This Week in Rheumatology — Aug 21, 2026

Welcome to This Week in Rheumatology. This week we're covering 10 notable papers spanning treatment strategy and drug delivery in rheumatoid arthritis, the lung as a shared target across connective tissue disease, and how classification criteria, immunogenicity and age shape the patients in front of us. Let's dive in.

We start with strategy in early rheumatoid arthritis, where RMD Open publishes the PRIMERA trial from Looijen and colleagues, a multicentre open-label randomised study asking whether we can do better than standard treat-to-target by personalising it. Three hundred and eight disease-modifying-drug-naive patients were randomised either to routine care, meaning methotrexate with glucocorticoid bridging for everyone and reassessment every three to four months, or to a tailor-made approach in which the first drug was chosen by autoantibody status, methotrexate for seropositive patients and hydroxychloroquine for seronegative patients, with the option to intensify as early as one month if disease activity remained above target. The result was plainly negative on both co-primary outcomes. Use of biologic or targeted synthetic drugs at ten months was around a fifth in each arm, numerically slightly higher in the tailored group, and mean disease activity trajectories were essentially superimposable. Patient-reported outcomes and adverse events did not differ either. So neither antibody-stratified drug selection nor very early intensification improved outcomes, and the practical message is that conventional treat-to-target with methotrexate and bridging steroids remains the benchmark. It also quietly reassures those treating seronegative patients that hydroxychloroquine-first did not produce worse disease control.

Staying with rheumatoid arthritis but moving to how we deliver and protect, Rheumatology reports a small randomised non-inferiority trial from Bovens and colleagues comparing subcutaneous with intravenous ultra-low dose rituximab. Thirty-five patients in stable low disease activity on ultra-low dose intravenous rituximab were randomised to 336 milligrams subcutaneously or 200 milligrams intravenously every six months. Drug exposure over the first six months was almost identical between routes, and the change in disease activity met the non-inferiority margin. The authors are appropriately cautious, because the exposure estimates carried moderate random error and the trial is tiny, so they stop short of a definitive claim. Still, for units already using ultra-low dosing, subcutaneous administration is a plausible route with obvious advantages in chair time and cost. Alongside that, a Norwegian modelling study in the same journal from Hagen and colleagues addresses something we can act on immediately, the adjuvanted recombinant zoster vaccine in rheumatoid arthritis. Over a lifetime horizon in a hypothetical cohort of ten thousand patients, vaccination prevented roughly two and a half thousand cases of herpes zoster and around thirteen hundred cases of postherpetic neuralgia, at an incremental cost of around eleven thousand euros per quality-adjusted life year, comfortably below the assumed threshold, with a 98 percent probability of being cost-effective. The model was most sensitive to the assumed risk of postherpetic neuralgia. If your clinic has been inconsistent about zoster vaccination in rheumatoid arthritis, this is health-economic support for making it routine.

The lung theme cuts across two diseases this week. In Rheumatology, Kanda and colleagues report real-world outcomes for biologics in rheumatoid arthritis-associated progressive pulmonary fibrosis. Out of nearly two and a half thousand patients starting a biologic, 443 had interstitial lung disease and 135 met criteria for progressive pulmonary fibrosis, treated with tumour necrosis factor inhibitors, abatacept or interleukin-6 receptor blockade. Across the whole group, joint disease activity improved substantially and lung function and computed tomography scores remained stable over 52 weeks, which is itself a reassuring signal. After propensity-score weighting there were no differences between classes in disease activity, imaging scores or drug retention, but forced vital capacity moved in opposite directions, declining about three percent with tumour necrosis factor inhibitors and improving about three percent with interleukin-6 receptor blockade. Interleukin-6 receptor blockade, a non-usual-interstitial-pneumonia pattern, and lower joint disease activity at one year all independently predicted improvement in forced vital capacity, while abatacept had the fewest adverse events. This is retrospective and the numbers per arm are small, so it is hypothesis-generating rather than definitive, but it supports tight articular control as part of lung management. Complementing that, Lupus Science and Medicine publishes a large systematic review from Semmler and colleagues pooling 98 studies and nearly 156 thousand patients on interstitial lung disease in systemic lupus erythematosus. Pooled prevalence was about 11 percent, rising to roughly 18 percent in studies that used high-resolution computed tomography, and clinical cohorts reported around a quarter while registry-based studies reported around 7 percent. Heterogeneity was extreme and the incidence estimate rested on only three registries, so treat that figure as exploratory. The clinically useful point is that lung involvement in lupus is likely commoner than we assume, and how hard you look determines what you find.

