From April 2015 to May 2024, this study enrolled 15 adult patients (≥18 years old) diagnosed with YAOS at our tertiary hospital. YAOS is diagnosed with the fulfillment of 2 major criteria, at least one minor criterion, the molecular criterion, and the exclusion criteria. Major clinical criteria: (1) periodic occurrence≥twice; (2) recurrent fever or dermatitis or both. Minor clinical criteria: (1) Oligo- or polyarthralgia/inflammatory arthritis, or distal extremity swelling; (2) abdominal pain or diarrhea or both; (3) sicca-like symptoms; (4) pericarditis or pleuritis or both. Molecular criterion: NOD2 IVS8⁺¹⁵⁸ or R702W or both, or other rare variants. Exclusion criteria: High titer antinuclear antibodies, inflammatory bowel disease, Blau syndrome, adult sarcoidosis, primary Sjögren syndrome, and monogenic autoinflammatory diseases based primarily on genetic testing results, clinical manifestations, and therapeutic response (e.g., no response to colchicine) (16). Informed consents were obtained from all participants. This study was approved by the Institutional Review Board of Peking Union Medical College Hospital and was performed according to the Declaration of Helsinki.

Clinical and demographic information was meticulously documented. Whole-exome sequencing based on the NovaSeq 6000 next-generation sequencing platform was performed in each patient. Sanger sequencing was conducted in probands’ families to identify and confirm NOD2 sites when feasible. The remission of clinical symptoms was defined as decreases in the duration or frequency of each episode, as well as severity of clinical symptoms after treatment. Response categories were classified as good, partial, or no response, based on the remission of clinical symptoms and normalization of inflammatory markers by more than 80%, 20%–80%, or less than 20%. Clinical data and mutation information were compared with the largest Caucasian cohort (3).

IBM SPSS Statistics (V.25) was used to analyze the data. Continuous variables were expressed as mean ± standard deviation or median and interquartile range. Categorical variables were described as frequency distribution. Student’s t-test was used to compare continuous variables between two patient groups. Chi-square was performed to compare categorical variables. Fisher’s exact test was used to compare the prevalence of different genotypes between our cohort and the general East Asian population. p<0.05 was considered statistically significant.

The demographic and clinical characteristics of Chinese YAOS patients are shown in Table 1 (details shown in Supplementary Table 1). Among these 15 Han patients with YAOS, the gender ratio of male to female was 7: 8, and the average age at the onset of the disease was 35 (_S.D.17) years old, with 2 patients (2/15,13.3%) experiencing disease onset in juvenile (≤14 years old). The average age of disease diagnosis was 42 (_S.D.14) years old, with a median duration of disease at diagnosis was 5 years (ranging from 1 to 30 years). Only 2 patients (1/15,13.3%) reported a family history of recurrent fever and/or autoimmune diseases.

All patients experienced unprovoked, self-limiting episodes of fevers (Tmax>38.5°C), lasting 1 to 2 weeks and occurring with frequencies ranging from once a week to once a year. 11 patients (11/15, 73.3%) presented with gastrointestinal symptoms, with abdominal pain (8/11,72.7%), diarrhea (7/11,63.6%), nausea and vomiting (7/11,63.6%), and constipation (2/11,18.2%). Additionally, 6 of those with gastrointestinal symptoms reported weight loss, and 4 exhibited anemia, though none had progressive anemia or malnutrition. The abdominal pain usually affects the left and upper abdomen. A history of upper gastrointestinal bleeding was noted in 2 patients. Among 8 patients who displayed abdominal pain, 7 underwent extensive examination including computed tomography (CT), abdominal ultrasound, positron emission tomography (PET)/CT, scanenterography, oesophagogastroduodenoscopy, capsule endoscopy, colonoscopy, and mucous biopsy with no evidence of inflammation bowel disease (IBD).

Arthralgia/arthritis was observed in 7 patients (7/15, 46.7%), affecting multiple joints across both upper and lower extremities, with 3 (3/7,42.9%) patients involving the upper extremities. 3 patients performed MRI or ultrasound of the joints, 2 presented with joint effusion, and 1 reported bone marrow edema. No bone erosions were reported radiographically. Bilateral lower extremity swelling was noted in 7 patients (7/15,46.7%) ( Figures 1A, B ). Myalgia was reported in 7 patients (7/15, 46.7%), with 2 (2/6,33.3%) only on the legs, and others are generalized. 3 patients underwent further examinations such as MRI, enzymes, electromyography, and muscle biopsies, and showed no typical inflammatory myopathy changes consistent with myositis. Fatigue and lymphadenopathy were also reported in 46.7% of patients.

