Dissociation of Vasoreactivity to Acetazolamide and Hypercapnia
Comparative Study in Patients With Chronic Occlusive Major Cerebral Artery Disease
Background and Purpose The aim of this study was to compare the effect of vasodilative stimuli for the measurement of cerebrovascular reactivity obtained by acetazolamide and hypercapnia in patients with chronic occlusive major cerebral artery disease.
Methods We examined 24 patients with unilateral occlusive lesions of a major cerebral artery using the 133Xe inhalation technique and single-photon emission CT. Regional cerebral blood flow (CBF) was measured during a resting state, during inhalation of 5% CO2, and 15 minutes after the administration of acetazolamide consecutively in the same patients. Normative values of resting CBF and acetazolamide reactivity were obtained in 21 normal subjects.
Results All patients with the exception of 1 showed an increase in CBF during hypercapnia ipsilateral to the occlusive lesion. Ipsilateral acetazolamide reactivity was preserved in 13 patients. Conversely, 11 patients showed an absent response or paradoxical CBF reduction. Ipsilateral CO2 reactivity did not correlate with acetazolamide reactivity when all 24 patients were considered. However, there was a significant correlation between acetazolamide and CO2 in the 13 patients who showed preserved acetazolamide reactivity (r=.60, P<.05). No significant correlation was present in the remaining 11 patients with reduced acetazolamide reactivity. Although significant blood pressure augmentation was observed in hypercapnia, we could not find a correlation between change of blood pressure and CO2 reactivity.
Conclusions Acetazolamide identified patients with reduced vasomotor reactivity who appeared to have preserved CO2 reactivity. Acetazolamide testing may be useful in the assessment of cerebral hemodynamics. However, further investigations are necessary to assess the clinical utility of these tests.
Measurement of cerebral vasoreactivity has been used to evaluate the capacity of perfusion reserve in patients with cerebral ischemia.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Two major cerebral vasodilatory methods, an inhalation of CO21 2 3 4 5 6 8 9 10 11 14 and an intravenous injection of acetazolamide,7 12 13 15 16 have been used to estimate the cerebral vasoreactivity. The mechanism of increase in CBF by hypercapnia has been attributed to either the vasodilatation due to the extracellular/intracellular acidification of vascular smooth muscle cells or the neurogenic mechanism through the stimulus of the vasomotor center of the brain stem.16 Meanwhile, acetazolamide selectively inhibits a carbonic anhydrase in erythrocytes, glial cells, capillary endothelium, and choroid plexus.17 Severinghaus and Cotev19 showed acetazolamide-induced extracellular acidosis even though the Pco2 was kept constant by moderate hyperventilation. The CBF increase that follows acetazolamide injection is dose dependent, and the effect of 1 to 2 g of acetazolamide is of similar magnitude as that observed with the inhalation of 7% CO2.16 20 It has been suggested that the magnitude of extracellular acidification derived from acetazolamide can quantitatively explain the increase in CBF.21 22 Therefore, in normal subjects a similar amount of elevation in CBF is expected when the same degree of extracellular acidification is obtained both by hypercapnia and by acetazolamide injection.
Although the benefit of extracranial-intracranial arterial bypass surgery is still arguable,23 several investigators have suggested that vascular reconstructive surgery is beneficial, in particular for patients with reduced perfusion reserve.2 7 24 Since then, the importance of appropriate evaluation of perfusion reserve has been stressed for the selection of a therapeutic strategy.2 5 7 8 15 16 24 However, the relationship between acetazolamide reactivity and CO2 reactivity has not been fully explored in a pathological condition, such as in cerebral tissues that may have reduced perfusion reserve stemming from an occlusive lesion located in the major cerebral arteries.24
In this study, we compared the CBF measurements of vasoreactivity to acetazolamide and to induced hypercapnia in patients with occlusive cerebrovascular disease. We performed quantitative CBF measurements using a 133Xe inhalation method and SPECT, both with induced hypercapnia and acetazolamide injection, consecutively in the same patients.
Materials and Methods
We studied 24 subjects with unilateral occlusive lesions of the major cerebral artery (17 men and 7 women; age range, 52 to 76 years [mean, 60 years]). These subjects were recruited from those referred to the Hokkaido Neurosurgical Memorial Hospital in the course of assessment for vascular reconstructive surgery.
