Blood samples were collected to explore potential serum biomarkers associated with neurocognitive function in small-cell lung cancer (SCLC) patients who received prophylactic cranial irradiation (PCI).This pre-specified study included patients with blood samples available, who participated in a phase III trial (NCT01780675). Blood samples were collected before PCI and 3-days post-initiating PCI. Neurocognitive decline was defined as a decrease of ≥ 5 points on total recall in the Hopkins Verbal Learning Test-Revised (HVLT-R) assessed from pre-PCI to 4-months post-PCI. Biomarkers were screened using univariate logistic regression analysis. P < .1 was considered statistically significant.Forty-eight enrolled patients who had blood samples at baseline were included and 27 were available for analysis as the other 21 did not assess neurocognitive function at 4-months. Lower levels of Tie-2 (OR = 0.999, 90% CI 0.998-1.000, P = .062), and higher levels of MIP-1b (OR = 1.022, 90% CI 1.000-1.044, P = .093), CCL-17 (OR = 1.004, 90% CI 1.001-1.006, P = .029), and IL-1α (OR = 1.597, 90% CI 1.077-2.367, P = .05) before PCI were correlated with neurocognitive decline at 4-months. Decrease of VEGF-C (OR = 0.972, 90% CI 0.949-0.996, P = .055), CCL-17 (OR = 0.993, 90% CI 0.988-0.999, P = .036), IL-1α (OR = 0.788, 90% CI 0.635-0.979, P = .071), and VEGF (OR = 0.981, 90% CI 0.965-0.997, P = .051) 3-days post-initiating PCI were also associated with neurocognitive decline at 4-months.Biomarker levels before PCI and changes in their levels 3-days post-initiating PCI may be linked to subsequent neurocognitive decline at 4-months. If validated, these biomarkers could be used to predict the risk of neurocognitive decline and act as a decision aid for personalized PCI in SCLC.Copyright © 2024 Elsevier Inc. All rights reserved.