June 2026 Research Update

Across these papers, adult-onset Still's disease (AOSD) is shown as a diagnosis that is often reached during a broader search for the cause of ongoing inflammation or fever, after other causes are checked. The overall pattern is that symptom timing (especially fever, sore throat, and rash) and repeat re-evaluation as new signs appear can influence how quickly AOSD is recognized, and imaging can contribute but is not specific. [1] [2] [3] [4]

In a retrospective study of adults evaluated for inflammation of unknown origin or fever of unknown origin, noninfectious inflammatory diseases were the most common final category, and AOSD was reported as more prevalent among the fever-of-unknown-origin group. In the same study, a whole-body inflammation scan (18F-FDG PET/CT) was reported to contribute to the diagnosis in 52.1% of cases, with 85% sensitivity and 17% specificity, meaning it often detected inflammation but also frequently flagged findings that were not ultimately diagnostic. [1]

In an 89-person AOSD cohort, initial symptoms related to diagnostic timing: fever and sore throat tended to rise quickly in the two weeks before the point when patients met diagnostic classification, while rash at the start was linked to delays in blood testing and reaching classification. The same study reported that when sore throat was the first symptom, ferritin tended to be higher and macrophage activation syndrome occurred more often, but long-term outcomes did not differ by the initial symptom group. [2]

Two case reports highlighted why diagnosis can be difficult in day-to-day practice: symptoms and test results can overlap with infections, immune system overactivation syndromes, kidney disease, or cancer concerns, so the working diagnosis may change over time. One patient was ultimately reclassified as AOSD only after multidisciplinary reassessment and new biopsy information as the clinical picture evolved. [3] [4]

For patients and caregivers, these studies mainly provide background on why getting to an AOSD diagnosis can take time: clinicians often start with a broad evaluation for fever or inflammation, and the diagnosis may depend on how symptoms unfold (for example, whether rash is present early) and whether complications appear that require reassessment. Imaging may help show where inflammation is occurring, but in the study here it was not specific enough to point to a single cause on its own, and case reports show that rare overlaps can further complicate the picture. [1] [2] [3] [4]

Across these reports (each describing one person), the main pattern was diagnostic confusion: some people had adult-onset Still's disease (AOSD) with uncommon features, while others had AOSD-like inflammation caused by a different underlying condition or trigger. Unusual signs involved the skin and the heart lining (pericardium), and several reports highlighted very high inflammation blood markers, including ferritin and interleukin-18 (IL-18). [5] [6] [7] [8] [9]

One report described AOSD with an atypical skin eruption where a skin biopsy showed neutrophil collections near the skin surface and many dying (necrotic) skin cells, after other diagnoses were ruled out. The person had a good initial response once corticosteroids were started, suggesting that AOSD can have skin changes that do not match the classic description. [5]

Another report described a patient whose most striking problem was a massive hemorrhagic pericardial effusion (blood-tinged fluid around the heart) and who did not have a rash. After infection and cancer were investigated, glucocorticoids led to a dramatic improvement within 48 hours, including fever resolution and less pericardial drainage, supporting AOSD despite an unusual presentation. [6]

Several papers highlighted situations where AOSD can be mistaken for, or occur alongside, other diagnoses. One patient treated as AOSD was later confirmed to have VEXAS syndrome through UBA1 genetic testing; the illness was glucocorticoid-refractory and improved rapidly after baricitinib, with remission reported by 3 months and no relapse at 6 months. [8]

Two reports focused on AOSD-like illness in complex clinical settings: one during clozapine titration (with rising IL-18 and ferritin and no clear infection) and another after PCR-confirmed scrub typhus in a patient with seronegative rheumatoid arthritis. In these single-patient descriptions, fever and other Still-like features improved after stopping clozapine plus corticosteroids and supportive care in one case, and improved rapidly after tocilizumab in the other. [7] [9]

For patients and caregivers, these reports are mainly background information showing that AOSD symptoms can be atypical (for example, unusual rashes or fluid around the heart) and that some other conditions or medication-related reactions can closely mimic AOSD. Because all the evidence here comes from individual cases, it cannot show how often these situations happen or which features reliably separate AOSD from a mimic; it mainly illustrates why careful evaluation and follow-up are often needed when the presentation is unusual or the response to steroids is unexpected.

Across these studies, a common pattern was that earlier disease control on biologic treatments was connected to a better chance of reducing or stopping glucocorticoids (steroids) later on. The clearest signals came from (1) meeting specific early response targets on tocilizumab and (2) starting canakinumab closer to the time symptoms began, but the studies were not designed to prove cause and effect.

In a small, post-hoc (exploratory) analysis of adults treated with tocilizumab, more than half reached inactive disease by the final follow-up, but fewer reached inactive disease without glucocorticoids. Patients who met an early week 4 target or were inactive by month 6 were more likely to later reach glucocorticoid-free inactive disease. [10]

In the international AIDA registry study of canakinumab, starting treatment within 3 months of disease onset was linked to faster steroid reduction by months 3 and 6 compared with starting later. Early starters were also more likely to stop canakinumab because of long-term remission, and they had a higher frequency of a monocyclic (one-episode) disease course than those who started later. [11]

One exploratory finding from the tocilizumab analysis was that people who were inactive by month 6 had higher swollen joint counts and lower interferon-gamma (a blood immune signal) at baseline. This suggests that baseline features might relate to who achieves longer-term steroid-free control, but it needs confirmation in larger studies. [10]

For patients and caregivers, these studies add context about steroid-sparing outcomes: some people can reach periods of inactive disease with reduced or even no glucocorticoids while using biologic treatments. The canakinumab results reflect real-world care in a large registry, while the tocilizumab results highlight that early response milestones may relate to longer-term steroid-free control; however, neither study on its own can guarantee what will happen for an individual, and safety details were not reported in these summaries.

