A. Hellberg et al., Monitoring of intrathecal oxygen tension during experimental aortic occlusion predicts ultrastructural changes in the spinal cord, J THOR SURG, 121(2), 2001, pp. 316-323
Citations number
30
Categorie Soggetti
Cardiovascular & Respiratory Systems","Cardiovascular & Hematology Research
Objective: To study the correlation between intrathecal PO2 and ultrastruct
ural changes in the spinal cord during thoracic aortic occlusion in pigs.
Material and methods: In 18 pigs, online intrathecal oxygenation was monito
red by a multiparameter Paratrend catheter (Biomedical Sensors, High Wycomb
e, United Kingdom) during 60 minutes' clamping of the proximal and distal d
escending thoracic aorta. The animals were randomly divided into 2 groups (
A and B) depending on the level of distal aortic clamping. Distal aortic pe
rfusion was restored through an aorto-iliac shunt, which also maintained lo
w thoracic segmental perfusion of the spinal cord in group B. Perfusion-fix
ation technique was used before harvesting the spinal cord specimens, which
later were evaluated with light and electron microscopy by an independent
observer. Intrathecal parameters were interpreted as normal if PO2 was more
than 0.8 kPa and PCO2 was less than 12 kPa, as intermediate ischemia if PO
2 was 0.8 or less or PCO2 was more than 12 kPa, and as absolute ischemia if
PO2 was 0.8 or less and PCO2 was more than 12 kPa.
Results: Among 6 animals with ultrastructural changes of absolute spinal co
rd ischemia-reperfusion injury, 5 also had absolute ischemia according to v
ariables derived by the Paratrend catheter, The 2 methods were in agreement
in 3 of 5 animals with intermediate ischemia-reperfusion changes and in 5
of 6 animals with normal findings. The accuracy of cerebrospinal fluid PO2
and PCO2 to predict electron microscopy-verified intermediate or absolute i
schemia-reperfusion injury was 94%.
Conclusions: Monitoring of intrathecal PO2 after clamping of the descending
aorta correlated with ultrastructural changes in the spinal cord in this p
ig model.