalmopathy

According to TAO eyes, Twenty-three eyes (57.5%) had active TAO (CAS ≥ 3), whereas 17 eyes (42.5%) had inactive TAO (CAS < 3). Hertel exophthalmometric readings in TAO group ranged between 16 to 23 with a mean 20.9 (Figure 1).

No statistically significant results were encountered when FAZ of superficial (p = 0.08) and deep retinal layers (p = 0.64) were compared between both groups (Figure 2, Table 3).

Figure 2A: FAZ area calculation in TAO study case

Figure 2B: FAZ area calculation in healthy control 

Figure 2 Shows insignificant differences between groups according FAZ

Table 3. Area of Foveal Avascular Zone. FAZ of superficial and deep retinal layers in thyroid-associated ophthalmopathy cases and control groups (n=80 eyes).

A statistically significant difference regarding total peripapillary thickness between both groups was exhibited, with lower level in TAO case group (105.03 ± 9.4). There is also a statistically significant difference between both groups regarding RNFL thickness-Superior, with lower level in TAO case group (128.9 ± 15.6). There was also significant difference between both groups regarding RNFL thickness-Nasal, with lower level in TAO case group (78.1 ± 10). There is no statistically significant difference between TAO case and control groups regarding RNFL thickness inferior or temporal, (Figure 3, Table 4).

Figure 3: Retinal Nerve Fiber Layer Thickness among thyroid-associated ophthalmopathy cases and control groups.

Table 4. Retinal Nerve Fiber Layer thickness among thyroid-associated ophthalmopathy cases and control groups (n=80 eyes).

TAO: thyroid associated- ophthalmology.          RNFL: Retinal Nerve Fiber Layer Thickness

Data expressed as mean ± SD.                             t: independent samples -t- test.

*: significant p ≤ 0.05.

There was statistically significant difference between active and inactive TAO patients as determined by one-way ANOVA regarding IOP (active vs inactive: 20.1 ± 2.4 vs 17.2 ± 1.4; p < 0.001) and Hertel exophthalmometric readings (active vs inactive: 21.8 ± 1.7 vs 19.5 ± 1.5; p < 0.001), with higher values in active group. While there is no statistical correlation observed between both groups regarding RNFL thickness, FAZ area, superficial and deep capillary plexus PD, optic nerve head and peripapillary PD values.

There is statistically significant difference between active, inactive and normal control groups regarding IOP with the highest level (20.1 ± 2.4) in active TAO group. There is also statistically significant difference between the groups regarding superficial capillary plexus, deep capillary plexus, optic nerve head and peripapillary perfusion density values with the highest level of perfusion among normal control group (44.8 ± 1.9, 43.7 ± 2.4, 53.7 ± 2.6, 53.7 ± 2.1) respectively (Table 5,6).

Table 5. Superficial and deep capillary plexus perfusion density among thyroid-associated ophthalmopathy cases and control groups (n=80 eyes).

  TAO: thyroid associated- ophthalmology.               PD: Perfusion Density

Data expressed as mean ± SD or number (%).            Optic Nerve Head and

Table 6. Peripapillary perfusion density among thyroid-associated ophthalmopathy cases and control groups (n=80).

TAO: thyroid associated- ophthalmology.                            PD: Perfusion Density

Data expressed as mean ± SD or number (%).                         t: independent samples -t- test.

*: significant p ≤ 0.05.

A significant positive moderate correlation was encountered between Hertel exophthalmometric readings and IOP (r = 0.64) (Table, Figure). There is also a significant positive strong correlation between Hertel exophthalmometric readings and Clinical Activity Score (r = 0.76). While a negative exophthalmometric readings and total peripapillary thickness (r = -0.35), and there is significant negative weak correlation between Hertel exophthalmometric readings and RNFL thickness-Superior (r = -0.39). Regarding FAZ-Superficial retinal layers, a positive moderate correlation (r = 0.33) was displayed with exophthalmometric readings. While there is no statistically significant correlation between Hertel exophthalmometric readings and other parameters.

