Ethambutol Optic Neuropathy
Ethambutol Optic Neuropathy
Ethambutol (EMB) is an antibiotic used to treat infection by Mycobacterium species, particularly Mycobacterium tuberculosis and non- tubercular infections like M. avium complex and M. kansasii. Unfortunately, one serious and vision threatening side effect of EMB is ethambutol-induced optic neuropathy (EON).  Less common side effects of EMB include peripheral neuropathy, cutaneous reactions, thrombocytopenia, and hepatitis. 
The prevalence of EON in patients treated for tuberculosis is estimated to be around 1-2%.  According to the World Health Organization (WHO), there are around 9.2 million new cases of tuberculosis each year, 55% of which will take ethambutol. Given that the incidence of EON is about 1-2% among treated patients, these statistics suggest there could be as many as 100,000 new cases of EON annually.  Furthermore, the risk of EON is highly dose dependent. The estimated prevalence of EON for ethambutol doses of ≤ 15, 20, 25, and > 35 mg/kg per day are < 1%, 3%, 5-6%, and 18-33%, respectively. At any of the therapeutic antimicrobial dosing regiments, there is a variable and sometimes idiosyncratic risk for EON and thus there is effectively no truly “safe” dose for EMB. Other than ethambutol dosage, risk factors for EON include age greater than 65 years and hypertension. Because ethambutol is excreted by the kidneys, renal disease can also increase the risk of optic neuropathy. Finally, multiple case reports have identified isoniazid, another first-line treatment for tuberculosis, as a cause of optic neuropathy similar to EON. Therefore, patients taking ethambutol in combination with isoniazid may be at increased risk for visual loss. 
Retrobulbar optic neuropathy (normal appearing optic disc on presentation) is the most common form of EOM.  While the exact mechanism of the neurotoxic effect of EMB on the optic nerve is unknown, it is believed that the metal chelating effects of this drug may be responsible. One theory is that the chelation of copper disrupts oxidative phosphorylation, as there is less copper available in human mitochondria. Another theory is that the chelation of zinc inhibits lysosomal activation. Moreover, in animal studies on rat optic nerves, zinc deficiency has been associated with destruction of myelin and glial cell proliferation, suggesting that there may be a similar effect in humans.  In addition, prolonged EMB use has been shown to be associated with deficiencies in vitamin E and vitamin B1 which may exacerbate the optic neuropathy.
Unlike other toxic optic neuropathies, EON can occur with a very short period of time following initiation of therapy. Symptoms may develop anywhere from 1 to 36 months after starting the drug. The majority of patients (> 60%) present with bilateral, painless, symmetric loss of central visual acuity, central visual field (central/cecocentral scotoma) and dyschromatopsia.  The visual acuity loss may vary from minimal (20/25) to severe (no light perception) and the severity of the visual loss at onset is often mild and insidious. Color vision loss may be the first sign of EON and typically involves loss of red and green color perception, but loss of blue and yellow is also possible.  Central vision loss is the most common visual field defect, including central or cecocentral scotoma (intra-axial form) on formal visual field testing.  However, other possible visual defects include bitemporal hemianopsia due to optic chiasm involvement and peripheral field constriction (extra-axial form). Pupil responses may be normal initially but then patents develop sluggish pupillary light reflex bilaterally with preservation of the near response (light near dissociation of the pupils). A relative afferent pupillary defect may not be seen due to the bilateral and symmetric nature of EON.  Initially, the optic nerve may be normal on fundoscopy, but optic disc pallor may eventually develop as the visual loss progresses. 
The diagnosis of EON is made clinically with visual acuity, formal perimetery and color vision testing. Fundoscopy is necessary to exclude other etiologies of visual loss in particular retinal disease. The presence of optic disc edema or macular pathology would argue strongly against EON. Patients should receive a baseline eye exam prior to initiation of EMB therapy in order to document pre-treatment ocular pathology. Unfortunately unlike other toxic maculopathies (e.g. hydroxychloroquine) there is no pre-toxicity finding in EON for screening. Nevertheless, high-risk patients for EON (e.g., long duration or high dose EMB, renal dysfunction) should have visual assessment every month and all patients on EMB should have central vision assessment (e.g., visual acuity, Amsler grid) monthly (although this screening does not require a full dilated ophthalmology exam). If vision changes are noticed, the patient should see an ophthalmologist and if there is a suspicion for EON, the prescribing physician must be contacted in order to make decision of whether to continue or discontinue EMB therapy. 
