The plateau phase is a slippery slope: raising blood pressure may lower brain perfusion

The brain is highly dependent on cerebral blood flow to maintain its extraordinary productivity. Just a few seconds of reduced perfusion can reduce our level of consciousness and precipitate an injurious fall. Too much perfusion raises intracranial pressure within the confines of the skull and may induce irreversible brain damage. Tight control of cerebral blood flow, therefore, is a precondition for human life. Cerebral autoregulation is the mechanism that maintains cerebral blood flow in the face of shifts in blood pressure. The clinical understanding of cerebral autoregulation is still strongly influenced by Lassen's original publication from 1959, which depicted the concept of autoregulation: within a wide range of blood pressure levels, cerebral blood flow remains stable (Lassen, 1959). Lassen clearly indicated that his plot, which illustrated this ‘plateau phase’, was in fact an interpolated, group average plot that was based on several groups with various clinical characteristics. Nevertheless, clinicians (and many scientists alike) have since worked with the assumption that this autoregulatory curve applies to each (healthy) individual. This has led to a prevailing conception that under normal conditions, cerebral blood flow is held perfectly constant. In the years following Lassen's publication, a modification to his curve was proposed for a highly prevalent clinical condition, hypertension. Without very much experimental evidence, it was hypothesized that the autoregulatory curve would shift to the right under conditions of hypertension, such that the plateau of stable cerebral blood flow would extend to higher blood pressure levels, but would no longer cover lower blood pressure. This would place hypertensive individuals (especially when elderly) at risk of hypoperfusion if their high blood pressure was lowered. These two clinical concepts of cerebral autoregulation (stable flow over a wide range of pressures, with a rightward shift with hypertension) continue to influence clinical management. For example, generations of physicians have been reluctant to treat hypertension in the elderly for fear of inducing cerebral hypoperfusion. In addition, the firm reliance on the self-stabilizng properties of cerebral autoregulation may have lowered interest in investigating optimal blood pressure levels in critical care medicine. In this issue of The Journal of Physiology, Jie Liu and colleagues challenge some of these widely held notions of cerebral autoregulation (Liu et al. 2016). In a group of relatively healthy, somewhat older adults (mean ± SD, 66 ± 6 years), who were normotensive or had well-controlled hypertension, the authors lowered and raised blood pressure pharmacologically to investigate the plateau phase of autoregulation. This design translates to a wide range of clinical settings where blood pressure is manipulated pharmacologically by vasoactive drugs. Three important findings emerge from their work. (1) Far fewer than half of participants had the hypothesized plateau-shaped relationship between blood pressure and cerebral blood flow. If flow is plotted against pressure, as did Lassen, the ideal slope for the plateau phase is 0. However, Liu et al. plotted cerebrovascular resistance against pressure, which yields an ideal slope of 1. Even when they applied a wide ‘normal range’ of slopes from 0.5 to 1.5, still half of subjects fell outside this range, showing that there is strong inter-individual variability in cerebral autoregulation. (2) A quarter of participants demonstrated an unexpected pattern of response: when their blood pressure was lowered, cerebral blood flow increased, whereas when blood pressure was raised, perfusion fell, by as much as 20%. Is this a spurious observation? Recent studies of cerebral blood flow in hypertenson and antihypertensive treatment indicate this may not be the case. Intensive blood pressure reduction unexpectedly raised cerebral blood flow in elderly hypertensive patients (Liu et al. 2016). Indeed, hypertension has been linked to reduced cerebral blood flow (Tryambake et al. 2013). Even though further validation is warranted, Liu et al. may have discovered an important and common variant behaviour of cerebral autoregulation that may have important clinical implications, for example in management of high blood pressure in hypertension and stroke. (3) Participants with signs of small vessel disease on magnetic resonance imaging were more likely to show this unexpected pattern of lower brain blood flow with induced hypertension (and higher flow with hypotension), even more so in the posterior circulation. Again, this challenges the traditional autoregulation concept. The a priori hypothesis was that participants with poorer autoregulation would show more signs of small vessel disease than participants with normal autoregulation. This hypothesis was not confirmed. Taken together, Liu et al.’s findings show us that we cannot extrapolate Lassen's curve to the individual patient and that without measurements of cerebral blood flow, we cannot predict which level of blood pressure will yield optimal brain perfusion for the individual patient. None declared.