- Email: [email protected]
Seizure susceptibility during recovery from hypercapnia in neonatal dogs
ELSEVIER
Seizure Susceptibility During Recovery From Hypercapnia in Neonatal Dogs Hiroshi Yoshioka, MD, PhD**, Shoko Nioka, MD, PhD*, Hidenori Miyake, MD*, Aziza Zaman, MD*, Tadashi Sawada, MD, PhDt, and Britton Chance PhD*
Seizure susceptibility during recovery from bypercapnia was investigated in seven anesthetized neonatal dogs; 13,20, or 30% CO, gas was administered for 30 min through a ventilator to result in three levels of hypercapnia in which measured PaCO, values were approximately 70, 100, and 140 mm Hg. Thereafter, the animals were allowed to recover for 45 min; during this recovery phase, electrocorticography was performed. In five of seven dogs, -1.5 Hz slow irregular spike and wave bursts appeared at 6 min after abrupt withdrawal from hypercapnia and lasted several minutes. This seizure activity was followed by a brief period of electrical suppression. This phenomenon was most often observed during the recovery from moderate hypercapnia and between the PaCO, values of 100 and 50 mm Hg. When seizure activities appeared in the electrocorticogram, arterial blood pressure increased -40 mm Hg from the preseizure level. These results suggest that neonatal seizures may occur during recovery from hypercapnia. Yoshioka H, Nioka S, Miyake H, Zaman A, Sawada T, Chance B. Seizure susceptibility during recovery from hypercapnia in neonatal dogs. Pediatr Neurol 1996;15:36-40.
Introduction Neonatal asphyxia may result in permanent brain injury, including mental retardation and cerebral palsy [l], and hypercapnia is an important component of neonatal asphyxia. Effects of hypercapnia on neonatal brain are still controversial. Some studies have suggested that hypercapnit acidosis with hypoxia would be deleterious for neonatal brain [2]. Studies of cerebellar neuronal proliferation in newborn mice after 30-min exposure to either pure CO, gas or N,,
From the *Department of Biochemistry and Biophysics; University of Pennsylvania; Philadelphia, Pennsylvania, U.S.A.; the +Department of Pediatrics; Kyoto Prefectural University of Medicine; Kyoto, Japan.
36
PEDIATRIC
NEUROLOGY
Vol. 15 No. 1
demonstrated that the generation time, the time needed for one cell to divide into two cells, of the external granular cells in the CO, group was significantly longer than that in the controls [3] but was the same in the N, group as in the controls [4]. In addition, the brain ATP concentration was reduced 30% during hypoxia in the CO, group, but remained almost normal in the N, group [5]. In the same mouse model, Hasegawa et al. [6] detected free radical generation during CO, hypoxia, but not during N, hypoxia, directly, by using electron spin resonance spectroscopy and suggested that the free radicals that appeared during CO, hypoxia might play a role in producing the differences in brain injury between the two kinds of hypoxia. On the other hand, in a previous study [7], we investigated the effects of steady-state hypercapnia on the electrocorticogram (ECoG) and oxidative metabolism in neonatal dog brain and observed that hypercapnia-induced acidosis reduced brain ADP concentration as well as ECoG power output. That these reductions in ADP concentration and ECoG power output were indicative of low ATP synthesis activity suggested that hypercapnia would be beneficial to neonatal brain. Recently, Vannucci et al [8] reported that in an immature rat model mild hypercapnic cerebral hypoxia/ischemia was associated with less severe brain damage than was normocapnic hypoxiaJischemia. Although it is still not known whether hypercapnia with hypoxia/ischemia is deleterious to the neonatal brain, it is recommended in clinical instances that infants with respiratory disturbance and hypercapnia be mechanically ventilated to normalize arterial PO, and PCO, [9]. In addition, effects of rapid correction of hypercapnic acidosis on neonatal brain are not well known. In the present study, we investigated ECoG changes during recovery from hypercapnia in neonatal dogs. Materials
and Methods
Animal Prepurarior~. The study was approved by the University of Pennsylvania Animal Care and Utilization Committee. Seven male
Communications should be addressed to: Dr. Yoshioka; Department of Pediatrics; Kyoto Prefectural University of Medicine; Kawaramachi, Kamikyo-ku; Kyoto 602, Japan. Received November IS, 1995: accepted April 22. 1996.
0 1996 by Elsevier Science Inc. All rights reserved. PII SO887-8994(96)00116-6 0887.8994/96/$15.00 ??
