Abstract

Serum VEGF level is regarded to be a biomarker for the diagnosis of stroke. Even though there have been published plethora of original articles describing higher blood VEGF concentrations since the 1970s, however, there is no any meta-analysis report for serum VEGF levels in the field of evidence-based medicine yet. A systematic review was performed by searching the online biomedical databases including retrieving 14 case–control studies including within-article subgroups after fulfilling the inclusion and exclusion criteria without the beginning date restriction, until 2020 for ischemic stroke patients. The Q quantity and I2 statistic index showed a high heterogeneity (84.895 and 84.687, respectively) and the random-effects model of meta-analysis was applied for further analyses. The meta-analysis on a total number of 769 stroke subjects and 621 controls found that the weighted pooled SMD for overall serum VEGF levels on different days of testing was 1.92 (95 CI, − 4.059–0.219, p value = 0.079) and the pooled SMD for overall serum VEGF levels on day 1 of testing was − 1.083 (95 CI, − 4.229–2.063, p value = 0.500). The meta-regression results demonstrated that different days of testing do not significantly affect serum VEGF concentrations in ischemic patients and actually their serum levels are time-independent. Based on the recently published studies, this meta-analysis showed that serum VEGF levels were not significantly associated with an ischemic stroke diagnosis. Thus, researchers may concern another ideal serum or cerebrospinal fluid-derived biomarker for stroke diagnosis. © 2020, Fondazione Società Italiana di Neurologia.