Measuring cerebral auto‐variability

The human brain is capable of regulating cerebral blood flow (CBF) during changes in cerebral perfusion pressure (CPP), which is calculated as the difference between mean arterial pressure (MAP; i.e. driving force) and intracranial pressure (ICP). The intrinsic ability of the brain to maintain a relatively constant cerebral perfusion in the presence of changes in CPP has been termed cerebral autoregulation (CA), and is calculated as the relative change in CBF or cerebral blood velocity (CBV) for a given relative change in MAP, while assuming ICP is constant.

Historically, however, the idea that changes in MAP and CPP can affect CBF has been around for over a century. For example, Bayliss & Hill (1895) stated that ‘In all physiological conditions a rise of arterial pressure accelerates the flow of blood through the brain, and a fall slackens it’. This ideology was largely accepted until 1959, when Lassen interpreted independent data sets to indicate that the range of static CA in humans is rather substantial (Lassen, 1959); that is, the human brain was capable of maintaining a constant CBF through a dynamic range of MAP from 60 to 150 mmHg. However, the CA curve Lassen proposed was determined based on 11 different experimental groups, and many participants had known pathology and were taking medications that can directly alter CA. Since then, contrary to Lassen (1959), and in agreement with Bayliss & Hill (1895), an overwhelming amount of research has been presented that suggests the human brain CA capacity is rather unexceptional, and a much smaller range of MAP is capable of penetrating the cerebrovasculature in healthy humans (see Willie et al. 2014). Despite recent research, these findings are often overlooked, and many physiologists and medical personnel revert back to the dogma that CA has a large dynamic range, and CPP remains constant during somewhat severe hypo‐ and hyper‐tension.

Regional differences in CBF regulation have been a topical area of interest in recent years. The posterior cerebral circulation (e.g. vertebral and posterior cerebral arteries; VA and PCA) and anterior cerebral circulation (e.g. internal carotid and middle cerebral arteries; ICA and MCA) account for ∼30% and ∼70% of total CBF, respectively. There has been contradicting evidence, supporting or opposing, regional differences in the brain in terms of CA. Embryonically, the rationale for differential CA between anterior and posterior circulations seems unconvincing as the vessel tissue is derived from the same mesodermic germ layer. This means that theoretically, these vessels should be regulating blood flow under the same governing principles, and therefore CA should be the same when normalized for vessel size. Regional CA has primarily been quantified using transcranial Doppler ultrasound (TCD) by insonating the MCA and PCA. But, there has been much debate surrounding the validity of TCD, which measures cerebral blood velocity, necessitating the use of Duplex ultrasound in its place, which measures cerebral blood flow. The ICA and VA are responsible for supplying blood that eventually arises in the MCA and PCA, respectively, and are the chief targets when measuring extra‐cranial CBF. In addition to exploring regional differences in CA in healthy humans, it is also important to explore differences in CA in pathology in order to grasp a thorough understanding of the mechanisms that control CA in order to gain valuable knowledge regarding how to potentially treat patients with cardiovascular and/or cerebrovascular conditions. For example, patients with hypertension, heart failure and CA impairment have been linked to cerebral ischaemic episodes; however, CA has yet to be explored in patients with white matter hyperintensity (WMH; also known as leukoaraiosis). Cerebral small vessel disease is thought to underscore WMH, and impairments in CA may therefore be present where WMH is shown. This potential pathophysiological link between CA and WMH was recently explored by Liu et al. (2016) in The Journal of Physiology.

In Liu et al. (2016), regional CA was quantified in 27 participants with various degrees of WMH using sodium nitroprusside and phenylephrine to induce steady‐state changes in MAP. The authors quantified CA using linear regression analysis of the relative change in CBF and cerebrovascular resistance to the relative change in MAP. The primary findings were as follows: (1) large individual variability in anterior and posterior CA, including some cases of ‘over‐reactive’ CA, where CBF was negatively correlated with MAP (e.g. increase in MAP resulted in a decrease in CBF), and (2) posterior CA was positively correlated with WMH. On the basis of these data, a number of pertinent and related points of discussion are explored in this Journal Club article.

