Visual Recovery Time in Patients with Ethambutol-induced Toxic Optic Neuropathy
Article information
Abstract
Purpose
We aimed to investigate the visual recovery time in patients with ethambutol-induced toxic optic neuropathy (EON) and identify the factors associated with the visual recovery time.
Methods
In this retrospective cohort study, we reviewed the medical records of 35 eyes from 35 patients with EON. Visual recovery was defined as a gain of three or more lines from the nadir.
Results
Patients were observed following discontinuation of ethambutol (EMB), with the mean follow-up period of 21.0 ± 16.0 months. The visual acuity at nadir was logarithm of the minimum angle of resolution 1.4 ± 0.4, and the final visual acuity was logarithm of the minimum angle of resolution 0.6 ± 0.5. Twenty-seven eyes (77.1%) showed significant visual recovery. In Kaplan-Meier survival, the mean estimated time for visual recovery was 15.2 ± 3.0 months, and 50% of the patients experienced visual recovery at 8.3 ± 2.2 months following EMB discontinuation. Multivariate Cox regression analysis identified several significant risk factors for delayed visual recovery, including duration of EMB medication ≤6 months, period from symptom onset to EMB discontinuation >14 days, and baseline peripapillary retinal nerve fiber layer thickness >98 μm.
Conclusions
Our study indicated a mean time of visual recovery of 15 months for EON cases. Therefore, patients diagnosed with EON should be followed up for more than 1 to 2 years to evaluate their visual recovery. Delayed EMB discontinuation, short duration of EMB use, and initial peripapillary retinal nerve fiber layer thickening were associated with delayed visual recovery. Therefore, patients taking EMB should be followed up regularly for early detection of EON and immediate discontinuation of EMB to prevent severe damage to the optic nerve.
Ethambutol (EMB) is one of the major antibiotics used in the treatment of Mycobacterium species. However, one of the main side effects of EMB is toxic optic neuropathy. The prevalence of EMB-induced toxic optic neuropathy (EON) among patients on EMB doses of 15 to 25 mg/kg/day is 1% to 2% [1–3]. The risk increases with high doses and has been reported to be up to 33% when taking more than 35 mg/kg/day [4–7]. Several risk factors of EON development other than high dose have been identified, such as old age, hypertension, renal dysfunction, and concomitant isoniazid (INH) therapy [1,2,8,9].
Since there is no effective treatment for EON, discontinuing EMB promptly remains the only effective strategy to prevent progression and facilitate visual recovery. Therefore, early recognition of risk factors and detection of EON is important. Although 30% to 64% of patients show some visual improvement over a period of several months with prompt discontinuation of EMB, the average visual recovery is two lines on the Snellen chart, and complete visual recovery is relatively rare [1,3,10–12]. The clinical factors associated with poor visual outcomes include old age, initial visual acuity (VA) worse than 6 / 60, and loss of temporal peripapillary retinal nerve fiber layer (pRNFL) [12–14].
However, the analysis of the visual recovery time in patients with EON is currently limited. Predicting the visual recovery time can offer valuable guidance to patients and clinicians who must wait for recovery after discontinuing the drug without any additional treatment. Therefore, in this study, we aimed to investigate the visual recovery time in patients with EON and identify the factors associated with this recovery.
Materials and Methods
Ethics statement
The current study is a retrospective cohort study, and the study protocol was reviewed and approved by the Institutional Review Board of Asan Medical Center (No. S2022-2642-0001). The study procedures were conducted in accordance with the tenets of the Declaration of Helsinki. The requirement for informed consent was waived due to the retrospective nature of the study.
Patients
We screened patients who visited the neuro-ophthalmologic clinic in Asan Medical Center (Seoul, Korea) from January 2003 to December 2022. The study included patients who were diagnosed with EON by a neuro-ophthalmologist. EON was diagnosed when the patients met the following criteria as described in previous studies: (1) emergence of visual symptoms following EMB treatment and (2) fulfillment of both major criteria (abnormal color vision test results with no other evident cause for abnormal color vision; central or paracentral scotoma on the visual field [VF] test), or one major criterion and two minor criteria (VF defect other than a central or paracentral scotoma; optic disc pallor) [1,15].
Exclusion criteria in this study were as follows: (1) systemic or other ophthalmological diseases that affect visual function, including color vision, VA, or VF test results; (2) follow-up period of less than 3 months after symptom onset; and (3) missing data on final visual outcomes.
Lastly, one eye with the worse VA at nadir was selected for the final analysis to eliminate any within-subject factors which may increase type 1 error. In cases where the final VA in both eyes was the same, we included the eye with worse VF test result in the analysis.