Two papers deal with the tools we use to label and to monitor patients. Rheumatology carries a meta-analysis from Riancho-Zarrabeitia and Riancho on the 2023 American College of Rheumatology and EULAR antiphospholipid syndrome criteria, pooling 24 studies. Specificity was essentially perfect at around 99 percent, but sensitivity was only about three quarters overall and, critically, phenotype-dependent. In thrombotic disease sensitivity was above 80 percent, whereas in obstetric antiphospholipid syndrome it fell to roughly a fifth to a third depending on how the data were pooled. Failure to classify was driven mainly by isolated IgM antibody positivity and by obstetric presentations. The takeaway is one criteria developers always make and clinicians often forget: these are classification criteria for research cohorts, not a diagnostic test, and using them at the bedside will systematically under-recognise obstetric disease. In RMD Open, Kimpton and colleagues systematically reviewed anti-drug antibodies in psoriatic arthritis across 28 studies. Prevalence varied enormously, highest with adalimumab, infliximab and golimumab, intermediate with ustekinumab, ixekizumab, certolizumab and guselkumab, and very low or absent with secukinumab and etanercept. For tumour necrosis factor inhibitors, antibodies consistently tracked with lower drug levels and poorer response, and concomitant methotrexate reduced their frequency. For interleukin-17, interleukin-12/23 and interleukin-23 inhibitors the data on drug levels and outcomes were sparse. Assay heterogeneity means the numbers are not comparable across studies, so the practical implication is limited to a familiar one: when a tumour necrosis factor inhibitor loses effect in psoriatic arthritis, immunogenicity is a real and methotrexate-modifiable explanation.

Finally, three papers on age and paediatric disease. In Rheumatology, Akay and colleagues describe 60 children with colchicine-resistant familial Mediterranean fever on canakinumab. Attack rates fell from a median of seven per patient-year to about one, and C-reactive protein dropped and stabilised after three months regardless of dosing interval. But inflammatory markers deceived: although the biochemical trajectories were identical across groups, complete remission was achieved in about 57 percent of those on a stable interval and only about 27 percent of those in whom interval extension failed, and every patient in that failed group had attacks. Arthritis at disease onset was the one independent predictor of successful extension, with roughly seven-fold higher odds, though the confidence around that estimate is wide. In short, do not use C-reactive protein alone to justify stretching the dosing interval. From the Royal Children's Hospital in Melbourne, Herman and colleagues report 57 children with chronic non-bacterial osteomyelitis, of whom 22 received zoledronic acid, averaging just under three doses. Just over three quarters of those treated showed significant clinical improvement, and among the ten with paired whole-body magnetic resonance imaging, eight improved radiologically. Transient flu-like symptoms occurred in about half, with no serious adverse events. Given six-monthly infusions, zoledronic acid looks like a practical option in refractory disease, albeit on retrospective single-centre evidence. And Aslan and colleagues compared 86 juvenile with 94 adult-onset dermatomyositis patients at one centre. Calcinosis and arthritis clustered in the juvenile group, with calcinosis in a third of children and a single adult case, anti-NXP2 dominated in children while antisynthetase antibodies dominated in adults, and steroid-free and drug-free remission were both more often achieved in juvenile disease. Malignancy occurred only in adults, in about a quarter, and each additional year of age at diagnosis carried a small but measurable increase in both malignancy and mortality risk.

If you only have time for one paper this week, make it the PRIMERA trial in RMD Open [1]. It is a properly randomised test of the personalisation strategies many of us have been tempted to adopt informally, and it tells us they do not improve outcomes over conventional treat-to-target.

Here are the key takeaways from this week in Rheumatology. Antibody-stratified first-line drug choice and one-month intensification did not beat standard treat-to-target in early rheumatoid arthritis, so stay with methotrexate, bridging steroids and three-monthly reassessment [1]. Recombinant zoster vaccination in rheumatoid arthritis is very likely cost-effective and should be routine [5]. Interstitial lung disease is commoner in lupus than most of us assume, particularly if you image, and in rheumatoid arthritis with progressive fibrosis, tight joint control matters, with a signal favouring interleukin-6 receptor blockade for lung function and abatacept for safety [7,4]. The 2023 antiphospholipid criteria are highly specific but miss most obstetric cases, so do not use them to rule out disease at the bedside [3]. And in colchicine-resistant familial Mediterranean fever, normal inflammatory markers do not license extending the canakinumab interval, because clinical remission rates diverge even when C-reactive protein does not [8].

That's your roundup for This Week in Rheumatology. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And a quick rating on Apple Podcasts or Spotify helps other physicians discover the show.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.