To the Editor:
The article by Zaharchuk et al1 raises a series of questions.
First, the authors are careful to delineate many problems concerning interpretation of data gathered with the BOLD technique and the variations they employ. These limitations might help explain some of the difference between the microvascular versus the global cerebral blood volume changes they observe during hypoperfusion. However, the authors never call our attention to a much more glaring problem: In their Figure 3, there is no decrease in either total cerebral blood volume (CBV) or microvascular volume over the whole autoregulatory “plateau.” This is impossible. There must be an increase in resistance—a decrease in diameter—somewhere to account for the good autoregulation over the plateau. Until we have a reasonable explanation for their finding, it seems premature to trust their techniques and accept their conclusion that “CBV changes during hemorrhagic hypotension are far less than … [those] … reported by some previous studies.”
Second, the authors talk about cessation of autoregulation as the pressure falls below the so-called lower limit of autoregulation. This way of expressing matters is the one usually used, but it is highly misleading. Beyond each end of the “plateau” the vessels continue to autoregulate—ie, to relax as pressure falls and to constrict as pressure rises. Thus, the resistance continues to go up at pressure above the high end of the “plateau” and to fall at pressure below the low end. However, these autoregulatory changes in tone are no longer sufficiently larger to compensate for the changes in pressure, and the flow changes more markedly for each change in pressure than it did for changes in pressure over the “plateau.” Autoregulatory responses are not absent, and the flow-pressure relationship is not truly passive until some much lower (or higher) pressure is reached. This is NOT seen in the authors’ Figure 2, in which the relationship between flow and pressure appears to be totally passive as soon as we fall off the low end of the “plateau.” Perhaps this, too, is a function of the problems in measuring CBF with the imaging techniques they use. Perhaps, also, the authors were not aware of the conundrum presented by the data in Figure 2 because they expected (erroneously) that flow would be “passive” as soon as the “plateau” was left.
Finally, the authors’ Figure 2 does show something that is in agreement with the results reflected in the work of Kontos, myself, and others cited in their Reference 14. The “plateau” is not flat (the reason I place the term in quotes or say so-called). Rather, it has a gentle slope. It is interesting that in spite of the problems I point out, their data still show this. In any case, it is correct that the famous “plateau” plotted by Lassen (see critique in the Handbook of Physiology, chapter by Heistad and Kontos on cerebral circulation) from a summary of heterogeneous papers in the literature, is not a plateau at all and animals that fail to show an absolutely flat flow over the range of maximally effective autoregulation are behaving normally.
- Copyright © 2000 by American Heart Association
Zaharchuk G, Mandeville JB, Bogdanov AA Jr, Weissleder R, Rosen BR, Marota JJ, Iadecola C, Kim SG. Cerebrovascular dynamics of autoregulation and hypoperfusion: an MRI study of CBF and changes in total and microvascular cerebral blood volume during hemorrhagic hypotension. Stroke.. 1999;30:2197–204.
Dr Rosenblum’s letter highlights some of the more surprising findings of our recent article reporting total and microvascular CBV changes during cerebrovascular autoregulation and hypoperfusion in the rat.R1 Our findings do challenge the conventional wisdom that CBV must increase dramatically to maintain CBF in the face of declining mean arterial blood pressure (MABP).R2 However, it is important to remember that our study was the first to use an imaging-based technique capable of sampling both the brain’s cortical surface and its deeper parenchyma. The cortical surface has a higher fraction of large vessels, so it is not entirely surprising that CBV in the parenchymal responds differently. As most of the brain lies below the cortical surface, understanding such parenchymal changes may yield important insight into how the brain as a whole reacts to MABP alterations.
