Control of end-tidal PCO₂ reduces middle cerebral artery blood velocity variability: Implications for physiological neuroimaging

Abstract

Breath-by-breath variability of the end-tidal partial pressure of CO₂ (Pet_CO2) has been shown to be associated with cerebral blood flow (CBF) fluctuations. These fluctuations can impact neuroimaging techniques that depend on cerebrovascular blood flow. We hypothesized that controlling Pet_CO2 would reduce CBF variability. Dynamic end-tidal forcing was used to control Pet_CO2 at 1.5 mm Hg above the resting level and to hold the end-tidal partial pressure of oxygen (Pet_O2) at the resting level. Peak blood velocity in the middle cerebral artery (MCA) was measured by transcranial Doppler ultrasound (TCD) as an index of CBF. Blood velocity parameters and timing features were determined on each waveform and the variance of these parameters was compared between Normal (air breathing) and Forcing (end-tidal gas control) sessions. The variability of all velocity parameters was significantly reduced in the Forcing session. In particular, the variability of the average velocity over the cardiac cycle was decreased by 18.2% (P < 0.001). For the most part, the variability of the timing parameters was unchanged. Thus, we conclude that controlling Pet_CO2 is effective in reducing CBF variability, which would have important implications for physiologic neuroimaging.

Introduction

Cerebral blood flow (CBF) is regulated by both blood pressure and the carbon dioxide in arterial blood (Mchedlishvili, 1980). For instance, by altering CO₂ levels, without any direct control of blood pressure, it is possible to manipulate blood flow (Mchedlishvili, 1980, Ide et al., 2003, Hetzel et al., 1999). Typically, a 2.5% increase in CBF is seen with a 1 mm Hg increase in the partial pressure of arterial CO₂ (Ide et al., 2003, Wise et al., 2004). CO₂ vasodilator effects appear to regulate flow in cerebral arterioles through hydrogen ion interactions.

Dynamic CBF regulation can be highly variable over short periods of time (Panerai et al., 2003). It has been shown that this variability is related to changes in CBF modulators, specifically, arterial CO₂ and blood pressure (Wise et al., 2004, Panerai et al., 1998, Panerai et al., 1999, Panerai et al., 2000, Panerai et al., 2003, Venkatesh et al., 2002). Panerai et al. (2000) examined CBF in relation to the beat-by-beat fluctuations in mean arterial blood pressure and the breath-by-breath variability of the end-tidal partial pressure of CO₂ (Pet_CO2) and found that CO₂ effects had much greater influence over variability than blood pressure. Furthermore, breath-to-breath Pet_CO2 variability has been shown to be a major source of this physiologic variability (Wise et al., 2004). We propose that the variability of CBF will decrease when the variability of breath-to-breath Pet_CO2 is decreased. This has important implications in magnetic resonance (MR) imaging as low-frequency noise due to physiologic variability has been shown to be the dominant source of noise in MR blood oxygenation level dependent (BOLD) imaging (Kruger and Glover, 2001). Additionally, physiologic BOLD noise has been correlated with Pet_CO2 fluctuations. Since the BOLD effect produces small amplitude changes in the MR signal, reducing noise has important implications for image quality and signal quantification.

To examine the role of CO₂ on CBF fluctuations, we employed the technique of dynamic end-tidal forcing (DEF) and transcranial Doppler ultrasound (TCD). DEF is a technique that controls the end-tidal respiratory gases accurately and continuously on a breath-by-breath basis (Robbins et al., 1982a, Robbins et al., 1982b). In the absence of respiratory disease, end-tidal gases reflect arterial blood gases (Robbins et al., 1990); therefore, manipulations of Pet_CO2 and Pet_O2 result in arterial blood gas changes. TCD has a high temporal resolution to measure cerebral blood velocity. The TCD blood velocity waveforms have various key features, which can be defined and used for an appropriate waveform analysis that accounts for temporal and velocity changes (Kurji et al., in press). Using the assumption that cross-sectional area of the insonated vessel does not change in mild hypercapnia (Poulin and Robbins, 1996a), cerebral blood velocity can be used as an index of CBF. The combination of DEF and TCD to examine CBF responses to varying CO₂ and O₂ conditions has previously been described (Ide et al., 2003, Poulin and Robbins, 1996a, Poulin et al., 1996b, Poulin et al., 1998).

Natural fluctuations in arterial CO₂ appear to be a major source of CBF variability (Wise et al., 2004, Panerai et al., 2000, Panerai et al., 2003, Venkatesh et al., 2002). We hypothesize that the variability of CBF, as measured by cerebral blood velocity in the middle cerebral artery (MCA), will decrease with control of Pet_CO2 with DEF, regardless of any blood pressure effects.