Among patients with ankylosing spondylitis or undifferentiated spondyloarthritis, risk for anterior uveitis may hinge on the choice ofbiologic disease-modifying antirheumaticdrug (bDMARD), a large Swedish cohortstudy suggests.
Study results were reported in the openingplenary abstract session at the congress.
“Randomised, controlled trials indicate
that compared to tumour necrosis factor
(TNF) inhibitors, secukinumab has similar
efficacy regarding axial inflammation in
spondyloarthritis and better efficacy re-
garding cutaneous psoriasis, but is inferior
in inflammatory bowel disease,” noted lead
investigator Dr. Ulf Lindström, of the depart-
ment of rheumatology and inflammation
research in the Institute of Medicine at the
University of Gothenburg (Sweden). “How-
ever, the efficacy of secukinumab, compared
The investigators used national registry
data to study 3,568 patients with ankylos-
ing spondylitis or undifferentiated spon-
dyloarthritis who started bDMARDs in
2005-2018. They considered four agents:
the anti–interleukin-17A antibody secuki-
numab and the TNF inhibitors etanercept,
adalimumab, and infliximab.
Analyses based on 4,523 treatment episodes showed that after excluding the 23%of patients who had previously experiencedanterior uveitis, merely 0.9% of patientsexperienced new-onset anterior uveitis whileon their bDMARD, Dr. Lindström reported.
There was confounding by indication,whereby patients with previous anterior uveitis were channeled toward adalimumab andinfliximab, and away from secukinumab andetanercept. In addition, there was confounding by line of treatment, with secukinumabusually used in the third line.
After excluding patients who had expe-
rienced anterior uveitis in the past year to
partly address confounding, the adjusted
risk for first on-treatment anterior uveitis
was about twice as high with secukinum-
ab and with etanercept as compared with
adalimumab. After additionally excluding
all biologic treatment episodes beyond the
third line, elevation of risk remained signif-
icant only for etanercept.
“There is probably a higher occurrence ofanterior uveitis on treatment with secukinumab, compared to adalimumab, but theremay still be residual confounding and biasthat we need to consider,” Dr. Lindströmconcluded.
Findings in context
“These results are not surprising as we
have known that secukinumab and etaner-
cept are not good for controlling recurrent
and chronic uveitis,” Dr. Nigil Haroon com-
mented in an interview. However, “a single
episode of uveitis or infrequent episodes
are not usually considered a contraindica-
tion to starting these drugs.”
Study caveats included lack of adjust-
ment for uveitis severity and potentially
missed uveitis episodes in patients who
treated it themselves with steroid eye-
drops, he said. “Standard practice is to
keep drops with them to start at the earli-
est possible time point.”
“It would be useful to know the number
of patients who stopped medications as a
result of uveitis,” added Dr. Haroon, who is
codirector of the spondylitis program at the
University Health Network and associate
professor of medicine and rheumatology at
the University of Toronto. “Time-to-event
“The study raises an important point re-
garding channeling bias, and this is import-
ant to consider when interpreting clinical
trial data as well. Investigators are unlikely
to include patients with history of uveitis
(or strong family history of inflammatory
bowel disease or personal history of gut
symptoms) in studies with IL- 17 inhibitors
and etanercept. Hence, the results have to
be interpreted with caution.”
Dr. Lindström and coinvestigators assessed incidences of any anterior uveitis(ascertained from outpatient ophthalmology visits having this diagnostic code)and of anterior uveitis flares (the subsetoccurring after a gap of at least 90 dayswithout the diagnosis).
When they excluded patients who hadexperienced anterior uveitis in the yearbefore starting therapy, secukinumab andetanercept carried the highest incidencesof anterior uveitis (6.8 and 7. 5 per 100patient-years, respectively) and anterioruveitis flares ( 2. 8 per 100 patient-years foreach), he reported.
With adalimumab as the comparator,adjusted risk for first on-treatment anterioruveitis in this population was significantlyhigher with secukinumab (hazard ratio,
2.23) and etanercept (hazard ratio, 1.80).
When the investigators additionally
excluded episodes of therapy beyond the
third line, only etanercept carried notably
higher incidences of anterior uveitis ( 7.0
per 100 patient-years) and anterior uveitis
flares ( 2.6 per 100 patient-years). “This
could imply that some of the higher inci-
dence rate seen for secukinumab could
be due to the fact that these patients are
harder to treat and have received more bi-
ologics before,” Dr. Lindström proposed.
With adalimumab again as the comparator, the adjusted risk for first on-treatmentanterior uveitis in this population wassignificantly higher only with etanercept(hazard ratio, 1.85).
A final analysis included all patients whostarted adalimumab in 2004-2018 and thenswitched to one of the other agents, dramatically reducing confounding by indication. Inthis population, the incidence rate ratio ofanterior uveitis flares was 3.05 for secukinumab, 1.79 for etanercept, and 0.53 forinfliximab, compared with adalimumab.
Dr. Lindström disclosed that he had norelevant conflicts of interest. The studydid not receive any specific funding. Dr.Haroon disclosed consulting for Amgen,AbbVie, Janssen, Lilly, Novartis, and UCB.
Some biologics may be better than others
for averting anterior uveitis
DR. LINDSTRÖM DR. HAROON