Fifteen patients showed minor neurological deficits, typically mild hemiparesis, related to previous ischemic attack. A reversible ischemic neuronal deficit was manifested in 2 patients. Seven patients had episodes of transient ischemic attack without fixed neurological deficit before the study. Patients with moderate to severe neurological deficit were excluded from this study. All 24 patients underwent conventional cerebral angiography involving three or four vessels that revealed unilateral occlusive lesion at the cervical ICA or horizontal portion of the MCA. Angiography showed a severe stenosis (>90%) of the cervical ICA in 4 patients. In these patients, intracranial collateral formation was not visualized on angiograms. In 5 patients, the angiograms showed an occlusion of the horizontal part of the MCA. In all of these cases, a collateral formation was revealed by a retrograde filling of MCA branches through the anterior or posterior cerebral artery via leptomeningeal anastomosis. A complete occlusion of the cervical ICA was seen in 15 patients. In these patients, the sources of collateral flow were ophthalmic, circle of Willis, or leptomeningeal anastomosis. Although an irregularity of the contralateral cervical ICA was observed in 4 of 24 patients, these findings were not of hemodynamic significance. None of the patients examined in this study presented angiographically detectable stenosis of the major cerebral artery on the contralateral side of the lesion. All patients underwent CT or MRI. To evaluate the hemodynamics in the noninfarcted area, we excluded patients with an infarct measuring >2 cm in diameter. However, patients with lacunae at the level of the basal ganglia or centrum semiovale were included in this study. The clinical features, CT/MRI, and angiographic findings are summarized in Table 1⇓.
All patients were examined at a point when their neurological status was stable and no sooner than 4 weeks subsequent to their last ischemic episode. We used the Headtome Set-031 (Shimadzu Co). The spatial resolution was 20 mm in the plane and 24 mm axially, measured as full width at half maximum. rCBF was calculated by the “sequential picture method” described by Kanno and Lassen.26 Each subject had three CBF measurements in succession that same day: (1) at the resting state, (2) during the induced hypercapnia (CO2 CBF), and (3) at 15 minutes after an intravenous injection of acetazolamide (10 mg/kg). The hypercapnia was induced by inhalation of 5% CO2 with 95% O2, which began at 3.5 minutes before the CBF measurement and continued throughout the CBF measurement. To avoid the possibility of induced hypercapnia influencing the upcoming acetazolamide test, a quiet interval of more than 1 hour was scheduled between the second CBF measurement and the scheduled injection of acetazolamide. Subsequently, each patient's head was accurately positioned using line markers on the patient's face. Images were obtained of three transverse brain slices situated at 2.0, 5.5, and 9.9 cm above and parallel to the orbitomeatal line. A 12.8-cm2 circular region of interest was placed on the second slice (5.5 cm) to quantify rCBF in the frontotemporal cortical area, which represents MCA perfusion territory.
Arterial blood was sampled to measure the Paco2 in each CBF measurement. In addition, we measured BP by auscultation two times for each CBF measurement. The average values of mean BP in each CBF measurement were used to assess the change of mean BP.
The Paco2 and mean BP at the resting state were compared with those of induced hypercapnia and those after intravenous injection of the acetazolamide.
The comparison of the effect of vasodilators was performed in two different ways: (1) within-subject assessment to evaluate the different effect of two vasodilators and (2) correlation analysis to assess the relationship between the parameters obtained by both acetazolamide and CO2 in each patient.
The first analysis was based on the hypothesis that if the effects of acetazolamide and induced hypercapnia were equivalent, they should have induced comparable increases in CBF in the hemisphere ipsilateral to the occlusive lesion. To further improve data analysis, we defined the asymmetry ratio as absolute CBF changes in ipsilateral hemisphere/contralateral hemisphere.
Assuming the same effect of both vasodilators on the contralateral hemisphere, changes of the asymmetry ratio will express differences in the response to acetazolamide and hypercapnia in the ipsilateral hemisphere.
The vasoreactivity to acetazolamide (V[AZ]) was calculated as follows: V(AZ)=100*(AZ rCBF−resting rCBF)/resting rCBF.