Statistical correlations have been explored between BCVA log MAR and other parameters. There is a negative weak correlation between BCVA log MAR and RNFL thickness -Nasal (r= -0.32). No statistical correlations were maintained between BCVA log MAR and superficial capillary plexus, deep capillary plexus, optic nerve head or peripapillary perfusion density values. Additionally; No statistical correlations were encountered between BCVA log MAR and CAS, Hertel exophthalmometric values or IOP (Table 7-9, Figure 4,5).

Table 7: Association between active and inactive TAO cases and other parameters in TAO case group (n=40 eyes).

TAO: thyroid associated- ophthalmology.                    RNFL: Retinal Nerve Fiber Layer Thickness

BCVA: Best Corrected Visual Acuity.          FAZ: Foveal Avascular Zone.      PD: Perfusion Density

t: independent samples -t- test.                      c²: chi-square test.    *: significant p ≤ 0.05.

Table 8. Comparison between active, inactive TAO cases and control groups regarding all parameters (n=80 eyes).

Table 9.  Spearman’s Correlation between BCVA and other parameters in TAO case group (n=40 eyes). 

Not significant: p > 0.05.                                                                r: correlation coefficient.

TAO: thyroid associated- ophthalmology.            RNFL: Retinal Nerve Fiber Layer Thickness

BCVA: Best Corrected Visual Acuity.          FAZ: Foveal Avascular Zone.     PD: Perfusion Density

Figure 4: Correlation between Hertel exophthalmometric readings and IOP in TAO case group.

 Figure 5A: Female patient aged 36 years, with thyroid-associated ophthalmopathy with Bilateral Eye lid edema, upper and lower lid retraction, conjunctival injection, inflammation of the caruncle and dryness. OCTA images of;  superficial capillary plexus and FAZ area. Deep capillary plexus and FAZ area (image below).

Figure 5B: OCTA images of; (A) optic nerve head. (B) Radial capillary peripapillary plexus (Left eye).

Figure 5C. OCTA Printout (A. Left macula & B. left optic Discs)

DISCUSSION

OCTA is a novel imaging modality that offers noninvasive fundus angiographs. It works without a contrast medium, so it protects against dye allergies and possible adverse reactions. OCTA relies on the intrinsic motion of blood in the retinal vessels as it separates stationary structures to identify the blood flow. OCTA can also provide 3D images compared to 2D images provided by fundus fluorescein angiography and indocyanine green angiography. It is reported that early changes in TAO could be well investigated and monitored with the objective parameters of OCT and OCTA [8,9].

In the current study, we investigated the structural and hemodynamic changes in twenty thyroid-associated ophthalmopathy patients (40 eyes) using OCT and OCTA. The patients were recently diagnosed and free from any associated comorbidities or other ocular diseases. They were subclassified according to clinical activity score (CAS) into two groups. Active group (n=23 eyes) with score equal to or more than 3, and inactive group (n=17 eyes) with score less than 3. Twenty healthy individuals (40 eyes) with nearly matched age were recruited and have been investigated with the same OCT and OCTA parameters.

Regarding OCT parameters, it is known that retinal nerve fiber layer (RNFL) thickness shows variations according to age and ethnicity [10]. That's why we used our own reference range from the healthy, age matched control group. As a result, we found a statistically significant difference in the average total peripapillary thickness in case group compared to healthy controls (P= 0.04) with lower levels in case group (105.03 ± 9.4). The superior and nasal quadrants showed decreased levels in TAO case group (P= 0.02) in both parameters. However, no statistically significant difference in RNFL thickness has been noted in our study between active and inactive TAO cases. Previous studies suggested that the enlarged extraocular muscles and increased orbital tissue volume might be the cause of direct compression on the optic nerve and thinning of the RNFL. Additionally, peripapillary RNFL thinning would be an early sign of optic nerve involvement [8,9]. Other causes of RNFL thinning in TAO patients have been documented in previous studies including; hypoxic ischemia from increased intraocular pressure and associated decrease in retinal vascular perfusion [11].

