May 2026 Research Update

These studies move the biomarker story in two directions at once: IL-18 looks promising for separating adult-onset Still's disease (AOSD) from similar conditions, but the exact number you get can depend strongly on how it is measured. That means IL-18 is becoming clearer as a signal of inflammation in AOSD, while the practical challenge of getting consistent results across labs is also becoming clearer. Together, they describe both the potential and the measurement problem that can affect how confidently a blood marker is interpreted across different settings. [1] [2]

In one study of Japanese adults, the blood level of active IL-18 was higher in AOSD than in rheumatoid arthritis, familial Mediterranean fever, and healthy controls, and it tracked with how active the disease seemed to be. Active IL-18 also went down after immunosuppressive treatment in the subset with follow-up samples, which fits with the idea that it reflects ongoing inflammation. The same study reported that higher active IL-18 levels were seen more often with rash and an enlarged spleen, tying the marker to recognizable clinical features. [1]

A separate laboratory-focused study compared IL-18 and other inflammation markers across common testing platforms and found large differences in measured values between platforms. Even when results moved in the same direction, the absolute numbers could differ substantially, and mathematical corrections only partly reduced these gaps. This helps explain why a cutoff value from one study may not transfer cleanly to another lab or hospital if the testing method is different. [2]

For patients, this month adds clarity that IL-18 is a marker that often runs high in AOSD and tends to fall when inflammation is treated, which may make it useful for diagnosis and tracking in research settings. At the same time, the measurement study shows that IL-18 numbers are not automatically interchangeable between different labs or test systems, which can create confusion if results are compared across places or over time using different methods. What remains uncertain is how best to standardize IL-18 testing so that a single threshold or target number means the same thing everywhere. This is mainly about improving understanding and test reliability, not a change in day-to-day care on its own. [1] [2]

This month adds a clearer picture of which complications and patterns may show up over years in Still's disease, based on a single-center cohort followed over a decade. The study reported that macrophage activation syndrome (MAS) was common in this group, and interstitial lung disease (ILD) occurred in a smaller subset, especially among those with relapsing disease. It also highlights that some early clinical features were linked to later course patterns, while the timing of starting anakinra (an IL-1 blocking medicine) was not independently tied to better long-term results in their analyses. [3]

In this cohort from Turkey (focused on pediatric Still's disease/sJIA), MAS was reported in 36.7% of patients and ILD in 8.2%, with ILD appearing only in the relapsing subgroup. Serositis at diagnosis (inflammation around organs, such as the lining of the lungs or heart) was reported as a strong predictor of MAS, and older age at onset plus rash were linked to a polycyclic (relapsing) course. The study also reported that ILD was strongly linked to MAS and relapsing disease, connecting these complications rather than treating them as isolated problems. [3]

When the authors compared earlier versus later anakinra starts, earlier treatment was used more often in severe cases, but it was not independently associated with improved long-term outcomes. In practical terms, their data suggested that starting anakinra within 3 months led to outcomes comparable to starting within 2 weeks, at least in this single center and with the limitations of retrospective data. This does not settle the best timing question, but it narrows what can and cannot be concluded from routine-care comparisons. [3]

For patients and caregivers, this study mainly helps set expectations about what researchers watch for over time in Still's disease: MAS and ILD were both seen, and certain early features (like serositis) lined up with higher MAS risk in this cohort. It also suggests that simply comparing who started anakinra earlier versus later in routine care may not clearly show a long-term advantage, partly because sicker patients may be treated sooner. What remains uncertain is how well these patterns apply to adult-onset Still's disease specifically, since this cohort was pediatric Still's disease/sJIA, and how these findings would look across multiple centers. This is useful context for understanding prognosis research, rather than something that directly changes care by itself. [3]

This month's case reports show that AOSD can be harder to recognize than the textbook description, especially in older adults, during pregnancy, or when severe inflammation complicates the picture. Several reports described unusual or misleading signs, such as persistent fever without the typical rash, prominent lymph node findings that mimicked cancer, or unusual skin changes. Across these individual stories, improvement was often reported after corticosteroids, and some cases added targeted immune medicines (such as anakinra, tocilizumab, or infliximab) when disease was severe or complicated. [4] [5] [6] [7] [8] [9] [10] [11]

Older-adult presentations in these reports often centered on long, unexplained fever and less typical symptom patterns. One 57-year-old woman had fever and serious neurologic symptoms without the characteristic rash and improved after steroids plus a single dose of tocilizumab (an IL-6 blocking medicine). A 72-year-old man had months of fever of unknown origin with broad negative testing, improved with prednisone, and started outpatient tocilizumab based on partial fulfillment of AOSD criteria. [8] [9]

Several reports focused on very high inflammation with MAS or HLH-like complications, where diagnosis and rapid treatment decisions were central to the course described. A 60-year-old woman initially improved on steroids but later developed MAS and improved after methylprednisolone and anakinra (an IL-1 blocking medicine). Another patient had lymph node biopsy findings that initially resembled lymphoma, but further review and HLH criteria supported reactive inflammation with secondary HLH, and she improved with high-dose steroids used in an HLH protocol. [4] [5]

Two reports highlighted pregnancy as a setting where treatment and safety questions come to the foreground, but details were uneven across papers. One pregnant woman diagnosed by Yamaguchi criteria was treated with oral steroids and infliximab for the rest of pregnancy, without pregnancy outcomes described. Another pregnancy case report described initial disease control with anakinra but suspected liver injury that improved after stopping it, use of intravenous immunoglobulin, delivery at 34 weeks with a healthy baby, and postpartum control with tocilizumab followed by drug-free remission after stopping therapy. [6] [10]

AOSD in these reports also included unusual skin and kidney involvement that could distract from the underlying diagnosis. One man had persistent itchy, hyperpigmented plaques along with fever and joint pain, met Yamaguchi criteria, and improved quickly on steroids plus anti-inflammatory pain medicine, with no recurrence noted at 3 months. Another man developed MAS along with a kidney disease pattern called membranoproliferative glomerulonephritis and improved after pulsed steroids followed by oral steroids and tacrolimus, alongside supportive treatments as described. [7] [11]

For patients, these reports mostly add awareness of the range of ways AOSD can show up, including cases without a classic rash, cases dominated by prolonged fever, and cases complicated by MAS or HLH-like syndromes. They also show the kinds of targeted medicines doctors may try in severe or special circumstances, including anakinra, tocilizumab, infliximab, tacrolimus, and intravenous immunoglobulin, but these are individual experiences rather than proof that one approach works best. What remains uncertain is how often these atypical patterns occur and which treatments are safest and most effective in situations like pregnancy, because the evidence here is based on single patients. This information may matter now mainly as context for understanding diagnosis delays and the range of reported complications, while stronger answers require larger studies. [4] [5] [6] [7] [8] [9] [10] [11]