METHODS

The animal protocols used were reviewed and approved by the Animal Care and Use Committee of the University of Texas at San Antonio. All animals were maintained in a facility accredited by the American Association for Accreditation of Laboratory Animal Care. Farm-bred newborn piglets (1–14 days old, weighing 1.5–3.0 kg) were anesthetized initially with ketamine hydrochloride (33 mg/kg im) and acepromazine (3.3 mg/kg im) and were maintained on α-chloralose (30–40 mg/kg iv initially, supplemented with 5–7 mg ⋅ kg⁻¹ ⋅ h⁻¹). Catheters were inserted into the femoral vein for maintenance of anesthesia and administration of medications and fluid, and into the femoral artery to record systemic blood pressure and sample arterial blood gases and pH. Animals underwent tracheostomy with insertion of an endotracheal tube (3.5-mm internal diameter) and were mechanically ventilated with room air. Core temperature was monitored with a rectal probe and was maintained between 37.5 and 38.5 °C via a warming blanket.

Cranial Window Placement

For insertion of the closed cranial window, the scalp was surgically removed with a cauterizing scalpel, and an opening was made in the skull over the parietal cortex. Incisions were made in the dura, which was retracted over the cut bone edge. A stainless steel and glass cranial window was fixed in place with dental acrylic. The space under the window was filled with artificial cerebrospinal fluid (aCSF) containing (in mg/l) 220 KCl, 132 MgCl₂, 221 CaCl₂, 7,710 NaCl, 401 urea, 665 dextrose, and 2,066 NaHCO₃ (pH = 7.33, $P{\text{CO}}_2$ = 46 mmHg,$P{O}_2$ = 43 mmHg). The fluid was maintained at body temperature. Pial arterioles (26–38 μm) were directly observed via a dissecting microscope with a mounted video camera. Vessel diameter was measured with a video microscaler.

Experimental Design

Characterization of postictal pial arteriolar responses.

All animals underwent an initial observation period of at least 30 min postinstrumentation with recording of baseline parameters including pial arteriolar diameter, mean arterial blood pressure (MAP), core temperature, and arterial blood gases and pH. These parameters were recorded for each tested response at 1, 3, and 5 min. Preseizure control responses were then quantified for hypercapnia (ventilation with a 10% CO₂-20% O₂-70% N₂ gas mixture), hypoxia (ventilation with a 10% O₂-90% N₂ gas mixture), and sodium nitroprusside (10⁻⁶ M). Hypercapnia and hypoxia were maintained for 5 min, and maximal vessel diameter was recorded. Sodium nitroprusside was topically applied under the cranial window and maintained for 5 min with the recording of maximum vessel diameter. At the end of each tested response, the area under the cranial window was gently flushed with aCSF to remove the previous stimulus and allow the vessels to return to baseline diameter.

Seizures were induced with intravenous injection of bicuculline (1 mg/kg), a GABA-receptor blocking agent. Bicuculline (Sigma, St. Louis, MO) was dissolved in 1 N HCl and brought to a pH of 6.0–6.5 with 1 N NaOH to a concentration of 1 mg/ml. The onset and termination of clonic movements defined the duration of the seizures. The cranial window was flushed with aCSF following seizure activity in all piglets before further testing. Maximum pial vessel dilation response to hypoxia, hypercapnia, and sodium nitroprusside was repeated in the immediate postictal period (∼10 min). Additional measurements were done at 60, 90, and 120 min in response to both hypoxia and sodium nitroprusside or until vessel dilation with hypoxia returned to preictal baseline.

Role of adenosine in altered responses.

A dose-response curve to determine the dilatory effects of topical adenosine was performed with concentrations of adenosine (Sigma) ranging from 10⁻¹² to 10⁻⁶ M as background data to this series of experiments. Pial vessel diameter and MAP were serially recorded during each 10-min exposure of each concentration tested. Between each tested concentration, the pial surface was flushed with aCSF, allowed to return to baseline diameter, and then observed for another 10-min period before another adenosine concentration was administered.

An initial group (group I) of newborn pigs was similarly prepared as described in Characteristics of postictal pial arteriolar responses. After preseizure control responses to both hypercapnia and hypoxia (5-min exposure) with recovery to baseline were measured, a dilating dose (as confirmed by the dose-response curve) of topical adenosine (10⁻⁵ M) was administered topically to the pial surface for a 10-min period of exposure with recording of vessel diameter at 1, 3, 5, and 10 min. The window was then gently flushed with aCSF, and the pial surface was allowed to recover for a 10-min period. Seizure activity was then induced with bicuculline as before, and pial arteriolar vessel response to hypoxia was again evaluated in the immediate postictal period. Topical adenosine (10⁻⁵ M) was then readministered for a 10-min period. The window was again gently flushed with aCSF, and the pial surface was allowed to recover for a 10-min period. The hypoxic stimulus was then repeated within 60 min postictal.

