Clinical characteristics and peripheral immune cell profile of children with Stevens-Johnson syndrome complicated by sepsis: an exploratory retrospective cohort study

SJS is a rare but life-threatening mucocutaneous hypersensitivity disorder, and sepsis is one of its most severe infectious complications. However, the clinical and immunological features of pediatric SJS complicated by sepsis remain unclear. This single-center retrospective cohort study included 47 children with SJS admitted between 6 November 2015 and 6 November 2025. Patients were reclassified into a sepsis group (n = 11) and a non-sepsis group (n = 36) according to the 2018 Chinese guideline definition of sepsis (infection plus life-threatening organ dysfunction), applying pediatric-corrected organ-dysfunction thresholds. Demographic characteristics, suspected pathogen distribution, comorbidities, EBV serology results, and 56 admission laboratory parameters (including T-lymphocyte subsets) were compared. Continuous variables were compared using the Mann-Whitney U test, and categorical variables using Fisher’s exact test or the chi-square test. Variables with P < 0.05 in univariate analysis were assessed for collinearity and multivariable logistic regression was attempted. Because the 11 sepsis events led to quasi-complete separation, the model was not estimable; therefore, no odds ratios are reported and associated factors are presented by non-parametric comparison. Of 47 children with SJS, 11 (23.4%) met the redefined sepsis criteria. No significant between-group differences were observed in sex (P = 0.722), age (P = 0.372), or suspected pathogen distribution. The sepsis group had a longer hospital stay (17.00 [12.50–22.50] vs. 9.00 [8.00–12.00] days, P = 0.002). Five of 56 laboratory parameters differed significantly: basophil percentage, monocyte count and CD19 + B-cell absolute count were lower, AST was higher, and the CD4+/CD8 + ratio was lower (all P < 0.05). C-reactive protein, fibrinogen and the previously reported T-lymphocyte subset alterations showed no significant between-group differences. Owing to quasi-complete separation in the small sepsis group, multivariable logistic regression was not estimable, and associated factors are therefore not reported as odds ratios. Under the 2018 guideline redefinition, children with SJS complicated by sepsis show a longer hospital stay and a peripheral immune-inflammatory profile characterized by decreased basophils, monocytes and CD19 + B-cell absolute count, a reduced CD4+/CD8 + ratio and elevated AST, whereas sex, age, pathogen distribution and traditional inflammatory markers did not differ significantly. These findings are preliminary and hypothesis-generating. Constrained by the retrospective single-center design, the small sepsis subgroup, the absence of adjustment for multiple comparisons and the lack of functional immune markers, they cannot serve as a clinical risk-stratification tool at present and require validation in larger prospective multicenter studies. All results should be interpreted as exploratory because 56 laboratory parameters were compared without multiplicity correction.