Adult-onset Still's disease: 2026 so far

Observational data increasingly linked biologics with less steroid exposure.

In January 2026 to September 2026, several non-randomized studies linked biologic-based strategies with faster early remission and/or greater steroid reduction. This was reported especially with tocilizumab (an IL-6 blocker) paired with a short-term high-dose steroid approach, and with biologic-first starting strategies. Longer-term follow-up of tocilizumab in a small extension cohort described maintained control for many people who stayed on treatment and steroid dose reductions. In that group, some serious adverse events and treatment stops were reported, but no new safety signals were reported. Registry and cohort reports also linked earlier canakinumab (an IL-1 blocker) use to faster steroid tapering and more discontinuations due to long-term remission, while one cohort reported similar response rates between anakinra (an IL-1 blocker) and tocilizumab. In these routine-care reports and selected follow-ups (not randomized comparisons), biologic use was repeatedly associated with periods of control that sometimes allowed glucocorticoids (steroids) to be lowered or stopped.

Research covered in: January 2026, February 2026, March 2026, April 2026, June 2026, July 2026, August 2026

Reports clarified common MAS escalation patterns and relapse risk.

In the months covered, routine-care data and a global physician survey described high-dose glucocorticoids (steroids) as the usual starting treatment for MAS. They also described frequent add-on use of anakinra and ciclosporin (immune-suppressing medicines). A multicenter cohort focused on glucocorticoid-refractory MAS (MAS that did not respond to steroids). This group was mostly systemic JIA (systemic juvenile idiopathic arthritis, a childhood inflammatory arthritis) with a smaller adult-onset Still's group. The cohort reported that many patients had at least partial responses by lab and clinician assessment, but MAS recurred in about a third and deaths occurred, showing that early improvement did not always mean MAS was fully resolved. Smaller reports described additional rescue approaches in severe cases, including therapeutic plasma exchange (a procedure that filters plasma from the blood) and, in single cases, interferon-gamma blockade and JAK inhibition (medicines that block inflammation signaling) after earlier treatments were insufficient. These reports show that MAS care in real life was often intensive and used multiple drugs, and that relapse after initial improvement was a documented risk in these datasets.

Research covered in: February 2026, March 2026, June 2026, September 2026

Ferritin and systemic scoring repeatedly anchored MAS prediction models.

From January through September 2026 reports, ferritin (a blood protein that often rises with inflammation) and measures of overall disease burden repeatedly showed up as key MAS risk signals. Several groups moved from single markers toward combined prediction models. A large registry analysis and a single-center cohort both highlighted ferritin and systemic score as linked to MAS, and the registry modeling estimated substantially higher predicted MAS probabilities when multiple risk factors clustered. Separate work proposed an admission-based bedside score using five routine items. Later reports showed very strong separation of MAS vs non-MAS cases using HScore (a scoring system used to estimate risk of severe immune over-activation) in one dataset, with a modified version still performing well without bone marrow findings. These studies point toward more structured, numbers-based ways to flag higher MAS risk or identify MAS earlier, although the reports did not show that using these tools in care pathways improves outcomes.

Research covered in: January 2026, March 2026, April 2026, August 2026

Multiple studies tested simpler, more responsive activity scores.

In the months covered, studies proposed and validated practical ways to label AOSD as active vs inactive and to measure change over time using structured scores. Early in 2026, a five-item symptom checklist plus CRP (C-reactive protein, a blood marker of inflammation) was reported to separate active from inactive disease well in a cohort and an external dataset. Later work compared several activity scores against physician assessment. It suggested some tools were more responsive to change over time (e.g., modified DAVID) while others aligned more with joint activity (e.g., modified SDAS). A separate validation study reported strong agreement of sJADAS10 (a 10-item joint and symptom activity score) with a reference activity definition across clinic visits. These reports suggest follow-up and research comparisons may become more consistent as shared activity definitions are used more often, but they did not identify a single best score for all adults in routine care.

Research covered in: March 2026, August 2026, September 2026

Reports added tools and examples for difficult diagnoses.

Across January to September 2026, diagnosis-focused reports showed that AOSD can present in less typical ways. This included fever-dominant illness without classic rash. Reports also showed it can be confused with infections, malignancy (cancer), drug reactions, or other inflammatory conditions, with very high ferritin often described as a clue in case reports. Cohort work described reasons diagnosis can be delayed, including symptom timing patterns, and found that rash at onset was linked to delays in testing/classification in one cohort. The same work noted that PET/CT (a scan that can show inflammation) often showed inflammation but was not specific. A record-based model showed high accuracy for identifying AOSD among people with fever of unknown origin when AOSD was already suspected. A small clinicopathology series reported a skin-biopsy pattern (superficial dyskeratotic keratinocytes, a specific change seen under the microscope) in many confirmed cases. In the months covered, these reports added decision-support tools and supportive clues for complex diagnostic workups, while also documenting that an AOSD-like picture can sometimes turn out to be a different underlying disease.

Research covered in: January 2026, February 2026, June 2026, July 2026, September 2026