In our current study, we documented that, there is no statistical significant difference between case and healthy control groups regarding FAZ area of superficial and deep retinal layers (P= 0.08), (P= 0.64) respectively. Furthermore, our study showed no statistical significant difference between active and inactive TAO cases regarding FAZ area in superficial (P= 0.11) and deep retinal layers (P= 0.59), respectively. In contradiction to our results, enlargement of FAZ area was documented in previous research indicating a sign of capillary non perfusion in TAO patients [8,12].

The perfusion density values of the superficial and deep retinal layers were quantitatively evaluated in our study. The perfusion density represents the total area of the perfused vessels per unit area within an OCTA measurement area [13].

Our results revealed statistically significant difference between TAO case group and healthy control group regarding superficial and deep capillary plexus perfusion density values (P ≤ 0.001) in both parameters, with lower levels in TAO case group (40.3 %± 3.2%), (38.2% ± 4.5%) respectively. Whereas our results did not show any difference between active and inactive cases regarding the perfusion density of either superficial (P= 0.47) or deep retinal layers (P= 0.53).

The study of retinal microvascular density changes in TAO patients revealed inconsistent results. Jamshidian et al. 2019, conducted a study to compare macular and peripapillary vasculatures in active TAO, not active and not compressive TAO, and control eyes [14]. They have concluded that whole macular and parafoveal superficial vessel density values as well as average peripapillary vessel density value were significantly lower in active TAO patients compared with the other groups. Similarly, Dave et al. [15] reported that peripapillary and macular area vascularization were reduced in patients with active TAO compared to healthy controls.

Akpolat et al. [16], recruited only inactive TAO patients in their study, they noted alterations in foveal and parafoveal vessel density in patients within active TAO. They have demonstrated a decrease in retinal vessel diameter and an increase in macular vessel density in inactive TAO patients.

Furthermore, Yu L et al. [9], have used OCT-A imaging to study the retinal microvasculature in active TAO patients. They demonstrated that the macular vessel density was altered in both superficial and deep retinal layers. They observed an increase in retinal vascular density in inactive TAO cases compared to healthy controls, along with significant changes in RNFL thickness, choroidal thickness and superficial retinal vessels [9].

The present study evaluated OCTA as a noninvasive modality in screening TAO patients for sub-clinical compressive optic neuropathy. The compression in the orbit puts pressure on the venous system, causing retro blood flow and increased episcleral venous pressure. It is a possible explanation for reduced macular and ONH perfusion in patients with TAO [12].

 In our current study, we found a statistically significant difference in optic nerve head and peripapillary perfusion density in TAO patients compared to healthy subjects with decreased levels in TAO cases (48.4 ± 5.0 vs 53.7 ± 2.6), (48.3 ± 4.6 vs 53.7 ± 2.1) respectively.

Previous studies have evaluated the optic nerve vascular perfusion in TAO patients. Mihailovic et al. [7], and Jian et al. [11] reported a decrease in the ONH perfusion and peripapillary vessel density in TAO cases which is consistent with our findings.

Conversely, Sabermoghaddam et al. 2022, demonstrated no difference between TAO cases and healthy controls regarding optic nerve headand radial peripapillary vessel density. However, they found that patients with higher CAS had a significant decrease in ONH microvasculature [17].

In another study by Del Noce et al. [18], a significant decrease in the peripapillary and deep macular capillary plexus vessel density was detected in TAO patients compared to healthy age and gender-matched controls. The study demonstrated a significant relationship between higher CAS scores and lower peripapillary vessel density [18]. Furthermore, Jian et al. [11] did not detect any significant difference between active and inactive TAO patients regarding the peripapillary perfusion density which is in line with our current results.

Peripapillary VD in TAO patients is easily affected by intraocular and orbital pressure and appears earlier than visual impairment [13]. The previous literature reported that optic disc function might be affected before clinical deterioration of vision occurs, and therefore early detection of optic nerve damage could be essential in preventing significant vision loss [17].

