James H. Halsey

Risks and benefits of shunting in carotid endarterectomy. The International Transcranial Doppler Collaborators.

BACKGROUND AND PURPOSE Controversy continues about the pathogenesis of perioperative stroke in carotid endarterectomy and the use of shunting. The purpose of this study was to determine, using transcranial Doppler ultrasonography, the severity of ischemia during clamping of the carotid artery as a basis for analysis of complications in patients operated on with and without shunting. METHODS In a retrospective study, 11 centers contributed 1,495 carotid endarterectomies monitored with transcranial Doppler. The cases were divided into groups with severe, mild, and no ischemia, and each group was subdivided according to shunt use. The perioperative rate of severe stroke attributable to intraoperative ischemia, in addition to total perioperative stroke, was determined for each subgroup. RESULTS Severe ischemia occurred in 7.2% of our cases but cleared spontaneously in about half of these. In those with persisting ischemia the rate of severe stroke was very high, while shunting protected against stroke in such cases. If ischemia did not occur, the stroke rate was higher with shunting, although not so high as in unshunted cases with severe ischemia. Slightly more than one third of the severe strokes were due to postoperative cerebral hemorrhage or carotid thrombosis, unrelated to clamp-induced ischemia or shunting. CONCLUSIONS Carotid endarterectomy complications might be reduced by selectively shunting only for severe persisting ischemia. Monitoring of cerebral ischemia would be essential to selective shunting.

Transcranial Doppler and rCBF compared in carotid endarterectomy.

In eight patients undergoing carotid endarterectomy, the mean velocity and an index of pulse amplitude in the middle cerebral artery were monitored continuously by transcranial doppler ultrasound. rCBF was measured by intracarotid injection of 133 Xenon shortly before and at the time of the carotid artery occlusion, and again a few minutes after carotid flow was reestablished. Comparison of the mean velocity in the MCA and the cortical convexity rCBF revealed relatively little hysteresis in their relationship from prior to after the occlusion. There was however, considerable variability in this relationship among patients. Both the rCBF and the velocity decreased substantially at occlusion in three cases, neither changed very much in three. While in two, though the rCBF decreased significantly, the velocity did not change. The index of pulse amplitude was somewhat more sensitive to the occlusion, decreasing in the seven cases in which it was recorded, including one in which the rCBF did not change.

Pressure distribution in the pial arterial system of rats based on morphometric data and mathematical models.

The objective of the present work was a theoretical evaluation of pial arterial pressures in normotensive rats and spontaneously hypertensive rats based on the geometry and topography of the pial arterial system as well as on various topological models of the vascular trees. Pial branches of the middle cerebral artery in the diameter range of 30–320 μm were selectively visualized by corrosion compound, and the diameter and length of vascular segments were measured. The vessels were classified into branching orders by the methods of Horsfield and Strahler. The steady-state pressure distribution in the pial arterial system was calculated assuming that the flow at the bifurcations was partitioned in proportion to a given power of the diameters of the daughter branches (diameter exponent). The maximum number of vascular segments along the longest branch varied between 16 and 33. The mean branching ratio was 4.14 ± 0.23 (SD). The mean diameter of vessels classified into Strahler orders 1–5 were: 50 ± 12, 71 ± 19, 106 ± 22, 168 ± 22, and 191 ± 7 μm, respectively. The calculated pressure drop in the pial trees of normotensive rats was approximately twice as large in proximal orders 3 and 4 than in distal orders 1 and 2. The mean pressure in arteries of order 1 ranged from 54.4 to 58.4 mm Hg in the normotensive rat (input pressure: 83 mm Hg), and from 77.2 to 89.0 mm Hg in the spontaneously hypertensive rat (input pressure: 110 mm Hg). The coefficient of variation of terminal pressures in vessels of order I increased linearly with the mean pressure drop in the system. The coefficient of variation in terminal pressure had a minimum as a function of the diameter exponent in case of each pial tree. At its minimum, it was higher in all spontaneously hypertensive rats (10.1–22.9%) than in any normotensive rats (6.0–8.5%). The corresponding diameter exponents were in most cases lower in the spontaneously hypertensive rat (1.3–2.5) than in the normotensive rat (2.5–3.0). Topologically consistent models of the pial arterial network predicted significantly less variation in intravascular pressures than was obtained by direct calculations. More idealized models suggested the dependence of coefficient of variation in terminal pressure on the total number of vascular segments contained by the tree. All models predicted the existence of the minimum of coefficient of variation in terminal pressure in function of the diameter exponent. From these, we conclude that (a) a hemodynamic configuration of the pial arterial system resulting in the smallest variation in cerebral perfusion pressure may exist, and (b) the pial vascular structure of spontaneously hypertensive rat allows less homogeneous terminal pressure distribution than does that of normotensive rats.

Middle cerebral artery transcranial doppler velocity monitoring during orthotopic liver transplantation: Changes at reperfusion—A report of six cases

STUDY OBJECTIVE To determine the effect of reperfusion of the grafted liver on transcranial Doppler blood flow velocity in the middle cerebral artery in humans during orthotopic liver transplantation. DESIGN Clinical study. SETTING University hospital. PATIENTS 6 patients scheduled for orthotopic liver transplantation. INTERVENTIONS Middle cerebral artery blood flow velocity (MCAVm) was monitored continuously using a transcranial Doppler (TCD) probe. The TCD measurements were noninvasive. MEASUREMENTS AND MAIN RESULTS The EME TC2000S TCD probe (Nicolet, Inc., Memphis, TN) was secured to the head using a strap providing continuous measurement of MCAVm. All other data were recorded by a patient monitoring system and a respiratory gas analyzer. Averaged MCAVm increased significantly in 5 of 6 patients (p < 0.001) when pre-reperfusion and post-reperfusion values were compared. Maximum post-reperfusion values for MCAVm, pulsatility index (PI), and systolic Doppler velocity (Vs) were greater than the corresponding immediate pre-reperfusion values (p < 0.05, p < 0.05, and p < 0.001, respectively). The increases in MCAVm cannot be explained on the basis of hypercarbia alone and were observed in the presence of systemic arterial hypotension and abrupt increases in central venous pressure, particularly at the time of graft reperfusion. CONCLUSIONS MCAVm increased with reperfusion of the grafted liver. These data suggest that multiple factors--including hypercarbia, lactic acidosis, or multiple vasoactive substances released by the grafted liver--may contribute to the observed increases in MCAVm, Vs, and PI.