Beagle dogs aged 14 days and weighing 640-1,260 g were anesthetized with an intravenous infusion of fentanyl (1 pg/kg/hr) and droperidol (50 kg/kg/hr) after anesthesia was induced intraperitoneally with fentanyl (3 pg) and droperidol (50 kg). They then were tracheostomized, paralyzed with gallamine triethiodide (1 mg/kg subcutaneously), and mechanically ventilated at the rate of 25-30 breaths/min with 33% 0, and balanced N,O to maintain arterial PO, at 150-200 torr and arterial PCOZ at 32-38 torr [7]. Arterial and venous catheters were inserted through femoral cutdowns for measurement of blood pressure, blood gas, and arterial pH (pHa); 10% dextrose was infused at a rate of 5 ml/kgihr throughout the experiment. The skull was exposed. and the temporal muscle was removed. Body temperature was maintained at 37” + 1°C with a water heat blanket. Arterial blood gases and pH were measured with a Radiometer blood gas analyzer ABL2. Experimentt~l Pmrocol. After normocapnic condition was stabilized at PaCOz between 32 and 38 torr, 13% CO, gas was given through the ventilator for 30 min (mild hypercapnia). Blood gas was measured at 7 and 20 min of the hypercapnic period. Thereafter, animals were allowed to recover to normocapnia for 45 min. Blood gas was measured at 25 and 45 min of the recovery period. and gas data confirmed the recovery from hypercapnic change in PaCO, [7]. Then the initial hypercapnia/recovery episode was followed by 20% CO, gas administration (moderate hypercapnia) and later by 30% CO, gas administration (severe hypercapnia), conducted m the same way and for the same duration. E&G. Small brass screws were used as bipolar leads. They were inserted through small burr holes in the skull onto the dura over the right lateral and occipital cortexes. A ground lead was inserted into the frontal cortex at the midline. The leads were encased in dental acrylic to prevent signal degradation due to fluid accumulation. The ECoG was recorded with a Gould polygraph recorder; the signal was also collected and analyzed by a fast Fourier transform to obtain the frequency amplitude of spectra (I-30 Hz) [IO]. The integrated voltage output per unit of time was calculated and expressed as a percent of the control value.
Results Changes in PaCO, during hypercapnia and recovery are depicted in Table I. PaCO, increased 70, 100, and 140 torr as inhaled CO, gas concentration was increased to 13, 20, Table 1. Effects of bypercapnia and rapid withdrawal hypercapnia on arterial PCO,and pH Parameter Control 13% co, 7 min 20 min Recovery 25 min 45 min 20% co, 7 min 20 min Recovery 25 min 45 min 30% co, 7 min 20 min Recovery 25 min 45 min Values are mean f SD
from
PaCO,
pHa
30.4 +3.1
7.3 1 f 0.09
58.9 + 11.6 68.5 f 7.6
7.10 i- 0.07 7.05+0.10
35.5 * 6.6 33.5 rt 7.8
7.25 It 0.08 7.29 f 0.04
88.2f11.6 97.1 f 13.9
6.99 + 0.05 6.94 + 0.05
36.7 ? 4.9 34.7 f 4.7
7.18 * 0.12 7.29kO.10
119.7 f 9.7 142.0f 14.1
6.88 f 0.06 6.8 I + 0.08
42.0 * 6.0 37.3 * 6.6
7.23 k 0.05 7.32 + 0.04
and 30% (P < .005, one-way repeated-measures analysis of variance, SYSTAT [ 111). pHa decreased to 6.80 during 30% CO2 administration from 7.3 1 during normocapnia (P < .005), while blood pressure and arterial PO, were maintained. An example of ECoG power spectral changes during graded hypercapnia and recovery is depicted in Figure 1. The more hypercapnia progressed, the more ECoG output was reduced [7] (Fig 1). Soon after the cessation of hypercapnia, ECoG power began to increase and usually recovered to the control condition within 5 min. However, in five of seven animals, abnormal high ECoG activities appeared between 3 and 10 min (mean 6.4 min) after the beginning of the recovery phase and lasted several minutes (1-7 min; mean 4.9 min) (Table 2). Thereafter, a brief period of electrical suppression always followed. In the animal depicted in Figure 1 (dog 5), an abnormal high ECoG burst appeared 10 min after the end of moderate hypercapnia and lasted for -4 min. Inspection of the original ECoG tracing revealed that these abnormal activities were -1.5 Hz slow irregular spike and wave bursts (Fig 2). Because the animals were completely anesthetized and had received muscle relaxant, no clinical convulsions were observed. Such a phenomenon was observed in one (14%) of seven animals after mild hypercapnia, in five (71%) after moderate hypercapnia, and in two (29%) after severe hypercapnia, respectively (Table 2). It is noteworthy that significant increase in arterial blood pressure was always associated with this abnormal ECoG burst. Mean arterial blood pressure was 77.3 +- 21.1 mm Hg (mean + S.D.) in the preseizure (recovery) condition and 120.4 + 27.7 mm Hg during seizure, respectively, indicating that mean increase in arterial blood pressure associated with the seizure activity was 43.1 + 15.0 mm Hg. In the case depicted in Figure 2, mean arterial blood pressure increased 56 mm Hg, from 57 to 113 mm Hg. Because we did not obtain any arterial blood sample during the occurrence of seizure activities, the accurate blood gas or electrolytes condition when seizure activities appeared was not detected. In three cases, however, blood gas was measured 2, 3 and 4 min after the cessation