On the other hand, CO2 reactivity can be calculated using two methods1 2 3 4 5 6 8 9 10 11 28 29 : (1) as absolute change in CBF per millimeter of mercury of Pco2 (mL·100 g−1·min−1·mm Hg−1 Pco2) and (2) as percent change in CBF over the baseline CBF per millimeter of mercury of Pco2 (%/mm Hg Pco2). In a physiological range of Paco2 (25 to 60 mm Hg), a linear relationship between the increase of CBF and change of Paco2 is assumed.28 In the present study, we used percent change of CBF per mm Hg Paco2 as an index of CO2 reactivity to compare with acetazolamide reactivity, which also reflects the percent change over the baseline. Thus, CO2 reactivity (V[CO2]) was calculated as follows: V(CO2)=(100*[CO2 rCBF−resting rCBF]/resting rCBF)/ΔPaco2. Normal values of the resting rCBF (mean±SD, 43.1±3.0 mL·100 g−1·min−1) and the vasoreactivity to acetazolamide (20.3±5.3%) were obtained from 21 age-matched normal volunteers with a mean age of 58.5±8.6 years (range, 45 to 78 years).
The resting rCBF and acetazolamide reactivities were rated as reduced when the values were less than 2 SDs below the normative values.
Mean comparisons were performed using Student's t test. Correlation analysis was performed using Pearson product-moment correlation coefficients.
The relationship between resting CBF and acetazolamide reactivity was assessed by a χ2 test. We also explored the possible influence of BP augmentation and resting CBF on the vasoreactivity to hypercapnia using multiple linear regression.
Physiological Data: Change of Paco2 and Mean BP
A summary of the change of Paco2 and mean BP is given in Table 2⇓. The mean increase of Paco2 in induced hypercapnia was 5.60±1.88 (1 SD) mm Hg, and the mean increase in mean BP was 9.07±6.32 (1 SD) mm Hg. Mean BP in hypercapnia was significantly higher than BPs during the resting state and after acetazolamide injection (P<.001; t test). There was no significant change in Paco2 or in mean BP after the injection of acetazolamide compared with the resting state.
Table 3⇓ shows the change of Paco2 and mean BP during CO2 inhalation, resting rCBF, and absolute increase in CBF and percent increase over the baseline in each stimulation in the 24 patients. Both percent and absolute CBF increase in CO2 were normalized by the change of Paco2. Resting CBF measurements showed normal blood flow in 11 patients and reduced blood flow in 13 patients. In the studied patient population, ipsilateral resting CBF was significantly lower than contralateral resting CBF (P<.01, paired t test). Twenty-three patients showed CBF increases during hypercapnia in both hemispheres (ipsilateral, 1.80%/mm Hg to 12.5%/mm Hg; contralateral, 2.11% to 10.46%). One (case 20) showed poor response to hypercapnia in both hemispheres (ipsilateral, 0.71%/mm Hg; contralateral, 2.11%/mm Hg). Acetazolamide reactivity was considered to be preserved in the contralateral hemisphere in all patients (11.3% to 36.7%). Ipsilateral acetazolamide reactivity was preserved in 13 patients. Conversely, the remaining 11 patients showed impaired acetazolamide reactivity. Eight of the 11 patients showed an absent response or paradoxical CBF reduction (presumably intracerebral steal phenomenon) ipsilateral to the occlusive lesion. Another 3 cases (cases 19, 21, and 23) revealed reduced response to acetazolamide (2.9% to 6.7%). In this study, 9 of 11 (73%) patients with normal resting CBF exhibited preserved vasoreactivity to acetazolamide, whereas 9 of 13 (69%) patients with reduced resting CBF exhibited reduced acetazolamide vasoreactivity; therefore, there is a significant relationship between resting CBF and vasoreactivity to acetazolamide (P<.05; χ2 test).
