Clinical profile and visual outcomes in ethambutol-induced toxic optic neuropathy

Abstract

Purpose:

To characterize the severity of vision loss, dose-related adverse effects, and visual outcomes in ethambutol-induced toxic optic neuropathy (ETON) patients.

Methods:

A prospective observational study was conducted in a tertiary eye center in South India between July 2019 and December 2021. Consecutive adults (age >18 years) taking revised antitubercular regimens and diagnosed with ETON were included. Data collected regarding patient demographics, presentation, dose-related severity of vision loss, and visual outcomes after drug discontinuation were analyzed.

Results:

We analyzed 214 eyes from 107 patients (mean age: 51.1 ± 13 years; male: female ratio = 2:1). The median follow-up time was 11 months (interquartile range: 4–14 months). Mean visual acuity improved from 1.2 ± 0.6 log of minimum angle of resolution (logMAR) at presentation to 0.8 ± 0.6 logMAR at the final follow-up. Patients above 60 years old had the worst visual acuity at presentation and had the least improvement at the last follow-up compared to those in younger age groups (1.3–1.2 logMAR, P = 0.2). The mean exposure time to ethambutol was 6.8 ± 3.3 months, and the mean dose was 1042 ± 247 mg/day, that is, 16.7 ± 4 mg/kg/day. Mean presenting best corrected visual acuity (BCVA; 1.3 ± 0.6 logMAR) was worse in patients taking a higher dose of ethambutol (≥16 mg/kg) compared to those taking <16 mg/kg (mean BCVA: 1 ± 0.5 logMAR).

Conclusion:

Higher age and higher doses of ethambutol were associated with an increased risk of vision loss and worse vision at presentation. Prolonged exposure to high-dose ethambutol, as in fixed drug dose combinations, leads to severe visual impairment which might be partially reversible.

Ethambutol is a synthetic bacteriostatic antitubercular drug used against mycobacterial infections. It is one of the least toxic first-line antituberculous agents.[1] However, numerous studies, including case reports and series, have reported partially reversible vision loss in patients exposed to ethambutol, as its chelating properties cause optic and chiasmal neuropathy. Its ocular side effects were first reported in the 1960s.[2] The Indian TB Report 2022 highlighted that India bears the world's highest tuberculosis (TB) burden and is implementing the largest TB control program to combat the disease effectively.[3] The Revised National Tuberculosis Control Program (RNTCP) administers antituberculosis therapy (ATT) in a fixed-dose combination (FDC) to patients enrolled under supervision for better compliance. In 2016, the ethambutol regimen was modified, extending its use from 2 to 6 months and increasing the frequency from thrice weekly to daily.[4] This has resulted in a significant rise in cases of subacute vision loss due to ethambutol-induced optic neuropathy.[5] Therefore, there is a need for a prospective study to look for risk factors associated with subacute vision loss and its severity in patients on ATT, to develop targeted preventive strategies and reduce ocular side effects. Previously published studies have been mostly retrospective in their designs and have provided limited information on the dose of ethambutol used, its duration, and its impact on visual outcomes. In addition, very few studies have reported the high-risk characteristics of patients having ethambutol-induced optic neuropathy. The current study aims to address these gaps by providing more comprehensive information on these critical aspects, including the relationship of final visual outcomes with initial structural changes in optical coherence tomography (OCT). In this study, we report the demographic profile, association with the duration and dose of ethambutol, clinical characteristics, and visual outcomes of patients diagnosed with ethambutol-induced toxic optic neuropathy (ETON) and treated under the revised RNTCP program for pulmonary and extrapulmonary TB.

Methods

Study design and ethical considerations

This prospective observational study was conducted between July 2019 and December 2021 after obtaining approval from our institute’s ethics committee. It adhered to the principles of the Declaration of Helsinki. Written informed consent for examination and study participation was obtained from all participants before enrollment.

