Patterns of bold signal responses to progressive hypercapnia enhance the interpretation of underlying cerebrovascular pathologies
Joseph A Fisher1,2,3, Olivia Sobczyk1, Adrian P Crawley4, Julien Poublanc4, Paul Dufort1, Lashmi Venkatraghavan5, David J Mikulis1,4, and James Duffin2,3

1Institute of Medical Sciences, University of Toronto, Toronto, ON, Canada, 2Department of Physiology, University of Toronto, Toronto, ON, Canada, 3Departments of Anaesthesia, University Health Network, Toronto, ON, Canada, 4Joint Department of Medical Imaging and the Functional Neuroimaging Laboratory, University Health Network, Toronto, ON, Canada, 5Department of Anaesthesia, University Health Network, Toronto, ON, Canada

Synopsis

We surveyed the varied patterns of BOLD changes in response to a ramp CO2 stimulus ranging from hypocapnia to hypercapnia in 10 healthy individuals and 10 patients with steno-occlusive disease. The patterns of response fell into 4 types, based on the two linear slopes fitted to each range. Maps of these types on a voxel by voxel basis were compared to cerebrovascular reactivity (CVR) calculated as the linear slope over the whole ramp. We suggest that for assessing cerebrovascular reactivity, CVR and type scoring enhance the interpretation of each other, and that modeling the possible underlying patho/physiologies to explain the type patterns is the portal to further work.

Purpose

To classify the varied BOLD responses to progressive hypercapnia in patients with known cerebrovascular steno-occlusive disease (SOD).

Introduction

Increasing the partial pressure of CO2 (PCO2) in the arterial blood reduces vascular resistance and increases cerebral blood flow (CBF). When the stimulated CBF exceeds the inflow capacity of the major arterial vessels, competition for flow is established between vascular regions. In healthy brains, the regional resistances are balanced and the pattern of the flow response to CO2 is sigmoidal in shape. However, in the presence of cerebrovascular disease, flow is preferentially redistributed to the healthier vessels, and produces various patterns of flow response to progressive hypercapnia (ramp) in various regions. We set out to classify these patterns of flow response to a ramp stimulus of increasing end-tidal PCO2 (PETCO2), for each voxel in patients with SOD, and map them to the corresponding location on the anatomical scan.

Methods

We monitored the BOLD response as a surrogate measure of cerebral blood flow, in 10 healthy subjects and 10 patients with SOD. The BOLD signals were divided into those corresponding to the upper and lower halves of a ramp PETCO2 from 5-10 mmHg below resting to10-15 mmHg above resting. A linear least squares fit of each part was used to calculate the two slopes. These slopes were then categorised as positive (+) or negative (-).

Results

We found that the changes in BOLD signal as a function of PETCO2 in each voxel could be categorised into one of four types based on the two slopes of the lower and upper range of the CO2 stimulus as follows: (A) +/+, (B) +/-, (C) -/-, (D) -/+, as illustrated in Figure 1. Each of the four types of BOLD response patterns shown in Figure 2 were color coded (A = red, B = light blue, C = dark blue, and D = yellow) and the colors were superimposed on the corresponding voxel of the anatomical scan to produce ‘type’ maps. Figure 2 shows an example. For comparison, we also calculated the CVR in the classical form where CVR = Δ BOLD / Δ PETCO2. The type map shows an area of abnormality over a greater extent than the steal alone. However C types indicate negative CVR and true steal.

Discussion/Conclusion

This is the first report of BOLD response patterns to a ramp PETCO2 stimulus. We found that all voxels in healthy subjects, and patients show one of four flow response patterns to a ramp CO2 stimulus that we call ‘types’, two of which are biphasic. CVR is a subset of ‘types’ with the best correlation to types A and C, which are linear. Biphasic types B and D have small values of CVR because of an averaging of the two slopes. We suggest that for assessing cerebrovascular reactivity, CVR and type scoring enhance the interpretation of each other, and that modeling the possible underlying patho/physiologies to explain the type patterns is the portal to further work.

Acknowledgements

No acknowledgement found.

References

No reference found.

Figures

Figure 1: Example plots of the 4 types of response patterns in BOLD MRI responses to a global ramp increase in PETCO2 from 30 to 55 mmHg. The voxel locations are indicated by the cursors on the inset CVR maps. These four patterns of response accounted for all those observed in the voxels of healthy subjects and patients.

Figure 2: A 18 year old male with a history of moyamoya disease affecting the right MCA territory. (a) the angiogram shows an occlusion of the right MCA. (b) The CVR shows steal predominantly in the right white matter. (c) The type map shows predominantly B C and D types on the right side (left of the figure) in a more extensive area than the steal (blue) in the CVR map.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4380