| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Department of Neurosurgery, Weill Medical College, Cornell University, New York, New York; and Departments of Anesthesiology, Neurological Surgery, and Internal Medicine, College of Physicians and Surgeons, Columbia University, New York, New York
Address correspondence to Shailendra Joshi, MD, Department of Anesthesiology, P&S P Box 46, College of Physicians and Surgeons of Columbia University, 630 W. 168th St., New York, NY 10032. Address e-mail to sj121{at}columbia.edu. No reprints will be available.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
Cranial window implants permit the measurement of vasodilator effects of drugs at multiple levels of the cerebral circulation. Topical administration of drugs within cranial windows has often been used to investigate the pharmacological effects of vasoactive drugs. Compared with IV or intraarterial administration, topical administration of drugs permits the investigation of pial arterial responses without the confounding systemic side effects of the drugs. Thus, topical administration is ideally suited for testing a wide range of molar concentrations. As reviewed by Iadecola (6), arterial size seems to affect the pharmacological response to drugs. However, the effect of arterial generation on arterial reactivity has not been reported (7). There is some evidence to suggest that certain characteristics of blood vessels, such as their innervation, may be a function of the arterial hierarchy (8–11).
Therefore, we hypothesized that the topical application of verapamil would increase pial arteriolar diameter independently of the arterial generation and size and, in contrast, that topical nitroglycerin would increase arterial diameter in way that was directly related to the size and inversely related to the arterial generation.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Animals were placed prone in a stereotactic frame supporting the head and body. We used a modified version of the chronic cranial window previously described by Levasseur et al. (13). A trephination (diameter 8 mm) was prepared over the left cerebral hemisphere, the dura mater was reflected, and a custom-built metal chamber with inflow and outflow channels was inserted into the cranial defect. The chamber was covered with a glass disk that was held in place with dental cement. The pial surface was suffused with mock cerebrospinal fluid (CSF) at 0.1 mL/min, and the outflow was adjusted approximately 3–4 cm above the window. The mock CSF contained sodium 150 mM, potassium 3 mM, calcium 1.4 mM, magnesium 0.8 mM, phosphate 1.0 mM, and chloride 155 mM (Harvard Bioscience, Holliston, MA). The dead space of the window and the connectors was between 1.2 and 1.5 mL.
After surgery, we waited for 15 min for the preparation to stabilize before testing it for CO2 response. According to initial ETco2, the response to CO2 was tested by inducing hypercapnia or hypocapnia by altering the respiratory rate to increase or decrease the ETco2 by 10 mm Hg. In our model, there was an excellent correlation between ETco2 and Paco2 (ETco 2 = 10.6 + 0.7 Paco2; n = 35; r = 9; P < 0.0001). Therefore, we relied on ETco2 measurements to adjust minute ventilation. The change in arterial diameters in response to the change in ETco 2 was visually assessed on the video monitor. It corresponded to a 5%–10% change in measured diameters when assessed with videomicroscopy and imaging software, as described below. The ventilatory challenge was used only as a screening test before the preparation was subjected to topical drug challenges. All 10 preparations demonstrated a positive response to ventilation change. After testing for CO2 reactivity, the ventilation was adjusted to maintain ETco 2 of approximately 40 to 45 mm Hg, and the preparation was allowed to rest for approximately 30 min. The preparation was then subjected to topical drugs.