The scatter diagram of the ipsilateral-to-contralateral ratio for absolute changes of CBF represents the different effect of acetazolamide and induced hypercapnia in the ipsilateral side in each patient (Fig 1⇓). Thirteen of 24 patients with preserved acetazolamide reactivity (cases 1 through 13) had comparable CBF increases in acetazolamide and CO2. Although the CBF increases were comparable in these subjects, the asymmetry ratio was significantly lower for acetazolamide than CO2 (P<.05; t test). In patients with reduced acetazolamide reactivity (case 14 to 24), the asymmetry ratio was also significantly lower for acetazolamide than CO2 (P<.001; t test). Except for case 20, these patients showed relatively preserved CO2 response but negative response to acetazolamide. One patient (case 18) in this group showed more than a twofold increase in ipsilateral CBF versus contralateral CBF.
In patients with occluded ICAs, collateral formation through the circle of Willis was predominantly observed in patients with normal resting flow and preserved acetazolamide reactivity, whereas ophthalmic or leptomeningeal collateral formations were frequently shown in the other subgroups. Although we excluded patients with extensive infarction, white matter lesions in the centrum semiovale were frequently observed in our patients. However, neither the occlusive lesion site nor the clinical status correlated with resting CBF, acetazolamide, or CO2 reactivity.
Absolute CBF change after the administration of acetazolamide or CO2 was not correlated with resting CBF in either hemisphere.
Fig 2⇓ shows the vasoreactivity relationship between acetazolamide and CO2 in the ipsilateral hemisphere. Data for all 24 subjects did not show a significant correlation between values for acetazolamide and CO2 vasoreactivities in either the ipsilateral or the contralateral hemisphere. However, in 13 patients with preserved acetazolamide reactivity, there was a significant correlation between vasoreactivity to acetazolamide and to CO2 (r=.60, P<.05) in the ipsilateral hemisphere. In patients with reduced acetazolamide reactivity, there was no significant correlation between acetazolamide and CO2 vasoreactivities.
Multiple regression was used to estimate the effect of resting CBF, ΔPaco2, and change of BP to predict CBF in hypercapnia. The multiple R2 was .90 (P=.0001), indicating that the model accounts for 90% of the variance in CBF in hypercapnia. The equation used to estimate CBF after CO2 was CBFCO2=−0.78+1.03 CBFrest+1.45 Paco2>2+0.09 BP.
Resting CBF and Paco2 were significant predictors (P=.0001 and P=.004, respectively); however, BP did not make a significant contribution to the model (P=.50).
Fig 3⇓ shows the illustrative case of a 60-year-old man (case 16) with chronic stage left MCA occlusion.
Our results indicate that patients with chronic occlusive major cerebral artery disease may show preserved CO2 reactivity but absent or inverse response to acetazolamide. In our study, only patients with no major neurological signs, no extensive cerebral infarction, and unilateral major cerebral artery occlusive disease were included. Ringelstein et al25 reported a significant association between acetazolamide and induced hypercapnia by measuring blood flow velocity using transcranial Doppler sonography. However, the use of transcranial Doppler may not be as sensitive as our direct SPECT measurements in the assessment of CBF changes. In addition, inclusion of patients with major neurological deficit and bilateral occlusion in that study may have contributed to the above-mentioned differences.
Our patient population with preserved acetazolamide response did show a significant vasoreactivity association with CO2 reactivity. This finding suggests that the compatibility may be maintained above a certain threshold of acetazolamide reactivity and that the presence of a correlation between acetazolamide and CO2 may be highly dependent on the patient selection. We acknowledge that our patient population may not be entirely representative of the general population with occlusive cerebrovascular disease. However, considering that approximately 50% of our patients presented impaired acetazolamide reactivity, we do not think that simple measurement error was responsible for our findings.
In the present data set, we found a large variability in absolute vasoreactivity values with both vasodilators. However, large individual variability in acetazolamide and CO2 response has been already reported by previous investigators.13 27 28 29 For this reason, we have also used side-to-side asymmetry ratio to compare the effect of these vasodilators on CBF. It is understood that the asymmetry ratio of vasoreactivity cannot reflect actual cerebral hemodynamics because the evaluation of asymmetry can overlook bilaterally compromised hemispheres.
Although our study was primarily designed to compare the utility of acetazolamide and hypercapnia in the screening of hemodynamically compromised patients, these results suggest that a different physiological effect may underlie the response to these two vasodilators distal to the occlusive lesion.
