In this study, we compared the thickness of retinal layers of patients with GD with GO, without GO and healthy controls. GCL, IPL, RPE of patients with GO were thinner than the layers of healthy controls, while RNLF, INL, OPL, ONL were similar in groups.
Casini and colleagues evaluated RNLF in patients with GO, patients with GD with no sign of orbitopathy and healthy controls. RNLF was similar in the three groups (p>0.05). Sayin and colleagues compared RNFL in GO and healthy controls . The inferior RNFL was slightly thinner in the patients with GO than controls (p=0.043). At the same time, superior, temporal and nasal RNFL were similar in GO and healthy controls (p>0.05). Also, Forte and colleagues  found no significant reduction in RNFL thickness between patients with GO and ocular hypertension and healthy controls. A recent study by Kurt and colleagues showed that RNLF was thinner only in the superior zone of the patients with GO when compared with healthy controls (p = 0.039). Similar values were noted in the temporal, nasal, and inferior areas . Meirovitch et al. demonstrated significant thickening of the RNFL in patients with GO compared to controls, unlike the other studies .
Casini et al. found a significant decline in mean central GCL thickness in patients with GO compared with healthy controls . Wang et al. compared RNFL and GCC (including the RNFL, ganglion cell layer and inner plexiform layer) between patients with active GO, Dysthyroid optic neuropathy (DON) and healthy controls . In active GO and DON, the blood supply to the superficial layer of the macular area is reduced; the smaller the blood vessel density, the thinner the RNFL and GCC. Romano and colleagues found the average of GCC was significantly lower (P=0.0005) in the patients with GO with optic nerve compression than in healthy controls. The average RNFL thickness was not different in patients with GO and healthy controls. Kurt and colleagues did not observe any statistically significant difference in the ganglion cell layer between the patients with GO and healthy controls . There are only a few studies that evaluate GCL separately from other layers in patients with GO. INL is evaluated within the GCC complex in only some studies, as shown above. To the best of our knowledge, this study is the first study that evaluates IPL alone apart from the GCC complex.
RPE has formed a monolayer of cells located between the retinal photoreceptors and the fenestrated choriocapillaris. RPE is essential for the maintenance and survival of the overlying photoreceptor cells and for organising the integrity of the choroidal capillaries. RPE was thinner in patients with GD (with and without GO) than in controls in our study. INL, OPL, and ONL were similar in three groups. No previous studies analysed the change in INL, OPL, ONL, and RPE patients with GD.
Our results showed no relationship between CAS and retinal layers in patients with GO. Also, there was no correlation between Hertel value and retinal layers, except ONL and RPE. Mugdha and colleagues found no correlation between CAS and RNFL . Casini et al. demonstrated no significant relationship between CAS, retinal thickness, GCL and Hertel value . There was no correlation between the severity of the orbital disease and the RNFL thickness in the study by Meirovitch et al. . ONL and RPE of the patients with GD are only evaluated in our study; further studies are needed to interpret this relationship.
In our study, TRAB, Anti-TPOAb, and Anti-TgAb positivity are not different in patients with and without GO. Lee and colleagues investigated the relationship between ophthalmopathy and TRAB, Anti-TPOAb, and Anti-TgAb in pediatric patients with GD . Similar antibody positivity was detected in patients with and without GO. The role of Anti-TPO Ab in autoimmune thyroid disease remains controversial. Wright-Pascoe et al. demonstrated that both TPO Ab and Tg Ab were correlated with the incidence of adulthood GO . In contrast, Khoo et al. demonstrated that TPO Ab negativity was associated with an increased risk of GO in adults . These studies assessed adult, adolescent and pediatric patients, in contrast to our study, which was limited to adult patients with GD.
This study has limitations. We have a relatively small number of cases because of our exclusion criteria and the limited GO range. However, our results reached a level of significance.
In summary, patients with GO have thinner GCL, IPL, and RPE than healthy controls, even without changes in visual acuity. We hope that evaluation of OCT images may yield early signs of optic neuropathy to allow for early treatment. Further investigations are needed to validate our findings and use these early findings in more timely treatment to prevent significant vision problems.