Triple therapy -- with methotrexate, hydroxychloroquine, and sulfasalazine -- is little used for the treatment of rheumatoid arthritis (RA) in the U.S. today, and clinical outcomes showed less efficacy than regimens combining methotrexate with a tumor necrosis factor (TNF) inhibitor, analysis of data from a large national registry found.
Among more than 15,000 patients enrolled in the Corrona RA registry from 2001 to 2019, only 2.5% ever initiated triple therapy compared with 97.1% who started a TNF inhibitor after being on methotrexate, according to Jeffrey R. Curtis, MD, of the University of Alabama at Birmingham, and colleagues.
After propensity score matching, the likelihood of being in a state of low disease activity, defined as a Clinical Disease Activity Index (CDAI) of 10 or lower, was greater in the TNF combination group than in the triple therapy group (49.2% vs 33.3%), with an odds ratio of 0.50 (95% CI 0.31-0.82), the researchers reported online in Arthritis Care & Research.
Several randomized studies have found similar outcomes for triple therapy and TNF inhibitor combination therapy. These results were seen for patients with early RA, methotrexate-naive patients, and in established RA. And given the expense associated with biologic therapies, triple therapy has been shown to be considerably more cost effective.
But clear information about the use of these regimens and clinical outcomes in real-world practice has been lacking, particularly in administrative claims databases that do not include data on disease activity.
To address this knowledge gap, Curtis and colleagues analyzed data from the Corrona registry, which has enrolled patients from 131 sites in 42 U.S. states. The outcomes of interest were discontinuation of treatment, changes in RA disease activity and physical function, and attainment of low disease activity or remission.
Many differences were seen before propensity score matching in the TNF inhibitor/methotrexate group versus the triple therapy group. Among those who were biologic-naive at the time of treatment initiation, those starting triple therapy were older (60 vs 57 years), had longer duration of disease (8.4 vs 6.5 years), and more often had a history of serious infections (6% vs 3%) and malignancy (13% vs 4%). However, the triple therapy group had lower disease activity, with mean CDAI scores of 16 versus 21, and had more comorbidities.
Similar differences were seen among patients who were biologic-"experienced" at baseline. Patients initiating triple therapy again were older (60.9 vs 57.1 years) and had longer disease duration (17 vs 11.5 years). They also had lower mean CDAI scores (17.4 vs 19.9) and more often had a history of cancer, serious infections, diabetes, and cardiovascular disease.
After propensity score adjustment, 118 patients initiating triple therapy were matched with 348 patients starting TNF/methotrexate treatment. Mean age in the matched groups was 61, three-quarters were women, and the majority were white. Baseline CDAI was 21 and the average number of tender and swollen joints was six to seven.
Among biologic-naive patients, treatment persistence at 1 year was lower in the triple therapy group (45% vs 69%). After further adjustment for factors including type of insurance and seropositivity, the likelihood of discontinuation was more than doubled in the triple therapy group (HR 2.17, 95% CI 1.63-2.88). Among biologic-experienced patients, the adjusted risk for discontinuation also was higher (HR 1.51, 95% CI 1.06-2.15).
Attainment of a change in CDAI greater than the minimally clinically important difference, which was a change in 12 units for those in high disease activity and six units for those with moderate activity, was less likely in the triple therapy group (33.3% vs 56.9%, OR 0.38, 95% CI 0.23-0.62). In addition, the mean decrease in CDAI was -5.5 units in the triple therapy group compared with -9.3 units in the TNF/methotrexate group, which represented a significant -3.8 (95% CI -1.5 to -6.1) unit difference.
Changes in the health assessment questionnaire also favored the TNF/methotrexate group, with an average change of -0.18 units.
The researchers offered several possible explanations for their findings. Noting that patients given triple therapy were older and had longer duration of disease, the team wrote: "We speculate that this may be that clinicians are more comfortable with biologics in younger patients with fewer comorbidities given safety concerns over their use in older, more ill patients, a fear that may or may not be grounded in evidence."
Regarding the differences observed in the earlier randomized clinical trials and real-world use, they noted that "patients participating in a clinical trial may have different motivations, and perhaps fewer alternative treatment options, than those receiving care in real-world settings."
Moreover, patient expectations may have been influenced by direct-to-consumer advertising, Curtis and co-authors suggested.
They concluded that there were multiple "meaningful" differences in patients given triple therapy versus anti-TNF/methotrexate treatment, but after propensity score matching, "biologic-naive TNF/methotrexate patients had superior persistence and clinical effectiveness outcomes."
"These real-world findings add to our understanding of the choice of triple versus TNF/methotrexate therapy," the researchers said.
A limitation of the study, they noted, was the possibility of residual confounding even after propensity score matching.
Disclosures
The authors reported financial relationships with Corrona (several co-authors are employees), AbbVie, Amgen, Bristol-Myers Squibb, Eli Lilly, Janssen, Myriad, Pfizer, Regeneron, Roche, UCB, GlaxoSmithKline, Gilead, and Sanofi.
Primary Source
Arthritis Care & Research
Source Reference: Curtis JR, et al "Real-world outcomes associated with triple therapy versus TNFi/MTX therapy" Arthritis Care Res 2020; DOI: 10.1002/acr.24253.