Our cohort study showed that the short- and long-term efficacy of IVMP in treating DON was 52.54% in the first week; additionally, we identified that older age (≥55 years), long duration of onset (≥1 month), poor baseline VA (≥1 logMAR), and proptosis at baseline were predictors of poor response to IVMP in the first week.
Intravenous glucocorticoids have been used as the first line of treatment for DON by several previous studies and per the EUGOGO 2021 recommendation. (8-10, 17) The criteria of improvement are varied based on different parameters including BCVA, color vision, or visual field.(9, 10, 18, 19) Our present study set the criteria based on previous studies.(8, 9) We found that IVMP 1 gm for 3 consecutive days improved visual parameters with respect to both VA and proptosis. The improvement in vision after 1 week was significantly better than the improvement after 6 months. As corticosteroid efficacy is dependent on the dosage, their actions have processed via genome by binding at glucocorticoid receptors or express via other transcription factors when using lower dose 100 mg and those are long-term effect. For high dose of IVMP act via non-genome process by acting on decrease number and activity of plasma cell and dendritic cell which is a quick response within 1 day and 8 days.(20, 21) This theory is supported by Bartalena et al who found recurrence of DON after successful medical treatment in more than 30% patients from 1 week to several months after three consecutive doses of IVMP.(22) Although our study observed no recurrence of DON, the benefit of IVMP slightly decreased in the long-term. The improvement in proptosis by at least 2 mm was 23.73% (14/59 eyes) in our study. Our results are consistent with those of another multicenter study which showed that IVGC can improve proptosis by approximately 1 mm in a European population.(23)
Aging increases the risk of severity in thyroid eye disease as found by researchers in both Western and Eastern populations.(1, 16, 24, 25) The current study observed that older age (≥55 years) is a predictive risk factor of poor response to IVMP, in line with Barczyński et al’s findings, wherein young age was a positive predictor for visual recovery in Polish patients with DON.(26) Anderson et al. (27) proposed that different ages showed a variety of manifestations in either fat or muscle enlargement, i.e., younger subjects have fat predominance and older subjects have muscle enlargement. In addition, a recent study by Guo et al (28) discovered that the fat ratio in the orbit is a predictor of good response to IVGC in active GO. Those findings support the idea that young patients tend to have fat predominance and better response to IVMP in our population. However, this result is inconclusive and should be prospectively studied by analyzing both visual and radiological parameters to predict treatment outcome in the future.
Previous studies discovered that duration of disease is a crucial factor influencing the treatment outcome in TED or DON.(5, 26, 28) Barczyński et al (26) revealed that DON patients presented with 2 months duration gain more response rate compared patients presented with 7 months disease duration. Another study in Asians observed that having symptoms <1 year obtained more benefits of treatment in active TED (Thyroid eye disease).(28) Tagami et al (5) found that a 3-month disease duration showed significantly better visual outcome than a 6.5-month disease duration in DON patients. Similarly, our series observed that longer duration of optic neuropathy significantly increased unresponsiveness to IVMP by approximately 5 times. The explainable theory supported by a study on optic neuritis proposed the benefit of IVMP in the sense that they can relieve inflammation and preserve undamaged nerve fibers.(29)
Poor baseline VA was identified as a risk factor for unresponsiveness to IVMP in many studies.(5, 26, 28) Mckeag et al (30) found mild visual loss had better visual outcome in term of VA after IVGC. Tagami et al (5) observed that baseline BCVA worse than 0.7 logMAR was more prone to requiring orbital surgery owing to failure of IVMP therapy. Garip Kuebler et al (11) in their retrospectively review stated that a baseline BCVA of 0.3 logMAR obtained better therapeutic benefit than BCVA of >0.6 logMAR. Correspondingly, our study observed that poor BCVA ≥1 logMAR increased the OR of unresponsive outcome by 8.87-times in a multivariable analysis model.