Ideally, prevention of EON involves stratifying at risk patients, screening for visual loss monthly, and detecting EON prior to onset of clinically significant and irreversible visual changes and optic atrophy. Different screening modalities have been proposed to detect subclinical EON, including vision evoked potentials (VEP), optical coherence tomography (OCT), and multifocal electroretinography (mfERG), none of which have been validated for large population screening purposes. VEP uses electrical signals generated in the occipital cortex following visual stimulation, with a p100 wave occurring about 100 milliseconds after the stimulus. EON may show slowed neural conduction to the occipital cortex and, thus, result in increased latency of the p100 wave. Measuring p100 latency is an established component in the diagnosis of optic nerve diseases like optic neuritis, and recent studies have suggested that VEP may be useful for detecting subclinical optic nerve damage in patients taking ethambutol.  In a 2016 study of 31 patients taking ethambutol, the mean p100 latency increased from 101 ms to 106.4 ms at two months and 115.1 ms at four months.  Another study found that 34.8% of patients taking ethambutol had an increased p100 latency above 107 ms, as compared to a control mean of 97 ms.  Studies have also shown an increased latency in patients diagnosed with EON. A 2015 study of 62 patients with EON, VEP was performed on nine patients. Five of the nine patients were found to have an increased latency of the p100 wave.  Unfortunately, any lesion in the visual pathway (including refractive error, media opacities, retinal lesions) other than optic nerve disease (including EON) can produce an abnormal VEP and thus VEP is not specific for EON.
While OCT is used to assess various optic neuropathies and will detect changes in the peripapillary retinal nerve fiber layer (pRNFL) thickness and ganglion cell inner plexiform layer in patients with clinical EON, it is unclear if OCT is an effective screening tool for subclinical EON. Studies have shown that OCT detects a 20-79% decrease in the pRNFL thickness in patients diagnosed with EON.  However, studies assessing the use of OCT in detecting subclinical EON have had contradictory results. Two studies, representing 31 and 37 patients taking ethambutol and without EON symptoms, found that the pRNFL thickness increased overall or in specific quadrants. Another study found that in a group of patients taking ethambutol, about 3% of 104 eyes had temporal pRNFL thinning.  In study of 20 patients taking ethambutol, researchers reported that on average, the RNFL thickness decreased by 5 µm over two months.  In addition to optic nerve damage, ethambutol toxicity also affects the retinal cell layers. Because mfERG can be used to detect occult retinopathies, it has been suggested that this modality may be used to identify subclinical EON which can affect macular function and be detectable on mfERG. 
Other than EMB, multiple drugs (including isoniazid) can cause a toxic optic neuropathy. The clinical features of toxic optic neuropathy from any source are characterized by similar symptoms of central visual loss, dyschromatopsia, a progressive painless, bilateral, and symmetric course, sluggish pupils with light near dissociation, with or without relative afferent pupillary defect, optic disc pallor, and central/cecocentral scotomas on visual field testing.  Common causes of toxic optic neuropathy are shown in Table 1. The main concern in patients treated for TB is that progressive visual loss despite discontinuation of EMB for presumed EON could represent another optic neuropathy including toxic optic neuropathy from isoniazid.
Table 1. Causes of toxic optic neuropathy 
|Alcohols||Methanol, ethylene glycol|
|Antibiotics||Chloramphenicol, sulfonamides, linezolid|
|Antitubercular drugs||Isoniazid, ethambutol, streptomycin|
|Heavy metals||Lead, mercury, thallium|
|Others||Carbon monoxide, tobacco|
Currently, there is no effective treatment for EON. However, if the condition is detected early and the drug is discontinued promptly (before the development of irreversible optic atrophy), between 30-64% of patients have been reported to show visual improvements over a course of several months.  However, full recovery is rare, and the average improvement is two lines on the Snellen chart. Older patients, especially those over 60 years of age, have been found to have poorer recovery compared to younger patients. 
