Background: Granulocyte-colony stimulating factor (G-CSF) shows promise as a treatment for stroke. This systematic review assesses G-CSF in experimental ischaemic stroke.
Methods: Relevant studies were identified with searches of Medline, Embase and PubMed. Data were extracted on stroke lesion size, neurological outcome and quality, and analysed using Cochrane Review Manager using random effects models; results are expressed as standardised mean difference (SMD) and odds ratio (OR).
Results: Data were included from 19 publications incorporating 666 animals. G-CSF reduced lesion size significantly in transient (SMD -1.63, p<0.00001) but not permanent (SMD -1.56, p=0.11) focal models of ischaemia. Lesion size was reduced at all doses and with treatment commenced within 4 h of transient ischaemia. Neurological deficit (SMD -1.37, p=0.0004) and limb placement (SMD -1.88, p=0.003) improved with G-CSF; however, locomotor activity (> or =4 weeks post-ischaemia) was not (SMD 0.76, p=0.35). Death (OR 0.27, p<0.0001) was reduced with G-CSF. Median study quality was 4 (range 0-7/8); Egger's test suggested significant publication bias (p<0.001).
Conclusions: G-CSF significantly reduced lesion size in transient but not permanent models of ischaemic stroke. Motor impairment and death were also reduced. Further studies assessing dose response, administration time, length of ischaemia and long-term functional recovery are needed.