Concomitant EEG, Lactate, and Phosphorus Changes by 1H and 31P NMR Spectroscopy During Repeated Brief Cerebral Ischemia

Pilots of high-performance aircraft are subject to transient loss of consciousness due to cerebral ischemia resulting from sudden high gravitational stress. To assess the effects of gravitational stress-induced blackout on cerebral metabolism and electrical function, we developed an animal model in which global cerebral ischemia is produced repeatedly at short intervals. Rats were prepared by ligation of subclavian and external carotid arteries and the right carotid artery was cannulated bidirectionally to measure circle of Willis and systemic pressures. Ischemia was induced by inflation of an occluder about the left carotid artery. Interleaved 31P and 1H NMR spectra were acquired on a 4.7-T Biospec system simultaneously with EEG recordings. We report results from 20 experiments of 30-min duration in which rats were subject to 30 1-min ischemia:reflow cycles of 10I:50R, 20I:40R, 30I:30R, and 40I:20R [numbers are seconds of ischemia (I) and reflow (R) during each 1-min cycle]. During ischemia the graded delivery of the ischemic insult permitted direct correlations between 2- to 5- and 7- to 20-Hz EEG activity and progressive changes in pH, lactate, ATP, phosphocreatine (PCr) and Pi. The best correlations were found between EEG activity and pH and PCr; correlation coefficients ranged from 0.93 to 0.95. A loss of EEG activity was observed without significant sustained energy loss in all but the most severe cycle.

Progressive lacunar infarction with demonstrated patency of the middle cerebral artery.

Two cases of progressive hemiplegia were closely followed by daily clinical examination. In both, the CT scan and CSF were normal on admission. In both, objective aggravation occurred in three or more steps over four days, progressing from minor finger clumsiness to total paralysis of the arm. In both cases a second CT scan a day after appearance of hemiplegia demonstrated a lacune in the corona radiata just above the internal capsule. In one case an intravenous digital subtraction angiogram demonstrated patency of the middle cerebral artery during the course of the progression. In the other case, serial study with transcranial Doppler ultrasound documented the continued patency of the middle cerebral artery. These two cases demonstrate that it is not necessary to postulate transient occlusion of the middle cerebral artery as an essential mechanism for progressive lacunar infarction.

Nitric Oxide as a Mediator of Cerebral Blood-Flow, Synaptic Plasticity, and Superoxide-Mediated Brain Injury

We propose that excessive production of nitric oxide initiated by activation of glutamate receptors after cerebral ischemia potentiates free radical injury to the brain. Nitric oxide, produced by neurons and possibly astrocytes, helps regulate local cerebral blood flow and plays an essential role in synaptic plasticity and normal development of the brain. Nitric oxide itself is a weak oxidizing agent, but after reaction with superoxide (O 2 - ), it forms the strong and relatively long-lived oxidant peroxynitrite anion (ONOO-) (Beckman et al., 1990). Peroxynitrite is sufficiently stable even in the presence of physiological concentrations of glutathione and other cellular antioxidants to diffuse for up to several cell diameters. Depending on what peroxynitrite reacts with, it can produce oxidants with the reactivity of hydroxyl radical (HO⋅), nitrogen dioxide (NO2), nitronium ion (NO 2 + ), and possibly singlet oxygen. Thus, the conversion of nitric oxide to peroxynitrite and related secondary oxidants may provide a common link between glutamate and free radical-mediated injury. Glutamate antagonists and free radical scavengers can reduce infarct volume by approximately the same extent in rat middle cerebral artery occlusion models of stroke (Oh and Betz, 1991). Recently, nitroarginine, a competitive inhibitor of nitric oxide synthesis, has also been shown to reduce infarct volume in a mouse middle cerebral ischemia model (Nowicki et al., 1991).

Evaluation of lactate production and clearance kinetics by 1H NMR in a model of brief repetitive cerebral ischemia

Pilots of high-performance aircraft are subject to repeated transient cerebral ischemia during high-gravitational stress maneuvers. Previously we have demonstrated that repeated episodes of transient cerebral ischemia and reflow are cumulative and lactate accumulations appear to be exponential. To evaluate the metabolic events determining the kinetics of lactate accumulation, and therefore the rates of substrate utilization, we have used in vivo 1H nuclear magnetic resonance with a 5-s time resolution to measure lactate production and clearance. The individual rates for each animal were then used to predict the accumulation of lactate in the same animal during 30 episodes of ischemia and reflow. Lactate accumulation was modeled as the balance between a zero-order production process during the ischemic period and a first-order clearance process. The predicted lactate accumulation showed excellent agreement with the observed time course, validating the predictive power of the simple model used. The highly reproducible nature of this model and its accuracy in predicting lactate accumulation should enable more accurate studies of the deleterious effects of lactate accumulation in cerebral ischemia by providing a highly reproducible means for generating a specific level of lactate accumulation.