Neuro-ophthalmologic examinations
We reviewed patients’ medical records for basic demographic data such as age, sex, and medical history. Data on mycobacterial infection and treatment were also obtained. Patients underwent a comprehensive ophthalmic examination at the initial visit, including best-corrected VA measurement using Jin vision chart, intraocular pressure measurement, slit-lamp biomicroscopy, and f undus examination. The optical coherence tomography (OCT) images were obtained using a Cirrus HD OCT system (Carl Zeiss Meditec). Images with signal strength ≥6 were included, and the quality of each image and segmentation of the retinal layers were manually checked by a single author (HRA). The thickness of pRNFL and macular ganglion cell-inner plexiform layer (mGCIPL) were measured.
Statistical analysis
All data with continuous variables are presented as mean ± standard deviation. VA was converted to a logarithm of the minimum angle of resolution (logMAR). We defined “visual recovery” as a gain of three or more lines of log-MAR VA from the nadir, which present a doubling of the visual angle, to identify a truly significant vision improvement. In the cases where the visual acuity at nadir was worse than logMAR 1.3, we defined the visual recovery as the visual improvement with VA ≤logMAR 1.0. We performed Kaplan-Meier survival and Cox regression analyses to investigate the visual recovery course and identify the factors associated with the visual recovery. For the Cox regression analysis, we classified the patients into two subgroups based on the median value of each factor. Variables with a p < 0.1 in univariate analysis were included in the multivariate analysis. We considered p < 0.05 to be statistically significant. All statistical analyses were performed using IBM SPSS ver. 23.0 (IBM Corp).
Results
In this retrospective study, 35 eyes from 35 patients diagnosed with EON were included (Table 1). The mean age at diagnosis of EON was 63.6 ± 14.5 years (range, 25.8–85.9 years), and 23 patients (65.7%) were female. Eleven patients (31.4%) had preexisting hypertension and seven (20.0%) showed abnormal renal function. Among three patients under the age of 40 years, mitochondrial DNA sequencing was performed on two and confirmed negative. The remaining patient, who is undergoing dialysis due to renal disease, was diagnosed with EON, which manifested as a significant impairment in visual function, following the use of EMB.
Baseline demographic and clinical characteristics of total study population (n = 35)
All patients were treated for mycobacterial infections with drugs including EMB. Among the 18 patients whose drug usage was investigated, 12 (66.7%) concurrently used INH. The mean daily EMB dose was 18.1 ± 3.3 mg/kg/day (range, 12.3–25.5 mg/kg/day), and the mean duration of EMB medication was 6.5 ± 3.1 months (range, 2.7–14.9 months). The average period from symptom onset to drug discontinuation was 33.8 ± 32.2 days and the maximum period was 154 days. Patients were observed after discontinuation of the drug, and the average follow-up period was 21.0 ± 16.0 months (range, 3.0–59.6 months). The average VA at nadir was logMAR 1.4 ± 0.4 (range, logMAR 0.5–2.0). At baseline, the thicknesses of pRNFL and mGCIPL were 98.1 ± 9.8 μm (range, 83–119 μm) and 68.4 ± 9.3 μm (range, 52–80 μm), respectively.
The final VA was logMAR 0.6 ± 0.5 (range, logMAR 0–1.7), with 27 eyes (77.1%) exhibiting significant visual recovery during the follow-up period. Fig. 1 shows the Kaplan- Meier survival curve for visual recovery of all study patients, with the mean estimated time for visual recovery calculated at 15.2 ± 3.0 months. The median estimated time for visual recovery, which indicates the time when 50% of the patients show visual recovery, was 8.3 ± 2.2 months.
Survival curve for the visual recovery of ethambutol-induced toxic optic neuropathy.
The outcomes of the Cox regression analysis for all 35 subjects are summarized in Table 2. In the univariate analysis, several factors were significantly associated with a longer recovery time, such as concurrent use of INH, the period from symptom onset to EMB discontinuation >14 days, VA worse than logMAR 1.0, and baseline pRNFL thickness >98 μm Preexisting hypertension, abnormal renal function, and duration of EMB medication ≤6 months were correlated with a delay in visual recovery with borderline statistical significance with p-values less than 0.1. In the multivariate Cox regression analysis, duration of EMB medication ≤6 months, period from symptom onset to EMB discontinuation >14 days, and baseline pRNFL thickness >98 μm were significant risk factors for delayed visual recovery.
Factors associated with visual recovery time in EMB-induced toxic optic neuropathy (n = 35)
Discussion
In the current study, we investigated the longitudinal clinical course of EON. About 77% of patients showed significant visual recovery, which is a comparable result to previous studies [1,13,16]. However, in contrast to previous studies, we performed survival analysis to identify the visual recovery time. This analysis revealed that the mean estimated time for visual recovery was 15 months, and half of the patients achieved visual recovery within 8 months following drug discontinuation. This suggested that patients with EON should be followed up for more than 1 to 2 years to evaluate their visual recovery.
