Background: The incidence, clinical manifestations, and contributory factors in ethambutol-induced optic neuropathy (EON) have not been widely studied.
Methods: A retrospective chart review of 857 patients who took ethambutol for treatment of tuberculosis identified 89 patients who complained of decreased vision after initiation of treatment and were therefore referred to the ophthalmology clinic of Seoul National University Hospital, Korea, between January 2003 and December 2005. Information was extracted in regard to systemic illnesses, duration and dose of ethambutol therapy, visual acuity, color vision, ophthalmoscopic examination, visual fields, and visual evoked potentials (VEPs).
Results: EON was diagnosed in 13 (1.5%) patients during a follow-up period of 12.54 ± 9.97 months. The average dose of ethambutol was 17.85 ± 2.21 mg/kg/day, and the duration of therapy was 9.38 ± 10.12 months. Ophthalmic findings included decreased visual acuity (65.4%), abnormal visual fields (65.4%), abnormal color vision (61.5%), optic disc pallor (38.5%), or increased latency on VEP tests (65.4%). Slightly less than one third of patients showed improvement in visual function after discontinuing ethambutol. The latency for recovery was 5.38 ± 1.71 months. No patient with optic disc pallor at the time of diagnosis of EON showed visual function improvement. Renal dysfunction and the daily dose of ethambutol, but not the duration of treatment, contributed to EON.
Conclusions: Based on this study, the incidence of EON in Koreans is estimated to be <2%. However, visual function after discontinuation of ethambutol is reversible in only a minority of patients and does not occur if optic disc pallor is present. Renal dysfunction and daily dose of ethambutol, but not duration of ethambutol treatment, seem to be related to development of EON.
Department of Ophthalmology (EJL, S-JK, HKC, JHK, YSY), Seoul National University College of Medicine, Seoul, Korea; and Seoul Artificial Eye Center (S-JK, HKC, JHK, YSY), Seoul National University Hospital Clinical Research Institute, Seoul, Korea.
Address correspondence to Seong-Joon Kim, MD, Department of Ophthalmology, Seoul National University College of Medicine, 28, Yongon-dong, Chongno-gu, Seoul 110-744, Korea; E-mail: email@example.com
This study was supported by Grant 21-2005-003-0 from the Seoul National University Hospital research fund.
Ethambutol hydrochloride is commonly used for the treatment of tuberculosis (TB) and Mycobacterium avium complex infections (1). Its exact mechanism is not known, but the drug may act as a chelating agent in disrupting one of several metal-containing enzyme systems in the nucleic acid structures of mycobacteria (2,3).
Ever since the introduction of ethambutol treatment, drug-related ocular toxicity has been well recognized, and the clinical manifestations of ethambutol-induced optic neuropathy (EON) have been reported in several studies (4-10). The Physicians’ Desk Reference even provides guidelines for clinicians to screen for optic nerve toxicity (11).
There have been several studies related to the incidence of EON, but all had small numbers of patients (5,6,9,10,12-14). The purpose of this study was to evaluate the incidence and clinical manifestations of EON in the Korean population and to identify the factors that influence the development of this side effect.
A retrospective chart review of 857 patients who started ethambutol treatment for TB infection between January 2003 and December 2005 at Seoul National University Hospital, Korea, identified 89 patients who complained of decreased vision and were referred to an ophthalmology clinic. Patients with other ophthalmologic problems related to visual loss (optic neuropathies other than EON, cataract, diabetic retinopathy, or glaucoma) were excluded. Charts were reviewed for the following data: age, gender, major target organ of TB infection, total duration and dose of ethambutol treatment, and the presence of other underlying diseases such as those of the kidney and liver, visual evoked potential (VEP) test results, tests of visual acuity, visual field, and color vision, and optic disc abnormalities.