The presence of proptosis in predicting response to IVGC is controversial. Wiersinga et al (31) proposed that the fibrotic stage in orbital fibroblasts of TED has poor response to corticosteroids. In addition, proptosis may not correlate with DON as proposed by Mckeag et al.(30) In a Chinese population, Guo et al (28) found that proptosis (>20.78 mm) was a significant positive factor to the response of IVGC in moderate-to-severe GO. Conversely, the present study found proptotic orbits (≥22 mm) significantly decreased the IVMP response by almost 5 times. It is difficult to draw a convincing conclusion, considering that Guo et al (28) studied active GO, while our study was on patients with DON; this may reflect the differences in severity and affect the treatment outcome.
Other predictive factors of poor response were analyzed. Sex, smoking, hyperthyroidism, type 2 DM, and disc swelling were not risk factors in our study. Smoking has a harmful effect on patients with GO due to several reasons. It increases oxygen free radicals in orbital tissues and lowers IL-1.(32, 33) Eckstein et al (34) found that smoking is a dose-dependent factor of unresponsiveness to immunosuppressant therapy, but smoking effect is no longer than 1 year. Bartalena et al (12) proved that tobacco use decreases the efficacy of fluocortolone; however, it has no impact on the final visual outcome. Recently, Xing et al (13) discovered that smoking increased the OR by 12.4 to poor response to IVGC in GO; however, they found no difference between current and ex-smokers. Interestingly, smoking was not a significant risk factor in our series. This is likely because the number of smokers in our study was few, which further reduced the analytical power.
Hyperthyroidism and hypothyroidism have been proposed as risk factors for the severity of GO owing to activation of TSH receptors and release of oxygen free radicals.(35, 36) Kung et al (37) found no relationship between T3 or T4 and severity of GO in an Asian population. Roy et al (38) and Wang et al (39) found contradictory results and showed that baseline increase in T4 and euthyroid status affected the efficacy of IVGC in active GO. Recently, a Japanese study discovered instability of thyroid status influenced worse response to IVMP and was more prone to requiring orbital decompression in DON patients.(5) Interestingly, our study found no association of euthyroid status in predicting treatment response. Although Balazs et al (40) proposed that methimazole has immunosuppressive benefits and patients who received it may gain better therapeutic outcome that those that did not. We did not study the effect or dosage of anti-thyroid medication, because most cases were on thyroid medication prescribed by their primary hospitals, rather than from our center.
Type 2 DM has been significantly observed in GO or DON in several studies.(1, 14) Its impact on final visual outcome is different among each study. Kalman et al (41) and Jeon et al (16) found no difference in visual function. By contrast, a study by Ramamurthy et al (42) revealed increased prevalence and severity of DM in TED and found that 9 out of 10 patients with DON had type 2 DM. Correspondingly, Rath et al (43) discovered that DM is a risk factor for DON in the Indian population and tended to have worse final visual outcome. Although type 2 DM was frequently observed in our series, it was not a significant predictor of poor response in the study.
Currò et al (9). discovered disc swelling at initial diagnosis was a risk factor for unresponsiveness to IVGC in DON patients. Surprisingly, although disc swelling did not reach statistical significance in binary logistic regression analysis in our study, it seemed to be a protective factor of poor response. This explanation may not be the actual biological reason, but it may be because of improved awareness of the ophthalmologist, given that the presence of disc swelling is a very specific finding in DON disease, and it is hence likely that those patients were received earlier diagnosis and treatment than those in the normal disc group.(30)
The strength of our study is its large sample size and single-center design. The following are some of the limitations of the study: first, because of its retrospective nature, the outcome of some visual parameters such as color vision and visual field were not studied after the treatment because of missing data. Second, since our referred patients already received anti-thyroid medication from their primary hospitals, we could not study the benefit of methimazole as an immunosuppressive agent. However, our study identified older age (≥55 years), longer onset duration before treatment (≥1 month), poor baseline BCVA (≥1 logMAR), and proptosis (≥22 mm) as risk factors for unfavorable treatment response with IVMP in DON in Thai populations.