CLUSTER ANALYSIS OF AUTOANTIBODIES IN PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS
Bibliographic record
Abstract
PV195 / #239 Poster Topic: AS22 - SLE Heterogeneity Background/Purpose Systemic lupus erythematosus (SLE) is an autoimmune disease with diversity of autoantibodies and clinical manifestation. The identification of patient clusters by autoantibody profile can predict prognosis and mortality. The aim of this study is to define and describe serological clusters and their clinical and epidemiological characteristics, as well as their association with comorbidities, disease activity, severity and damage. Methods Descriptive, observational and multicenter study that includes patients with SLE from the Spanish national registry RELESSER. 1740 patients were included in the cross-sectional study and 718 in the prospective study (with annual follow-up during 5 years). The autoantibodies selected for cluster analysis were anti-DNA, anti-Sm, anti-RNP, anti-Ro, anti-La and antiphospholipid antibodies. Cluster analysis was carried out using Gower distance. To compare the distributions of categorical variables, Chi-square tests were used or Fisher’s exact test in cases of low expected frequencies. For continuous variables, nonparametric tests such as Kruskal-Wallis or ANOVA were applied, depending on the data distribution and homogeneity of variances. Results Four serological clusters were defined (Table 1). Cluster 1 (absence of anti-extractable nuclear antigen antibodies; 44.54% of the patients) was characterized by lower frequency of vasculitis (6.6%), leukopenia (49.1%) and lymphopenia (48.0%) (Table 2). Cluster 2 (antiphospholipid antibodies; 12.24%) was represented by higher frequency of high blood pressure (35.5%), hemolytic anemia (13.5%), thrombocytopenia (39.9%), vasculitis (12.5%), visual disturbances (9.1%) and higher use of immunoglobulins (10.3%) and oral anticoagulants (39.0%). Cluster 3 (anti-Ro and anti-La; 26.67%) had the lowest frequency of lupus nephritis (24.9%). Patients from cluster 4 (anti-Smith and anti-ribonuclear proteins; 16.55%) were younger at disease onset (median age of 29.2 years) and had the highest frequency of lupus nephritis (38.5%), leukopenia (66.0%), lymphopenia (62.2%), hypocomplementemia (89.5%), myositis (5.6%) and cutaneous manifestations. Besides, they had a higher frequency of osteoporosis (11.0%) and severe infections (26.4%) and a higher use of glucocorticoids (93.5%), azathioprine (42.0%), cyclophosphamide (25.5%) and mycophenolate mofetil (24.0%) (Table 3). Regarding disease activity assessed by SLEDAI (Systemic Lupus Erythematosus Disease Activity Index), at visit 1 of the longitudinal study, patients in cluster 4 had the highest scores: 2.5 ± 3.37 (1.70 ± 3.09 in cluster 1 (p=0.18), 2.18 ± 4.26 in cluster 2 (p=0.4), 1.86 ± 2.50 in cluster 3 (p=0.011). After 5 years of follow-up, no differences were observed between clusters. Concerning damage assessed by SLICC/ACR DI, patients in cluster 2 exhibited the highest scores at visit 1: 1.93 ± 2.30 (1.42 ± 1.81 in cluster 1 (p=0.018), 1.13 ± 1.72 in cluster 3 (p<0.001), 1.67 ± 1.92 in cluster 4 (p=0.4)). After 5 years of follow-up, a significant increase was observed across all clusters (p<0.001), with differences persisting at the end of follow-up (p=0.049). Regarding severity measured by Katz, patients in cluster 4 had the highest scores at visit 1: 5.24 ± 2.09 (4.45 ± 2.13 in cluster 1 (p<0.001), 4.57 ± 2.20 in cluster 2 (p=0.019), 4.35 ± 1.69 cluster 3 (p<0.001)). The differences persist between clusters after follow-up (p=0.005). As for mortality, 21 deaths were recorded: 5 in cluster 1 (1.74%), 6 in cluster 2 (5.50%), 6 in cluster 3 (2.97%) and 4 in cluster 4 (3.36%), with no significant differences between clusters (p=0.427). Table 1. Epidemiological characteristics and comorbidities Table 2. Clinical characteristics Table 3. Treatments Conclusions In our cohort, the serological profile constitutes a crucial factor for the clinical stratification of patients and predicting their prognosis. Nevertheless, further studies are required to facilitate a more precise identification and comprehensive understanding of these patients.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".