Biomarker Patterns Track 90-Day Recovery After Thrombolysis in Acute Ischemic Stroke
In patients with acute ischemic stroke, a simple panel drawn from routine blood work may help identify who is likely to do poorly after intravenous thrombolysis. Serum D-dimer, the neutrophil-to-lymphocyte ratio (NLR), and the C-reactive protein to albumin ratio (CRP/ALB) each separated the poorer-outcome group from the better-recovery group in this 90-patient cohort, and the combination performed better than any single marker alone.
The clinical logic is straightforward. D-dimer reflects thrombin generation and fibrin turnover, so a higher value can signal a more active prothrombotic state. NLR captures the balance between innate immune activation and relative lymphocyte suppression, a pattern often seen when systemic inflammation is intense enough to worsen vascular injury. CRP/ALB mixes two biologically meaningful signals in one ratio: rising CRP points to inflammatory burden, while falling albumin often parallels illness severity, catabolism, and poor physiologic reserve.
90 Patients, 2 Outcome Groups, and a 90-Day Functional Endpoint
Ninety patients admitted between June 2021 and October 2023 were followed after thrombolysis and divided by 90-day modified Rankin Scale outcome into a poor prognosis group of 29 cases and a good prognosis group of 61 cases. The study design used binary logistic regression to identify prognostic factors and receiver operating characteristic analysis to test how well each marker distinguished the two recovery trajectories.
Age emerged as a risk factor, along with D-dimer, albumin, CRP/ALB, neutrophil count, lymphocyte count, and NLR. The signal is clinically plausible. Older patients often arrive with more comorbidity, lower physiologic reserve, and less tolerance for ischemic injury, while a pro-inflammatory and prothrombotic blood profile can accompany larger infarcts, more endothelial dysfunction, and a less favorable response to reperfusion.
A single sentence captures the core of the dataset: the poorest outcomes clustered where coagulation activation and systemic inflammation overlapped.
Discrimination Improved When the Three Markers Were Combined
The performance of the individual markers was moderate, but the combined model carried more useful separation. The area under the curve was 0.683 for D-dimer, 0.769 for NLR, and 0.728 for CRP/ALB, while the combined AUC reached 0.803. In practical terms, that is the kind of incremental improvement clinicians look for when several incomplete biologic signals are fused into one bedside estimate.
Sensitivity was highest for NLR at 100.0%, with specificity of 68.9%. D-dimer also showed very high sensitivity at 96.6%, but specificity was only 39.3%, which limits its value as a stand-alone discriminator. CRP/ALB reached 100.0% sensitivity as well, with specificity of 45.9%. The pattern suggests these markers are better suited to screening and risk enrichment than to making a definitive prognosis on their own.
The combined AUC of 0.803 matters because post-thrombolysis care often depends on early triage decisions. Patients at higher risk for poor functional recovery may need closer neurologic monitoring, aggressive management of complications, and more deliberate rehabilitation planning. A biomarker bundle that integrates coagulation activity, inflammatory burden, and nutritional or systemic reserve could therefore serve as a practical adjunct to imaging and neurologic examination.
Why Inflammation and Coagulation Move Together in Stroke
The correlations among the three markers were all positive: D-dimer with NLR, D-dimer with CRP/ALB, and NLR with CRP/ALB. Spearman coefficients were 0.367, 0.482, and 0.462, respectively, each with statistical significance. Those relationships fit the biology of ischemic stroke, where clot formation, endothelial injury, and immune activation do not occur in isolation.
Inflammation can amplify thrombosis by promoting platelet activation, endothelial dysfunction, and leukocyte recruitment. Coagulation activation can, in turn, intensify inflammatory signaling and microvascular obstruction. Low albumin may mark a more fragile systemic state, reducing the body’s ability to buffer acute stress and recover after the ischemic event. In that framework, these biomarkers are not redundant; they reflect overlapping but distinct pieces of the same pathophysiologic cascade.
What This Adds to Thrombolysis Risk Stratification
The study supports using D-dimer, NLR, and CRP/ALB as prognostic indicators after intravenous thrombolysis in acute ischemic stroke. The results do not replace the modified Rankin Scale or clinical examination, and they do not provide a treatment target. They do, however, suggest that readily available laboratory values can sharpen early prognosis when interpreted together rather than in isolation.
Two practical limits are built into the data itself. First, the cohort was small, with only 90 patients, which restricts how far the findings can be generalized. Second, this is a single prognostic snapshot tied to 90-day outcomes; the markers were assessed as predictors, not as tools proven to change management or improve recovery when acted upon. Even with those caveats, the direction of effect is coherent: higher inflammatory and coagulation burden lined up with worse functional outcome after reperfusion therapy.
For stroke teams, the immediate clinical use would be selective rather than universal. A patient arriving for thrombolysis with an unfavorable profile across these markers may deserve closer follow-up than one whose laboratory pattern is less inflammatory and less prothrombotic. The numbers are not large enough to justify overstatement, but they are consistent enough to make the biomarker trio worth considering in future validation cohorts and in broader prognostic models that already include age, neurologic severity, and imaging findings.