We diagnosed EON according to the following criteria (12-17). The visual symptoms had to have appeared only after the start of ethambutol therapy. The diagnosis had to satisfy more than 1 of the major criteria and more than 2 of the minor criteria, including minor criterion 1. Major criteria are: 1) abnormal results on the Hardy-Rand-Rittler (HRR) color vision test and no other reasonable cause for abnormal color vision, and 2) central or paracentral scotoma on the Goldmann or Humphrey perimeter. Minor criteria are: 1) visual field defects other than central or paracentral scotomas, and 2) optic disc pallor. Exclusion criteria included other ophthalmologic problems that could cause visual loss (diabetic retinopathy, cataract, or glaucoma) or optic nerve head abnormalities (glaucoma, hereditary or acquired optic neuropathy, or optic neuritis). Visual acuity was excluded from the criteria because there was little information regarding its baseline value before the medication was begun.
As soon as the patients developed ophthalmic symptoms, they were referred to the ophthalmology clinic, with the time interval between the onset of symptoms and the ophthalmologic examination being no greater than 6 weeks. The basic ophthalmologic evaluation, including visual acuity, color vision, and optic disc evaluation, was performed every 3 months. Visual fields were checked at least every 6 months. The VEP study was performed only at the initial visit.
The relationship between the daily dose of ethambutol and renal function and the incidence of EON was analyzed with the statistical software program SPSS V13.0 (SPSS Inc, Chicago, IL) using the Mann-Whitney test. The 76 patients with visual complaints who did not meet criteria for a diagnosis of EON were used as control subjects.
Incidence of EON
Of 857 patients identified as having been treated with ethambutol for TB, only 13 (1.5%) who were referred to the ophthalmology clinic satisfied our diagnostic criteria for EON. The ages of affected patients ranged between 31 and 86 years (mean 58.23 ± 16.68 years); 9 were men and 4 were women. The mean dose and duration of ethambutol were 17.85 ± 2.21 mg/kg/day (800 mg/day) over 9.38 ± 10.12 months. The period of follow-up was 12.54 ± 9.97 months (Table 1).
Visual acuity loss was counted if it had exceeded 2 Snellen chart lines between the last ophthalmologic examination before ethambutol was started and the first examination after the medication was started or if it was 20/40 or worse in the absence of other causal factors. The symptoms of decreased visual acuity included cloudy or blurry vision, difficulty reading, faded color, or frequent changes in prescription for eyeglasses or contact lenses.
Visual acuity loss was documented at 7.31 ± 9.45 months (range 1-36 months) after ethambutol treatment was started. Of 26 eyes, 17 (65.4%) showed decreased visual acuity. In 4 patients, visual acuity loss was bilaterally symmetric; in 6 patients, it was asymmetric. It varied from 20/15 to counting fingers, with the mean being 20/120 ± 0.85 (log minimum angle of resolution [MAR] −0.77 ± 0.85).
Five eyes of 3 patients showed improvement in visual acuity after ethambutol was discontinued. On average, the visual acuity improved from 0.53 ± 0.65 to −0.50 ± 0.62 (visual acuity by log MAR ± SD, P < 0.05), but improvement was documented in only 6 eyes of 4 patients.
The initial appearance of the optic disc was found to be related to visual recovery (P = 0.02, 2 test) in that none of the 6 patients with optic disc pallor recovered visual acuity in follow-up. The initial visual field and color vision had no significant relationship with visual recovery. The daily and cumulative dose of ethambutol, duration of medication treatment, and glomerular filtration rate (GFR) were not significantly related to recovery of visual acuity either (P = 0.94, 0.83, 0.94, and 0.26, respectively, Mann-Whitney test).
Seventeen (65.4%) eyes of 9 patients displayed central or paracentral scotomas on the Goldmann or the Humphrey perimeter. The defects were bilateral in 8 patients and unilateral in 1 patient. Nine eyes had central scotomas; the remaining eyes exhibited peripheral constriction. One patient showed bilateral altitudinal defects in the superior hemifield. The visual fields of 7 of 9 patients with defects had improved during the follow-up period; 5 of these 9 patients stopped ethambutol medication 5.8 ± 3.56 months before the resolution of visual defects was documented. The other two patients did not stop ethambutol because of disease severity (Case 11) or flouting of the doctor's recommendation (Case 6). However, the routine anti-TB regimen was discontinued 3 and 10 months before the last visual field examination, respectively.