Since zinc and copper deficiencies induced by the metal-chelating effects of EMB are thought to lead to the development of EON, supplementing these minerals has been proposed as a method to reduce the likelihood of EON.  Likewise, vitamin deficiencies (e.g., vitamins E and B1, 9, 12) may exacerbate EON, these vitamins can be supplemented.  However, more research is still required to verify the reduction of EON risk with supplementation of these micronutrients. 
Most patients who discontinue EMB after onset of visual changes (without optic atrophy) will recover their vision over a period of weeks to months.  OCT can measure pRNFL and may provide prognostic information regarding visual recovery however structure-function correlation between EON prognosis and OCT is somewhat variable. About 30-64% of patients will improve visual function if EON is detected early and EMB promptly discontinued. However, some patients will fully recover their visual acuity while others will be left with permanent residual visual defect. On average, patients who recover vision will improve two lines on the Snellen chart.  While most patients will experience some visual improvement, some patients may not recover their vision or will continue to experience vision loss even after stopping the EMB.  Additionally, while optic disc pallor occurs later in the course of EON, presence of optic disc pallor at the onset of visual symptoms is associated with poor prognosis. 
In summary, clinicians should be aware of the risk of EON in any patient on EMB. A baseline eye examination is recommended prior to starting EMB to document pre-existing ocular pathology and baseline visual field testing. Patients on EMB should be risk stratified with high risk patients (e.g., high dose (>15 mg/kg), long duration, other ocular co-morbidities, vitamin deficient, or renal failure) being evaluated monthly. EON can occur within weeks to months of starting EMB and thus high vigilance and clinical suspicion is warranted. All asymptomatic patients on EMB however should undergo some type of central vision (e.g., visual acuity, color vision, central visual field) assessment monthly but these screening tests do not necessarily require a formal ophthalmology visit each time. Patients on EMB should be counseled on the risk of EON and any complaint of visual loss in a patient on EMB should be taken seriously and these symptomatic patients should undergo full ophthalmologic exams. Although there is no pre-toxicity screening protocol or testing that has been validated for EMB, formal visual acuity, central visual field testing, fundus examination, and OCT pRNFL, as well as possibly electrophysiologic testing (e.g., VEP, mgERG) should be considered in patients suspected of having EON. Prompt discontinuation of EMB in EON is critical to prevent permanent visual loss and irreversible optic atrophy but the ophthalmic provider should directly contact the prescribing physician before discontinuation of EMB. Vitamin and mineral supplementation remains unproven in EON but could also be considered and the visual prognosis while generally good is variable and depends in part on the presence and severity of optic atrophy.
- Chamberlain, P. D., Sadaka, A., Berry, S., & Lee, A. G. (2017). Ethambutol optic neuropathy. Current Opinion in Ophthalmology, 28(6), 545-551. doi:10.1097/ICU.0000000000000416
- Chan, R., & Kwok, A. (2006). Ocular toxicity of ethambutol. Hong Kong Medical Journal, 12(1), 56-60.
- Sadun, A. A., & Wang, M. Y. (2008). Ethambutol optic neuropathy: How we can prevent 100,000 new cases of blindness each year. Journal of Neuro-Ophthalmology, 28(4), 265-268.
- Kass, I., Mandel, W., Cohen, H., & Dressler, S. H. (1957). Isoniazid as a cause of optic neuritis and atrophy. Journal of the American Medical Association, 164(16), 1740-1743.
- Nair, K. G. (1976). Optic neuritis due to INH complicating tuberculous meningitis. The Journal of the Association of Physicians of India, 24(4), 263-264.
- Kulkarni, H. S., Keskar, V. S., Bavdekar, S. B., Gabhale, Y. (2010). Bilateral optic neuritis due to isoniazid (INH). Indian Pediatrics, 47, 533-535.
- Kim, K. L., & Park, S. P. (2016). Visual function test for early detection of ethambutol induced ocular toxicity at the subclinical level. Cutaneous and Ocular Toxicology, 35(3), 228-232.
