Original Research Comparison of the Effects of Independently- Controlled End-Tidal PCO 2 and PO 2 on Blood Oxygen Level–Dependent (BOLD) MRI Eitan Prisman, MD, MA, 1 Marat Slessarev, MSc, 1,2 Jay Han, BSc, 1,2 Julien Poublanc, MSc, 3 Alexandra Mardimae, BSc, 1,2 Adrian Crawley, PhD, 3 Joseph Fisher, MD, 1,2 and David Mikulis, MD 3 * Purpose: To assess the effect of changes in end-tidal par- tial pressure of O 2 (PETO 2 ) on cerebrovascular reactivity (CVR) estimated from changes in blood oxygen level– de- pendent (BOLD) signal during cyclic changes in end-tidal partial pressure of CO 2 (PETCO 2 ). Materials and Methods: BOLD response to fixed cyclic step changes in PETCO 2 (range = 30.4 – 48.8 mmHg) and PETO 2 (range = 100.6 – 444.0 mmHg) was studied in four healthy volunteers. Results: The BOLD reactivity to PETCO 2 and PETO 2 were 0.283 (0.188 – 0.379) (median, range) and 0.004 (0.003– 0.006)%/mmHg, respectively, in the whole brain; 0.438 (0.382– 0.614) vs. 0.006 (0.004 – 0.009)%/mmHg, respec- tively, in the gray matter; and 0.075 (0.065– 0.093) vs. 0.002 (0.001– 0.002)%/mmHg, respectively, in the white matter. Conclusion: The BOLD reactivity to PETO 2 was much smaller than that to PETCO 2 . However, BOLD reactivity can be significantly distorted by CO 2 -induced changes in PETO 2 . We conclude that PETO 2 should be carefully con- trolled during studies that use BOLD reactivity as an indi- cator of CVR. Key Words: BOLD MRI; cerebrovascular reactivity; end- tidal CO 2 ; end-tidal O 2 ; mapping J. Magn. Reson. Imaging 2008;27:185–191. © 2007 Wiley-Liss, Inc. CEREBROVASCULAR REACTIVITY (CVR), defined as a change in cerebral blood flow per unit change in stim- ulus, can be used as an index of the brain’s capacity for perfusion autoregulation. CVR is impaired in diseases that affect the cerebral vasculature, such as carotid stenosis (1), stroke (2), moyamoya disease (3), and can- cer (4 – 6). Quantitative mapping of CVR requires a method of quantifying blood flow with adequate spatial resolution to detect changes in brain areas of interest. Magnetic resonance imaging (MRI) has the necessary spatial resolution and is noninvasive compared to positron emission tomography (PET), which requires injection of radioactive tracers, or computed-tomogra- phy (CT), which produces ionizing radiation. Several MRI methods of measuring cerebral blood flow exist. For example, arterial spin labeling MRI is an estab- lished method of measuring regional cerebral blood flow. However, it is still primarily used in research set- tings and is not widely available to clinicians. In con- trast, most clinicians have access to blood oxygen level– dependent (BOLD) MRI that is dependent on changes in deoxyhemoglobin (dHb) concentrations, which in turn are inversely proportional to changes in cerebral blood flow. BOLD MRI is therefore commonly used as a sur- rogate indicator of changes in cerebral blood flow to determine CVR in clinical settings (1– 6). Also required is a means of administering a quantifiable and repro- ducible stimulus. The manipulation of end-tidal partial pressure of CO 2 (PETCO 2 ) causes a change in arterial partial pressure of CO 2 (PaCO 2 ), thereby providing a suitable stimulus to induce changes in cerebral blood flow. Table 1 summarizes common methods of inducing changes in PETCO 2 during CVR studies. It is well known, however, that an increase in PETCO 2 stimulates ventilation, resulting in a significant in- crease in end-tidal partial pressure of O 2 (PETO 2 ) (7). An increase in PETO 2 can have two opposing effects on the BOLD signal. On the one hand, hyperoxia may cause vasoconstriction (8), thereby decreasing cerebral blood flow, increasing venous dHb levels (given unchanged metabolic O 2 consumption) and ultimately decreasing BOLD signal intensity. On the other hand, hyperoxia increases arterial oxygen content by either increasing the O 2 saturation of hemoglobin and the fraction of O 2 physically dissolved in the plasma. For a given brain O 2 consumption and cerebral blood flow, the increase in 1 Department of Anesthesiology, University Health Network, University of Toronto, Toronto, Canada. 2 Department of Physiology, University of Toronto, Toronto, Canada. 3 Department of Medical Imaging, University Health Network, University of Toronto, Toronto, Canada. Contract grant sponsor: Ontario Research and Development Challenge; Contract grant number: BRAIN Project #01-MAR-0936. *Address reprint requests to: D.J.M., MD, Department of Medical Im- aging, Toronto Western Hospital, 3MC-431, 399 Bathurst Street, To- ronto, ON, Canada M5T 2S8. E-mail: mikulis@uhnres.utoronto.ca Received September 21, 2006; Accepted July 3, 2007. DOI 10.1002/jmri.21102 Published online 29 November 2007 in Wiley InterScience (www. interscience.wiley.com). JOURNAL OF MAGNETIC RESONANCE IMAGING 27:185–191 (2008) © 2007 Wiley-Liss, Inc. 185