Arterial blood gas analysis of samples directly obtained beyond cerebral arterial occlusion during endovascular procedures predicts clinical outcome

Abstract

Real-time intra-procedure information about ischemic brain damage degree may help physicians in taking decisions about pursuing or not recanalization efforts.

Methods: We studied gasometric parameters of blood samples drawn through microcatheter in 16 stroke patients who received endovascular reperfusion procedures. After crossing the clot with microcatheter, blood sample was obtained from the middle cerebral artery (MCA) segment distal to occlusion (PostOcc); another sample was obtained from carotid artery (PreOcc). An arterial blood gas (ABG) study was immediately performed. We defined clinical improvement as National Institutes of Health Stroke Scale (NIHSS) decrease of ≥4.

Results: The ABG analysis showed differences between PreOcc and PostOcc blood samples in mean oxygen partial pressure (Pre-PaO2: 78.9 ± 16 .3 vs. 73.9 ± 14 .9 mmHg; P < .001). Patients who presented clinical improvement had higher Post-PaO2 (81 ± 11 .4 vs. 64.8 ± 14 .4 mmHg; P = .025). A receiver-operator characteristic (ROC) curve determined Post-PaO2 > 70 mmHg that better predicted further clinical improvement. Patients with Post-PaO2 > 70 mmHg had higher chances of clinical improvement (81.8% vs. 0%; P = .002) and lower disability (median mRS:3 vs. 6; P= .024). In the logistic regression the only independent predictor of clinical improvement was Post-PaO2 > 70 (OR: 5.21 95% CI: 1.38-67.24; P = .013).

Conclusion: Direct local blood sampling from ischemic brain is feasible during endovascular procedures in acute stroke patients. A gradient in oxygenation parameters was demonstrated between pre- and post-occlusion blood samples. ABG information may be used to predict clinical outcome and help in decision making in the angio-suite.