Why is cerebral autoregulation so variable?

The authors comment that high variability in regional CA is not fully understood, but do provide some valuable suggestions as to why this was observed, such as differences in the endothelial, myogenic and metabolic response to changes in MAP. Furthermore, the authors meritoriously measured aortic pulse wave velocity and the resulting data indicated that individual differences in arterial stiffness and/or cerebrovascular tone also contributed to CA variability. The authors also addressed some important methodological limitations that may have contributed to CA variability. First, there was a possibility that the drugs administered to alter CPP could have been responsible for the increased variability in CA. Second, arterial carbon dioxide tension, which acts as a very potent stimulus on cerebral arteriolar diameter, was not controlled throughout experimentation. Third, by design, the participants included in their study were aged, many were on hypertensive medications, and some possibly going through the menopause, each of these also being likely to contribute to CA variability. Although the study design and interventions were impressive, it would have been interesting to compare the results against a healthy population to ascertain ‘normal’ CA responses.

There are several other potential contributors to CA variability in the present study that were not addressed. Transcranial Doppler ultrasound has adopted a negative reputation in recent years due to the possibility of intracranial vessel diameter being altered in the face of most physiological stimuli, specifically during changes in arterial carbon dioxide tension. Nevertheless, the stability and ease of use of TCD is extremely appealing, whereas obtaining adequate measurements with peripheral ultrasound can be more of a challenge, particularly with the VA, where the most variability in CA was presented in this study. In addition, in most people, VA flow is not uniform between both left and right VA's, which could contribute to CA variability. Because of this, it would have been beneficial for the authors to additionally measure the PCA using TCD to examine if the same results found with the VA were reproducible. Since the validity of TCD is still under debate, it is worthwhile to measure regional CBF using both TCD (MCA and PCA) and Duplex ultrasound (ICA and VA) when investigating CA. Other sources of CA variability could be attributed to the analysis of CA. In Liu et al. (2016) CA was not examined in both hypo‐ and hypertensive interventions separately, as there is some evidence of hysteresis between these two conditions (Numan et al. 2014). Also, analysis of CBF would have been more robust over a greater amount of cardiac cycles (i.e. >5) in order to dampen the measuring variability associated with respiration.

Cerebral autoregulation and white matter hyperintensity

The authors provided novel information on regional CA and WMH, a poorly understood disease that is common in the ageing population. The slope of posterior CA measured in the VA positively correlated with WMH, suggesting that CA was impaired in participants with WMH. This finding meaningfully contributes to the literature that supports the notion that differential regional CBF regulation is observed not only to acute physiological stimuli (e.g. carbon dioxide), but also in cerebral pathological conditions (i.e. WMH). One of the authors’ interesting findings was that participants with greater amounts of WMH presented with over‐reactive CA. For example, if MAP increased, then CBF subsequently decreased. Explanation of this novel finding is unclear, but over‐reactive CA could augment the development of WMH as transient increases in MAP could result in cerebral hypo‐perfusion or intracranial vascular steal. Nevertheless, future studies are needed to confirm these findings, perhaps using an MRI approach to measure global CBF as a substitute for ultrasound, and using non‐pharmacological methods to manipulate blood pressure such as lower‐body differential pressure, as it is unknown whether the drugs used in the current study directly affect CA.

Do we continue?

Our ability to measure and quantify CA is still limited, partly due to large variability within and between participants, and the clinical implications of impaired CA are still unclear. Yet, due to recent advancements in science and technology in the past decade, our knowledge of the cerebrovasculature has increased exponentially. Because of this, perhaps it is plausible to suggest that in the near future we will be able to firmly grasp to what extent impaired CA negatively affects humans, and determine whether CA has clinical implications for predicting or treating cerebrovascular disease. The study by Liu et al. (2016) elegantly demonstrates why CA research is important. Indeed, elucidation of the mechanistic properties of CA is paramount to our comprehension of pathologies associated with cerebrovascular disease.

Additional information

Competing interests

None declared.