In the current study, we demonstrated a statistically significant difference between TAO case group and normal control group regarding BCVA log MAR with reduced values in TAO case group (0.5 vs 1) respectively. Moreover, we investigated the association between OCT and OCTA parameters and the visual outcome. The results revealed no association except for a negative weak correlation between BCVA log MAR and RNFL thickness - Nasal (r = -0.32, P= 0.04).

Consistent with our findings, Del Noce et al., 2022 observed changes in peripapillary microvasculture in TAO patients which did not interfere with visual acuity, [18] as well as Akpolat et al. [2] who didn't find any association between visual acuity and increased vessel density in their study on inactive TAO cases. Whereas, Yu et al. [8] suggested that visual acuity might be affected by altered vessel density.

High intraocular pressure is a common finding in patients with TAO. Irreversible orbital changes such as extraocular muscle fibrosis and fatty degeneration can lead to an increase in the orbital volume with restriction of the exophthalmos process. As a result of that, aqueous humor outflow resistance and secondary elevation of the IOP can occur [11].

Consistent with the previous literature, IOP elevation is noted in TAO group in our present study. IOP levels showed a statistically significant difference between TAO case group and healthy controls (18.9 ± 2.5 vs 15 ± 2.2) respectively. Additionally, the study reported higher levels of IOP among the active TAO case group compared to the inactive group (20.1 ± 2.4 vs 17.2 ± 1.4) respectively.

Proptosis is a result of the expansion of orbital fat and muscle tissue into the bone structure. In the previous literature, proptosis was reported higher in TAO patients with dysthyroid optic neuropathy (DON) than those who do not have DON (Yang et al., 2023). In the present study, Hertel exophthalmometryvalues were significantly higher in active TAO patients than inactive group (21.8 mm ± 1.7mm), (19.5mm ± 1.5mm) respectively.

Moreover, a statistical significant positive moderate correlation between Hertel exophthalmometric readings and IOP (r= 0.64, P ≤ 0.001) was documented in our results. As well as a statistical significant positive strong correlation between Hertel exophthalmometric readings and Clinical Activity Score (r= 0.76, P ≤ 0.001). While there is significant negative moderate correlation between exophthalmometric readings and total peripapillary thickness (r = -0.35, P= 0.03), and there is a statistical significant negative weak correlation between exophthalmometric readings and RNFL thickness– Superior (r = -0.39, P= 0.01).

Differences in sample size, age and sex, racial diversity, parameter settings of OCTA system and inclusion criteria for TAO patients may partially explain the discrepancy among previous studies. In addition, technical differences and variations in patient activity status from moderate to severe may also contribute to the inconsistent results. Future research with more meticulous design is needed to validate the predictive and diagnostic potential of OCT and OCTA in Thyroid associated orbitopathy patients.

In conclusion, the current study revealed a reduction in retinal microvascular perfusion in superficial and deep retinal layers as well as optic nerve head and peripapillary region in TAO patients using OCTA.

Reduced peripapillary perfusion density is considered a promising prognostic indicator of optic neuropathy in TAO, which can also identify early subclinical dysthyroid optic neuropathy (DON). The study revealed no changes in perfusion density values between active and inactive TAO patients. TAO Patients who have high IOP are at risk for reduced retinal vascular perfusion and nerve fiber layer thickness

