BLOG: What we need to know about hydroxychloroquine retinal toxicity during COVID-19
About a year ago I wrote a piece on hydroxychloroquine retinal toxicity, in which I explored the drug’s mechanism and why it yields its characteristic bullseye.
As we all know, hydroxychloroquine has been in the news lately as a possible treatment for COVID-19, so I thought it a good time to revisit what we know about hydroxychloroquine and how it affects the retina.
I want to iterate here that this will not be an opinion on the usefulness of the medicine against COVID-19, but rather a primer for eye care providers to use when and if a patient presents with a recent history of hydroxychloroquine use. As I write this, there is a shortage of the medicine in some pharmacies, and it would stand to reason that if there is a spike in the number of patients using hydroxychloroquine, then there may be a spike in the number of patients with hydroxychloroquine retinal toxicity, especially if these patients are taking it at higher-than-recommended doses.
Recall that the medication was first derived in the 1930s as chloroquine, a schizonticide (a killer of protozoa) used to treat malaria, which was itself derived from quinine, the primary treatment for malaria at the time. Soon after in the 1940s, a hydroxy-group was added to the chloroquine molecule to create hydroxychloroquine. This addition lessens (but does not eliminate) the ability of hydroxychloroquine to cross the blood-retinal barrier, thereby decreasing the retinal toxicity compared with chloroquine.
Hydroxychloroquine acts against malaria mainly by elevating the pH level in lysosomes, making it difficult for protozoa to break down hemoglobin in red blood cells. As a friendly reminder, lysosomes are organelles within a cell that help digest material and break down waste. Lysosomes contain a lot of enzymes and typically operate in a fairly acidic (low pH) environment. So, when hydroxychloroquine raises the pH in parts of a malaria-infected cell, it inhibits polymerases necessary for the parasite’s protection and it causes an inability of the parasite to carry out hemoglobin digestion.
The mechanism of action of hydroxychloroquine for inflammatory conditions like rheumatoid arthritis and lupus is less certain but is thought to involve the same elevation of lysosomal pH. If the lysosomes are mistakenly breaking down self-antigens in the joints of a patient with rheumatoid arthritis, then raising the lysosomal pH would be a good thing, and this is why hydroxychloroquine works for some autoimmune diseases. But this is also exactly why it creates retinal toxicity. In the normal process of sight, lysosomes in our retinal pigment epithelium (RPE) are constantly breaking down the photoreceptor outer segment waste that is accumulated. If these lysosomes are impaired, then outer segments will build up in the form of lipofuscin. This lipofuscin is toxic to the outer retina and RPE, and eventually the patient will suffer from loss of photoreceptors and subsequent blind spots in the affected areas.
Hydroxychloroquine-related changes have been found in many layers of the retina, including the ganglion cells, photoreceptors and RPE. So why does the retinopathy start in the RPE? Because hydroxychloroquine binds to melanin with a strong affinity and is, thus, present in higher concentrations in the RPE. That’s also the reason why hydroxychloroquine toxicity is so scary: because of the strong affinity to melanin, the drug will be retained in the RPE for a time even after the patient stops taking it; the risk for toxicity continues even after the medicine is stopped. This toxicity is permanent. If the photoreceptors become so damaged that they die, we have no way to restore them.
So as eye care providers, what should we be looking for in patients that admit to taking hydroxychloroquine? Well, for starters, don’t just look for pigment changes in the RPE. Once you see RPE clumping and the classic bullseye pattern, the damage to the photoreceptors is fairly advanced. In fact, the American Academy of Optometry doesn’t even recommend using fundus examination/photography as an adequate screen for hydroxychloroquine toxicity for that exact reason. They recommend automated visual fields (usually 10-2) and spectral domain OCT (not the older TD-OCT). Fundus autofluorescence and microperimetry also add value if available. I would highly recommend reading AAO’s screening recommendation, which was updated in 2016 and is included in the references (Marmor MF, et al).
The accompany figures are images from one of my patients with hydroxychloroquine retinal toxicity. She barely has any detectable fundus pigment changes, but the abnormalities are evident on fundus autofluorescence and SD-OCT. And we actually caught her damage early; most of this OCT damage progressed after we stopped her hydroxychloroquine.
Remember that if our future patients are taking hydroxychloroquine for non-standard reasons, they might not be taking non-standard dosing. Further, research shows retinal toxicity will happen earlier when the dose is significantly higher than 5 mg/kg. So, let’s be alert when we finally get back out there, and let’s have a low threshold to work up a patient for hydroxychloroquine toxicity. A well-scrutinized OCT might save some vision.
Fox RI. Semin Arthritis Rheum. 1993;doi:10.1016/s0049-0172(10)80012-5.
Jorge A, et al. Nat Rev Rheumatol. 2018;doi:10.1038/s41584-018-0111-8.
Lee DH, et al. Ophthalmology. 2015;doi:10.1016/j.ophtha.2015.01.014.
Marmor MF, et al. Ophthalmology. 2016;doi:10.1016/j.ophtha.2016.01.058.
Melles RB, Marmor MF. Ophthalmology. 2015;doi:10.1016/j.ophtha.2014.07.018.
Wetterholm DH, Winter FC. Arch Ophthalmol. 1964;doi:10.1001/archopht.1964.00970010098016.
Yam JCS, et al. Hong Kong Med J. 2006;12(4):294-304.
Yoon YH, et al. Invest Ophthalmol Vis Sci. 2010;doi:10.1167/iovs.10-5278.