If the maximal vasodilatation occurred in the region with absent response or paradoxically fell in response to acetazolamide, the increase in CBF with hypercapnia may not merely stem from hypercapnic vasodilative stimuli. Hypercapnia leads to hypertension, as demonstrated by the mean BP increase of approximately 10 mm Hg in our subjects. This finding is also consistent with previous reports.29 Although our data do not show a significant association between the change of BP and CBF increase in hypercapnia, we suggest that in subjects with severely impaired perfusion reserve, in particular those with decreased resting CBF, part of the increase in CBF may stem from elevated BP in hypercapnia.
One may argue that, since the CBF increase following the injection of acetazolamide is dose dependent,17 21 the brain delivery of acetazolamide may be decreased proportionally to the degree of ischemia. However, in view of the existence of patients with normal resting flow and reduced acetazolamide reactivity,7 12 13 we believe that the contribution of ischemia in explaining our results is negligible.
Another possibility may be that the time delays in reaching the maximum effect of acetazolamide may cause an overestimation of decreased cerebrovascular reactivity. In normal subjects, the CBF increase starts 2 minutes after acetazolamide injection, attaining maximal effect at 20 minutes, which lasts for 30 minutes after injection.16 However, Hayashida et al30 investigated the cerebrovascular reactivity to acetazolamide using [15O]H2O PET in patients with ICA occlusion or stenosis and reported that the maximal response was observed at 10 minutes after injection. More recently, Kuwabara et al31 reported that the percent increase in CBF at 5 minutes was lower than at 20 minutes after intravenous acetazolamide in the territory with chronic cerebrovascular occlusive lesion, and they emphasized that the steal phenomenon was more evident at 5 minutes after injection. Although these recent studies suggest a delayed effect of acetazolamide in the region with hemodynamic fluctuation, a 15-minute duration before the CBF measurement (as in our study) can be considered adequate to obtain a consistent response to acetazolamide.
In a clinical setting, acetazolamide is more easily administered than CO2 and does not change arterial CO2 tension or significantly alter BP. According to our results and to previous reports, differences in CO2 reactivity between normal subjects and subjects with reduced reactivity are relatively small.2 8 The large variability of individual CO2 reactivity values also makes identification of specific subgroups difficult. It is important that in our study all subjects except 1 showed as great, or even greater, an increase in CBF in the ipsilateral as in the contralateral hemisphere during hypercapnia. This finding may have several different explanations. First, although the utility of the measurements of CO2 reactivity has been reported by previous investigators,1 2 3 4 5 6 8 9 10 11 14 our results suggest rather that induced hypercapnia may not always provide adjunctive information to resting CBF measurements. Second, this phenomenon may reflect a lack of autoregulatory vasoconstriction in response to augmented BP.
The purpose of vasoreactivity measurements may be to identify the reduced perfusion reserve, which is not suggested by the existence of watershed-zone infarct or poor angiographic collateral formation. Also, a reduction of resting flow can result from either failure of autoregulatory compensation or reduced metabolic demand. In our study, 4 of 13 patients showed reduced resting flow but normal acetazolamide reactivity. This is in keeping with results of other investigators.7 12 13 However, the finding of reduced resting flow in these patients may express decreased metabolic demand. We think that this patient subgroup has reduced metabolic demand but preserved vasoreactivity, as also has been suggested by evaluation of oxygen extraction fraction in PET studies.7 12 13 15 16
In conclusion, the clinical significance of the reduced acetazolamide reactivity should be further validated, in particular by exploring its correlation with hemodynamic parameters obtained by PET and the association with long-term follow-up.
Selected Abbreviations and Acronyms
|(r)CBF||=||(regional) cerebral blood flow|
|ICA||=||internal carotid artery|
|MCA||=||middle cerebral artery|
|PET||=||positron emission tomography|
|SPECT||=||single-photon emission CT|
|V(AZ)||=||vasoreactivity to acetazolamide|
|V(CO2)||=||vasoreactivity to CO2|
The authors thank the technical staff of the Department of Radiology, Hokkaido Neurosurgical Memorial Hospital, for data acquisition and Francine Mandel, PhD, for statistical analysis.
- Received December 29, 1995.
- Revision received August 1, 1996.
- Accepted August 1, 1996.