Study participants: inclusion and exclusion criteria

Consecutive patients with ethambutol-related toxic optic neuropathy, experiencing acute/subacute vision loss while on FDC ATT per RNTCP 2017 criteria, were enrolled. Patients with central nervous system TB, recurrent pulmonary/extrapulmonary TB, or coexisting ocular conditions (e.g., diabetic macular edema, ocular TB) were excluded.

Data collection and follow-up

Demographic data, ethambutol dosage, duration, and patient weight were recorded. Visual assessments included acuity, color vision, visual fields (Humphrey Visual Field [HVF] analyzer), and OCT of retinal nerve fiber layer (RNFL) and macular ganglion cell layer (mGCL). Dosage was calculated based on the number of ATT tablets (275 mg ethambutol per tablet). Upon diagnosis, ethambutol was discontinued, alternative regimens were advised, and vitamin supplements were prescribed. Patients were followed up at least three times post-discontinuation. All were included regardless of follow-up duration, though only those with at least three follow-up visits were analyzed for visual outcomes.

Sample size calculation

Based on the previous reported studies, the test–retest variability of the visual acuity (VA) charts is >2 lines, therefore we chose an effect size (Δ) of 0.3 log of minimum angle of resolution (logMAR). Given the effect size of 0.3 logMAR and a standard deviation (σ) of 0.6 logMAR in VA across the highest and lowest ethambutol doses, with 80% power (Zβ) and significance (Zα) at 0.05, the sample size was calculated as 110 participants across groups.

Formula:

Specifically, in this formula, σ (standard deviation) reflects variability in VA among patients with similar conditions as per previous retrospective studies[9,12] and Δ (effect size) represents the threshold considered meaningful in terms of visual loss.

Given the variability in patient cohorts and the lack of robust previous data on ETON patients, these values were chosen conservatively to ensure statistical power without underestimating the potential effect.

Statistical analysis

Data were analyzed using Microsoft Excel 2019. The Shapiro–Wilk test assessed normality. Parametric tests were used to compare means between two independent/dependent groups, with one-way analysis of variance for comparisons among more than two groups. A P value <0.05 was considered statistically significant. Multivariate linear regression analyzed the relationships between age, VA, ethambutol dose, mGCL, temporal RNFL, and final visual outcome in patients with at least 3 months of follow-up.

Results

Demographic characteristics: Two hundred and fourteen eyes from 107 patients were included in the study during the study period. Male patients represented a significantly larger proportion of the study population, and the male-to-female ratio of the cohort was 74:33. The mean age at presentation was 51.1 ± 13.5 years (range: 18–80 years), and 58% (62/107) of the participants were middle aged (36–55 years) [Table 1]. The mean best corrected visual acuity (BCVA) at presentation was comparable among all age groups statistically (P = 0.2), but was clinically worse in the elderly age group (>60 years, 1.3 logMAR) [Fig. 1]. Similarly, at the last follow-up (minimum of 3 months), the improvement in VA was comparable in all age groups (P = 0.09), though clinical improvement was least in the age group above 60 years (1.3 to 1.2 logMAR) [Fig. 1].

Table 1.