of the abnormal high-voltage bursts after moderate hypercapnia, and at that time PaCO, values were 48.1, 41.5, and 37.5 mm Hg, respectively. Discussion Woodbury et al. [ 121, investigating effects of hypercapnia on brain excitability in adult rats, reported that many rats exhibited spontaneous repetitive seizures during exposure to atmospheres containing 30-40% COZ. At the same time, they observed that clonic seizures also occurred within 30 s after abrupt removal of rats from a CO, concentration 335% and that all rats showed withdrawal seizures after exposure to 45% CO,. At 30 s after abrupt discontinuation of 50% CO, atmosphere, average plasma CO, level obtained from cardiac puncture was 133 mm Hg [ 131. Since that study [13], however, no study has been made of whether such a seizure would occur even in neo-
Yoshioka
et al: Hypercapnia
Withdrawal
Seizures
37
CONTROL STATE
RECOVERY FRM
CO2 203
Figure 1. Fourier-transformed ECoG data with use of a condensed spectral arra). In this display format, frequency is on the ordinate and ranges from 0 to 30 Hz (left to right). Time is shown at right (hr:min:s), with earlier times toward the top. Amplitude of the ECoG signal is represented by intensity (eight-level gray scale). Each line represents 2 s of ECoG data. The left column shows decreases in both frequency and power (amplitude of the ECoG signal) with progression of hypercapnia. At 18:27:16, recovery from moderate hypercapnia was started, and the ECoG power and frequency recovered at -l&31:09. At -18:37:15, ubnormal high ECoG activities appeared and lasted 4 min. Note the electrical suppression at 18:41:47.
natal animals. In a previous study of effects of steady-state hypercapnia on ECoG in neonatal dogs, we observed no seizure activity during hypercapnia [7]. In the present study, we clearly demonstrated that neonatal dogs also have seizure susceptibility during recovery from hypercapnia. ECoG measurements revealed that abnormal high-voltage spike and wave bursts appeared approximately 6 min after the beginning of the recovery phase and lasted several minutes. It is likely that the experimental animals did not reveal any convulsive movements because they were anesthetized and receiving muscle relaxant when spike and wave bursts appeared in the ECoG. Because a brief period of elec-
38
PEDIATRIC
NEUROLOGY
Vol. 15 No. I
trical suppression always followed, the seizure activities during recovery phase from hypercapnia may be associated with energy disturbance in the brain. The underlying mechanism of CO, withdrawal seizure is not known. With regard to excitability of the adult rat brain, hypercapnia exerts a triphasic effect [ 141. In adult rats, mild hypercapnia (FiCO, 520%) decreases cortical excitability, whereas moderate hypercapnia (FiCO, 25 40%) increases it. Severe hypercapnia (FiCO, 240%) has an anesthetic effect due to the marked cortical and subcortical depression. Therefore, Woodbury and Karler [ 141 considered that on abrupt withdrawal of animals from
Table 2. Occurrence of seizure activity during withdrawal from hypercapnia in neonatal dogs Withdrawal Dog
From Hypercapnia
Mild
Moderate
+ (7*, 4’)
+ (7, 7)
Severe
1 2 3 4 5 6 I
+ (9,4.5) + (1% 4) + (6, 5) + (4, 4)
*Interval (min) from beginning of withdrawal activity. ‘Duration (min) of seizure activity.
200
ABP
ImmHgl
+ (5, 7) + (3, 1)
to onset of seizure
anesthetizing concentrations of CO,, the brain concentration of CO, rapidly decreases to the level that induces hyperexcitability and that seizures then occur. However, in our study of ECoG changes during graded hypercapnia in neonatal dogs in a previous study, electrical activity in newborn brain demonstrated a monophasic response to hypercapnia and progressively decreased with increasing concentrations of CO, [7]. Therefore, brain excitability due to hypercapnia itself is not likely to be the cause of CO, withdrawal seizures in neonatal dogs. Abrupt changes in concentration and distribution of brain electrolytes may be related to the withdrawal seizures from hypercapnia, although we did not measure serum and brain electrolytes in the present study. Woodbury et al. [ 121 reported that the brain intracellular sodium concentration increased rapidly on abrupt withdrawal of adult
*
100
0
EC00 5OpV
Figure 2. Original polygraphic record of the arterial blood pressure (ABP) and ECoG in the same animal depicted in Figure 1. Recording speeds of ABP (first and fourth raws) and ECoG (second andfijih raws) are 6 mm’min. In the third and sixth raws, high-speed (50 mm/s) short runs of ECoG are pulled oat from the indicated portions. (A) control condition. (B) Beginning of mild hypercapnia (FiCO, 13%). (C) Beginning of withdrawal from mild hypercapnia. (D) Beginning of moderate hypercapnia (RCO, 20%). (E) Beginning of withdrawal from moderate hypercapnia. (F) Onset of seizure activity. (G) End of seizure activity. (H) Beginning of severe hypercapnia (FiCO, 30%). (I) Beginning of withdrawal from severe hypercapnia. (J) Onset of seizure activity. (K) End of seizure activity. Note abrupt appearance of irregular spike and wave bursts in the ECoC and simultaneous increase in ABP during the withdrawal from moderate hypercapnia.