Among 22 patients who exhibited abnormal results in the HRR color vision test, 16 (61.5%) eyes of 8 patients satisfied the diagnostic criteria for color vision loss attributable to EON. Average color vision for each patient at initial and final examinations was 13.0 ± 7.08/20 (range 0-20) and 16.11 ± 6.45/20 (range 0-20), respectively. The condition was bilateral in all patients. Ten eyes showed impairment only in red-green color perception, and the others showed impairment in blue-yellow color perception as well. The color vision in 7 eyes of 4 patients normalized or improved at a mean of 5.33 ± 0.58 months after discontinuation of ethambutol.
Optic disc abnormalities were observed in 17 (65.4%) eyes. The most common finding was pallor in 10 (38.5%) eyes. Optic disc edema was present in 2 eyes. Among the patients who stopped taking ethambutol, only 1 patient later demonstrated a normalized optic disc. The pallor was unchanged in 1 patient and worse in 5 patients in the follow-up period.
Seventeen (65.4%) eyes of 8 patients showed delayed p100 VEP latency (127.70 ± 18.46 ms in pattern VEP and 140.90 ± 12.78 ms in flash VEP). In 2 patients (4 eyes) there was no response in pattern VEP, whereas p100 latency was delayed in flash VEP, and in 1 patient (2 eyes), the amplitude of the flash VEP was extinguished, whereas there was p100 delay in pattern VEP.
The daily dose and cumulative dose of ethambutol were 17.85 ± 2.21 and 5,224.39 ± 5,776.50 mg/kg, respectively. The latency between onset of treatment and diagnosis was 9.38 ± 10.12 months. The daily dose of ethambutol was significantly higher in patients with EON (17.85 ± 2.21 mg/kg) than in the 76 patients who were treated with ethambutol and had visual symptoms but did not meet the criteria for a diagnosis of EON (14.47 ± 3.46 mg/kg) (P < 0.001, Mann-Whitney test). The duration of ethambutol therapy was not significantly different between the patients with EON (9.38 ± 10.12) and those without EON (8.64 ± 8.52) (P > 0.05). All but 3 patients (Case 6, 7, and 11) stopped ethambutol medication as soon as possible after the diagnosis of EON. One patient (Case 11) had multidrug-resistant TB, so it was considered too risky to stop the ethambutol. The other 2 patients (Case 6 and 7) ignored the recommendation to stop the medication.
GFR was significantly reduced in patients with EON (44.33 ± 17.21 mL/min) compared with that for patients without EON (71.4 ± 27.8 mL/min) (P < 0.05). EON incidence was significantly higher in patients with decreased renal function (P = 0.002, Mann-Whitney test).
An increase in visual acuity of ≥2 Snellen lines was considered “visual acuity recovery.” An expansion of the visual field by >5 was considered “visual field recovery.” According to these criteria, 6 (60%) of 10 eyes (4 patients) that had visual loss showed improvement in visual acuity at 5.38 ± 1.71 months after discontinuation of ethambutol. The visual fields of 7 (77.8%) of 9 patients with defects had improved 5.8 ± 3.56 months after discontinuation of ethambutol. Clinical manifestations in the patients with EON are summarized in Table 2.
In this study, the incidence of EON was 1.5%. The daily dose but not the duration of ethambutol treatment contributed to EON. Visual function improved in less than one third of patients after discontinuation of the medication. If visual function was to improve, it did so within 9 months after ethambutol was discontinued. However, none of the patients with optic disc pallor at time of diagnosis of EON showed visual function improvement during follow-up.
In past studies, the incidence of EON was reported to be approximately 1% (6,8,18). In 1998, a report stated that its incidence was in the range of 1%-5% (19). Although we found an incidence of 1.5%, it may actually be higher if our accrual methods overlooked some patients who had EON.
The incidence of EON has been reported to be correlated with daily dose and duration of ethambutol treatment (6,20). Although we could not find any correlation between the duration of treatment and the development of EON, the daily dose of ethambutol was significantly higher in those who developed EON, averaging 17.85 ± 2.21 mg/kg.