Demographic details of patients diagnosed of ethambutol-induced optic neuropathy

Characteristics		No.(%)
Mean age of the population (n=214 eyes of 107 patients)		51.1±13 years
Male: female		74:33
Age groups across the study cohort
a. Less than 40 years     b. 40–60 years     c. More than 60 years		8 (7%) 62 (58%) 37 (35%)
Mean duration of ethambutol taken before presentation (months)		6.8±3.29
Mean duration between start of ethambutol and onset of DOV (months)		5.5±2.5 months
Mean cumulative dose of ethambutol consumed before the onset of decrease in vision (mg)		1042±247
Average dose of ethambutol/kg body weight		16.7±4.03
Bilateral involvement		214 eyes (100%)
Mean BCVA at the time of presentation (logMAR)		RE: 1.2±0.6 OS: 1.2 ± -0.6
Mean BCVA at the last visit (minimum 3 months follow-up) (logMAR)		0.81 ± 0.7 (range: 0–3)
Median follow-up duration (months) (IQR)		11 (4–14)
Visual field defects
a. Central/paracentral     b. Bitemporal     c. Advanced		62 eyes (42%) 38 eyes (26%) 48 eyes (32%)
HVF MD
a. Right eye     b. Left eye		-13.4±8.3 dB -13.2±9.1 dB
Optic nerve at the time of presentation
a. Normal     b. Pallor     c. Pseudo-edema		48 eyes, (24%) 162 eyes (78%) 4 eyes
OCT characteristics
a. Mean RNFL thickness at presentation     b. Mean mGCL thickness at presentation		RE: 95.03±22 μm|LE: 95.31±21 μm RE: 64.76±13 μm|LE: 65.2±13 μm
Ocular findings
a. Normal fundus     b. Diabetic retinopathy     c. Disc hemorrhage		87% 9% 4%
MRI findings, n=130 eyes (65 patients)
a. Normal     b. Optic nerve hyperintensities T2W     c. T2W optic nerve and chiasmal hyperintensities     d. Only chiasmal hyperintensities T2W		40 (62%) 11 (17%) 10 (15%) 4 (6%)

BCVA=best corrected visual acuity, DOV= Diminution of vision, HVF=Humphrey Visual Field, IQR=interquartile range, LE=left eye, logMAR=log of minimum angle of resolution, MD=mean deviation, mGCL=macular ganglion cell layer, MRI=magnetic resonance imaging, OCT=optical coherence tomography, RE=right eye, RNFL=retinal nerve fiber layer, T2W=T2 weighted

Figure 1.

The mean best corrected visual acuity at presentation and the last follow-up visit (minimum 3 months follow-up) in different age groups

Ethambutol dose: All the patients included in the study were on a fixed dose combination of ATT, as per the new RNTCP guidelines (2017). The mean duration of exposure to ATT, including ethambutol, taken before presentation was 6.8 ± 3.3 months (range: 2–20 months). However, the average duration between the start of ATT and the onset of a decrease in vision was 5.5 ± 2.5 months. The average dose of ethambutol (by 71 patients for whom exact dose schedule was available) taken by the study participants before presentation was 1042 ± 247 mg/day. The average weight (n = 71) of the cohort at presentation was 61.8 ± 14 kg. Thus, the calculated dose per kilogram body weight of the patients was 16.7 ± 4.03 mg/kg body weight/day (range: 5.7–25 mg/kg body weight/day). Based on the dose per kilogram body weight of ethambutol, the proportion of patients on >25 mg/kg, 16–<24.9 mg/kg, and <15.9 mg/kg was 2%, 60%, and 38%, respectively. Mean presenting BCVA (1.3 ± 0.6 logMAR) was worse in patients taking a higher dose of ethambutol (≥16 mg/kg) compared to those taking <16 mg/kg (mean BCVA: 1 ± 0.5 logMAR). In addition, there was a significant difference in the final VA between the groups at their last visit (0.9 ± 0.7 logMAR vs. 0.8 ± 0.7 logMAR in >16 and <16 mg/kg, respectively; P = 0.02).

Vision loss due to ethambutol toxicity: The mean BCVA at presentation was 1.2 ± 0.6 logMAR in either eye (range: 0–3 logMAR). All patients presented with bilateral diminution of vision; however, asymmetrical decrease (over two-line difference) in vision was seen in 24 patients (22%).

Profound visual impairment (BCVA ≤20/200) was seen in 81% (87/107) of participants. However, there was no statistically significant difference in BCVA among the different age groups at the time of presentation (P < 0.2) [Fig. 1].

Ocular findings at presentation

Optic nerve head findings: Temporal or diffuse optic disc pallor was observed in 78% of patients. four eyes in two patients presented with grade 1 disc edema. Forty-eight eyes (22%) showed a normal optic nerve at presentation.

