Elevated platelet-to-lymphocyte ratio is associated with stroke-associated pneumonia in acute ischemic stroke: a...

Stroke-associated pneumonia (SAP) is a common and serious complication of acute ischemic stroke (AIS) that substantially worsens clinical outcomes. The platelet-to-lymphocyte ratio (PLR), an accessible composite index derived from routine complete blood counts, may reflect post-stroke immune imbalance. However, its role in predicting SAP in AIS patients remains unexplored.

This retrospective cohort study included 1,745 patients with acute ischemic stroke (AIS) after systematic screening of 2,150 patients. After the original screening, 1,766 eligible patients remained; 21 additional patients with extreme PLR values (>1,000) were excluded. HbA1c missing values were imputed once using multiple imputation by chained equations (MICE) with a random forest estimator. PLR was calculated as platelet count divided by lymphocyte count and log2-transformed. Multivariable logistic regression, generalized additive modeling (GAM), and two-piecewise logistic regression were used to examine associations between PLR and stroke-associated pneumonia (SAP). Subgroup analyses were performed across seven clinical variables with interaction tested by likelihood ratio test. Predictive performance was evaluated by receiver operating characteristic (ROC) analysis and compared with C-reactive protein (CRP) and A2DS2 score using the DeLong test.

SAP occurred in 359 patients (20.6%). PLR was higher in patients with SAP than in those without SAP (median 136.17 vs. 113.69, p < 0.001). In the fully adjusted model, log2-PLR remained associated with SAP (OR = 1.36 per 1-unit increase, 95% CI: 1.12-1.65, p = 0.0021). Compared with the lowest quartile, the highest PLR quartile was associated with increased SAP odds (OR = 1.62, 95% CI: 1.03-2.54, p = 0.0386). GAM and two-piecewise regression identified a threshold at PLR = 224.8 (log2-PLR = 7.812). Below this threshold, the association was not significant (OR = 1.21, 95% CI: 0.93-1.57, p = 0.1551), whereas above it the association was significant (OR = 2.06, 95% CI: 1.10-3.87, p = 0.0240). The AUCs were 0.584 for PLR, 0.825 for CRP, and 0.757 for A2DS2. The AUCs of the PLR + CRP and PLR + A2DS2 models were 0.803 and 0.763, respectively.

Elevated PLR is associated with SAP in AIS patients, exhibiting a non-linear threshold-dependent relationship. Although standalone discriminatory performance is modest, PLR is universally available from routine blood counts and may serve as a cost-effective tool for early SAP risk stratification. Prospective multicenter validation is warranted.