Yoshioka
et al: Hypercapnia
Withdrawal
Seizures
39
rats from hypercapnia and that seizures appeared at that time. Serum concentration of calcium ion may be decreased with abrupt withdrawal from hypercapnia, possibly resulting in seizures [ 11. Another noteworthy finding of the present study is that blood pressure abruptly increased with the appearance of seizure activities in ECoG. Mean increase in arterial blood pressure associated with the seizure activity was -40 mm Hg. Experimental evidence indicates that abrupt increase in arterial blood pressure causes intraventricular hemorrhage, which is a most important cause of brain injury in prematurely born infants. Goddard et al. [15] indicated that increases in mean blood pressure of only 20-40 mm Hg, when induced rapidly, were able to initiate choroid plexus, subependymal, and intraventricular hemorrhages in puppies. Because the dog brains were not pathologically investigated in the present study, we do not know whether intracranial hemorrhage was present or not in the brains. Fourteen-day-old puppies, which we used in this study, still possess the periventricular germinal matrix in their brains [ 161, and the maturation of the 16day-old dog brain appears to be equivalent to that of the human fetal brain between 32 and 40 weeks of gestation [ 16,171. Accordingly, recovery phase from hypercapnia may be a serious period of high risk for induction of neonatal seizures with brain energy depression and intraventricular hemorrhage. This work was supported by grants from the Ministry of Health and Welfare in Japan and Grant No. 07671278 from the Ministry of Education in Japan. References [l] Volpe JJ. Neurology of the Newborn, 3rd ed., Philadelphia: W.B. Saunders, 1995:21 l-59,314-69. [2] Rehncrona S, Rosen L, Siesjo BK. Excessive cellular acidosis: An important mechanism of neuronal damage in the brain? Acta Physiol Stand 1980;110:435-7.
40
PEDIATRIC
NEUROLOGY
Vol. 15 No. 1
[3] Yoshioka H. Mino M, Morikawa Y. Kasubuchi Y. Kusunoki T. Changes in cell proliferation kinetics in the mouse cerebellum after total asphyxia. Pediatrics 1985;76:965-9. [4] Yoshioka H. Neonatal asphyxia and subsequent brain dcvclopment in the mouse. In: Yabuuchi H, Watanabe K, Okada S, eds. Nconatnl Brain and Behavior. Nagoya: Nagoya University Press. 1987:27-34. [S] Yoshioka H. Fujiwara K, Ishimura K, et al. Brain energy metabolism in two kinds of total asphyxia: An in vivo phosphorus nuclear magnetic resonance spectroscopic study. Brain Dev (Tokyo) l988;lO: 88-9 1. [6] Hasegawa K, Yoshioka H, Sawada T, Nishikawa H. Direct measurement of free radicals in the neonatal mouse brain subjected to hypoxia: An electron spin resonance spectroscopic study. Brain Res 1993; 607:161-6. [7] Yoshioka H, Miyake H, Smith DS, Chance B, Sawada T, Nioka S. Effects of hypercapnia on ECoG and oxidative metabolism in neonatal dog brain. J Appl Physiol 1995;78:2272-8. [S] Vannucci RC. Towfighi J. Heitjan DF, Brucklacher RM. Carbon dioxide protects the perinatal brain from hypoxic-ischemic damage: An experimental study in the immature rat. Pediatrics 1995;95:868-74. [9] Ballard RA. Resuscitation in the delivery room. In Taeusch HW, Ballard RA, Avery ME, eds. Schaffer and Avery’s Diseases of the Newborn, 6th ed. Philadelphia: W.B. Saunders. 199 1: 193-206. [lo] Myers RR, Stockard JJ, Saidman LJ. Monitoring of cerebral perfusion during anesthesia by time-compressed Fourier analysis of the electroencephalogram. Stroke 1977;8:33 l-7. [ll] Wilkinson L. SYSTAT. Evanston, IL: SYSTAT, 1986. [12] Woodhury M, Rollins LT. Gardner MD. et al. Effects of carbon dioxide on brain excitability and electrolytes, Am J Physiol 1958;192: 79-90. [13] Brodie DA, Woodbury DM. Acid-base changes in brain and blood of rats exposed to high concentrations of carbon dioxide. Am J Physiol 1958;192:91-4. [14] Woodhury DM. Karler R. The role of carbon dioxide in the nervous system. Anesthesiology 1960;21:686-703. [15] Goddard J, Lewis RM, Armstrong DL, Zeller RS. Moderate. rapidly induced hypertension as a cause of intraventricular hemorrhage in the newborn beagle model. J Pediatr 1980;96: 1057-60. [16] Yoshioka H, Goma H, Nioka S. et al. Bilateral carotid artery occlusion causes periventricular leukomalacia in neonatal dogs, Dev Brain Res 1994;78:273-8. [17] Goddard J, Lewis RM, Zeller RS. Intraventricular hemorrhage-An animal model. Biol Neonate 1980;37:39-52.