Visual fields: Visual fields were tested for 148 eyes from 74 patients. Central scotoma (42%, 62 eyes) was the most common type of field defect, followed by advanced field loss (32%, 48 eyes) and bitemporal field defect (26%, 38 eyes). At presentation, the average mean deviation for HVF performed was -13.4 and -13.2 dB in the right eye (RE) and left eye (LE), respectively [Table 1].

Other ocular examination findings: All patients (100%) had normal intraocular pressure at presentation. The retinal examination was normal for 186 eyes from 93 patients (87%); mild–moderate diabetic retinopathy without diabetic macular edema was seen in 28 eyes from 14 patients (13%), and isolated disc hemorrhage was seen in two eyes from two patients.

OCT, RNFL, and mean mGCL thickness: The average peripapillary RNFL thickness was 95.03 ± 22.23 and 95.31 ± 21.17 µm in RE and LE, respectively. The mean mGCL thickness was 64.76 ± 13 and 65.2 ± 13 µm in RE and LE, respectively. The average peripapillary RNFL thickness was within the normal range for age; however, the mean mGCL thickness was significantly lesser compared to age-matched controls, as per published literature.

Visual evoked potential (VEP): VEP was performed for 46 eyes from 23 patients at the initial presentation or during evaluation. There was increased P100 implicit time and reduced amplitudes in nine patients. Pattern VEP with extinguished responses was seen in seven patients (those with severe vision loss).

Follow-up and improvement in BCVA post-discontinuation of ethambutol: The median follow-up time was 11 months (interquartile range: 4–14 months; range: 1–29 months). Only 64 patients (60%) followed up for at least 3 months. This was mainly due to travel restrictions caused by the coronavirus disease 2019 (COVID-19) pandemic. The mean VA at the last follow-up visit for patients (minimum 3 months follow-up) was 0.8 ± 0.7 logMAR (range: 0–3 logMAR). At least two-line improvement was seen in 72 eyes from 36 patients (56%) at their last visit. Moreover, a five-line improvement in the early treatment diabetic retinopathy study (EDTRS) chart was seen in 40 eyes from 20 patients (31%).

Multivariate linear regression analysis of factors affecting the final VA in patients with ETON

In our study, we assessed the impact of presenting BCVA, temporal RNFL thickness, mean ganglion cell-inner plexiform layer (GCIPL) thickness, and ethambutol dose per kilogram body weight on the final BCVA [Table 2]. The only significant predictor was the mean VA at presentation, which showed a strong correlation with the final visual outcome (coefficient = 1.074, P < 0.01). Other factors, including age, weight, ethambutol dosage per kilogram body weight, mean RNFL and mean mGCL thicknesses, had no significant association with the outcome. This suggests that initial VA is a key determinant in predicting visual recovery in ETON patients.

Table 2.

Multivariate linear analysis comparing various dependable (predictor) factors affecting final visual outcome in patients with ethambutol-related toxic optic neuropathy

Variable		Coefficients		Standard error		P		95% CI
								Upper limit		Lower limit
Age		0.004		0.01		0.65		0.022		-0.014
Ethambutol dose/body weight		-0.028		0.03		0.34		0.033		-0.090
Mean VA at presentation		1.074		0.18		0.00		1.456		0.692
Mean temporal rnfl thickness at presentation		-0.002		0.01		0.64		0.009		-0.015
Mean mgcl thickness at presentation		-0.00191		0.01		0.61		0.010		-0.016

CI=confidence interval, mGCL=macular ganglion cell layer, RNFL=retinal nerve fiber layer, VA=visual acuity