Skin rash was found in 5 patients (5/15,33.3%), mostly (3/5,60%) manifesting as erythematous patches and plaques, 2 (2/5,40%) as papule, 1 as purpura, and 1 as acne. The rash mainly affected the limbs (4/5,80%) ( Figures 1C, D ), with only 1 on the trunk and face. There were no reports about genital and perianal areas involved, vasculitis, erythema nodosum, or pyoderma gangrenosum.

Headaches occurred in 26.7% of patients. Proteinuria/hematuria, chest pain, periorbital edema, oral ulcers, and throat pain occurred in 20%. Scca-like symptoms (xerostomia), pleuritis, and anxiety and depression were reported in 13.3% of patients. Pericarditis was reported in 6.7%. There were no reports of uveitis.

The acute-phase reactants including erythrocyte sedimentation rate (ESR) and/or C reactive protein (CRP) were increased during episodes in all patients [29 (IQR 51) mm/h, and 78.76 (_S.D.68.2) mg/L, respectively]. Leucocytosis was observed in 10 patients (10/15,66.7%). The mean values for white blood cell (WBC) count, hemoglobin (Hb), and platelets (PLT) were 13.67 (_S.D.6.27)*10⁹/L,115.93g/L (_S.D.23.93), and 266.8 (_S.D.98.18)*10⁹/L, respectively. 6 patients have transient elevation of liver enzymes. Elevated tumor necrosis factor (TNF) was observed in 9 patients (9/10,90%) with a median value of 13.3 (IQR 12.55) pg/mL and interleukin (IL)-6 in 7 patients (7/11,63.6%) with a median value of 10.1 (IQR 57.1) pg/mL. Ferritin was elevated in 5 patients (5/8,62.5) with a median value of 363.2 (IQR 708.50) ng/mL (shown in Table 2 ). Most patients tested negative for autoantibodies, except for 3 patients (3/13,23.1%), including 2 patients who exhibited positive antinuclear antibody (ANA) (1:80-1:160, and 1:80, respectively), and 1 had antiphospholipid antibodies. Abnormal immunoglobulin and complement were found in 4 patients ( Supplementary Table 2 ). It suggests that patients with YAOS have a robust inflammatory response at the onset of the disease.

12 heterozygous NOD2 variants were detected including Q902K in 4 patients, R471C in 3, and single of c.-14C>T, A110T, S127L, R311W, A432V, Y514H, R541P, A661P, K818Q, A886V ( Figure 2 ). Additionally, 2 patients carried compound heterozygous involving variants Q902K and R471C as well as R471C and c.-14C>T. Besides, other than NOD2, 4 patients carried MEFV variants, including 3 with the E148Q variant and one each with R202Q and L110P, which were all considered benign. All the gene variants were rare among the East Asian population, with an allele frequency of less than 0.02 based on a database of Genome Aggregation Database (gnomAD), Exome Aggregation Consortium (ExAC), dPSNP, and 1000 Genomes Project (China). Allele frequencies of NOD2 variants found in our cohort were significantly elevated in our patients in comparison with healthy individuals (23) and the general East Asian population, except for variants R311W and c.-14C>T. According to the American College of Medical Genetics and Genomics (ACMG) guideline (24), 6 variants were categorized as Variant of Uncertain Significance (VUS), three as likely benign or benign, with the most frequent variants Q902K and R471C classifying as likely benign and benign respectively. c.-14C>T, A110T, S127L, A661P, K818Q, A886V, R471C, and A432V were identified as novel variants in YAOS. Out of the 15 patients with YAOS, families of 11 underwent whole exome sequencing (WES) to detect NOD2 variants, and 9 were found to be inherited from their parents without de novo variants. Details of NOD2 variants in 15 patients with YAOS are listed in Table 3 .

In this study, among the patients who underwent timely follow-up, the median duration of follow-up was 6 (IQR 15) months, and the median interval between visits was 4 (IQR 3) months. As for the treatment, most patients received a combination therapy of glucocorticoids, disease-modifying anti-rheumatic drugs (DMARDs), and/or sulfasalazine. Only 3 patients were once treated with sulfasalazine alone. Overall, 9 patients with YAOS responded well to prednisone (9/10,90%) which showed a good response in fever, gastrointestinal symptoms, arthralgia, and normalization of inflammatory indicators. 5 patients responded well to sulfasalazine (5/9,55.6%) and 1 had a partial clinical response. Oral sulfasalazine contributes to tapering down the dose of glucocorticoids, symptom improvement (fever, gastrointestinal symptoms, rash, lower extremity swelling, and arthritis), extending the interval between flare-ups, and decreasing the severity of attacks. One patient reported significant recovery in her physical functionality and resumed her work. One patient reported adverse reactions including nausea and vomiting and one experienced thrombocytopenia. 3 patients discontinued sulfasalazine due to drug allergies or intolerance. Additionally, most patients were treated with colchicine (7/9,77.8%), and antibiotics (5/8,62.5%) or underwent tonsillectomy without clinical response.

