September 07, 2018
12 min read

Despite Plaquenil dosing recommendations, retinal toxicity remains

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

The American Academy of Ophthalmology has published several dosing and screening recommendations for hydroxychloroquine to avoid potential retinal toxicity, yet some patients still experience permanent vision loss resulting from hydroxychloroquine retinopathy due to improper dosing of the drug and improper screening.

A cookie-cutter approach to dosing Plaquenil (hydroxychloroquine, Sanofi-Aventis), an anti-malarial drug that has been used as a treatment for systemic lupus erythematosus and rheumatoid arthritis, by rheumatologists and internal medicine doctors can lead to an increased risk for hydroxychloroquine (HCQ) retinopathy, according to OSN Retina/Vitreous Section Editor Andrew A. Moshfeghi, MD, MBA.

“Plaquenil comes in 200 mg tablets, and many doctors just off-handedly put everyone on 200 mg twice a day. And they feel like they’re giving good care if they remember to send that patient to an ‘eye doctor’ every year or even more frequently, without knowing what screening is needed or if that eye doctor knows what to do. Screening is good, but they could significantly reduce the likelihood of morbidity or the actual need for these patients to stop using the drug if they had appropriately dosed these patients in the first place,” he said.

A multitude of tests are needed to detect the earliest possible signs of hydroxychloroquine toxicity, according to Andrew A. Moshfeghi, MD, MBA.

Image: USC Roski Eye Institute

AAO revised guidelines

The AAO published dosing and screening recommendations for chloroquine and hydroxychloroquine in 2016, revising previous recommendations published in 2011. The current maximum daily HCQ dose recommended by the AAO is 5 mg/kg of real weight. The 2011 guidelines suggested a maximum dose of 6.5 mg/kg of ideal weight.

The biggest difference between the 2011 and current recommendations is the need for patients to be dosed by their real weight instead of their ideal weight. The ideal weight dosing guidelines were based on a set of articles written about scientific experiments done on rats and “a few scrawny monkeys” nearly half a century ago, according to Michael Marmor, MD, professor of ophthalmology at Stanford University and chair of the committee that revised the AAO guidelines in 2016.

Michael Marmor

“Those articles concluded that HCQ doesn’t bind to fatty tissue. So therefore, you’d want to measure its effective weight in the body from non-fatty tissue, which is what ideal weight is, more or less,” Marmor said. However, other evidence shows clearly that the pattern of toxicity in monkeys taking long-term chloroquine, a similar drug to HCQ, causes damage to their cells that is quite different from the human disease, he said.


In other words, animal experiments can be powerful and useful, but they are not proof of human disease, he said.

“If you read those old HCQ studies carefully, what they’re really showing is that HCQ binds primarily to pigmented (melanotic) tissues and glandular tissues and liver, and that’s about it. It doesn’t bind much to muscle, brain or anything else, including fat. It’s not that it isn’t just binding in fat, it’s that it isn’t binding to much of anything. Those experiments are not very convincing for using ideal weight,” Marmor said. “And now we have a large human demographic study that compares the predictive value of real- and ideal-weight dosing. Conclusion: Real weight is better.”

Until 2011, there were essentially no good demographic studies on this drug due to the number of years that it takes to accumulate toxicity and the lack of sensitive tests such as OCT to recognize early retinal damage. Until then, the only way toxicity was demonstrated was by looking in the fundus for a bull’s-eye, which is now considered to be irreversible damage, Marmor said. But the situation has now changed.

Real body weight

Based on new scientific evidence, which Marmor and Ronald B. Melles, MD, published in JAMA Ophthalmology, it was clear that new dosing guidelines were needed, Marmor said.

“We had almost 2,500 patients who were using the drug for more than 5 years in whom we could document toxicity with OCT as well as visual fields and autofluorescence in different patients. We could really document it,” he said.

The study showed real body weight predicted risk for toxicity better than ideal body weight for all patients. Furthermore, the co-authors found that the prevalence of toxicity relative to real body weight dosing was independent of body habitus, while thin individuals dosed by ideal body weight had a much higher risk for developing toxicity.

“When you use real weight to calculate dose, the risk is exactly the same whether you weigh 90 pounds or 250 pounds. To me, that’s how drug prediction should work,” Marmor said. “Giving less to heavy patients by calculating their dose by height makes no sense even if ‘safer’ — we give drugs to have an effective dose on board, no matter what the patient weighs.”

Dosing based on ideal weight should be dismissed completely because it was based on old science that was misread and propagated in the literature, according to Marmor.

“Ophthalmology now has good human evidence that shows a better prediction comes from real weight, and we should all practice evidence-based medicine. When someone comes up with a new comparable series that shows ideal weight or another formula is better, fine, I’ll change. But we need to use the best evidence we have,” he said.


