There is no reason to avoid the use of TNF inhibitors in patients with RA-ILD, according to a study that found no difference in survival and respiratory outcomes compared to other ILD therapies.
Speaking at the ACR Convergence 2023 meeting in San Diego, Associate Professor Bryant England presented data from a direct comparison of TNF inhibitors and non-TNF inhibitor biologics/JAK inhibitors which showed no difference in deaths or respiratory-related hospitalisations between the two groups of agents.
Associate Professor England, from the University of Nebraska Medical Center, said there were a lot of evidence gaps in the treatment of RA-ILD and particularly in the use of advanced therapies.
Some earlier comparative studies between TNFis and csDMARDs and between TNFis and rituximab had suggested some differential outcomes.
“In both of those studies, there was some concern that survival was poor in RA-ILD patients who were treated with TNF inhibitors, although neither of these findings were significant.”
“So this raises the question that I think we as providers are struggling with: are TNF inhibitors safe to be used in people with RA-ILD? In fact, there have been some organisations over time that have suggested we should avoid these therapies in this population.”
To provide more evidence, his study compared new users of advanced therapies in a data set of RA-ILD patients from the US Veterans Health Administration between 2006 and 2021 and using linkages to Medicare data and the National Death Index.
From a data set of over 1,000 veterans, the study propensity matched 237 initiated on TNFis and 237 initiated on non-TNFs/JAKi. Most patients were male (91%), white (80%), and about half (48%) were current smokers.
Given strong temporal trends in the use of the different agents, the investigators constructed time-specific propensity score models for an early study period, a mid study period and a late study period.
The study found no significant difference between treatment groups in the primary outcome of a three-year composite of death or respiratory-related hospitalisation (aHR 1.21).
Similarly, both outcomes separately – death (aHR 1.11) and respiratory-related hospitalisation (aHR 1.27) – were not significantly different between the treatment groups.
In the 36% of deaths which were respiratory-related, there was also no difference between the TNFi and non-TNFi/JAKi groups.
Similar findings were seen in one-year secondary outcomes – no significant differences between the treatment groups in the composite outcome, death and respiratory-related hospitalisations individually and in respiratory-related deaths.
Associate Professor England said different subgroup and sensitivity analyses largely supported the overall findings for no increased risk of death or respiratory hospitalisation with TNFis.
“If we go back to the question we asked at the beginning – are TNFis safe to be used in RA-ILD – the answer is, it’s complex.”
“So our findings don’t suggest that we should be systematically avoiding TNF in every single person with RA-ILD. But that’s different than whether there are specific subpopulations of RA-LD for which the choice of these therapies may differ.”
He said the study was unable to address issues such as ILD pattern and trajectory which could be important in such a heterogeneous disease.
Another limitation of the study was the male-dominated cohort which may impact the generalisability of the findings.
He called for more clinical trials to generate robust evidence on which to base clinical decisions regarding advanced therapies in RA-ILD.
Impressive and important study supporting safety of TNF inhibitors among patients with RA-ILD. Elegant methods of target trial emulation. Congrats to the authors on plenary presentation!#ACR23 pic.twitter.com/P7keAoOX7M
— Jeffrey Sparks MD MMSc (@jeffsparks) November 13, 2023