According to a balanced randomization table, animals received superfusion of either nitroglycerin or verapamil. Escalating doses of each drug were used (10–8, 10–6, 10–4, and 10–3 M). Mock CSF and all solutions were warmed to 35°C–36°C in a water bath. For each drug concentration, the window was primed by infusion of the new drug concentration at 30 mL/h for 4 min. This was sufficient to clear the dead space volume of the system, which was estimated to be 1.2–1.5 mL. After the window dead space was cleared, the drug superfusion rate was decreased to 6 mL/h; this ensured a continuous gentle outflow from the window. For each concentration, the drug superfusion lasted for 5 min. Thus, the test for each concentration of drug took 10 min: 1 min to change the syringe, 4 min to prime the cranial window, and 5 min for the steady-state superfusion during which the imaging data were collected. Between the two drug challenges, the brain was superfused with artificial CSF for 30 min or until the arterial diameters returned to near-baseline values. Thus, the experiment lasted for a total of 2.5 h: 40 min for CO2 response and recovery and 110 min for the 2 drug challenges, including 30 min of intervening rest. In theory, both ketamine and propofol have direct and indirect effects on CBF (14–16). The crossover study design and balanced randomization minimized the effects of time-related changes, if any, in preparation.
A custom-built Leitz epi-illumination microscope equipped with a Dage charge-coupled device camera was used for videomicroscopy. Video images were obtained during superfusion of mock CSF before drug superfusion and were then observed in real time and captured every 15 s. The images were magnified 150x and displayed in real time on a 12-in. high-resolution monitor. With this range of magnification, subtle changes (5%–10%) in arterial diameters can be tracked on the screen. In parallel, the images were stored on a Macintosh computer using Global Village imaging software. Eight vessels were selected in each cranial window for measuring arterial diameters. The vessels were chosen on the basis of their relationship to the main branches of the MCA. These main branches can be easily identified on the brain surface as one to three large vessels approximately 150 µm in diameter running from anterior-lateral to posterior-medial. Two main arterial segments were selected and named M-1. As the artery branched, subsequent generations of branching vessels were identified and named M-2, M-3, M-4, and so forth. Two arterial segments were identified for each of the 4 generations. Hence, the diameters of 8 arteries were measured in each window. Thus, in 10 animals, a total of 20 measurements were made for each of the 4 generations, i.e., a total of 80 arterial diameter measurements. Subsequently, these arteries were stratified as large (>60 µm) and small (<60 µm) vessels according to their diameters measured after CO2 challenge. This dichotomy was based on the separation of arteriolar diameters in in vitro studies (7). The stratification based on size was secondary to the arterial generations; therefore, we assessed changes in diameters of 30 large (>60 µm at baseline) and 50 small (<60 µm at baseline) pial arteries.
Unless stated otherwise, all data are given as mean ± sd. Data were analyzed by repeated-measures analysis of variance for within-group comparisons. Between-group (nitroglycerin and verapamil) differences were compared by factorial analysis of variance. Post hoc testing for repeated measures was performed by a Bonferroni-Dunn test. Drug responses were judged by the percentage of change in arterial diameters from predrug measurements and log transformation of the diameters measuring relative changes.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
There was a correlation between the arterial diameter recorded after recovery from the CO2 challenge and the arterial generation (Table 2). No difference in predrug diameters was recorded before superfusion of nitroglycerin or verapamil. Main branch MCA vessels had a diameter of 149 ± 25 µm before nitroglycerin superfusion and 154 ± 41 µm before verapamil superfusion (Table 3). These diameters were comparable to the diameters recorded after the CO2 challenge that are shown in Table 1. At baseline, small arteries (<60 µm; n = 50) had a mean diameter of 34 ± 15 µm before nitroglycerin and 32 ± 14 µm before verapamil, whereas large arteries (>60 µm; n = 30) had a baseline mean diameter of 129 ± 36 µm before nitroglycerin and 131 ± 47 µm before verapamil. It should be noted that there was a slight and statistically insignificant difference between the predrug diameters and baseline (posthypercapnia) measurements, in part because of randomized drug challenges.
|
|
Figure 1 shows the change in arterial diameters after superfusion with verapamil and nitroglycerin. Table 3 shows the percentage of change from predrug arterial diameters in response to topical verapamil and nitroglycerin. The vessels have been classified by both generation (M-1 to M-4) and size (<60 and >60 µm). Collectively for all arterial measurements, superfusion of nitroglycerin even at the largest concentration of 10–3 M caused no significant increase in arterial diameters (predrug 69 ± 50 µm, postdrug 76 ± 60 µm, n = 80). However, subgroup analysis revealed that there was a significant increase in M-1 and large arterial diameters (>60 µm) after exposure to 10–4 and 10–3 M nitroglycerin concentrations (Table 3).