Across this study, relapse risk in adult-onset Still's disease appeared to be partly predictable from what was present at diagnosis. Higher levels of inflammation in bloodwork and a joint-dominant pattern of symptoms stood out, along with the type of first treatment used, but the results come from a retrospective cohort and may not apply to every setting. [12]

In 87 adults treated at two rheumatology centers in Turkey, 31 people (35%) had a relapse during follow-up. Three baseline features independently predicted relapse: C-reactive protein (CRP) above 100 mg/L, a joint-dominant phenotype at baseline, and using a conventional synthetic disease-modifying antirheumatic drug (csDMARD) as first-line therapy. [12]

When the researchers combined these three predictors into one model, it separated relapse from non-relapse reasonably well (area under the curve 0.84). This suggests that a small set of baseline features may provide a useful snapshot of relapse risk, although treatment choice could also reflect how severe or joint-focused the disease was (a type of confounding). [12]

For patients and caregivers, this research mainly adds background about which early features sometimes travel together with later relapse, rather than offering a clear rule for what will happen to any one person. It is most relevant for people whose early disease is strongly joint-focused or who have very high CRP at diagnosis, but it remains uncertain how well the same predictors work in other countries, care systems, or treatment approaches because the study was small and looked back at records. [12]

Across these papers, the overall picture is that severe complications in Still's disease can involve multiple organs and body systems at once, including the blood-clotting system, liver inflammation pathways, and the lungs. The evidence ranges from a single-patient report to small observational clinical data and lab and animal research, so the certainty about what applies broadly is still limited. Together, they highlight that severe complications often overlap with macrophage activation syndrome, a dangerous over-activation of the immune system. [13] [14] [15]

One case report described a 22-year-old man with adult-onset Still's disease who developed macrophage activation syndrome along with early signs of a serious clotting disorder (pre-disseminated intravascular coagulation). After receiving high-dose steroids plus the JAK inhibitor tofacitinib, along with plasma transfusion, anticoagulation, and supportive care, he improved quickly and remained in remission during follow-up. Because this is one person, it shows what can happen in an individual case but cannot predict results for others. [13]

A translational study suggested that a molecule made by immune cells (itaconate) may have a double-edged role during macrophage activation syndrome: it reduced some inflammatory signals but increased another signal called CXCL10, which was linked to CD8 T cell buildup in the liver and signs of liver injury in a mouse model. In mice engineered to lack the enzyme that produces itaconate (Acod1), disease severity was reduced and liver CXCL10 and CD8 T cell infiltration were lower. This is mainly background research that helps explain possible pathways for organ injury, not a test of a treatment in people. [14]

A multicenter retrospective study in children with systemic juvenile idiopathic arthritis described a serious lung complication (systemic juvenile idiopathic arthritis-associated lung disease) that often followed earlier macrophage activation syndrome, with cough in all patients and shortness of breath in many. Imaging commonly showed ground-glass changes in the lungs, and most patients had treatment changes after lung disease was recognized; some improved on JAK inhibitors, but one patient died from fulminant macrophage activation syndrome. Although this is a pediatric condition, it adds context about how severe inflammation can extend to the lungs and how outcomes can vary. [15]

For patients and caregivers, this group of studies reinforces that severe Still's-related complications can involve more than joints and fever, including clotting abnormalities, lung symptoms (like cough and shortness of breath), and pathways linked to liver injury during macrophage activation syndrome. What is most directly relevant now is the clinical description of these complications and their real-world severity; the laboratory work on itaconate is mainly important for understanding biology and may influence future research. Big uncertainties remain because one report is a single case and the other clinical study is a small retrospective series in children, so it is not clear how often these complications occur in adults, which factors predict them, or which approaches work best for different people. [13] [14] [15]

Across the paper in this group, the overall message was that adult-onset Still's disease (AOSD) did not show a clear genetic signal pointing to a direct increase in risk of idiopathic thrombocytopenic purpura (ITP), a condition where the immune system attacks platelets. At the same time, the study suggested that some other autoimmune diseases may have different (higher or lower) genetic links to ITP, but the authors noted that confirmation in larger genetic datasets is needed. [16]

Using a two-sample Mendelian randomization approach (a genetics-based method meant to help separate correlation from cause), the researchers found no evidence that genetic risk for AOSD causes ITP. In the same analysis, Sjogren's syndrome showed a higher genetic risk for ITP, while polymyositis and primary sclerosing cholangitis (an autoimmune disease affecting the bile ducts) showed a lower genetic risk for ITP. [16]

The study also highlighted an important uncertainty: for several diseases, there were limited genetic instruments available, which can make results less stable. Because of this, the authors recommended larger genetic studies to verify whether these patterns hold up. [16]

For patients and caregivers, this is mainly background research about long-term autoimmune risk patterns rather than something that changes day-to-day care. It may be reassuring that, in this genetics-based analysis, AOSD itself did not appear to be a direct genetic cause of ITP; however, the study cannot rule out non-genetic causes of low platelets or effects related to inflammation, infections, or medications. What remains uncertain is whether future, larger genetic studies will confirm the same results, especially for conditions where the current genetic data were limited. [16]