Better dosing practices

“Many of our patients are overdosed. A 2018 study published in Arthritis Research & Therapy found that already about a third of patients with normal body mass index were overdosed with the dosing based on ideal body weight. However, when calculated with the updated dosing guidelines using real body weight, more than half of normal BMI individuals were excessively dosed. Therefore, more patients will need to reduce dosing under the new guideline,” Judy E. Kim, MD, OSN Retina/Vitreous Board Member, said.

“This means we should make a practice of inquiring our patients about their body weight at every visit for HCQ screening and calculating the safe dosage based on real body weight, even in those who previously were thought to be in the safe range based on ideal body weight. The change from using ideal body weight to real body weight actually makes this calculation easier to do. It also calls for further research to find ways to give physicians consistent reminders to accurately dose, such as through electronic health records, as well as research to decide safe and effective dosing,” she said.

Judy E. Kim

In addition to dosage, long duration of HCQ usage can put patients at higher risk for developing retinal toxicity. Other risk factors, such as underlying maculopathy, renal disease and use of tamoxifen, also increase risk for toxicity, Kim said.

However, prior risk factors such as age, liver disease and obesity were not included in the revised recommendations, she said.

Giving a patient a higher dose than necessary for a long duration is the most significant risk factor for developing HCQ retinal toxicity, according to the AAO guidelines.

Calculating a correct dose using real weight is not difficult, but HCQ pills come in only 200 mg tablets. “If a patient needs 265 mg of HCQ, how can the drug be accurately prescribed for a patient?” Marmor asked.

“In fact, it’s very easy because the drug is slow to stabilize in the body, so it does not have to be taken in the exact same dose every day. All you do is calculate the weekly dose; in other words, if you want 265 mg per day, you multiply by seven to get the weekly dose. You then divide by 200, and it will tell you how many pills you have to take for the week (in this case nine) — it is easy to figure what days to take two pills and what days to take one. If it comes off as an uneven number, round it off one day or the other. In our example, 5 days a week the patient would take one tablet and 2 days a week they would take two,” he said.


An ophthalmologist just has to establish a relatively stable HCQ level in the blood, so it is “very easy to give people the proper and accurate dose,” Marmor said.

Risks for toxicity

According to the AAO, patients who take an appropriate dose based on real weight have a less than 1% risk for developing HCQ retinal toxicity during the first 5 years of use and less than 2% during the first 10 years of use. The risk rises to nearly 20% after 20 years of use, but a patient who screens as normal after 20 years has only a 4% risk for converting to toxicity in the subsequent year.

HCQ retinopathy is typically asymptomatic in patients who are in the early stages of the disease, according to a 2015 study published in the Indian Journal of Ophthalmology by Hemang K. Pandya, MD, and colleagues, so continued screening of patients to monitor for signs is important.

However, when the disease advances, it can result in a deterioration of visual acuity, peripheral vision and night vision. When patients present with a classic bull’s-eye maculopathy, the disease is usually advanced and has caused irreversible damage, according to the study.

“When you take this drug and you take it wrongly or too long or too high a dose, you develop an atrophy of the macula in a ring-like pattern surrounding the center of the macula initially. If you keep taking it and don’t stop, it will continue to cause progressive damage that will involve the very center of the macula, similar to the type of atrophy that one gets with dry age-related macular degeneration with the advanced form of geographic atrophy. Similar to dry AMD, whatever damage you get at this point is irreversible,” Moshfeghi said.

Moreover, severe damage will continue to progress and worsen even if the drug is stopped.

Proper screening practices

Because of the extensive and irreversible damage that can stem from HCQ retinal toxicity, proper screening practices take on an elevated level of importance.

Ophthalmologists should use Humphrey visual field test and OCT — and fundus autofluorescence if available can help, too — to detect signs of HCQ toxicity, Moshfeghi said.

“The bottom line is, we use a multitude of tests at each visit. The reason why we do that is because we want to try to pick up the earliest possible signs of toxicity. If you just did one test, an OCT for example, which is very specific at picking up toxicity, it might be that you’re seeing age-related issues that just look like HCQ toxicity. Or a field test, which is very sensitive, may not have been reliably performed. You don’t want to stop the drug unnecessarily because it might be the only drug that is helpful for that patient’s rheumatological condition, such as lupus or rheumatoid arthritis. And of greater importance, Plaquenil has less systemic side effects than other drugs used for lupus, so there is a real advantage in using it as long as one can. You want a high threshold for saying you should stop this medication,” he said.


Additionally, knowing the differences among toxicity symptoms in different ethnicities can help validate HCQ toxicity and find signs of it before it causes irreversible damage, Brian Toy, MD, of USC Roski Eye Institute, said.

