A Clinical Cut-Off Value for the Systemic Immune-Inflammation Index to Predict Frequent Exacerbations in Stable COPD

Objective: Acute exacerbations (AECOPD) are primary determinants of clinical instability in chronic obstructive pulmonary disease (COPD), and the "frequent exacerbator" (≥2/year) phenotype markedly increases morbidity and healthcare utilization. In this study, we evaluated the association between the Systemic Immune-Inflammation Index (SII), calculated from routine hemogram parameters during the stable period, and the occurrence of frequent exacerbations within the subsequent 1 year, and aimed to define a clinically applicable SII threshold (cut-off). Materials and Methods: In this retrospective observational cohort study conducted at a tertiary care center, patients who attended the outpatient clinic between January 2020 and February 2025 and had COPD confirmed by post-bronchodilator spirometric criteria (FEV₁/FVC < 70%) were identified through electronic medical records. The index date was defined as a routine outpatient visit during stable COPD; patients were followed for AECOPD for 365 days after the index date. The stable period was defined as a visit occurring ≥4 weeks after the last exacerbation and without signs of acute infection. Patients with positive COVID-19 PCR results were excluded due to the uncertainty in distinguishing exacerbation from COVID-19. The primary endpoint was the development of frequent exacerbations (≥2 AECOPD) within 365 days. AECOPD was defined as an acute worsening of dyspnea, cough, and/or sputum requiring additional pharmacotherapy (systemic corticosteroids and/or antibiotics). SII, NLR, PLR, LMR, and PPN were calculated using hemogram parameters. Groups (<2 vs. ≥2 exacerbations) were compared; a ROC-Youden analysis was performed to determine cut-offs. After ROC-based dichotomization, univariate and multivariable logistic regression analyses were used to evaluate associations; multicollinearity was assessed using the VIF. To address potential optimism bias, diagnostic performance metrics (AUC, sensitivity, specificity) were internally validated using 1000 stratified bootstrap replicates. Results: A total of 159 patients were included. The cohort was predominantly male (91.2%). Demographic characteristics and most spirometric parameters were similar between groups; a trend toward lower absolute FVC was observed in the ≥2 exacerbation group (p = 0.051). Platelet counts were higher in the ≥2 exacerbation group (p = 0.029). In the ROC analysis, AUC values ranged from 0.505 to 0.591 across indices. For the SII, the AUC was 0.591 (95% CI: 0.500-0.677; p = 0.049), and the optimal cut-off was 1082.79. The LMR cut-off was 1.76; however, the LMR did not demonstrate statistically significant discriminatory performance in the ROC analysis (AUC 0.535; p = 0.448). In univariate analyses, SII > 1082.79 (OR = 3.028, 95% CI: 1.522-6.027; p = 0.002) was associated with frequent exacerbations. In a multivariable logistic regression adjusted for cardiovascular disease and overall comorbidity status, SII > 1082.79 remained independently associated (OR = 3.029, 95% CI: 1.485-6.179; p = 0.002). Other hemogram-derived indices did not retain independent prognostic significance in this outpatient cohort. Conclusion: SII measured during stable COPD was independently associated with frequent exacerbations over the subsequent 1 year. The SII > 1082.79 threshold may offer a practical risk stratification approach to flag "high-risk" patients in outpatient care. However, given the modest discriminative performance and the single-cohort derivation, this cut-off should be considered exploratory despite the use of bootstrap internal validation. Because this was a single-center study with a predominantly male cohort, the generalizability-particularly to female patients and other settings-requires prospective external validation.