As ethambutol is rapidly absorbed and mostly excreted via the urine, the dose must be adjusted according to the creatinine clearance. Some investigators believe that patients with impaired renal function should not be treated with ethambutol because of the potential for toxicity and the difficulty in regulating appropriate serum drug levels (21). In the present study, the development of EON was related to decreased renal function as GFR was significantly reduced in patients who developed EON (44.33 ± 17.21 mL/min). These results imply that patients with decreased renal function should be monitored more frequently and carefully by ophthalmologists.
According to previous reports, EON develops from 15 days to 2 years after initiation of ethambutol treatment (5,10,13). The mean interval between starting medication and documentation of EON has been 3.4 (22) to 5 (23) months. In the present study, visual symptoms developed at 7.31 ± 9.45 months (range 1-36 months) after treatment began. This interval could be considered the latency between the initiation of treatment and development of EON because patients were referred to the ophthalmology clinic not more than 6 weeks after the onset of ocular symptoms. The longer latency of EON development revealed in this study may be related to the relatively larger number of patients who needed a longer duration of anti-TB medication. For example, Case 2 had been treated with ethambutol for 3 years to control recurrent pulmonary TB, and her visual symptoms started 36 months after initiation of treatment.
The ophthalmic toxicity of ethambutol is thought to be reversible by discontinuation of the drug (24). However, only 6 eyes of 4 subjects-less than one third of our cohort-demonstrated visual acuity recovery, whereas others showed stationary or worsened vision. This finding is similar to that of previous reports, which have emphasized the irreversibility of visual loss associated with EON (21-23). Kumar et al (22) described 7 patients treated with 25 mg/kg/day ethambutol, among whom only 3 had a documented gain in visual acuity to better than 20/200 after 8.3 ± 2.1 months. Tsai and Lee (23) described a series of 10 patients with EON, only 5 of whom had any degree of improvement in visual acuity with a follow-up period of 21.8 ± 8.8 months. DeVita et al (21) described 2 patients with irreversible vision loss. Lim (14) reported that only 1 of 3 patients regained visual acuity to a near-normal level 9 months after discontinuation of ethambutol. In our study, we could find no correlation between cumulative dose of ethambutol and development of EON. Therefore, we conclude that apart from daily dose and renal dysfunction, unknown factors appear to account for the irreversibility of this toxic side effect.
Ethambutol toxicity affects the small-caliber papillomacular bundle axons, and optic disc pallor does not develop until months after the fibers are lost (17,25-27). The clinical signs of damage from ethambutol toxicity are said to be frequently absent on examination of the fundus (19,25,28). However, nearly half of our patients already had optic disc pallor at the time of diagnosis of EON. At the final ophthalmic examination, optic disc pallor was generally stable or worse even though ethambutol had been discontinued. The single patient whose optic disc pallor improved (Case 6) had bilateral optic disc edema at the initial examination and promptly stopped ethambutol therapy. About 3 months later, the visual field had normalized. The fact that none of the 6 patients with optic disc pallor had visual improvement during follow-up suggests that this finding indicates a poor prognosis.
Some authors have proposed the use of electrophysiological testing to screen for EON. A multifocal electroretinogram (ERG) may be valuable for diagnosing and monitoring EON (29), and full-field ERG and electro-oculograms have also revealed abnormalities (29,30). VEP can detect subclinical optic nerve damage, although some authors say that this technique is insensitive and has a high false-positive rate (31).
This study has some limitations. Although we included a large number of patients who had undergone ethambutol therapy at a large referral hospital in Korea, our study was not a multicenter one. Moreover, some patients might have dropped out during follow-up, and there might be unreported cases. For this reason, we describe our incidence as a minimum number.
1. Jorgensen JH, Turnidge JD. Antimicrobial agents and susceptibility testing. In: Murray PR, Baron EJ, Pfaller MA, et al., eds. Manual of Clinical Microbiology
. 7th ed. Washington, DC: ASM Press; 1999: 1467-662.