Discussion

This is the first major prospective observational study on patients with ETON in southern India. Previous studies were retrospective and lacked detailed information on how ethambutol dosage and duration affected visual outcomes [Table 3]. Garg et al.[6] conducted a prospective trial in northern India, with patients enrolled in revised national tuberculosis control programme directly observed treatment short-course (RNTCP-DOTS). Over a 2-month follow-up period, they found that 4% (6/127) of eyes experienced reduced vision due to ethambutol toxicity and 6% (8/127) had field defects. However, this study predates the revised RNTCP guidelines where ethambutol is administered daily in FDCs for a prolonged period. In addition, their study included much younger individuals with a mean age of 34 years and poor baseline VA, which the authors did not explain. In contrast, the mean age of patients in our study was 51 ± 13.5 years, with most of them in the age group of 40–60 years (58%). A few studies in the past showed a similar age distribution. Talbert Estlin and Sadun[7] reported that 40% of patients who developed ETON were over 65 years old, even though there was a declining trend in the incidence of TB worldwide.

Table 3.

A review of literature of baseline characteristics and factors affecting visual outcomes in patients with ethambutol-related toxic optic neuropathy

Author (no. of patients)		Year		Mean age group (years)		Ethambutol dose (mg/kg)		Visual field defects		Baseline VA (logMAR)		Change in VA after stopping ethambutol		Follow-up (months)		Ethambutol duration (months)		Avg. weight/cumulative dose		Remarks		Outcomes/comment
Chai and Foroozan[8] (n=8)		2007		69.5		NA		Centrocecal scotomas		20/40, CF at 3 ft		Improvement in six patients		Mean 8.1		Mean 8.9		NS		Disc changes observed; decrease in RNFL thickness		EON shows significant RNFL thickness decrease even after discontinuation
Lee et al.[9] (n=89)		2008		31–86		17.85±2.21		Central or paracentral scotomas		Loss at 7.31±9.45 months		Improvement in six eyes; no improvement in optic disc pallor		12.54±9.97		9.38±10.12		NS		Renal dysfunction, dose contributes to EON; incidence <2%		Recovery possible in minority, influenced by renal function and dosage
Menon et al.[10] (n=52)		2009		11–56		15–20 (2 months)		Peripheral isopter contraction (Goldmann perimetry) (7.69%)		NS		Three lines improvement		2		NS		Increased latency, RNFL thinning		Reversal of observed toxicity in 80% of eyes after 1 month of stoppage; sensitive tests needed		Pattern-VER and visual field examinations are sensitive tests to detect early toxicity; OCT may help identify patients at risk
Estlin and Sadun[7] (n=16, + meta-analysis n=54)		2010		59±15.25		14.5–21.4		Central scotoma		NS		NA		NA		NA		Suggests dose optimization		Optimization of dosing protocols needed		Customizing the dose of ethambutol can decrease the incidence of toxicity
Garg et al.[6] (n=64)		2015		34.23±15.54		>20		Acuity loss, field defects		0.26±0.39		Loss noted after 2 months		NA		2		As per 2015 RNTCP guidelines		Ocular symptoms led to discontinuation; improvement observed		Ethambutol under DOTS regimen causes ocular toxicity; early recognition crucial; guidelines needed for routine checkups
Chen et al.[11] (n=62)		2015		70.02±14.26		16.06±4.36		Field loss (19/21 tested)		0.95		0.62		6		5.94±4.06		3479.46±2189.16		Emphasizes early detection; EON in Taiwan 1.29%		Half showed visual improvement post-ethambutol cessation; no prognostic factors identified
Lee et al.[12] (n=84)		2019		Age 45.50±17.17 years		14.72±3.07		Peripheral Constriction, altitudinal defect		NA		NA		4.31±2.42		Lower cumulative dose associated with subclinical toxicity		Increased mean temporal RNFL thickness post-administration		Subclinical EON in 13% eyes; cumulative dose and duration are risk factors; recovery in 73% post-stoppage		Lesser cumulative dose, more likelihood of recovery
Varan et al.[13] (n=10)		2020		38.77±14.86		15		Cecocentral scotoma		NA		NA		4.44±1.94		2–8 (range)		NA		Long-term use causes RNFL thinning; SD-OCT useful		Significant thinning in temporal quadrant observed
Ambika et al.[14] (n=128)		2022		55.5±12.9		19.7±5.3		Cecocentral scotomas, defects		0.25±0.25		0.45±0.36		21.0±16.0		9.6		NA		Early diagnosis crucial; gradual recovery observed		Early detection and cessation prevents significant vision loss
Kumar et al.[27] (n=14)		2023		29–76 years		1078.5 mg (21 mg/kg)		Central scotoma, hemianopia, coNAtriction		Normal to blindness		Poor recovery post-ethambutol		3		4.81±1.99		NA		Comorbidities increase risk; fundus examination findings		Dose-dependent EON; patients with comorbidities at higher risk; need for ophthalmological examination
Sharma et al.[15] (n=50)		2023		18–72 years		15–20 mg/kg/day		Affected, but not described in detail		0.00		1.08 (±0.40)		6 + additional 3 months for observed reversibility		4–5		NA		6% of patients aged 65 and above developed irreversible optic nerve damage		Need for careful monitoring of visual function NA in older patients