In contrast to nitroglycerin, superfusion of verapamil resulted in a more profound concentration-dependent increase in arterial diameters (predrug 69 ± 57 µm; postdrug 96 ± 71 µm; n = 80; P < 0.001). Collectively, in the 10–3 to 10–8 M range, the concentration-dependent increase in pial arterial diameter was significantly larger for verapamil than nitroglycerin (P < 0.0001). As shown in Figure 2, predrug diameters were comparable before nitroglycerin and verapamil superfusion. Suffusion of 10–3 M nitroglycerin resulted in a minimal increase in arterial diameters. However, a far more pronounced effect at all levels of the arterial tree was evident after 10–3 M superfusion of verapamil. As shown in Table 3, topical verapamil caused a significant dose-dependent increase in arterial diameter across all arterial sizes. Smaller cerebral arterioles were more responsive to verapamil compared with the larger arterioles. At 10–3 and 10–4 M verapamil concentrations, compared with M-1, the M-3 arterioles demonstrated a more robust response to topical verapamil—e.g., 10–3 M verapamil increased the M-1 diameter by 27% ± 19%, whereas the M-3 diameter increased by 60% ± 47% (n = 20 each; P < 0.0087).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In most cranial window studies, pial arteriolar diameter measurements are made on a selected arteriole rather than on multiple vessels. In contrast, cranial window preparations allow the investigation of multiple arteriolar vessels with different diameters simultaneously. By using such a technique, Mchedlishvili (17) demonstrated that during seizures induced by strychnine or after an ischemic event, small pial arterioles (<100 µm) show a much more significant increase in diameter than larger arterioles. The advantage of multiple measurements is the elimination of bias to a particular arteriolar type. Hence, multiple measurements provide more precise information regarding the effect of the drug at different levels of cerebral microcirculation. The limitation of this approach is that the measurements of arteriolar diameters in different segments of the microcirculation may not be truly independent of each other. For example, relaxation of the proximal arteriole may increase the perfusion pressure in the distal arteriole and thus affect its radius. In theory, it could be argued that as long as the autoregulatory capacity of the downstream arterioles is intact, then changes in the downstream arteriolar tone should compensate for perfusion pressure changes due to alterations in tone of the upstream arteries.
Drugs for this study were selected on the basis of their potential clinical usefulness in situations in which manipulation of cerebrovascular resistance and blood flow may be necessary. Topical calcium channel blockers have been shown to cause increases in the diameter of cerebral arteries (18,19). Our results are consistent with observations of other calcium channel-blocking drugs. For example, the response of pial arterioles to the vasodilating effect of the topically applied calcium antagonists nifedipine and nimodipine is more pronounced in small arterioles as compared with large arterioles, especially in those <70–100 µm (20,21). The arteriolar vasodilation observed after superfusion is comparable to the 50% dilation observed by Takayasu et al. (22) in isolated murine cerebral arterioles with a mean diameter of 52 µm.
However, the effect of nitroglycerin on cerebral circulation is not clearly understood. Simultaneous measurements of CBF by single photon emission tomography and transcranial Doppler flow velocity measurements suggest that during IV nitroglycerin infusion, the diameter of the MCA increases but that this does not increase CBF (5). This observation is supported by at least one other study that directly measured the increase in MCA diameter by using magnetic resonance imaging after nitroglycerin infusion (23). When nitroglycerin is directly infused into angiographically normal cerebral arteries of patients, it fails to increase CBF (2). However, in contrast to healthy volunteers, patients with significant cardiac disease undergoing angiography demonstrated a significant increase in blood flow after IV nitroglycerin (24). Taken together, these data suggest that nitroglycerin may not have a profound effect on the distal resistance arterioles; however, nitroglycerin can dilate large pial cerebral arteries such as the MCA.
