The effect of granulocyte-colony stimulating factor on endothelial function in patients with myocardial infarction

Objective: The effects of granulocyte-colony stimulating factor (G-CSF) on endothelial function are unknown. Therefore, we investigated the effects of G-CSF on endothelial function. Methods: 76 patients participating in the MAGIC-Cell-3-DES trial were enrolled. These were patients with acute myocardial infarction (AMI) or old MI (OMI) who underwent percutaneous coronary intervention (PCI), and were prospectively randomised into a G-CSF group (G-CSF (10 mg/kg/day) injection for 3 days after PCI) or a control group. Additionally, 20 healthy volunteers were also enrolled. These subjects were categorised into five groups: AMI-control (n = 18), AMI-G-CSF (18), OMI-control (20), OMI-G-CSF (20) and healthy-G-CSF (20). Baseline flow-mediated dilation (FMD) of the brachial artery and serum inflammatory biomarkers were performed on day 1, and repeated on day 4 in all groups. G-CSF was injected for 3 days between days 1 and 4 in the AMI-G-CSF, OMI-G-CSF and healthy-G-CSF groups. Results: In both the healthy-G-CSF and OMI-G-CSF groups, G-CSF increased serum high sensitivity C-reactive protein (hsCRP) (0.3 (0.5) mg/l vs 6.1 (3.5) mg/l and 5.6 (3.8) mg/l vs 13.0 (7.7) mg/l, baseline vs post-G-CSF in the healthy and OMI-G-CSF groups, respectively, p<0.001). In the AMI-G-CSF group, G-CSF hindered the decline of hsCRP during the recovery phase, resulting in a relative increase in hsCRP. However, in all three groups, G-CSF did not significantly alter FMD. Conclusion: Despite an associated increase in systemic inflammation, G-CSF treatment does not lead to acute impairment of brachial artery endothelial function in either healthy subjects or patients with MI.