Seizure Susceptibility During Recovery From Hypercapnia in Neonatal Dogs Hiroshi Yoshioka, MD, PhD**, Shoko Nioka, MD, PhD*, Hidenori Miyake, MD*, Aziza Zaman, MD*, Tadashi Sawada, MD, PhDt, and Britton Chance PhD*
Seizure susceptibility during recovery from bypercapnia was investigated in seven anesthetized neonatal dogs; 13,20, or 30% CO, gas was administered for 30 min through a ventilator to result in three levels of hypercapnia in which measured PaCO, values were approximately 70, 100, and 140 mm Hg. Thereafter, the animals were allowed to recover for 45 min; during this recovery phase, electrocorticography was performed. In five of seven dogs, -1.5 Hz slow irregular spike and wave bursts appeared at 6 min after abrupt withdrawal from hypercapnia and lasted several minutes. This seizure activity was followed by a brief period of electrical suppression. This phenomenon was most often observed during the recovery from moderate hypercapnia and between the PaCO, values of 100 and 50 mm Hg. When seizure activities appeared in the electrocorticogram, arterial blood pressure increased -40 mm Hg from the preseizure level. These results suggest that neonatal seizures may occur during recovery from hypercapnia. Yoshioka H, Nioka S, Miyake H, Zaman A, Sawada T, Chance B. Seizure susceptibility during recovery from hypercapnia in neonatal dogs. Pediatr Neurol 1996;15:36-40.
Introduction Neonatal asphyxia may result in permanent brain injury, including mental retardation and cerebral palsy [l], and hypercapnia is an important component of neonatal asphyxia. Effects of hypercapnia on neonatal brain are still controversial. Some studies have suggested that hypercapnit acidosis with hypoxia would be deleterious for neonatal brain [2]. Studies of cerebellar neuronal proliferation in newborn mice after 30-min exposure to either pure CO, gas or N,,
From the *Department of Biochemistry and Biophysics; University of Pennsylvania; Philadelphia, Pennsylvania, U.S.A.; the +Department of Pediatrics; Kyoto Prefectural University of Medicine; Kyoto, Japan.
36
PEDIATRIC
NEUROLOGY
Vol. 15 No. 1
demonstrated that the generation time, the time needed for one cell to divide into two cells, of the external granular cells in the CO, group was significantly longer than that in the controls [3] but was the same in the N, group as in the controls [4]. In addition, the brain ATP concentration was reduced 30% during hypoxia in the CO, group, but remained almost normal in the N, group [5]. In the same mouse model, Hasegawa et al. [6] detected free radical generation during CO, hypoxia, but not during N, hypoxia, directly, by using electron spin resonance spectroscopy and suggested that the free radicals that appeared during CO, hypoxia might play a role in producing the differences in brain injury between the two kinds of hypoxia. On the other hand, in a previous study [7], we investigated the effects of steady-state hypercapnia on the electrocorticogram (ECoG) and oxidative metabolism in neonatal dog brain and observed that hypercapnia-induced acidosis reduced brain ADP concentration as well as ECoG power output. That these reductions in ADP concentration and ECoG power output were indicative of low ATP synthesis activity suggested that hypercapnia would be beneficial to neonatal brain. Recently, Vannucci et al [8] reported that in an immature rat model mild hypercapnic cerebral hypoxia/ischemia was associated with less severe brain damage than was normocapnic hypoxiaJischemia. Although it is still not known whether hypercapnia with hypoxia/ischemia is deleterious to the neonatal brain, it is recommended in clinical instances that infants with respiratory disturbance and hypercapnia be mechanically ventilated to normalize arterial PO, and PCO, [9]. In addition, effects of rapid correction of hypercapnic acidosis on neonatal brain are not well known. In the present study, we investigated ECoG changes during recovery from hypercapnia in neonatal dogs. Materials
and Methods
Animal Prepurarior~. The study was approved by the University of Pennsylvania Animal Care and Utilization Committee. Seven male
Communications should be addressed to: Dr. Yoshioka; Department of Pediatrics; Kyoto Prefectural University of Medicine; Kawaramachi, Kamikyo-ku; Kyoto 602, Japan. Received November IS, 1995: accepted April 22. 1996.
0 1996 by Elsevier Science Inc. All rights reserved. PII SO887-8994(96)00116-6 0887.8994/96/$15.00 ??
