Hydroxychloroquine fails to prevent COVID-19 in patients with rheumatic disease
Hydroxychloroquine fails to prevent COVID-19 among adults with rheumatoid arthritis, systemic lupus erythematosus or other related diseases, according to data published in The Lancet Rheumatology.
“The best intervention to improve outcomes of a potentially-serious viral infection is to prevent it from happening in the first place,” Chris A. Gentry, PharmD, of the Oklahoma City Veterans Affairs Healthcare System, told Healio Rheumatology. “In the absence of a safe and effective vaccine, antiviral agents are a necessary alternative. Early reports of hydroxychloroquine’s purported activity against COVID-19 offered hope for this, and randomized prospective trials were planned.”
“However, given low event rates, it will likely take a substantial amount of time before these trials are able to achieve adequate power to provide useful results,” he added. “In the meantime, our approach to gather and analyze observational data for a propensity-matched analysis from a large clinical administrative database in thousands of veterans with rheumatologic conditions who may or may not be receiving chronic hydroxychloroquine can help rapidly assess any effect on prevention of COVID-19.”
To analyze whether patients who take hydroxychloroquine for various rheumatic conditions have a lower risk for COVID-19 than those not receiving the antimalarial, Gentry and colleagues conducted a retrospective cohort study of individuals in the U.S. Veterans Health Administration clinical administrative database. The researchers included all adults in the database with either RA, SLE or associated rheumatic conditions, based on ICD-10 codes, who were alive March 1. In all, 10,703 patients treated with hydroxychloroquine, and 21,406 not receiving the drug, were included in the primary analysis.
Gentry and colleagues calculated a propensity score for each patient, with each patient treated with hydroxychloroquine matched to two patients not receiving it, who were regarded as controls. The primary endpoint was the proportion of patients with confirmed COVID-19 among those treated with hydroxychloroquine, compared with matched control individuals, between March 1 and June 30. Secondary outcomes included COVID-19 related hospital admission, mortality and intensive care use, as well as all-cause hospital admission and mortality. The researchers used multivariate logistic regression analysis to examine independent variables for COVID-19 infection.
According to the researchers, the incidence of active COVID-19 infection throughout the study period did not differ between the two groups, with a rate of 0.3% among those treated with hydroxychloroquine and 0.4% in the control group (OR = 0.79; 95% CI, 0.52-1.2). In addition, among the patients who developed COVID-19, there were no significant differences in secondary outcomes between the two groups. For all individuals included in the study, overall mortality was lower in the hydroxychloroquine group compared with those who did not receive the drug (OR = 0.70; 95% CI, 0.55-0.89).
In the multivariate logistic regression analysis, the researchers found that treatment with hydroxychloroquine was not associated with COVID-19 infection (OR = 0.79; 95% CI, 0.51-1.42).
“Unfortunately, our findings indicate that hydroxychloroquine did not appear to prevent COVID-19 infection in these veterans with rheumatologic conditions compared to similar veterans not taking hydroxychloroquine,” Gentry said. “Multivariate analysis supported this finding as well. For the small number of active infections found in each group, secondary outcomes may point researchers to other potential important questions to solve.”
“Had our study found a positive signal indicating hydroxychloroquine prevented COVID-19 in these veterans, conversations would have likely begun to assess whether hydroxychloroquine should be initiating in selected populations at higher risk of developing COVID-19 — i.e., health care workers, long term care facility residents, immunocompromised individuals,” he added. “The actual primary outcome of our study tells us that our priorities for finding effective preventive interventions should quickly shift to other strategies.”