REFERENCES

1. Abrishami M, Sabermoghaddam A, Salahi Z, Bakhtiari E, Motamed Shariati M. (2023). Macular microvasculature in patients with thyroid associated orbitopathy: a cross-sectional study. Thyroid Res. 16(1):31.
2. Akpolat C, Kurt MM, Yılmaz M, Ordulu F, Evliyaoglu F. (2020). Analysis of Foveal and Parafoveal Microvascular Density and Retinal Vessel Caliber Alteration in Inactive Graves' Ophthalmopathy. J Ophthalmol. 2020:7643737.
3. Al-Sheikh M, Ghasemi Falavarjani K, Akil H, Sadda SR. (2017). Impact of image quality on OCT angiography based quantitative measurements. Int J Retina Vitreous. 3:13.
4. Anvari P, Ashrafkhorasani M, Habibi A, Falavarjani KG. (2021). Artifacts in Optical Coherence Tomography Angiography. J Ophthalmic Vis Res. 16(2):271-286.
5. Aumann S, Donner S, Fischer J, Müller F. (2019). Optical Coherence Tomography (OCT): Principle and Technical Realization. In: Bille JF, editor. High Resolution Imaging in Microscopy and Ophthalmology New Frontiers in Biomedical Optics [Internet]. Cham (CH): Springer; 2019. Chapter 3.
6. Baeg J, Choi HS, Kim C, Kim H, Jang SY. (2023). Update on the surgical management of Graves' orbitopathy. Front Endocrinol (Lausanne). 13:1080204.
7. Mihailovic N, Lahme L, Rosenberger F, Hirscheider M, Termühlen J, Heiduschka P, et al. (2020). Altered retinal perfusion in patients with inactive Graves ophthalmopathy using optical coherence tomography angiography. Endocr Pract. 26(3):312-317.
8. Yu L, Jiao Q, Cheng Y, Zhu Y, Lin Z, Shen X. (2020). Evaluation of retinal and choroidal variations in thyroid-associated ophthalmopathy using optical coherence tomography angiography. BMC Ophthalmol. 20(1):421.
9. Yanni Y, Qian W, Wenbin W. (2024). Peripapillary RNFL cross-sectional area and its association with other parameters in a Chinese population. BMC Ophthalmol. 17;24(1):260.
10. Mugdha K, Kaur A, Sinha N, Saxena S. (2016). Evaluation of retinal nerve fiber layer thickness profile in thyroid ophthalmopathy without optic nerve dysfunction. Int J Ophthalmol. 9(11):1634-1637.
11. Jian H, Wang Y, Ou L, He W. (2022). Altered peripapillary vessel density and nerve fiber layer thickness in thyroid-associated ophthalmopathy using optical coherence tomography angiography. Int Ophthalmol. 42(3):855-862.
12. Xu B, Wang S, Chen L, Tan J. (2023). The early diagnostic value of optical coherence tomography (OCT) and OCT angiography in thyroid-associated ophthalmopathy. Ther Adv Chronic Dis. 20406223231166802
13. Zhang T, Xiao W, Ye H, Chen R, Mao Y, Yang H. (2019). Peripapillary and Macular Vessel Density in Dysthyroid Optic Neuropathy: An Optical Coherence Tomography Angiography Study. Invest Ophthalmol Vis Sci. 60(6):1863-1869.
14. Jamshidian Tehrani M, Mahdizad Z, Kasaei A, Fard MA. (2019). Early macular and peripapillary vasculature dropout in active thyroid eye disease. Graefes Arch Clin Exp Ophthalmol. 257(11):2533-2540.
15. Dave TV, Laghmisetty S, Krishnamurthy G, Bejjanki K, Ganguly A, Jonnadula GB, et al. (2022). Retinal vascularity, nerve fiber, and ganglion cell layer thickness in thyroid eye disease on optical coherence tomography angiography. Orbit. 41(2):170-177.
16. Akpolat C, Kurt MM, Yılmaz M, Ordulu F, Evliyaoglu F. (2020). Analysis of Foveal and Parafoveal Microvascular Density and Retinal Vessel Caliber Alteration in Inactive Graves' Ophthalmopathy. J Ophthalmol. 2020:7643737.
17. Sabermoghaddam A, Shoeibi N, Jafarzadeh H, Bakhtiari E, Salahi Z, Saeidi Rezvani T, et al. (2022). Optic nerve head optical coherence tomography angiography findings in patients with thyroid eye disease: a case-control study. Thyroid Res. 15(1):17.
18. Del Noce C, Roda M, Ferro Desideri L, Traverso CE, Vagge A. (2022). Evaluation of macular blood flow after intermittent intravenous infusion of high-dose corticosteroids (pulse therapy) in patients with thyroid-associated orbitopathy (TAO) using angio-OCT. Graefes Arch Clin Exp Ophthalmol. 260(2):571-576.