Predictive Value of Systemic Inflammatory Indices in Drug Hypersensitivity Severity

Drug hypersensitivity reactions (DHRs) range from mild cutaneous symptoms to life-threatening anaphylaxis, necessitating practical biomarkers for severity prediction and risk stratification. This study aimed to evaluate the predictive value of Systemic Immune-Inflammation Index (SII), Systemic Inflammation Response Index (SIRI), and Aggregate Index of Systemic Inflammation (AISI) measured during a stable period in patients with drug allergy.

This study included 200 patients with DHRs, categorized into severe (anaphylaxis; n = 67) and non-severe (n = 133) groups. To reflect steady-state inflammatory status, hemogram parameters were obtained at least 1 month after resolution of acute symptoms. Diagnostic performances were assessed using age- and sex-adjusted logistic regression and ROC analysis.

Baseline SII, SIRI, and AISI were all significantly higher in the severe group (SII: 516.8 vs. 438.1, p = 0.009; SIRI: 1.08 vs. 0.88, p = 0.012; AISI: 286.6 vs. 221.9, p = 0.001). In age- and sex-adjusted models, each index independently predicted anaphylaxis per 1-SD increase: SII OR 1.55 (95% CI 1.13-2.20), AISI OR 1.50 (95% CI 1.12-2.06), and SIRI OR 1.45 (95% CI 1.08-1.97). Using Youden-optimized thresholds, AISI ≥197 yielded the highest adjusted odds of anaphylaxis (aOR 4.11, 95% CI 2.00-9.07; p < 0.001), followed by SII ≥333 (aOR 3.56, 95% CI 1.55-9.28; p = 0.005) and SIRI ≥1.37 (aOR 3.11, 95% CI 1.59-6.15; p < 0.001).

Systemic inflammatory indices are independently associated with reaction severity in patients with immediate-type drug hypersensitivity. With AUC values ranging from 0.61 to 0.64, these indices demonstrate modest discriminatory performance and should be regarded as adjunctive screening tools rather than standalone diagnostic markers. Their elevation during the stable period may reflect a chronic pro-inflammatory state that predisposes individuals to more severe reactions upon drug exposure, a hypothesis requiring prospective validation in larger, multicenter cohorts.