In a separate group (group II) of newborn pigs, the preseizure control response to both hypoxia and hypercapnia was again initially assessed to determine baseline vessel response. This was compared with the immediate postictal hypoxic response following bicuculline-induced seizures. Adenosine (10⁻¹¹ M) at a nondilating concentration (>2 logarithms less than the threshold dilator concentration) was then topically applied to the pial surface. After 10 min of exposure to the adenosine, the hypoxic stimulus was repeated with adenosine still present. As for thegroup I adenosine experiment, all postictal responses were completed within the first 60 min postictal.

Statistical Analysis

Percent change in pial vessel diameter, systemic arterial pressure, and arterial blood gas values were compared by analysis of variance (ANOVA) for repeated measures or by unpairedt-test. Comparisons were made between control responses measured before seizure activity and postictal responses at determined time points. If the value was significant when compared by ANOVA, a Tukey-Kramer post hoc test was performed.P < 0.05 was considered significant. All measurements are presented as means ± SE of absolute values or as means ± SE of percent change from baseline values.

RESULTS

The arterial blood gases, pH, and MAP measurements recorded at baseline and in response to hypoxia, hypercapnia, and sodium nitroprusside (10⁻⁶ M) before seizure induction are shown in Table 1. Postseizure values are shown in Table 2. The preictal MAP was similar to control except during hypoxia, during which systemic pressure decreased (94 ± 3 to 63 ± 7 mmHg). Systemic blood pressure in the postictal period did not vary significantly from control during any of the dilator stimuli including hypoxia (Table 2). MAP did increase in the immediate postictal period versus preseizure control (94 ± 3 to 102 ± 4 mmHg), although the change was not statistically significant (P > 0.05).

Both pre- and postseizure values for arterial blood gases and pH were similar to control values except where predicted during exposure to hypoxic and hypercapnic stimuli. Preseizure$P{O}_2$ decreased during 5 min of ventilation with fractional inspired O₂ of 0.10. Similar findings occurred postictally with $P{O}_2$significantly decreasing from control after 10, 60, 90, and 120 min of hypoxia. Both pH and $P{\text{CO}}_2$ did not vary in response to hypoxia throughout the duration of the postictal period; however, pH did decrease predictably during hypercapnia with an increasing $P{\text{CO}}_2$.

Topical sodium nitroprusside, as expected, had no effect on arterial blood gas values and systemic pressure both preseizure and in the immediate and subsequent postictal period. The mean seizure duration was 13 ± 1 min. The majority of piglets had significant vasodilation of pial arterioles with seizure activity; however, it was not possible to consistently measure dilation accurately due to cranial window movement during clonic activity.

Hypercapnia and Sodium Nitroprusside Responses

Pial arteriolar dilation was not significantly altered in the immediate postictal period versus preseizure control values in response to hypercapnia (53.0 ± 5 vs. 55.0 ± 6%,P > 0.05; Fig.1). Similarly, vessel dilation to sodium nitroprusside (10⁻⁶ M) remained consistent both preseizure and throughout the postictal period measured both immediately and at 60 and 90 min (60.0 ± 6, 67.1 ± 9, 62.0 ± 6, and 61.2 ± 7%, respectively;P > 0.05).

Hypoxia Response

Preseizure vessel response to hypoxia resulted in consistent pial arteriolar vasodilation in newborn piglets (56.5 ± 6%; Fig.2). In the immediate postictal period (∼10 min) following bicuculline-induced seizure activity, pial vasodilation in response to hypoxia was significantly attenuated (6.3 ± 2%, P < 0.001). In addition, this altered response persisted into the postictal period measured at 60 and 90 min (21.7 ± 6 and 21.6 ± 5%, respectively;P < 0.01). Pial arteriolar response returned to near preseizure control dilation by 120 min (37.0 ± 9%), but the difference did not reach statistical significance (P > 0.05).

Adenosine

Before a dose-response curve to topical adenosine was determined, pial arteriolar dilation consistent with previous responses was noted in response to both hypercapnic and hypoxic stimuli. Figure3 illustrates a dose-response curve for topically administered adenosine. A topical adenosine concentration of 10⁻⁸ M was the threshold dilator concentration. MAP and blood gas analysis did not change during exposure to adenosine.

