To the Editor:
In the May 1998 issue of Stroke, Dumville et al1 showed that the classic CO2 test to assess cerebrovascular reactivity (CVR) may produce misleading results unless simultaneous changes in arterial blood pressure (ABP) are taken into account. A similar finding has recently been reported by Hetzel et al.2
The possibility of interactions among CO2, ABP, and cerebral blood flow (CBF) has been raised previously.3 4 5 The fact that 2 recent studies1 2 focusing on CVR arrived at similar conclusions despite significant differences in the populations studied reinforces the message that ABP has to be included as a significant covariate in any attempts to quantify the effects of CO2 on CBF velocity (CBFV), as usually measured with Doppler ultrasound in the middle cerebral artery. Dumville et al1 examined 56 patients with carotid artery disease (CAD) with a mean age of 67±8 years; the group studied by Hetzel et al2 comprised 81 healthy volunteers with ages ranging from 19 to 74 years. CVR in the latter group was slightly higher than in the former (3.6±1.6%/mm Hg versus 3.4±1.5%/mm Hg), but multiple regression analysis has shown that when the effect of ABP is taken into account, the CVR of the CAD patients dropped to 2.76±1.2%/mm Hg.1 Although Hetzel et al2 have obtained a significant correlation between ABP and end-tidal CO2, with a slope of 0.55 mm Hg/mm Hg CO2, they have not reported on corrected values of CVR when the influence of ABP is removed.
By studying healthy volunteers, Hetzel et al2 did not have the opportunity to observe the effects of ABP on the misclassification of subjects in relation to preestablished thresholds of CVR. In their patient population, Dumville et al1 reported that inclusion of ABP led to 14 patients showing a compromised CVR instead of only 8 when the conventional measure of CVR was adopted. Moreover, we have shown that in 4 patients the observed increase in CBFV during the CO2 test was primarily caused by ABP, a phenomenon that is also mentioned by Hetzel et al,2 as illustrated in their Figure 3.
Different methodological approaches are probably behind the variability in CVR values observed in the literature.6 Attempts to quantify and to compensate for the contribution of ABP are likely to produce an even greater disparity of results unless greater care is taken by thoroughly testing differing methodological alternatives,6 performing sensitivity analysis of parameters,1 and assessing reproducibility of results.2 Inclusion of ABP as a covariate also requires careful consideration of dynamic changes.7 8 In the study of Hetzel et al2 it is not clear whether a plateau in CBFV was established at each of the 4 phases of their study or whether the time delay between step changes in CO2 and the rise in CBFV was corrected for, as described by Dumville et al.1 Changes in ABP, either spontaneous or induced by the CO2 test, will produce different effects on CBFV, depending on the status of cerebral pressure-autoregulation.8 9 Because it is not possible to exert absolute control on the time course of these variables during clinical tests, dynamic modeling of the interaction between CO2, ABP, and CBFV is likely to play a major role in bringing further refinements to clinical applications of CVR testing.
- Copyright © 1999 by American Heart Association
Dumville J, Panerai RB, Lennard NS, Naylor AR, Evans DH. Can cerebrovascular reactivity be assessed without measuring blood pressure in patients with carotid artery disease ? Stroke. 1998;29:968–974.
Hetzel A, Braune S, Guschlbauer B, Dohms K. CO2 reactivity testing without blood pressure monitoring? Stroke. 1999;30:398–401.
Harper AM, Glass HI. Effect of alterations in the arterial carbon dioxide tension on the blood flow through the cerebral cortex at normal and low arterial blood pressures. J Neurol Neurosurg Psychiatry 1965;28:449–452.
Ackerman RH, Zilkha E, Bull JWD, Du Boulay GH, Marshall J, Russell RWR, Symon L. The relationship of the CO2 reactivity of cerebral vessels to blood pressure and mean resting blood flow. Neurology. 1973;23:21–26.
Kleiser B, Scholl D, Widder B. Assessment of cerebrovascular reactivity by Doppler CO2 and Diamox testing: which is the appropriate method? Cerebrovasc Dis. 1994;4:134–138.
Shapiro W, Wasserman AJ, Patterson Jr. JL. Mechanism and pattern of human cerebrovascular regulation after rapid changes in blood CO2 and tension. J Clin Invest. 1966;45:913–922.