For instance, wide-angle field imaging needs to be used on Asian patients in addition to the smaller-field imaging used on Caucasian patients. Asian patients tend to have toxicity that begins a little farther from the fovea compared with Caucasian patients, often at the arcade region, so a wide-field imaging test is needed, Toy said.

Brian Toy

“Widefield fundus autofluorescence and OCT can be helpful for detecting toxicity in Asian patients who can present with a pericentral pattern. Also, a Humphrey visual field 24-2 in addition to a 10-2 is necessary. This can add significantly to testing time, but once patients are told what it’s for, they are generally OK with it,” he said.

The signs can be missed easily in Asian patients if a widefield test is not used, Moshfeghi agreed, and if ophthalmologists are not up to date with the latest AAO guidelines, they may not know the test is necessary.

“If you only do what was previously recommended, such as what you learned in residency, the 10-2 Humphrey visual field, it will only test the central 10° of a patient’s vision. If you do that in an Asian patient who actually has damage, you’ll miss it a lot of times. You weren’t taught this at your residency or fellowship because this is a recent guideline — you have to do a 24-2 instead for those patients,” he said.

Visual field tests take a long time to complete and can be difficult for patients. A traditional 10-2 test takes at least 15 minutes per eye, and if a 24-2 is added, a patient is taking the tests for an hour or more, Marmor said.

“By the time you do this, people are falling asleep, they’re angry, they’re ready to walk out. It doesn’t work. However, there is a variation of the field test called SITA Fast. SITA Fast takes half as long, and even most glaucoma people are realizing it’s just about as good a test. So, for my Asian patients, I do both a 10-2 and a 24-2 SITA Fast; doing both fields takes no more time than the standard 10-2,” Marmor said.

Using electronic health records

Because of the need for continual screenings for patients who take HCQ, Toy said it is important for patients to not be lost to referral. An electronic medical record can be an interesting tool to track screenings and ensure the proper dosing is utilized, he said.


“The [electronic health record] system enables us to make a registry of patients taking HCQ to keep track of when they need to come in for their screening. I think that is particularly helpful in implementing the AAO recommendation to defer screening for the first 5 years when there is little risk of toxicity developing, but we don’t want patients being lost to follow-up for 10 to 15 years,” Toy said.

To improve screening rates, Toy and colleagues are working with Cerner to improve their EMRs. The programmers are working to include an automatic dose checker in the EMR of a patient being prescribed HCQ to automatically calculate the appropriate dose of the drug based on the patient’s real weight.

The second idea is to develop a registry of patients in the Los Angeles County health system who are currently on HCQ to keep track of them for future screening needs. Like the IRIS Registry, this would include a dashboard that ophthalmologists could bring up to automatically schedule patients for future screenings, he said.

“The third thing is to standardize our workflow. I think for any clinic, but particularly for our county clinic where we have limited resources in terms of technicians and equipment, we need to optimize our use of resources. So we have adopted a telemedicine approach where a patient comes in for OCT, visual field testing and photography, without seeing a provider at that visit. The testing results are interpreted by one of our retina doctors, and we can send recommendations about dose and follow-up to our medicine colleagues,” Toy said.

Toy said the screenings would be evaluated and appointments made for patients who are showing signs of developing HCQ toxicity.

Improved HCQ retinopathy screenings can be achieved with better education and better communication with rheumatologists and internists, Kim said.

“We need to get our message out to others at their meetings and magazines. We need to be sending letters to them after evaluating their patients,” she said.

Educating patients who are at higher risk for toxicity on the importance of eye examinations and follow-up can also be helpful. While follow-up typically is not necessary in the first 5 years, Kim said she will have higher-risk patients, such as those on tamoxifen or who have maculopathy or kidney disease, be evaluated sooner than the recommended 5 years.

Progressive disease

New data presented by Marmor at this year’s Association for Research in Vision and Ophthalmology meeting show that once a patient has retinal pigment epithelium (RPE) damage from HCQ toxicity, the disease will never stop progressing. For patients with early toxicity, before a bull’s-eye is observed, the disease will stabilize and the risk for visual loss is low. The drug can be stopped and everything will be OK, he said.


“But once there is RPE damage, once there is some evidence of a bull’s-eye, those patients never stop getting worse, and that was something we didn’t expect. We thought it would maybe take a few years, but it didn’t. It’s a progressive disease. Something is destabilized, and it never stops getting worse,” Marmor said.

HCQ toxicity can be avoided by using a proper dose and screening properly, he said.

“Bull’s-eye retinopathy is a scourge that should be eliminated along with leprosy and small pox. If you’re screening people, you should never see it. It’s severe damage,” Marmor said. – by Robert Linnehan

Disclosures: Kim, Marmor, Moshfeghi and Toy report no relevant financial disclosures.

Click here to read the POINTCOUNTER, "Should a patient with signs of retinal toxicity be taken off HCQ completely, or are a reduced dose and additional screening more appropriate?"