In Response: Pattinson et al.1 and Ramani2 raise interesting methodological issues that should be useful for future investigators doing similar research on anesthetics and other drugs which may alter Paco2 and thus confound evaluation of regional cerebral blood flow (rCBF) changes. In our report,3 we found widespread increases in rCBF with increasing remifentanil dose, but also noted a significant increase in Paco2 with remifentanil. There are published guidelines on normal rCBF response to changes in Paco2. However, concerned that such correction factors may not apply to humans receiving remifentanil and that they may not apply homogeneously across the brain, we rejected that approach. Rather, we chose to normalize the cerebral blood flow (CBF) measures by dividing each rCBF region of interest by global CBF. This takes into account some of the aforementioned issues, but not all, as detailed in our report. Nonetheless, this approach did allow us to ascertain some regional effects of remifentanil on rCBF. Notably, in a similar human remifentanil study by Wagner et al.4 using positron emission tomography, they chose to deal with the CO2 issue by coaching subjects to breathe more as the remifentanil depressed breathing, producing a rather flat CO2 change with increasing doses. This produced other potential issues but they clearly addressed the issue of remifentanil-induced hypercapnea. Notably, they also observed an element of limbic system activation from sedative dose remifentanil in humans. Pattinson et al.1 suggest that administering CO2 to the subjects separate from remifentanil administration would have solved the problem. Indeed, their suggestion would allow generation of regional CO2 reactivity maps, specific to the individual patient which could be used for a voxel-by-voxel correction for CO2 effect. We agree that such an approach would have been a meaningful improvement, particularly with respect to possible regional variations in CO2 reactivity, and should be considered in future similar studies. However, we suggest two caveats leading to caution in fully embracing this approach. The first is the assumption that the CO2 response will be unchanged by the drug being evaluated. The second is that the logistics of doing the CO2 reactivity map could be problematic. Is the CO2 administered in the same setting? If so does this produce any sensations of anxiety or dyspnea that may impact on subsequent tests? Is the CO2 given in a separate session producing other issues in reproducibility? Dr. Ramani suggests that the BOLD fMRI technique should be employed concurrently with the arterial spin labeling rCBF technique.2 He suggests that it is little affected by changes in Paco2. This is a puzzling statement to us given previous reports that hypercapnea does affect BOLD measures.5,6 BOLD contrast reflects a complex interaction between CBF, cerebral blood volume, and oxygen utilization. Thus, the suggestion to combine arterial spin labeling rCBF studies with BOLD is potentially useful for future studies, as also suggested by Davis et al.5 Such an approach could help separate metabolically induced flow increases from those due solely to CO2 changes. Thus and although perhaps for different reasons, we welcome Ramani’s suggestion and agree that it could be a meaningful improvement to the methods we employed. W. Andrew Kofke, MD, MBA FCCM Professor, Director of Neuroanesthesia Co-Director Neurocritical Care Departments of Anesthesia and Neurosurgery University of Pennsylvania Philadelphia, PA [email protected] John A. Detre, MD Associate Professor Departments of Neurology and Radiology Director, Center for Functional Neuroimaging University of Pennsylvania Philadelphia, PA Jiongjiong Wang, PhD Research Assistant Professor Departments of Neurology and Radiology University of Pennsylvania Philadelphia, PA Patricia A. Blissitt, RN, M.S.N., PhD, CCRN, CNRN, CCM, APRN, BC Staff Nurse, Neuroscience Intensive Care Unit Duke University Hospital Durham, NC