Aaslid R, Lindegaard KF, Sorteberg W, Nornes H. Cerebral autoregulation dynamics in humans. Stroke. 1989;20:45–52.
Studies by Dumville and colleaguesR1 and our groupR2 both report on the influence of CO2 on ABP during Doppler CO2 testing in humans. This variable CO2 effect on CBFV and ABP can lead to a misinterpretation of the results of cerebrovascular reactivity testing. Dumville and colleaguesR1 reported 4 cases in a group of 56 patients undergoing carotid endarterectomy with false overall pictures of CO2 reactivity, and weR2 reported 1 case with the same phenomenon in our study. Both studies emphasize the covariance of CO2 effects on ABP and CBFV. Therefore, Doppler CO2 testing monitors not only the CO2-induced vasodilation of cerebral arterioles with increase in CBFV but also the efficacy of cerebral autoregulation in maintaining stable cerebral blood flow with respect to increases in ABP. In patients and controls nonlinear fluctuations, especially rapid changes in ABP, correspond to similar changes in CBFV. This phenomenon, illustrated in Figures 2 and 3 in our article,R2 is characteristic of a high-pass filter response.R3 Variations in ABP of >0.1 Hz produce similar variations in CBFV.R3 R4 In our opinion, such variations of ABP, and not the linear increase in ABP, influence predominantly the CBFV. Multiple linear regression analysis without respect to the frequency of ABP variations will neglect this physiological fact of such a threshold-related influence. In patients with high-grade obstructions of brain-supplying arteries, cerebral autoregulation will be compromised because of missing effective collateral supply. Therefore, CO2- and ABP-induced vasomotor response must be differentiated from passive ABP-induced changes in CBFV.
In our study neither the dynamics of ABP variations nor stable hypercapnic state were considered. Steady-state conditions could not occur with our rebreathing method. For that reason, we refrained from correcting CO2 reactivity results for changes in ABP.
The time delay between dynamic changes in Petco2 and the rise and fall in CBFV is a characteristic that provides additional information about the vasomotor response.R5 This was not considered in our study, which could explain the most pronounced increase in ABP at the end of our CO2 test. We perform an ongoing study with respect to time delays between Petco2, ABP, and CBFV. First results in patients with severe carotid stenosis showed a highly significant correlation between CO2 reactivity and time delay of the fall in CBFV at the end of hypercapnia. The realignment of this time delay as performed by Dumville et alR1 may miss some insights into the pathophysiological dynamics of cerebrovascular responses intime. Their multiple linear regression analysis is the correct approach to consider linear interactions between the multimodal measured parameters. Contrary to the usual analysis of CO2 reactivity, this might prevent underestimating the degree of hemodynamic compromise, especially in patients with already-diminished vasomotor response. Beat-to-beat analysis is necessary to quantify and report corrected values of CO2 reactivity, and the analysis must also allow for a nonlinear relationship between parameters. Dynamic modeling of the interaction between the parameters will be the next step in the analysis of CO2 reactivity with respect to the dynamics of cerebral autoregulation by considering amplitudes and time delays.
Dumville J, Panerai RB, Lennard NS, Naykor AR, Evans DH. Can cerebrovascular reactivity be assessed without measuring blood pressure in patients with carotid artery disease? Stroke.. 1998;29:968–974.
Hetzel A, Braune S, Guschlbauer B, Dohms K. CO2 reactivity testing without blood pressure monitoring? Stroke.. 1999;30:398–401.
Blaber AP, Bondar RL, Stein F, Dunphy PT, Moradshani P, Kassam MS, Freeman R. Transfer function analysis of cerebral autoregulation dynamics in autonomic failure patients. Stroke.. 1997;28:1686–1692.
Diehl RR, Linden D, Lücke D, Berlit P. Phase relationship between cerebral blood flow velocity and blood pressure: a clinical test of autoregulation. Stroke.. 1995;26:1014–1019.
Poulin MJ, Liang PJ, Robbins PA. Dynamics of the cerebral blood flow response to step changes in end-tidal PCO2 and PO2 in humans. J Appl Physiol.. 1996;81:1084–1095.