- Copyright © 1996 by American Heart Association
Dyken ML, Campbell RL, Frayser R. Cerebral blood flow, oxygen utilization, and vascular reactivity. Neurology. 1970;20:1127-1132.
Gibbs JM, Wise RJS, Leenders KL, Jones T. Evaluation of cerebral perfusion reserve in patients with carotid-artery occlusion. Lancet. 1984;11:310-314.
Herold S, Brown MM, Franckowiak RSJ, Mansfield AO, Thomas DJ, Marshall J. Assessment of cerebral hemodynamic reserve: correlation between PET parameters and CO2 reactivity measured by the intravenous 133 xenon injection technique. J Neurol Neurosurg Psychiatry. 1988;51:1045-1050.
Levinne RL, Sunderland JJ, Lagreze HL, Nickles RJ, Rowe BR, Patrick A, Turski PA. Cerebral perfusion reserve indexes determined by fluoromethane positron emission scanning. Stroke. 1988;19:19-27.
Norrving B, Nilsson B, Risberg J. rCBF in patient with carotid occlusion: resting and hypercapnic flow related to collateral pattern. Stroke. 1982;13:155-162.
Olsen TS, Larsen B, Herning M, Skriver EB, Lassen NA. Blood flow and vascular reactivity in collaterally perfused brain tissue. Stroke. 1983;14:332-341.
Powers WJ, Press GA, Grubb RL, Garo M, Raichle ME. The effect of hemodynamically significant carotid artery disease on the hemodynamic status of the cerebral circulation. Ann Intern Med. 1987;106:27-35.
Rogg J, Rutiglino M, Yonas H, Johnson DW, Pentheny S, Latchaw R. The acetazolamide challenge: imaging techniques designed to evaluate cerebral blood flow reserve. AJNR Am J Neuroradiol. 1989;10:803-810.
Tsuda Y, Kimura K, Yoneda S, Etani H, Asai T, Nakamura M, Abe H. Comparison between TIAs due to the ICA occlusion and ICA mild stenosis. Neurol Res. 1983;106:197-222.
Volstrup S, Brun B, Lassen NA. Evaluation of cerebral vasodilatory capacity by the acetazolamide test before EC/IC bypass surgery in patients with occlusion of the internal carotid artery. Stroke. 1986;17:1291-1298.
Volstrup S. Tomographic cerebral blood flow measurements in patients with ischemic cerebrovascular disease and evaluation of the vasodilatory capacity by the acetazolamide test. Acta Neurol Scand. 1988;(suppl 114):vol 77.
Severinghaus JW, Cotev S. Carbonic acidosis and cerebral vasodilation after diamox. Scand J Clin Lab Invest. 1968;22(suppl 102):1E.
Volstrup S, henriken L, Paulson DB. Effect of acetazolamide on cerebral blood flow and cerebral metabolic rate of oxygen. J Clin Invest. 1984;74:1634-1639.
Bickler PE, Litt L, Banville PL, Severinghaus JW. Effect of acetazolamide on acid-base balance. J Appl Physiol. 1988;65:422-427.
Kanno I, Lassen NA. Two methods for calculating regional cerebral blood flow from emission computed tomography of inert gas concentrations. J Comput Assist Tomogr. 1992;3:71-76.
Davis SM, Ackerman RH, Correia JA, Alpert NM, Chang J, Buonanno F, Kelley RE, Rosner B, Taveras JM. Cerebral blood flow and cerebrovascular CO2 reactivity in stroke-age normal controls. Neurology. 1983;33:391-399.
Reivich M. Arterial PCO2 and cerebral hemodynamics. Am J Physiol. 1964;206:25-35.
Hayashida Y, Hirose Y, Ishida Y, Nishimura T. Visualization of drug reaction with acetazolamide using sequential PET O-15 measurement. J Nucl Med. 1995;36:242. Abstract.
Kuwabara Y, Ichiya Y, Sasaki M, Yoshida T, Masuda K. Time dependency of the acetazolamide effect on cerebral hemodynamics in patients with chronic occlusive cerebral arteries: early steal phenomenon demonstrated by [15O] H2O positron emission tomography. Stroke. 1995;26:1825-1829.