Preictal response in both group I(10⁻⁵ M) andgroup II(10⁻¹¹ M) adenosine-treated piglets resulted in predicted control vessel dilation to hypercapnic (57.0 ± 3 and 54.4 ± 2%, respectively) and hypoxic stimuli (46.1 ± 7 and 51.0 ± 8%, respectively; Figs.4 and 5). In addition, significant attenuation of the dilatory response to hypoxia occurred in the immediate postictal period in both groups [5.9 ± 2% (P < 0.001) and 11.7 ± 6% (P < 0.01), respectively; Figs.4 and 5]. In group I, pre- versus postictal vessel dilatory response to adenosine (10⁻⁵ M) was unaltered by bicuculline-induced seizure activity (18.7 ± 2 and 18.2 ± 3%, respectively; P = 0.32; Fig.6). Postictal dilatory response to hypoxia in group I was restored to control dilation following exposure to adenosine (10⁻⁵ M) in the immediate postictal period (43.3 ± 7%, P> 0.05; Fig. 4). Similarly, in group II piglets repleted with a nondilating dose of adenosine (10⁻¹¹ M), pial arteriolar dilation to hypoxia was also restored (46.7 ± 8%,P > 0.05; Fig. 5).

DISCUSSION

The new findings from the present study in newborn pigs are that1) pial arteriolar dilatory responses to hypercapnia and nitric oxide (NO) donors are not altered in the postictal period following bicuculline-induced seizure activity;2) hypoxia-induced cerebral vasodilation is significantly impaired in the postictal period;3) acute depletion of available adenosine appears to play a role in the impairment of hypoxia-induced cerebral vasodilation in the immediate postictal period; and4) adenosine has a permissive role in hypoxia-induced cerebral vasodilation.

CBF autoregulation is impaired in the postictal period in both perinatal animal models and human newborns. Pressure autoregulation is lost similarly following perinatal asphyxia. In addition, asphyxia impairs cerebral vasodilation to both hypercapnia and hypoxia depending on the severity of the insult (26, 27, 29). The loss of CBF autoregulation coupled with poor vessel reactivity to hypercapnia, as well as hypoxia, has been correlated with poor neurological outcome in asphyxiated neonates (26, 27). Cerebral vasodilatory reactivity to hypoxia has been shown to be preferentially preserved compared with hypercapnic dilation, although the loss of autoregulation precedes both (26, 36). In contrast to that in asphyxiated infants, we observed in our study that postictal pial arteriolar reactivity to hypercapnia remains intact, whereas vasodilation to hypoxia is significantly altered.

The mechanisms involving altered cerebral autoregulation following seizures, perinatal asphyxia, or other insults remain unclear. It is thought that, in the asphyxiated human newborn, cerebral vasoparalysis may be secondary to a state of maximum vasodilation resulting from elevated perivascular H⁺concentration or release of other factors such as adenosine, prostaglandins, and/or free radicals (3, 14, 26, 36). Studies using Doppler ultrasound in asphyxiated neonates have demonstrated cerebral hyperemia secondary to vasodilation coupled with decreased cerebral vascular resistance persisting for as long as several days postinjury (1, 3, 14, 34). Monin et al. (18), however, noted return of CBF to baseline within 30 min following seizure activity, thus suggesting that vasodilation did not contribute to the loss of autoregulation. In our study, pial arteriolar diameter also returned to preseizure control values in the immediate postictal period. Seizure duration was relatively short in our piglets (<15 min). It is possible that a longer duration of seizure activity causing significant metabolic acidosis may have resulted in persistent cerebral vasodilation similar to that observed in the human newborn postasphyxial state. Isolated metabolic acidemia with increased perivascular H⁺ concentration in several newborn animal studies has been shown to alter autoregulation (5, 19).

Hypercapnia-induced cerebral vasodilation in newborn animal models occurs via a permissive role of vasoactive prostanoids released by the endothelium in response to increased perivascular H⁺ concentration (12, 37). Given the unaltered pial arteriolar dilatory response to hypercapnia in piglets in the present study, it appears that this mechanism of endothelial prostanoid release and subsequent cerebral vasodilation with increased CO₂ remains intact following bicuculline-induced seizures. Cerebral vasodilation to topical sodium nitroprusside remained consistent with the preseizure dilatory response throughout the measured postictal period (90 min). Nitrovasodilators such as sodium nitroprusside release NO that stimulates soluble guanylate cyclase in the vascular smooth muscle cell, resulting in relaxation (17). An intact dilator response to sodium nitroprusside demonstrates that vascular smooth muscle response to a NO donor remains unaltered postictally.

The results of the present study demonstrate a marked reduction in cerebral vasodilation with the induction of hypoxia in the immediate postictal period. This altered response persisted 90 min into the postictal period and showed a trend toward incomplete dilation versus control at 120 min, although the difference was not statistically significant. Whereas pH and $P{\text{CO}}_2$ were unchanged during the 5 min of exposure to hypoxia preseizure, MAP was moderately diminished during this period, possibly contributing to the observed vasodilation. However, baseline dilation was similar to that observed in other studies with stable or minimally changed blood pressure during hypoxia in newborn pigs (2, 13). Furthermore, the moderate decrease in systemic blood pressure that we observed is within the expected range of autoregulation and should not have altered CBF in otherwise healthy piglets (11).