Beagle dogs aged 14 days and weighing 640-1,260 g were anesthetized with an intravenous infusion of fentanyl (1 pg/kg/hr) and droperidol (50 kg/kg/hr) after anesthesia was induced intraperitoneally with fentanyl (3 pg) and droperidol (50 kg). They then were tracheostomized, paralyzed with gallamine triethiodide (1 mg/kg subcutaneously), and mechanically ventilated at the rate of 25-30 breaths/min with 33% 0, and balanced N,O to maintain arterial PO, at 150-200 torr and arterial PCOZ at 32-38 torr [7]. Arterial and venous catheters were inserted through femoral cutdowns for measurement of blood pressure, blood gas, and arterial pH (pHa); 10% dextrose was infused at a rate of 5 ml/kgihr throughout the experiment. The skull was exposed. and the temporal muscle was removed. Body temperature was maintained at 37” + 1°C with a water heat blanket. Arterial blood gases and pH were measured with a Radiometer blood gas analyzer ABL2. Experimentt~l Pmrocol. After normocapnic condition was stabilized at PaCOz between 32 and 38 torr, 13% CO, gas was given through the ventilator for 30 min (mild hypercapnia). Blood gas was measured at 7 and 20 min of the hypercapnic period. Thereafter, animals were allowed to recover to normocapnia for 45 min. Blood gas was measured at 25 and 45 min of the recovery period. and gas data confirmed the recovery from hypercapnic change in PaCO, [7]. Then the initial hypercapnia/recovery episode was followed by 20% CO, gas administration (moderate hypercapnia) and later by 30% CO, gas administration (severe hypercapnia), conducted m the same way and for the same duration. E&G. Small brass screws were used as bipolar leads. They were inserted through small burr holes in the skull onto the dura over the right lateral and occipital cortexes. A ground lead was inserted into the frontal cortex at the midline. The leads were encased in dental acrylic to prevent signal degradation due to fluid accumulation. The ECoG was recorded with a Gould polygraph recorder; the signal was also collected and analyzed by a fast Fourier transform to obtain the frequency amplitude of spectra (I-30 Hz) [IO]. The integrated voltage output per unit of time was calculated and expressed as a percent of the control value.
Results Changes in PaCO, during hypercapnia and recovery are depicted in Table I. PaCO, increased 70, 100, and 140 torr as inhaled CO, gas concentration was increased to 13, 20, Table 1. Effects of bypercapnia and rapid withdrawal hypercapnia on arterial PCO,and pH Parameter Control 13% co, 7 min 20 min Recovery 25 min 45 min 20% co, 7 min 20 min Recovery 25 min 45 min 30% co, 7 min 20 min Recovery 25 min 45 min Values are mean f SD
from
PaCO,
pHa
30.4 +3.1
7.3 1 f 0.09
58.9 + 11.6 68.5 f 7.6
7.10 i- 0.07 7.05+0.10
35.5 * 6.6 33.5 rt 7.8
7.25 It 0.08 7.29 f 0.04
88.2f11.6 97.1 f 13.9
6.99 + 0.05 6.94 + 0.05
36.7 ? 4.9 34.7 f 4.7
7.18 * 0.12 7.29kO.10
119.7 f 9.7 142.0f 14.1
6.88 f 0.06 6.8 I + 0.08
42.0 * 6.0 37.3 * 6.6
7.23 k 0.05 7.32 + 0.04
and 30% (P < .005, one-way repeated-measures analysis of variance, SYSTAT [ 111). pHa decreased to 6.80 during 30% CO2 administration from 7.3 1 during normocapnia (P < .005), while blood pressure and arterial PO, were maintained. An example of ECoG power spectral changes during graded hypercapnia and recovery is depicted in Figure 1. The more hypercapnia progressed, the more ECoG output was reduced [7] (Fig 1). Soon after the cessation of hypercapnia, ECoG power began to increase and usually recovered to the control condition within 5 min. However, in five of seven animals, abnormal high ECoG activities appeared between 3 and 10 min (mean 6.4 min) after the beginning of the recovery phase and lasted several minutes (1-7 min; mean 4.9 min) (Table 2). Thereafter, a brief period of electrical suppression always followed. In the animal depicted in Figure 1 (dog 5), an abnormal high ECoG burst appeared 10 min after the end of moderate hypercapnia and lasted for -4 min. Inspection of the original ECoG tracing revealed that these abnormal activities were -1.5 Hz slow irregular spike and wave bursts (Fig 2). Because the animals were completely anesthetized and had received muscle relaxant, no clinical convulsions were observed. Such a phenomenon was observed in one (14%) of seven animals after mild hypercapnia, in five (71%) after moderate hypercapnia, and in two (29%) after severe hypercapnia, respectively (Table 2). It is noteworthy that significant increase in arterial blood pressure was always associated with this abnormal ECoG burst. Mean arterial blood pressure was 77.3 +- 21.1 mm Hg (mean + S.D.) in the preseizure (recovery) condition and 120.4 + 27.7 mm Hg during seizure, respectively, indicating that mean increase in arterial blood pressure associated with the seizure activity was 43.1 + 15.0 mm Hg. In the case depicted in Figure 2, mean arterial blood pressure increased 56 mm Hg, from 57 to 113 mm Hg. Because we did not obtain any arterial blood sample during the occurrence of seizure activities, the accurate blood gas or electrolytes condition when seizure activities appeared was not detected. In three cases, however, blood gas was measured 2, 3 and 4 min after the cessation