In the multivariate Cox regression analysis, discontinuing EMB more than 2 weeks after symptom onset delayed visual recovery. The hazard ratio of visual recovery in patients who stopped EMB more than 2 weeks after symptom onset was 0.34, signifying that if the patient continued to take EMB for more than 2 weeks after the onset of symptoms, the possibility of visual recovery would be reduced by 66%. Therefore, clinicians and patients should be aware that early detection of EON and immediate discontinuation of EMB after symptom onset are important factors for a favorable visual prognosis.
While VA at nadir was significantly associated with the visual recovery time in the univariate analysis, this relationship did not persist in the multivariate analysis. Previous studies on visual recovery in patients with EON also reported that worse initial VA was correlated with poor visual outcomes [13,16]. In this study, 16 of 21 eyes (76.2%) with VA at nadir worse than logMAR 1.5 discontinued EMB more than 2 weeks after symptom onset, suggesting that delayed drug discontinuation may have caused more severe visual impairment and worsened the final visual outcomes.
Although EON has dose-dependent toxicity [7,17], the daily EMB dosage was not a significant factor for visual recovery, as most patients had taken EMB at doses less than 25 mg/kg/day. Conversely, the duration of medication was a significant factor for visual recovery; a shorter duration of EMB medication was associated with delayed visual recovery. However, a longer duration of EMB was previously known as a risk factor for EON development [8,18]. The patients who developed EON despite the short duration of EMB use were probably vulnerable to EON with multiple risk factors related to the development of EON. This would have made them less likely to have visual recovery.
Several studies have investigated the change of pRNFL and mGCIPL thickness in patients with EON, with initial mGCIPL thinning suggested as a predictive factor for visual recovery [14]. In this study, the estimated recovery time for eyes with baseline mGCIPL thickness ≤70 μm was 13.7 ± 4.3 months and for the remaining participants was 7.1 ± 1.3 months. However, the observed difference between the two groups was not significant (p = 0.437). Conversely, we found a significant difference in visual outcome according to the baseline pRNFL thickness; initial pRNFL thickening was a risk factor for delayed visual recovery. Additionally, initial pRNFL thickness was the only difference between patients with and without visual recovery, regardless of the timing of visual recovery (Supplementary Table 1). In EON, the initial pRNFL thickness increased followed by a gradual decline [19,20]. Although the exact mechanism of EON remains unknown, the chelation of copper by EMB disrupts oxidative phosphorylation in mitochondria has been hypothesized [21]. In support of this theory, an increase in the initial pRNFL thickness has also been confirmed in one of the mitochondrial optic neuropathies, Leber hereditary optic neuropathy (LHON) [22,23]. Previous studies on visual recovery in LHON reported that the visual prognosis was worse in cases where changes in the peripapillary structures, such as telangiectasia and congestion, were prominent in the early stage [24]. The authors hypothesized that significant mitochondrial dysfunction could lead to respiratory chain defects that impair axonal plasma transport and cause marked peripapillary structural changes. Similarly, we found that the initial pRNFL thickening can delay visual recovery in EON. Peripapillary structural changes including pRNFL thickening may serve as surrogate markers of mitochondrial dysfunction, reflecting the severity of EON.
To the best of our knowledge, this study is the first of its kind to investigate the longitudinal course of visual recovery using survival analysis. However, the study had some limitations. First, owing to the retrospective design, some clinical data including the types of concomitant drugs and the results of OCT were missing. The timing of initial OCT was also inconsistent among the patients. However, since the timing of initial OCT was not correlated with the initial thickness of pRNFL (p = 0.590) and mGCIPL (p = 0.231), the timing of initial OCT is considered not to affect the results in this study (Supplementary Fig. 1). Second, the small number of patients due to the low incidence of EON may have affected the statistical significance. In particular, further studies with large number of patients are needed to determine the factors whose statistical significance was borderline in this study, such as hypertension and abnormal renal function.
In conclusion, the mean visual recovery time for EON was 15 months in our study, suggesting that patients with EON should be followed up for more than 1 to 2 years to evaluate their visual recovery. Delayed EMB discontinuation, short duration of EMB use, and initial pRNFL thickening were associated with delayed visual recovery. Therefore, patients taking EMB should be followed up regularly for early detection of EON, and immediate discontinuation of EMB should be done to prevent severe damage to the optic nerve.
Acknowledgements
None.
Notes
Conflicts of Interest: None.
Funding: None.
Supplementary Materials
Supplementary materials are available from https://doi.org/10.334/kjo.2023.0095.
Supplementary Table 1.
Comparison between patient with and without visual recovery
Supplementary Fig. 1.
Correlation between the timing of initial optical coherence tomography (OCT) with the initial thickness of (A) peripapillary retinal never fiber layer (pRNFL) and (B) macular ganglion cell-inner plexiform layer (mGCIPL).