For frequent disease flares, steroid dependence, or no response to multiple therapies, 6 patients were given tumor necrosis factor (TNF) inhibitors, with 4 (4/6, 66.7%) of these patients experiencing a resolution of symptoms (fever, myalgia, and abdominal pain) and a reduction in attack frequency. Canakinumab was effective in 1 patient (n=1/1,100%) for successful tailoring of glucocorticoids. Tocilizumab was effective in 1 patient (n=1/2,50%) for resolution of fever and myalgia. Biologic agents contributed to tapering down the dose of glucocorticoids, reducing attack frequency, improving symptoms, and normalizing inflammatory markers ( Table 1 ). 6 patients have a partial response to nonsteroidal anti-inflammatory drugs (NSAIDs) which mainly helped normalize the temperature and alleviate chills but did not alter the interval between disease attacks.

The average levels of acute-phase reactants, and scores of visual analog scale (VAS), physician global assessment (PGA), and 36-item short-form survey (SF-36) during follow-ups were collected to assess disease activity ( Figure 3 ), except for 2 patients who were lost to follow-up, 1 patient didn’t receive treatment after enrollment. Most patients experienced varying degrees of relapse during the follow-up, but after treatment, their SF-36, VAS, and PGA scores as well as WBC, hypersensitive C-reactive protein (hs-CRP), and ESR, showed improvement to some extent.

Among the 4 patients carrying the Q902K variant, all exhibited lower extremity swelling. This differed from patients carrying other variants (n = 11), where the proportion of lower extremity swelling was 27.3%. In the case of the R471C variant, which was identified in 3 patients, 2 of 3 (66.7%) suffered skin rashes involving limbs, compared to 25% in patients with other variants. Furthermore, MEFV was the most frequent co-existing gene. Patients with combined NOD2 and MEFV variants were identified in 4 patients who experienced a higher proportion of headaches (75% vs. 9.1%), chest pain (50% vs. 9.1%), myalgia (75% vs. 36.4%), and weight loss (50% vs. 36.4%). It suggested that MEFV may play a modifier role in disease phenotypes rather than acting as a causative gene, supported by no response to colchicine in these patients.

To date, the largest cohort of YAOS was reported in 2015 (3), which enrolled 54 patients who were all Non-Jewish White. This research found that IVS8⁺¹⁵⁸ was the most frequent mutation (85%, n = 46). Other common variants included R702W (n = 15,28%), and rare variants (12, 22%). In our cohort, The most common variant was Q902K (n = 4, 26.7%). The gender ratio of male to female in our Chinese patients was nearly balanced, compared to a female predominance in the Caucasian cohort (7:8 vs. 17:37). The average age at disease onset was comparable between our cohort and the Caucasian cohort (35 vs. 33.5 years old). Moreover, the mean disease duration at diagnosis was shorter in our cohort (5 vs. 10.7 years). In terms of symptoms, recurrent fever (100% vs. 61%, p<0.05) was more common in our cohort than the Caucasian cohort, as well as poly arthralgia/arthritis (100% vs. 68%), lower extremity swelling (46.7% vs. 32%), myalgia (46.7%% vs. 35%), lymphadenopathy (26.7% vs. 9%), sore throat (20% vs. 5%). Our cohort also reported occurrences of proteinuria/hematuria (3, 21.4%) which was not reported in the Caucasian cohort. One patient is concurrent with nephrotic syndrome and exhibited unexplained intermittent gross hematuria. One patient experienced microscopic hematuria and one patient reported proteinuria. However, frequencies of arthralgia/arthritis (46.7% vs. 87%, p<0.05), skin rashes (33.3% vs. 91%, p<0.05), headaches (26.7% vs. 56%, p<0.05), and sicca-like symptoms (13.3% vs. 56%, p<0.05) were significantly less prevalent in our cohort. The frequencies of gastrointestinal symptoms, oral ulcers, chest pain, pleuritis and/or pericarditis, periorbital edema, and weight loss were similar between the two cohorts. In laboratory data, leucocytosis (66.7% vs. 14%, p<0.05), elevated acute phase reactants (100% vs. 39%, p<0.05), and anemia (66.7% vs. 15%, p<0.05) were significantly more common in our cohort ( Table 4 ). Concerning treatment response, a greater proportion of patients in our cohort responded with show good response to TNF inhibitors (66.7% vs. 1 partial response), but less to sulfasalazine (55.6% vs. 100%).