of the abnormal high-voltage bursts after moderate hypercapnia, and at that time PaCO, values were 48.1, 41.5, and 37.5 mm Hg, respectively. Discussion Woodbury et al. [ 121, investigating effects of hypercapnia on brain excitability in adult rats, reported that many rats exhibited spontaneous repetitive seizures during exposure to atmospheres containing 30-40% COZ. At the same time, they observed that clonic seizures also occurred within 30 s after abrupt removal of rats from a CO, concentration 335% and that all rats showed withdrawal seizures after exposure to 45% CO,. At 30 s after abrupt discontinuation of 50% CO, atmosphere, average plasma CO, level obtained from cardiac puncture was 133 mm Hg [ 131. Since that study [13], however, no study has been made of whether such a seizure would occur even in neo-
Yoshioka
et al: Hypercapnia
Withdrawal
Seizures
37
CONTROL STATE
RECOVERY FRM
CO2 203
Figure 1. Fourier-transformed ECoG data with use of a condensed spectral arra). In this display format, frequency is on the ordinate and ranges from 0 to 30 Hz (left to right). Time is shown at right (hr:min:s), with earlier times toward the top. Amplitude of the ECoG signal is represented by intensity (eight-level gray scale). Each line represents 2 s of ECoG data. The left column shows decreases in both frequency and power (amplitude of the ECoG signal) with progression of hypercapnia. At 18:27:16, recovery from moderate hypercapnia was started, and the ECoG power and frequency recovered at -l&31:09. At -18:37:15, ubnormal high ECoG activities appeared and lasted 4 min. Note the electrical suppression at 18:41:47.
natal animals. In a previous study of effects of steady-state hypercapnia on ECoG in neonatal dogs, we observed no seizure activity during hypercapnia [7]. In the present study, we clearly demonstrated that neonatal dogs also have seizure susceptibility during recovery from hypercapnia. ECoG measurements revealed that abnormal high-voltage spike and wave bursts appeared approximately 6 min after the beginning of the recovery phase and lasted several minutes. It is likely that the experimental animals did not reveal any convulsive movements because they were anesthetized and receiving muscle relaxant when spike and wave bursts appeared in the ECoG. Because a brief period of elec-
38
PEDIATRIC
NEUROLOGY
Vol. 15 No. I
trical suppression always followed, the seizure activities during recovery phase from hypercapnia may be associated with energy disturbance in the brain. The underlying mechanism of CO, withdrawal seizure is not known. With regard to excitability of the adult rat brain, hypercapnia exerts a triphasic effect [ 141. In adult rats, mild hypercapnia (FiCO, 520%) decreases cortical excitability, whereas moderate hypercapnia (FiCO, 25 40%) increases it. Severe hypercapnia (FiCO, 240%) has an anesthetic effect due to the marked cortical and subcortical depression. Therefore, Woodbury and Karler [ 141 considered that on abrupt withdrawal of animals from
Table 2. Occurrence of seizure activity during withdrawal from hypercapnia in neonatal dogs Withdrawal Dog
From Hypercapnia
Mild
Moderate
+ (7*, 4’)
+ (7, 7)
Severe
1 2 3 4 5 6 I
+ (9,4.5) + (1% 4) + (6, 5) + (4, 4)
*Interval (min) from beginning of withdrawal activity. ‘Duration (min) of seizure activity.
200
ABP
ImmHgl
+ (5, 7) + (3, 1)
to onset of seizure
anesthetizing concentrations of CO,, the brain concentration of CO, rapidly decreases to the level that induces hyperexcitability and that seizures then occur. However, in our study of ECoG changes during graded hypercapnia in neonatal dogs in a previous study, electrical activity in newborn brain demonstrated a monophasic response to hypercapnia and progressively decreased with increasing concentrations of CO, [7]. Therefore, brain excitability due to hypercapnia itself is not likely to be the cause of CO, withdrawal seizures in neonatal dogs. Abrupt changes in concentration and distribution of brain electrolytes may be related to the withdrawal seizures from hypercapnia, although we did not measure serum and brain electrolytes in the present study. Woodbury et al. [ 121 reported that the brain intracellular sodium concentration increased rapidly on abrupt withdrawal of adult
*
100
0
EC00 5OpV
Figure 2. Original polygraphic record of the arterial blood pressure (ABP) and ECoG in the same animal depicted in Figure 1. Recording speeds of ABP (first and fourth raws) and ECoG (second andfijih raws) are 6 mm’min. In the third and sixth raws, high-speed (50 mm/s) short runs of ECoG are pulled oat from the indicated portions. (A) control condition. (B) Beginning of mild hypercapnia (FiCO, 13%). (C) Beginning of withdrawal from mild hypercapnia. (D) Beginning of moderate hypercapnia (RCO, 20%). (E) Beginning of withdrawal from moderate hypercapnia. (F) Onset of seizure activity. (G) End of seizure activity. (H) Beginning of severe hypercapnia (FiCO, 30%). (I) Beginning of withdrawal from severe hypercapnia. (J) Onset of seizure activity. (K) End of seizure activity. Note abrupt appearance of irregular spike and wave bursts in the ECoC and simultaneous increase in ABP during the withdrawal from moderate hypercapnia.
