Journal of Applied Physiology

Invited Editorial on “Fast and slow components of cerebral blood flow response to step decreases in end-tidal Pco 2 in humans”

Christian J. Lambertsen

the cited study by Poulin et al. (10) focuses on modeling dynamic cerebral vascular system responses to abrupt changes in its own environment. For the arterial blood PCO2 in such studies, an external environment is the lung alveolar gas. For the cells of the entire brain arterial and arteriolar system of contractile tubules, it is the PCO2 of the arterial blood itself, within the vascular lumen.

There is now broad awareness that the critical agent for influence on overall brain circulation is a CO2-mediated change in H+ activity in the internal environment of the arteriolar walls (4-7,12). This “control” influence is evidently not effected by change in H+ activity within the lumen of the brain vascular tree, as CO2delivers its authoritative influences freely through the surrounding barrier to the vascular walls (5, 6).

The early “hard point”1of dynamic human brain circulation study was provided by W. G. Lennox, F. A. Gibbs, and E. L. Gibbs sixty-three years ago (8, 9), as a heated thermocouple inserted in the jugular bulb for continuous monitoring of blood flow velocity and sequential sampling of arterial and brain venous blood during transition to unconsciousness on breathing nitrogen (9). Their skilled determinations of brain arteriovenous blood differences for O2 and CO2 content, hemoglobin O2 saturation, pH, and PCO2 provided information concerning exposures to hypoxia, hypercapnia, hypocapnia, and O2, relevant to brain competence (8). The well-recognized limitation was absence of a quantitative measure of brain blood flow and the consequent inability to compute brain O2 consumption, CO2 production, and vascular resistance from the extensive data being obtained.

Fifty-two years ago, development of the nitrous oxide method by Kety and Schmidt (3) was the hard point that provided the valuable quantitative index of blood flow, albeit averaged over a 15-min period and using about two person-days for one or two measurements. Situations of stable rates of brain O2consumption were established, including hypoxia, hypercapnia, and hypocapnia. Therefore, it became sensible to perform promptly repeatable arteriovenous O2difference [(a-v)Do 2] content measurements as indexes of brain blood flow (7), each requiring more than an hour.

The (a-v)Do 2content “index” of brain blood flow remains important in relating influences of blood gases, acid-base factors, and drugs to degrees of change between stable conditions and rates of change of overall respiration and brain blood flow (12-14). With naturally integrated respiratory and blood flow responses, the intrinsic rapid reactivities of brain vasculature to an increase or decrease in arterial blood PCO2 are obscured by the slower rates of the damped respiratory control and pulmonary gas exchange. The index still has retrospective application to classic early studies of brain (a-v)Do 2(8).

An imaginative hard point in the effort to determine maximum rates of cerebral blood flow response to CO2 was the use of (a-v)Do 2, saturation, and PCO2 measurement by Severinghaus and Lassen (12) in repetitive paired samplings of arterial and brain venous blood, beginning with an abrupt voluntary reduction of end-tidal PCO2 and then a prolonged stable maintenance of the hypocapnia. The data established a distinctly closer relation of the time constant for decrease in brain blood flow to the rate of PCO2 change in arterial blood than in jugular venous blood (12).

Poulin et al. (10) have now determined inherent rate constants of the cerebral blood flow system in response to the most rapid practically attainable reduction in intra-arterial PCO2 . The aggregate of methods described includes modern rapid-response engineering developments in computer controls and computations, simultaneous gas analysis and control of end-tidal O2 and CO2, transcranial Doppler flowmeter technology, and a data-acquisition system. The original handicap of uncertain variations in cross-sectional area of the jugular bulb has been replaced by ultrasonic sensing of change in cross-sectional area of the 3-mm-diameter middle cerebral artery and a single measure of brain blood flow that requires the duration of a heartbeat. Extremely rapid decrease in “arterial” PCO2 is accomplished by essentially eliminating both the lung and respiratory control from the experiment process through the use of sustained voluntary hyperventilation and breath-by-breath end-tidal forcing (11). The resulting preparation then resembles an isolated, thermally stable human brain vascular perfusion system. This can all count as a current hard point.

The measured fast time constant of brain blood flow response, for abrupt lowering of PCO2 , has been reduced by this system to 6.8 from the 20 s derived by (a-v)Do 2sampling thirty years ago (12). The modeling of rapid and overlapping slow components of brain blood flow for hypocapnia and for abrupt restoration of normocapnia all provide baseline descriptions of intrinsic brain vascular system response, with relevance to its normal links with respiration. The dynamic methods should have value in exploring other aspects of CO2roles in brain blood flow regulation, including influences of intrinsic NO (1, 2, 15).

With metabolically formed CO2 as the authoritative transmitter agent in both normal brain blood flow and normal respiratory control, the rapid intrinsic responsiveness of brain circulation to PCO2 is an asset to providing prompt reaction to sudden external influences. It contributes assurance in eupneic rest or exercise of retention of the stable high level of CO2 selected by the central respiratory sensor mechanism for its preferred local H+ environment, as respiratory increases or decreases are gently countered through the opposing influence of the altered arterial PCO2 on brain blood flow. Until substantial acute hypoxia is encountered, brain blood flow in humans can be considered passively responding in opposition to changes in respiration and not separately regulated. In the absence of hypoxia, it would be sensible to consider even large changes from less than to greater than eupneic PCO2 , and the reverse, to represent directional continua of CO2 effect, whether spontaneously induced by slow or rapid changes in ventilation or experimentally imposed. Such changes between “hypercapnia” and “hypocapnia,” being only designations of excess or deficiency of the same CO2 effector, acting on the same site, should be expected to be without a “threshold” in either direction. With PCO2 extensions of severe degree, the effector itself may progressively modify the site and, thereby, modify the action generated.

In the face of well-designed and fortunate interconnections of the influences of respiration and brain blood flow and metabolic gas exchange, the described system for rapidly repetitive measurements should find abundant opportunity to explore, dissect, and define the quantitative dynamic interrelations and their systemic relevance.


  • 1 “Hard point” is an engineering term denoting a point of reinforcement for secure attachment of related structures.