Cardiotomy suction: a major source of brain lipid emboli during cardiopulmonary bypass.

BACKGROUND: Brain injury remains a significant problem in patients undergoing cardiac surgery assisted by cardiopulmonary bypass (CPB). Autopsy brain specimens of patients after cardiac operations with CPB reveal numerous acellular lipid deposits (10 to 70 microm) in the microvasculature. We hypothesize that these small capillary and arterial dilatations result from a diffuse inflammatory response to CPB or from emboli delivered by the bypass circuit. This study was undertaken to determine which aspect of CPB is most clearly associated with these dilatations. METHODS: Thirteen dogs were studied in four groups: group I (n = 3), right-heart CPB; group II (n = 2), lower-extremity CPB; group III (n = 3), hypothermic CPB; and group IV (n = 5), hypothermic CPB with cardiotomy suction. All dogs in all groups were maintained on CPB for 60 minutes and then euthanized. Brain specimens were harvested, fixed in ethanol, embedded in celloidin, and stained with the alkaline phosphate histochemical technique so that dilatations could be counted. RESULTS: All dogs completed the protocol. The mean density of dilatations per square centimeter for each group was as follows: group I, 1.77 +/- 0.77; group II, 4.17 +/- 1.65; group III, 4.54 +/- 1.69; and group IV, 46.5 +/- 14.5. In group IV (cardiotomy suction), dilatation density was significantly higher than in group III (hypothermic cardiopulmonary bypass) (p = 0.04) and all other groups (p = 0.04). CONCLUSIONS: Blood aspirated from the surgical field and subsequently reinfused into dogs undergoing CPB produces a greater density of small capillary and arterial dilatations than CPB without cardiotomy suction, presumably because of lipid microembolization.

Cerebral microvascular alterations in aging, leukoaraiosis, and Alzheimer's disease.

We have been using alkaline phosphatase (AP) histochemical staining, formerly a research tool for the study of cerebral cortical vascular morphology, to examine pathological changes in the cortex and deep cerebral structures. Deep structures stain similarly to the cortex. The AP stain is found in the afferent vessels (small arteries, arterioles, and capillaries), but not in venules and veins. The stain is also present in leaky vessels, such as those in the area postrema. The vascular supply to the cerebrum is not homogeneous. Supply to the deep white matter, for instance, derives from the leptomeningeal border zone, and then medullary arterioles must wind their way for up to 4 cm before arriving at their ultimate destination. Adding to the difficulties, tortuosities develop in some of these vessels with aging. According to some calculations, hypertensive levels of blood pressure would be required to maintain irrigation through some of these vessels. We have identified a venous alteration that attends aging: periventricular venous collagenosis (PVC) is a previously unrecognized, noninflammatory, mural disease of the periventricular veins. In severe cases, examples can be found of veins that are completely occluded by this process. PVC is found in 65% of subjects over 60 years old, and it strongly correlates with leukoaraiosis. In addition to previously mentioned aging-related changes, we have found extreme tortuosity, multiplications, and aneurysms of the smallest arterioles and lumpy-bumpy capillaries in the deep structures of patients with Alzheimer's disease.

Human germinal matrix: venous origin of hemorrhage and vascular characteristics.

PURPOSE: To examine the vascular supply and architecture of the germinal matrix in the preterm neonatal brain and to determine whether veins or arterioles are the source of germinal matrix hemorrhage. METHODS: Brains from eight preterm neonates (24 to 35 weeks' gestation) and two full-term infants were fixed in alcohol, embedded in celloidin, sectioned at 100- and 500-micron thicknesses, stained for alkaline phosphatase, and examined with light microscopy. High-resolution contact radiographs of 500-micron-thick sections were also mounted on glass slides for microscopic examination. RESULTS: The upper and middle regions of the germinal matrix are supplied by branches of the lateral striate arteries, whereas the inferior part is supplied by branches of the recurrent artery of Heubner. In brain sections from four of the preterm infants, we found 15 circumscribed hemorrhagic foci within the germinal matrix. The largest was 5 mm in diameter; the smallest, 1 mm. All hemorrhages but one were closely associated with veins, with significant involvement of the perivenous space. The other hemorrhage appeared to be associated with an arteriole. In term and preterm infants, we found no arteriolar-to-arteriolar shunts, precapillary arteriolar-to-venules shunts, or vascular rete. At all gestational ages, the terminal vascular bed had only conventional branchings and connections. CONCLUSION: In preterm neonates, staining for endogenous alkaline phosphatase allows visual differentiation between afferent and efferent vessels. Germinal matrix hemorrhage in preterm neonates is primarily venous in origin. A hemorrhage can tunnel along the venous perivascular space, collapsing the vein and rupturing the tethered connecting tributaries. Extravasation of blood from the arterial circulation appears to be much less common.