Results
The overall cohort examining reduction of HCQ dose included 564 patients who reduced their dose, contributing 1,063 person-years of data, and 778 matched patients who started HCQ at the same time but continued HCQ maintenance therapy without a dose reduction, contributing 1,242 person-years. The average duration of HCQ use since time zero in this cohort was 3.4 years.
Before stratifying for disease activity, the group who reduced their therapy experienced 40 first flares per 100 person-years, compared with 31.9 first flares per 100 person-years on maintenance therapy. Those who reduced HCQ had a 20% greater risk of flares than did those who continued it (adjusted hazard ratio, 1.2). However, when those in remission were compared with those not in remission – independent of disease activity level – patients in remission were twice as likely to experience a flare if they reduced their HCQ dose (aHR, 2.14).
In the discontinuation cohort, 389 patients who stopped HCQ therapy contributed 657 person-years, and 577 matched patients who continued HCQ maintenance therapy contributed 924 person-years. The average duration of HCQ use since time zero in this cohort was 4.2 years. Before stratifying for disease activity, the average number of first flares per 100 person-years was 41.3 in the HCQ discontinuation group and 30 in the HCQ maintenance group, resulting in a 56% higher risk of flares for those who stopped HCQ, compared with patients who continued HCQ (aHR, 1.56). Looking only at those in remission, patients were nearly three times more likely to experience a flare if they stopped HCQ than were patients not in remission who continued a maintenance dose (aHR, 2.77).
Patient age is an important consideration
Overall, these findings are not surprising, said Jill P. Buyon, MD, director of the division of rheumatology and of the Lupus Center at NYU-Langone Health in New York. Dr. Buyon is not involved in the current study but is studying discontinuation of HCQ in older adults with lupus.
Dr. Jill P. Buyon
“It has been already shown that when lupus patients discontinue HCQ, flares are more likely, but does this apply to all age groups?” Dr. Buyon asked in an interview. “Data are essential to more accurately weigh the balance between accumulating ocular exposure, the explosion of new tools to assess retinal injury, and the risk of disease flare in a population that may have more stable/quiescent disease than younger patients.”
Although HCQ’s track record with infection risk is consistently better than that of more immunosuppressive drugs and is very safe during pregnancy, Dr. Buyon said her “ophthalmology colleagues persistently emphasize the risk of retinal accumulation of drug and ocular toxicity over time.” She referenced a recent case-control study in which overall prevalence of HCQ retinopathy was 7.5%, and greater for patients taking more than 5 mg/kg of HCQ or who used HCQ for more than 10 years.
”Risk escalates with continued use, and evaluation by sensitive approaches such as multifocal electroretinography suggests nearly a third of patients accrue retinal damage,” Dr. Buyon said. “As the longevity of patients improves and comorbidities such as renal insufficiency (which affects HCQ clearance) may increase, the ratio of efficacy to toxicity would be expected to decrease.” Further, the fact that disease activity may wane as people age means that rheumatologists treating older adults need to address a critical question, she said: “Can HCQ be safely withdrawn? This question is important in the context of an even broader concern regarding management of SLE in the elderly population, a topic which has received minimal attention.”
The study is limited by its observational design and the fact that the intervention was not randomly allocated, although the researchers attempted to adjust for confounders. Dr. Bernatsky also noted that mild flares might have been missed, and the researchers did not evaluate HCQ levels or adherence, nor did the data set include physicians’ or patients’ explicitly stated reasons for HCQ reduction or discontinuation.
”We estimated that 5% of patients may have reduced HCQ therapy as result of the AAO [American Academy of Ophthalmology] guidelines, 55% because of low disease activity state, and the remainder (40%) for other reasons, possibly intolerance or patient preference,” the researchers noted in their abstract. “Among those who discontinued HCQ, 4% had retinal changes of concern, 15% were in clinical remission, and the remainder stopped for unknown reasons, possibly intolerance or patient preference.”
Dr. Buyon also pointed out that the cohort was initially intended for studying cardiovascular risk and not designed to capture all visits during each year of follow-up.
“Thus, while hospitalizations would be well captured, not all flares, particularly those not severe, would be captured, and thus we may not have the complete picture,” she said, reiterating Dr. Bernatsky’s point that mild flares may have been missed.
”Clearly, this is a very important topic for the management of our patients, particularly those who are elderly and may have already reaped the benefits of hydroxychloroquine,” Dr. Buyon said. “Of course, we have to be mindful of the potential benefit with regard to blood clotting and lipid lowering. Nevertheless, accumulated ocular toxicity and cardiac issues such as cardiomyopathy may emerge to tip the balance after years of accumulated drug exposure.”
The research was funded by the Canadian Institute of Health Research, the Singer Family Fund for Lupus Research, and the SLICC Group. Dr. Bernatsky had no disclosures. Dr. Buyon noted that she has an R34 NIH planning grant to study the safety of withdrawal of hydroxychloroquine in elderly lupus patients that is relevant to this study.