www.elsevier.com/locate/ynimgNeuroImage 38 (2007) 1–2

Commentary

Anaesthetic use in animal models for neuroimaging

Chris Martin

University of Sheffield, Department of Psychology, Western Bank, Sheffield, South Yorkshire S10 2TN, UK

Received 12 March 2007; revised 10 April 2007; accepted 10 April 2007Available online 19 April 2007

We read with interest the recent paper by Fuqiang Zhao andcolleagues on improving the spatial localisation of the BOLDfMRI signal (Zhao et al., 2007). Briefly, the paper demonstratesthat the commonly observed post-stimulus undershoot in theBOLD signal may provide added spatial specificity in thelocalisation of functional responses in cortex. Their experimentswere conducted in an anaesthetised cat. Zhao et al. cite a paperthat we recently published in NeuroImage (Martin et al., 2006a)which compared the neurovascular coupling and the haemody-namic response function in awake and anaesthetised rat prepara-tions. In their discussion, Zhao et al. cite our article in support ofthe argument that their findings from anaesthetised animals arelikely to be applicable to awake humans. We believe thesecitations misrepresent somewhat the main findings and implica-tions of our paper. We would stress that this does not detract fromthe validity or importance of the main findings of Zhao et al.Nevertheless, we feel it important to clarify the results of ourstudy in the context of the important scientific issues raised withintheir discussion.

In Martin et al. (2006a), we used a combination of opticalimaging spectroscopy, laser Doppler flowmetry and electrophy-siological techniques to record both neural and haemodynamicresponses in awake rats and in rats anaesthetised with urethane.All recordings are made from the whisker (barrel) cortex andbrief (2 s) stimuli were delivered to the contralateral whisker padat frequencies ranging from 1 to 40 Hz. This design enabled us tofirstly elucidate and compare the neurovascular coupling relation-ship between the two anaesthetic states and secondly, to comparethe temporal haemodynamic response function across a range ofevoked response magnitudes. From our data we concluded that:(a) the neurovascular coupling relationship in awake andanaesthetised animals was different, and (b) the temporal structureof the haemodynamic response is affected by anaesthesia. Wefurther suggest that “in identifying specific temporal models ofthe haemodynamic response function, data from anaesthetisedanimal preparations should be used with some caution. It is clearthat in refining biophysical models of the fMRI BOLD signal,

E-mail address: c.j.martin@shef.ac.uk.Available online on ScienceDirect (www.sciencedirect.com).

1053-8119/$ - see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.neuroimage.2007.04.022

and in researching neurovascular coupling, data from awakeanimal preparations will be important.”

Though we do not wish to challenge the arguments of Zhao etal. that their findings are likely to be applicable to the awakehuman, we do not agree that the results reported in our papersupport their argument that “the general pattern of the BOLDfMRI response under anaesthesia is similar to that of [the] awakecondition”. Whilst the temporal response profile shares somesimilarities between anaesthetic states (we are able to fit gammavariates to both with a reasonable level of accuracy: R2 for fits tocerebral blood flow changes: awake—0.73; anaesthetised—0.95),there were clear differences in the temporal dynamics of theresponse; the anaesthetised animals displayed a lower frequency,more sluggish temporal response. These important differencesbetween the awake and anaesthetised animals, together with otherdifferences in neurovascular coupling, constitute the main findingsof our paper. Additionally, in a subsequent study, we comparedhaemodynamic responses to hypercapnic challenge in awake ratswith those from anaesthetised animals (Martin et al., 2006b).Most striking were the differences in temporal dynamics: theresponses in awake animals again evolved much more quicklythan those of the anaesthetised animals. This suggests that thereare fundamental differences in cerebral vascular reactivitybetween our awake and anaesthetised animals.

The cascade of neurophysiological, metabolic and haemody-namic processes that link the effects of a sensory stimulus tomeasurements of a change in the BOLD signal is obviously highlycomplex and susceptible to the differing effects of the variousanaesthetic agents. Many research reports, including Zhao et al.(2007), appropriately address this issue as part of their discussion.Indeed, it is our experience that journal referees frequently drawattention to the subject. However, there is very little synthesis ofunderstanding with respect to either the use of anaesthesia in theseanimal models or the differential effects of different anaesthetics.Each laboratory is well able to justify their choice of anaestheticregime, some because they appear to preserve normal neuronalfunction, others because they have minimal cardiovascular and/orhaemodynamic effects. Discrepancies in results between studiesare commonly ascribed to differences in anaesthetic regime andwhilst this explanation is plausible, it is not illuminating.

Where the data reported from studies using anaesthetisedanimal models have direct implications for the design, analysis and

2 C. Martin / NeuroImage 38 (2007) 1–2

interpretation of studies using awake humans, as is the case inmuch of the research into the physiological and biophysical basisof the fMRI BOLD signal, it is as important to consider the effectsof anaesthesia as it is species differences. However, awake animalmodels do not provide a complete solution. As well as beingtechnically challenging to develop, there are problems associatedwith movement artefacts and the danger of supplanting the adverseeffects of anaesthesia with those of restraint stress. Nevertheless wefeel that studies in animals free from the potentially confoundingeffects of anaesthesia are critical in this field.

References

Martin, C., Berwick, J., Martindale, J., Mayhew, J.E., 2006a. Investigatingneural-haemodynamic coupling and the haemodynamic responsefunction in the awake rat. NeuroImage 32 (1), 33–48.

Martin, C., Jones, M., Martindale, J., Mayhew, J., 2006b. Haemodynamicand neural responses to hypercapnia in the awake rat. Eur. J. Neurosci.24, 2601–2610.

Zhao, F., Jin, T., Wang, P., Kim, S.G., 2007. Improved spatial localization ofpost-stimulus BOLD undershoot relative to positive BOLD. Neuro-Image 34 (3), 1084–1092.