2. Forbes M, Peets EA, Kuck NA. Effect of ethambutol on Mycobacteria. Ann N Y Acad Sci
3. Addington WW. The side effects and interactions of anti-tuberculosis drugs. Chest
4. Phillips PH. Toxic and deficiency optic neuropathies. In: Miller NR, Newman NJ, Biousse V, et al., eds. Walsh and Hoyt's Clinical Neuro-Ophthalmology
. 6th ed. Baltimore: Lippincott Williams & Wilkins; 2005:447-64.
5. Carr RE, Henkind P. Ocular manifestations of ethambutol, toxic amblyopia after administration of an experimental antituberculous drug. Arch Ophthalmol
6. Barron GJ, Tepper L, Iovine G. Ocular toxicity from ethambutol. Am J Ophthalmol
7. Nair VS, LeBrun M, Kass I. Peripheral neuropathy associated with ethambutol. Chest
8. Citron KM. Ethambutol: a review with special reference to ocular toxicity. Tubercle
9. Roussos T, Tsolkas A. The toxicity of Myambutol on the human eye. Ann Ophthalmol
10. Karmon G, Savir H, Zevin D, et al. Bilateral optic neuropathy due to combined ethambutol and isoniazid treatment. Ann Ophthalmol
11. Physicians' Desk Reference
. 56th ed. Montvale, NJ: Medical Economics Company; 2002.
12. Melamud A, Kosmorsky GS, Lee MS. Ocular ethambutol toxicity. Mayo Clin Proc
13. Choi SY, Hwang JM. Optic neuropathy associated with ethambutol in Koreans. Korean J Ophthalmol
14. Lim SA. Ethambutol-associated optic neuropathy. Ann Acad Med Singapore
15. Fraunfelder FW, Sadun AA, Wood T. Update on ethambutol optic neuropathy. Expert Opin Drug Saf
16. Zoumalan CI, Agarwal M, Sadun AA. Optical coherence tomography can measure axonal loss in patients with ethambutol-induced optic neuropathy. Graefes Arch Clin Exp Ophthalmol
17. Sadun AA, Win PH, Ross-Cisneros FN, et al. Leber's hereditary optic neuropathy differentially affects smaller axons in the optic nerve. Trans Am Ophthalmol Soc
18. Citron KM, Thomas GO. Ocular toxicity from ethambutol. Thorax
19. Sivakumaran P, Harrison AC, Marschner J, et al. Ocular toxicity from ethambutol: a review of four cases and recommended precautions. N Z Med J
20. Leibold JE. The ocular toxicity of ethambutol and its relation to dose. Ann N Y Acad Sci
21. DeVita EG, Miao M, Sadun AA. Optic neuropathy in ethambutol-treated renal tuberculosis. J Clin Neuroophthalmol
22. Kumar A, Sandramouli S, Verma L, et al. Ocular ethambutol toxicity: is it reversible? J Clin Neuroophthalmol
23. Tsai RK, Lee YH. Reversibility of ethambutol optic neuropathy. J Ocul Pharmacol Ther
24. Gorbach SL, Barglett JG, Blacklow NR. Infectious Diseases
. 2nd ed. Philadelphia: WB Saunders Co; 1998:413-4.
25. Carelli V, Ross-Cisneros FN, Sadun AA. Optic nerve degeneration and mitochondrial dysfunction: genetic and acquired optic neuropathies. Neurochem Int
26. Sadun A. Acquired mitochondrial impairment as a cause of optic nerve disease. Trans Am Ophthalmol Soc
27. Sadun AA. Mitochondrial optic neuropathies. J Neurol Neurosurg Psychiatry
28. Kozak SF, Inderlied CB, Hsu HY, et al. The role of copper on ethambutol's antimicrobial action and implications for ethambutol-induced optic neuropathy. Diagn Microbiol Infect Dis
29. Behbehani RS, Affel EL, Sergott RC, et al. Multifocal ERG in ethambutol associated visual loss. Br J Ophthalmol
30. Hennekes R. Clinical ERG findings in ethambutol intoxication. Graefes Arch Clin Exp Ophthalmol
© 2008 Lippincott Williams & Wilkins, Inc.
31. Srivastava AK, Goel UC, Bajaj S, et al. Visual evoked responses in ethambutol induced optic neuritis. J Assoc Physicians India