- Srivastava, A. K., Goel, U. C., Bajaj, S., Singh, K. J., Dwivedi, N. C., & Tandon, M. P. (1997). Visual evoked responses in ethambutol induced optic neuritis. The Journal of the Association of Physicians of India, 45(11), 847-849.
- Chen, S. C., Lin, M. C., & Sheu, S. J. (2015). Incidence and prognostic factor of ethambutoil-related optic neuropathy: 10-year experience in southern Taiwan. The Kaohsiung Journal of Medical Sciences, 31(7), 358-362.
- Kim, Y. K., & Hwang, J. M. (2009). Serial retinal nerve ﬁber layer changes in patients with toxic optic neuropathy associated with antituberculosis pharmacotherapy. Journal of Ocular Pharmacology and Therapeutics, 25(6), 531–535.
- Zoumalan, C. I., Agarwal, M., Sadun, A. A. (2005). Optical coherence tomography can measure axonal loss in patients with ethambutol-induced optic neuropathy. Graefe’s Archive for Clinical and Experimental Ophthalmology, 243, 410-416.
- Han, J., Byun, M. K., Lee, J., Han, S. Y., Lee, J. B., & Han, S. H. (2015). Longitudinal analysis of retinal nerve fiber layer & and ganglion cell-inner plexiform layer thickness in ethambutol-induced optic neuropathy. Graefe’s Archive for Clinical Experimental Ophthalmology, 253(12), 2293-2299.
- Menon, V., Jain, D., Saxena, R., & Sood, R. (2009). Prospective evaluation of visual function for early detection of ethambutol toxicity. The British Journal of Ophthalmology, 93(9), 1251-1254.
- Gümüş¸ A, & Öner, V. (2015). Follow up of retinal nerve ﬁber layer thickness with optic coherence tomography in patients receiving antitubercular treatment may reveal early optic neuropathy. Cutaneous and Ocular Toxicology, 34(3), 212–216.
- Kardon, R. H., Morrisey, M. C., & Lee, A. G. (2006). Abnormal multifocal electroretinogram (mfERG) in ethambutol toxicity. Seminars in Ophthalmology, 21(4), 215-222.
- Sharma, P., & Sharma, R. (2011). Toxic optic neuropathy. Indian Journal of Ophthalmology, 59(2), 137-141. doi:10.4103/0301-4738.77035
- Lee, E. J., Kim, S. J., Choung, H. K., Kim, J. H., & Yu, Y. S. (2008). Incidence and clinical features of ethambutol-induced optic neuropathy in Korea. Journal of Neuroophthalmology, 28(4), 269-277.
- Ezer, N., Benedetti, A., Darvish-Zargar, M., & Menzies, D. (2013). Incidence of ethambutol- related visual impairment during treatment of active tuberculosis. International Journal of Tuberculosis and Lung Disease, 17(4), 447-455.
- Kumar, A., Sandramouli, S., Verma, L., Tewari, H. K., & Khosla, P. K. (1993). Ocular ethambutol toxicity: is it reversible? Journal of Clinical Neuroophthalmology, 13(1), 15-17.
- Tsai, R. K., & Lee, Y. H. Reversibility of ethambutol optic neuropathy. (1997). Journal of Ocular Pharmacology and Therapeutics, 13(5), 473-477.
- Kumar, A., Dada, T. Visual evoked response (VER) in ocular ethambutol toxicity. (1999). National Medical Journal of India, 12, 193-194.
- Rasool, M., Malik, A., Manan, A., Aziz, K., Mahmood, A. Zaheer, S., … Karim, S. (2015). Determination of potential role of antioxidative status and circulating biochemical markers in the pathogenesis of ethambutol induced toxic optic neuropathy among diabetic and non-diabetic patients. Saudi Journal of Biological Sciences, 22(6), 739-743. doi:10.1016/j.sjbs.2014.09.019
- Koul, P. A. (2015). Ocular toxicity with ethambutol therapy: Timely recaution. Lung India, 32(1), 1-3. doi:10.4103/0970-2113.148395