Patients with rheumatoid arthritis (RA) treated with rituximab (Rituxan) or JAK inhibitors had more severe COVID-19 disease courses, analysis of data from a large registry found.
In an adjusted multivariate analysis compared with patients on tumor necrosis factor (TNF) inhibitors, those on rituximab had a fourfold higher risk for worse disease (OR 4.15, 95% CI 3.16-5.44), while those on JAK inhibitors had a twofold greater risk (OR 2.06, 95% CI 1.60-2.65), reported Jeffrey Sparks, MD, of Harvard Medical School in Boston, at the European League Against Rheumatism virtual congress.
“As we all know, there has been intense interest during the pandemic in repurposing immune modulating drugs for COVID-19 treatment,” he said.
Some targeted disease-modifying antirheumatic drugs (DMARDs) can dampen the hyperinflammatory response in COVID-19, resulting in a less severe clinical course. “However, some DMARD targets may impair viral immune defenses, leading to a more severe course. This has been shown in observational studies for rituximab with an increase in mortality and ICU stay versus methotrexate or non-use,” he said.
Previous efforts to analyze risks for the various medications used in rheumatic diseases during COVID-19 illness have been limited by including a variety of diseases and often using only the single outcome of hospitalization.
To more clearly quantify these risks, Sparks and colleagues analyzed physician-reported cases of COVID-19 in patients with RA enrolled in the international COVID-19 Global Rheumatology Alliance registry from March 2020 to April 2021.
This ongoing registry thus far has included 15,127 patients, 6,132 of whom had RA.
The primary outcome was a mutually exclusive ordinal COVID-19 severity scale modeled after WHO outcome measures. Items on the severity scale were:
- No hospitalization
- Hospitalization without supplementary oxygen
- Hospitalization with oxygen or ventilation
“Interpretation of the effect size was the odds of being one level higher on the ordinal scale compared with the reference group of TNF inhibitors,” Sparks explained.
Covariates included age, sex, smoking, obesity, comorbidities such as cardiovascular disease, interstitial lung disease, and cancer, concomitant nonbiologic DMARD or corticosteroid use, disease activity, location, and calendar time.
The cohort included 1,388 patients on TNF inhibitors, 563 on JAK inhibitors, 364 on rituximab, 317 on IL-6 inhibitors, and 237 on abatacept.
Cancer and interstitial lung disease were more common among rituximab users, but the number of patients using corticosteroids and who had high levels of disease activity were similar across the groups.
The analysis showed that 85% of the TNF inhibitor group were not hospitalized, which was similar to what was seen for IL-6 inhibitors. However, patients on rituximab frequently were hospitalized requiring oxygen or mechanical ventilation (22%), and 15% died. The JAK inhibitor group also had a relatively high proportion of hospitalizations requiring oxygen or ventilation (15%).
Unlike rituximab or the JAK inhibitors, no increased risk for severe disease course was seen for abatacept (OR 1.26, 95% CI 0.88-1.80) or IL-6 inhibitors (OR 0.81, 95% CI 0.56-1.18).
For each level of the ordinal scale, rituximab users had worse outcomes compared with TNF inhibitor users:
- Hospitalization: OR 4.53 (95% CI 3.32-6.18)
- Hospitalization with oxygen or ventilation: OR 2.87 (95% CI 2.03-4.06)
- Death: OR 4.57 (95% CI 3.32-9.01)
In a sensitivity analysis excluding patients with interstitial lung disease or cancer, risks were still elevated for both rituximab (OR 4.34, 95% CI 3.23-5.82) and JAK inhibitors (OR 2.14, 95% CI 1.64-2.79).
And in a propensity-score matched analysis similar results were seen, with ORs of 4.70 (95% CI 3.31-6.65) for rituximab and 2.09 (95% CI 1.50-2.90) for JAK inhibitors. Again, no associations were seen for abatacept or IL-6 inhibitors.
“Rituximab was strongly associated with poor outcome. The very elevated odds for poor COVID-19 outcomes in rituximab users highlights the urgent need for risk-mitigation strategies, such as the optimal timing of vaccination,” he concluded.
Study limitations included the voluntary reporting of cases, and the fact that cases were collected by class rather than by individual drug, which is especially relevant for the JAK inhibitors, which could have different biologic effects. In addition, TNF inhibition could have protective effects against COVID-19.
Sparks and co-authors disclosed multiple relevant relationships with industry, including Bristol-Myers Squibb, Gilead, Inova, Janssen, Optum, Amgen, Viela Bio, Principia/Sanofi, AbbVie, UCB, Pfizer, Merck Sharp & Dohme, Celltrion, Eli Lilly, Mylan, HEXAI, Samsung, Sanofi-Aventis, Biogen, Celgene, Grunenthal, Medac, Pharmakern, Boehringer Ingelheim, Chugai, Horizon, Sobi, and AstraZeneca.