Contrary to our hypothesis, we failed to observe a robust increase in large pial arterial diameters after topical nitroglycerin. Subgroup analysis of our data revealed that the increase in arterial diameters after topical nitroglycerin was significant only for the M-1 and the large pial arterioles. The relatively selective effect of nitroglycerin on large and small cerebral arterioles remains controversial. Inoue et al. (25) and Kawaguchi et al. (26) found a greater effect on small arterioles, whereas Ishiyama et al. (27) reported a greater effect on larger arterioles. In this study, even at the largest dose, the arterial diameters of M-1 and arteries >60 µm increased by approximately 10%. This effect is considerably less than that reported with topical nitroglycerin by other investigators (25–29). The relative lack of significant vasodilation after topical nitroglycerin could be due to the temperature difference between our study and those by other investigators. The core temperature in our study was 35°C, and the temperature of the superfused drugs was maintained between 35°C and 36°C, which was slightly less than that in the previously cited studies (37°C–38°C). The response to nitric oxide (NO)-dependent vasodilators can be temperature dependent (25,26). It is possible that a slightly lower temperature may have attenuated the response to nitroglycerin in our study.
Our results demonstrate that at an equimolar concentration, between 10–8 and 10–3 M, topical verapamil is a much more powerful vasodilator than nitroglycerin. Our group (2) previously reported the effect of superselective intraarterial injection of nitroglycerin and verapamil on CBF in structurally normal vascular territories of patients undergoing cerebral angiography. Using the xenon-133 technique, we found that at clinically tolerable doses, verapamil significantly increased CBF, whereas nitroglycerin and nitroprusside had no such effect. Despite the difference in administration of both drugs in these studies (intraarterial versus topical), the current experiments suggest that the effect of verapamil on CBF is due to a prominent vasodilatory effect in the cerebral microcirculation, particularly in the small arterioles. More recently, we (4) have developed a method for measuring segmental vascular resistance in human cerebral circulation in response to intraarterial drugs in vivo. Using this approach, we have observed that intraarterial verapamil decreases both proximal and distal cerebral arterial resistance. The differences between intraarterial and topical studies is that topical applications are relatively independent of drug kinetics and systemic side effects. Furthermore, during topical applications, the drugs are applied on the extraluminal surface. Thus, the intraarterial and topical studies complement each other. In case of verapamil, they seem to yield a similar result (4).
One of the more obvious clinical implications of this study could be the use of topical verapamil in treating cerebral vasospasm. Intrathecal NO donors have been used successfully in reversing cerebral vasospasm, but the results of these studies suggest that calcium channel blockade might be superior to the use of NO donors. However, the pathophysiology of cerebral vasospasm might be more complex, and these two drugs will need to be further evaluated in a vasospasm model.
In conclusion, our results suggest that there are significant differences in the pial arterial response to topical verapamil and nitroglycerin in the 10–3 to 10–8 M range. Compared with nitroglycerin, topical verapamil results in a more profound increase in arteriolar diameters that is more significant for the small distal arterioles.
Thanks are due to the following staff members of the College of Physicians and Surgeons of Columbia University: Jodi Wagman and Sulli J. Popilskis, DMV (Institute of Comparative Medicine).
|
Footnotes |
|---|
Accepted for publication September 22, 2004.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Wang and S. Joshi Electrocerebral Silence After Intracarotid Propofol Injection Is a Function of Transit Time Anesth. Analg., June 1, 2007; 104(6): 1498 - 1503. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
)nitroglycerin (ntg; 
) on the percentage increase in arterial diameters compared with predrug measurements. Data from all vessels were pooled (n = 80). Error bars indicate 1 se. *Significant difference from 10–8 M; #significant difference from 10–6 M; @significant difference between drugs.