The mechanisms that produce cerebral vasodilation in response to hypoxia remain incompletely understood. Multiple mediators and dilator systems have been proposed to be involved in this process, including adenosine. Adenosine is considered to be an important mediator in both the adult and newborn cerebral circulation, and brain adenosine levels are known to increase during ischemia and hypoxia in several animal models (4, 16, 21, 30, 37, 38, 39). Additional studies have suggested that adenosine has a role in the mediation of dilator responses to hypotension and insulin-induced hypoglycemia as well in newborn pigs (11, 31). Cerebral blood flow is known to increase significantly during seizure activity in newborn animal models (6, 7,10, 18, 25), and evidence suggests that a similar elevation in CBF occurs in neonates (24). Coupled with this increased CBF with seizure activity, Winn et al. (40) showed increased brain adenosine levels during the initial minutes of seizure activity in an adult rat model. Specifically, Park et al. (22) found a significant increase in brain interstitial fluid adenosine levels in association with increased CBF during bicuculline-induced seizures in newborn pigs.

Recent studies have produced mixed evidence concerning adenosine in its role as a potential mediator of cerebral vasodilation to hypoxia. Pelligrino et al. (23) demonstrated in adult rats that severe hypoxia is dependent on neuronal activation promoting the additive release of both NO and adenosine, with NO having increasing significance going from moderate to severe hypoxia. In his paper regarding newborn pigs, Armstead (2) presented data suggesting a more complex role of adenosine in hypoxia-induced cerebral vasodilation also involving NO, cGMP, cAMP, and activation of ATP-sensitive K⁺(K_ATP) channels. Taguchi et al. (32), using adult rabbits, showed that hypoxic-induced cerebral vasodilation appears to involve activation of K_ATP channels. Adenosine-induced dilation of cerebral arterioles was not mediated by this process in their study, suggesting that adenosine may not contribute to activation of K_ATP channels during hypoxia (32). In the hypoxia model used by Leffler et al. (13), the authors were unable to demonstrate a role for adenosine in hypoxic cerebral vasodilation, but they did find evidence suggesting a mechanism involving endothelium-derived cytochromeP-450 epoxygenase metabolites of arachidonic acid.

In our study, we speculated that alterations of adenosine activity with seizure ictus may be involved with the loss of pial arteriolar vasodilation to hypoxia acutely in the postictal period. We therefore needed to determine whether this altered adenosine activity was due to a diminished response to adenosine or an acute depletion of available adenosine. Piglets that were repleted with a dilating topical dose of adenosine (10⁻⁵ M) in the period immediately following bicuculline-induced seizures had return of baseline preseizure pial arteriolar vasodilation to hypoxia within 60 min postictal. Cerebral vessel responsivity to adenosine (10⁻⁵ M) was not altered by seizure activity. In addition, similar results were obtained using a nondilating concentration of adenosine (10⁻¹¹ M), with return of cerebral vasodilation to baseline again by 60 min postictal. From these results, we can make several conclusions about the role of adenosine:1) the altered postictal response to hypoxia is not due to a decreased sensitivity to adenosine but, rather, to a depletion of available adenosine; and2) adenosine appears to have a permissive role in hypoxic cerebral vasodilation, and therefore concentration is not the determinant factor of the dilatory response.

Other potential mechanisms that may be involved in hypoxia-induced cerebral vasodilation in piglets postictally cannot be ruled out by the present study because they were not examined. Future studies to further define the role of adenosine and other possible mechanisms involved in this response may be important to help prevent resulting secondary brain injury following seizures in newborn infants.

Perspectives

We speculate that the loss of vasodilation to hypoxia in the postictal period may result in inadequate oxygen and substrate delivery to those margins of ischemic tissue in asphyxiated neonates. Furthermore, cardiovascular compromise with decreased cardiac output secondary to hypoxia may further exacerbate injury in high-risk infants. It is possible that preserved cerebrovascular dilation to hypercapnia may be protective of brain injury in those infants that are poorly ventilated, at least in the immediate postictal period, until normocapnia is established. However, depending on the gestational age of the infant, increased CBF during this period may in fact promote injury in those areas such as the vulnerable germinal matrix of premature infants.

In conclusion, the present study suggests that, after seizure activity in newborn pigs, cerebrovascular dilation to hypercapnia and sodium nitroprusside is preserved, whereas hypoxia-induced cerebral vasodilation is significantly impaired in the immediate postictal period. Adenosine depletion with seizure activity is a potential mechanism involved in the loss of postictal pial arteriolar dilation to hypoxia, and adenosine appears to have a permissive role in hypoxia-induced cerebral vasodilation.