CF=Counting finger, DOTS=Direct Observed Therapy Short course, ETON=Ethambutol induced toxic optic neuropathy, logMAR=log of minimum angle of resolution, NA=Not applicable, NS=Not specified, OCT=optical coherence tomography, RNFL=retinal nerve fiber layer, RNTCP=Revised National Tuberculosis Program, SD-OCT=Spectral domain-optical coherence tomography, VA=visual acuity, VER= Visual evoked response

Chen et al.[11] reported that final VA was worse in patients who were over 65 years old compared to younger age groups. In contrast to our findings, a retrospective study from South India reported that only 27% of patients with ETON were over 65 years.[14] Post-2000 population studies in India show TB incidence declines with age, peaking in adolescents and in the 25–44 years age group. Therefore, age may influence the risk and outcomes in patients with ETON.

In our study, ethambutol dosage and duration were identified as significant risk factors for both the occurrence and severity of visual loss (mean BCVA: 1.2 ± 0.7 logMAR). Most of our patients received doses ranging from 16 to 25 mg/kg body weight over an average treatment duration of 5.28 ± 1.87 months. Chen et al.[11] reported an increased risk of ETON with doses exceeding 800 mg/day. In addition, Fraunfelder et al.[16] reported a positive correlation between the duration of administration and dosage of ethambutol and the risk of ETON. While previous studies reported longer durations before ETON onset,[14] our patients experienced early visual deterioration, possibly due to a higher proportion of elderly individuals in our cohort. Moreover, our study participants received larger ethambutol doses, exceeding World Health Organization recommendations, for longer durations, as guided by the revised RNTCP guidelines.[5]

ETON may be partially reversible, especially with early drug discontinuation. Previous studies report a 30%–64% improvement in VA. Chen et al.[17] found a 50% recovery (≥2 Snellen lines) in half of their 16 patients after 6 months of follow-up. Ambika et al.[14] reported that 63% of their patients had an improvement of two lines or more after discontinuation of ethambutol. They reported better visual outcomes in patients with a follow-up of over 1 year, suggesting that vision may continue to improve even after prolonged discontinuation of ethambutol. In our study, 56% (n = 36) of patients improved by over two lines on the logMAR chart after 3 months of follow-up. However, patients over 60 years of age showed no improvement (P = 0.66), while those aged 40–60 years and under 40 years had significant improvement (P < 0.0001). These findings align with Srithawatpong et al.,[18] who reported better visual recovery in patients under 60 years, though their results were not statistically significant.