Dr Rosenblum’s first point concerns our finding of no significant increase in either total or microvascular CBV on the autoregulation plateau. Far from not calling attention to this fact, we had hoped to emphasize this most interesting finding. As Dr Rosenblum points out, it seems logical that vasodilation with concomitant CBV increase must occur to decrease cerebrovascular resistance as MABP falls within the autoregulatory range. We agree that such changes occur. However, we attribute the change in resistance (and diameter) to a very small subset of vessels, the cerebral arterioles, thought to comprise less than 5% of overall cerebral blood volume. The feasibility of such an interpretation is supported by our cerebrovascular model (more detail may be found in Reference 3), as well as previous evidence in other organ systems that CBF is regulated by changes in arteriolar diameter.R4 Because even large diameter changes in these relatively few vessels lead to only small overall increases in total CBV, it is possible to control CBF with minimal total CBV increases; as we have pointed out, such a system is particularly well-suited for the brain, where large changes in CBV are problematic due to the fixed overall volume of the bony cranium. Lastly, we reiterate that we cannot rule out CBV increases during autoregulation by as much as 20% due to uncertainties about the systemic concentration of the contrast agent during the course of the experiment. However, the conclusion remains that CBV does not increase dramatically (by >100%) as reported by previous methods that have sampled small amounts of tissue near the cortical surface.
Dr Rosenblum’s second point questions whether the CBF changes observed when MABP is below 50 mm Hg are consistent with previous reports and suggests that arterial spin label MRI methods may be at fault. He points out that previous studies demonstrate that the process of autoregulation is not “all-or-none”; ie, that the corners of the autoregulation curve are not sharp. He contends that we have documented “passive” CBF changes below a MABP of 50 mm Hg. We regret any confusion that may have been caused by placing a line based on a least-squares fit of CBF between 50 and 140 mm Hg onto our Figure 2 or by calculating the linear fits to our data within and below the autoregulation range in the Results section. It was never our intention to imply that autoregulation is lacking during mild hypotension; this line was placed arbitrarily only to support our selection of a range of MABP for comparison. Taking into account the error bars of the data in Figure 2, we do not believe that it is possible to argue that the CBF curve we show is inconsistent with a rounded lower limit of autoregulation. Also, the observation that CBV increases occur during mild hypoperfusion further suggests that autoregulatory processes continue to be active but are insufficient to curtail decreases in CBF.
Dr Rosenberg’s last comment points out that the autoregulation “plateau” should have a slight slope, as shown by his data.R5 Our data are in agreement with this. Rather than suggesting that the CBF measurement method is inaccurate, we believe that this finding argues that it is sufficiently sensitive and accurate to replicate this subtle physiological effect, which is likely secondary to anesthesia.R6 Unlike the CBV findings, the CBF results support previous observations during autoregulation. Because of this, we feel that the argument that the CBF measurement method is flawed cannot be based on the data presented in our article.
Last, we take issue with the contention that errors in the measurement techniques may be sufficient to “explain some of the difference between the microvascular versus the global cerebral blood volume changes” during hypoperfusion. As we have mentioned in the paper, there is a long history addressing the accuracy and robustness of susceptibility contrast MRI methods to measure CBV changes. Because of this, we believe that differences between total and microvascular CBV are entirely physiological and may have important implications during stroke. The ability to measure changes in microvascular CBV with spin echo susceptibility contrast MRI opens a new and fascinating window into events occurring at the capillary level during ischemia, by avoiding the potentially confounding effects of large vessel diameter changes.
Zaharchuk G, Mandeville JB, Bogdanov AA Jr, Weissleder BR, Rosen BR, Marota JJA. Cerebrovascular dynamics of autoregulation and hypoperfusion: an MRI study of CBF and changes in total and microvascular cerebral blood volume during hemorrhagic hypotension. Stroke.. 1999;30:2197–2205.
Powers WJ. Cerebral hemodynamics in ischemic cerebrovascular disease. Ann Neurol.. 1991;29:231–240.
Mandeville JB, Marota JJA, Ayata C, Zaharchuk G, Moskowitz MA, Rosen BR, Weisskoff RM. Evidence of a cerebrovascular post-arteriole Windkessel with delayed compliance. J Cereb Blood Flow Metab.. 1999;19:679–689.
Zweifach WB. Quantitative studies of microcirculatory structure and function. Circ Res.. 1974;34:834–866.
Kontos HA, Wei EP, Navari RM, Levasseur JE, Rosenblum WI, Patterson JL. Responses of cerebral arteries and arterioles to acute hypotension and hypertension. Am J Physiol.. 1978;243:H371–H383.
Michenfelder JD. Anesthesia and the Brain. New York, NY: Churchill Livingstone; 1988.