The Mechanism of Increased Blood Flow in the Brain and Spinal Cord During Hemodilution

Crystal, George J. PhD*†; Czinn, Edward A. MD*‡; Salem, M. Ramez MD*‡

Anesthesia & Analgesia:
doi: 10.1213/ANE.0000000000000078
Neuroscience in Anesthesiology and Perioperative Medicine: Research Report

BACKGROUND: Hemodilution is accompanied by an increase in cerebral blood flow, but whether this is due to vasodilation in response to reduced arterial oxygen content, reduced blood viscosity, or a combination of these mechanisms is a matter of debate. We performed the current study to gain insight into this question by evaluating the effect of hemodilution on (1) vasodilator reserve and (2) the level of blood flow during hypercapnia-induced vasodilation in regions of the brain and spinal cord.

METHODS: Sixteen mongrel dogs were anesthetized with halothane 0.9% (1 minimum alveolar concentration) while their lungs were mechanically ventilated. Radioactive microspheres (15 μm) were used to measure regional blood flow (RBF) in the cerebral cortex, cerebellum, pons, medulla, and spinal cord (cervical, thoracic, and lumbar segments). Arterial blood pressure was measured via an aortic catheter. Vasodilator reserve was assessed from the ratio of RBF during hypercapnia (PaCO2 approximately 65 mm Hg) to RBF before hypercapnia. PaCO2 was increased by the addition of dead-space tubing without changing the ventilator settings. The dilating effects of hypercapnia within the central nervous system (CNS) were assessed with hematocrit normal (group 1; n = 8) and after induction of isovolemic hemodilution to a hematocrit of 19 ± 4 (SD) with 5% dextran (group 2; n = 8).

RESULTS: Hemodilution increased RBF (P < 0.0001) and decreased the vasodilator reserve ratio (P < 0.05) in all regions of the brain and spinal cord; the ratios during hemodilution (group 2) were only 48% to 68% of those without hemodilution (group 1). The level of RBF during hypercapnia was not significantly different in the absence and presence of hemodilution (cerebral cortex: mean, 122 mL/min−1/100 g−1 vs mean, 108 mL/min−1/100 g−1; 95% confidence interval of the difference (95% CID), −53 to 26; P = 0.46; cerebellum: mean, 117 mL/min−1/100 g−1 vs mean, 100 mL/min−1/100 g−1; 95% CID, −52 to 18; P = 0.32; pons: mean, 83 mL/min−1/100 g−1 vs mean, 73 mL/min−1/100 g−1; 95% CID, −12 to 31; P = 0.35; medulla: mean, 96 mL/min−1/100 g−1 vs mean, 82 mL/min−1/100 g−1; 95% CID, −11 to 40; P = 0.25; cervical spinal cord: mean, 61 mL/min−1/100 g−1 vs mean, 52 mL/min−1/100 g−1; 95% CID, −18 to 34; P = 0.51; thoracic spinal cord: mean, 35 mL/min−1/100 g−1 vs mean, 46 mL/min−1/100 g−1; 95% CID, −30 to 8; P = 0.24; lumbar spinal cord: mean, 54 mL/min−1/100 g−1 vs mean, 58 mL/min−1/100 g−1; 95% CID, −25 to 15; P = 0.61). Neither hypercapnia alone nor combined with hemodilution affected mean arterial blood pressure (P = 0.78 and P = 0.81, respectively).

CONCLUSIONS: Hemodilution caused recruitment of the vasodilator reserve, suggesting that vasodilation played a role in the increase in RBF throughout the CNS. Although the mean values for RBF during hypercapnia were similar with and without hemodilution, a large variation in the responses precluded a conclusive determination of whether or not reduced blood viscosity also contributed to the hemodilution-induced increases in RBF. A dependence on vasodilation would limit autoregulatory capability throughout the CNS during hemodilution, which would enhance the risk for ischemia if hypotension was superimposed.

Author Information

From the *Department of Anesthesiology, Advocate Illinois Masonic Medical Center; Departments of Anesthesiology and of Physiology and Biophysics, University of Illinois College of Medicine; and Department of Anesthesiology, University of Illinois College of Medicine, Chicago, Illinois.

Edward A. Czinn, MD, is currently affiliated with Broward Health Medical Center, Fort Lauderdale, Florida and Florida International University, Miami, Florida.

Accepted for publication November 15, 2013.

Funding: Departmental Funds.

The authors declare no conflicts of interest.

Reprints will not be available from the authors.

Address correspondence to George J. Crystal, PhD, Department of Anesthesiology, Advocate Illinois Masonic Medical Center, 836 West Wellington Ave., Chicago, IL 60657. Address e-mail to

© 2014 International Anesthesia Research Society