Optic atrophy at presentation predicts poor visual recovery in ETON patients.[1] In our study, 78% had disc pallor, consistent with minimal visual improvement reported in the literature.[9,14] Common visual field defects include central scotomas (42% in our cohort) and bitemporal hemianopia (26%), which aligns with previous findings.[10,14,19,20] Neuroimaging helps rule out other causes of optic neuropathy; 60 of our patients underwent magnetic resonance imaging, with 18 showing optic nerve hyperintensities and three showing chiasmal involvement.[19,21,22,23]

Table 3 summarizes key findings from published studies on ethambutol-induced optic neuropathy over the past two decades, with sample sizes ranging from 10 to 84 participants. Despite being retrospective in design, these studies show comparable trends to our cohort, especially regarding the higher susceptibility of older age groups to ethambutol toxicity. Given that six studies predate the revised 2017 RNTCP guidelines and only three of the more recent studies originate from the Indian subcontinent,[12,14,15] direct comparisons with our prospective study are limited. Our study provides new insights into predictive factors for visual recovery and a detailed analysis of the dose–effect relationship, with nearly two-thirds of patients having documented ethambutol doses before presentation. Unlike any previously published studies, our study highlights the importance of structural parameters such as OCT RNFL and mGCL at presentation and their correlation to the final visual outcome.

Linear regression analysis showed that mean BCVA at presentation had a strong positive and statistically significant association with the outcome (coefficient = 1.074, P < 0.01). Other variables like age, weight, ethambutol dose/body weight, and thickness measures (RNFL, mGCL) showed weak associations and were not statistically significant (P > 0.3); the associations were not meaningful, given the lack of significance.

Ethambutol treatment is linked to long-term RNFL thinning, as noted by Varan et al.[13] and Chai and Foroozan.[8] In our cohort, eyes without disc pallor had thicker temporal RNFL than those with pallor (P = 0.02). Geyer et al.[24] found the GCIPL thickness to be a better early ETON marker than RNFL. Han et al.[25] and Teng et al.[26] also reported GCIPL thinning during treatment. In our study, GCIPL thickness was significantly lower than normal (64 ± 13 µm). A study from eastern Asia indicated that GCIPL thickness below 64 µm suggests ETON severity.[12] We found a negative correlation between final VA and mGCL/RNFL thickness, though no significant association was found in multivariate linear regression analysis [Figs 2 and 3].

Figure 2.

Correlation between visual acuity (BCVA) in the last visit and average mGCL thickness at the time of presentation. BCVA = best corrected visual acuity, mGCL = macular ganglion cell layer

Figure 3.

Correlation between visual acuity (BCVA) at the last visit and average RNFL thickness at the time of presentation. BCVA = best corrected visual acuity, RNFL = retinal nerve fiber layer

Conclusion

Our study is one of the foremost to prospectively observe patients with ETON in FDC as per the revised RNTCP guidelines. The study has a few limitations, including high attrition rates and COVID-19–related travel restrictions that hindered regular interval investigations and outcome analysis. The lack of accurate ethambutol dosage data for all patients, inconsistent data on the dose of ethambutol per kilogram body weight, and non-emphasis on compliance further limited our analysis, precluding consideration of confounding factors, such as hypertension, diabetes mellitus, renal function, or baseline health parameters. Furthermore, it was not feasible to correlate the effect of body weight and other concomitant risk factors such as age and renal dysfunction. Another limitation of the study was the varied duration of follow-up within the cohort, mainly due to COVID-19 restrictions. This resulted in a wide range of follow-up periods (1–29 months, median of 11 months), which may have contributed to variability and potentially impacted the assessment of the final visual outcomes. Although our eye care center serves patients from across the country, the study population is skewed toward South and East India, limiting its representation of diverse demographics and practice patterns. Despite these limitations, the current study with a large sample size, with relatively longer follow-up, sheds light on the impact of ethambutol administered in FDC on vision loss severity and recovery. OCT findings of RNFL and macular GCIPL thickness may serve as crucial indicators for early diagnosis and good prognosis of ETON. The shift toward FDC-based TB management has increased the prevalence and severity of ETON, warranting prudence in ethambutol dosing adjustments based on weight to mitigate cumulative dose effects. Early screening and prompt drug discontinuation remain essential for preventing severe vision loss.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Hyderabad Eye Research Foundation (HERF) provided the financial support.