Yoshioka
et al: Hypercapnia
Withdrawal
Seizures
39
rats from hypercapnia and that seizures appeared at that time. Serum concentration of calcium ion may be decreased with abrupt withdrawal from hypercapnia, possibly resulting in seizures [ 11. Another noteworthy finding of the present study is that blood pressure abruptly increased with the appearance of seizure activities in ECoG. Mean increase in arterial blood pressure associated with the seizure activity was -40 mm Hg. Experimental evidence indicates that abrupt increase in arterial blood pressure causes intraventricular hemorrhage, which is a most important cause of brain injury in prematurely born infants. Goddard et al. [15] indicated that increases in mean blood pressure of only 20-40 mm Hg, when induced rapidly, were able to initiate choroid plexus, subependymal, and intraventricular hemorrhages in puppies. Because the dog brains were not pathologically investigated in the present study, we do not know whether intracranial hemorrhage was present or not in the brains. Fourteen-day-old puppies, which we used in this study, still possess the periventricular germinal matrix in their brains [ 161, and the maturation of the 16day-old dog brain appears to be equivalent to that of the human fetal brain between 32 and 40 weeks of gestation [ 16,171. Accordingly, recovery phase from hypercapnia may be a serious period of high risk for induction of neonatal seizures with brain energy depression and intraventricular hemorrhage. This work was supported by grants from the Ministry of Health and Welfare in Japan and Grant No. 07671278 from the Ministry of Education in Japan. References [l] Volpe JJ. Neurology of the Newborn, 3rd ed., Philadelphia: W.B. Saunders, 1995:21 l-59,314-69. [2] Rehncrona S, Rosen L, Siesjo BK. Excessive cellular acidosis: An important mechanism of neuronal damage in the brain? Acta Physiol Stand 1980;110:435-7.
40
PEDIATRIC
NEUROLOGY
Vol. 15 No. 1
[3] Yoshioka H. Mino M, Morikawa Y. Kasubuchi Y. Kusunoki T. Changes in cell proliferation kinetics in the mouse cerebellum after total asphyxia. Pediatrics 1985;76:965-9. [4] Yoshioka H. Neonatal asphyxia and subsequent brain dcvclopment in the mouse. In: Yabuuchi H, Watanabe K, Okada S, eds. Nconatnl Brain and Behavior. Nagoya: Nagoya University Press. 1987:27-34. [S] Yoshioka H. Fujiwara K, Ishimura K, et al. Brain energy metabolism in two kinds of total asphyxia: An in vivo phosphorus nuclear magnetic resonance spectroscopic study. Brain Dev (Tokyo) l988;lO: 88-9 1. [6] Hasegawa K, Yoshioka H, Sawada T, Nishikawa H. Direct measurement of free radicals in the neonatal mouse brain subjected to hypoxia: An electron spin resonance spectroscopic study. Brain Res 1993; 607:161-6. [7] Yoshioka H, Miyake H, Smith DS, Chance B, Sawada T, Nioka S. Effects of hypercapnia on ECoG and oxidative metabolism in neonatal dog brain. J Appl Physiol 1995;78:2272-8. [S] Vannucci RC. Towfighi J. Heitjan DF, Brucklacher RM. Carbon dioxide protects the perinatal brain from hypoxic-ischemic damage: An experimental study in the immature rat. Pediatrics 1995;95:868-74. [9] Ballard RA. Resuscitation in the delivery room. In Taeusch HW, Ballard RA, Avery ME, eds. Schaffer and Avery’s Diseases of the Newborn, 6th ed. Philadelphia: W.B. Saunders. 199 1: 193-206. [lo] Myers RR, Stockard JJ, Saidman LJ. Monitoring of cerebral perfusion during anesthesia by time-compressed Fourier analysis of the electroencephalogram. Stroke 1977;8:33 l-7. [ll] Wilkinson L. SYSTAT. Evanston, IL: SYSTAT, 1986. [12] Woodhury M, Rollins LT. Gardner MD. et al. Effects of carbon dioxide on brain excitability and electrolytes, Am J Physiol 1958;192: 79-90. [13] Brodie DA, Woodbury DM. Acid-base changes in brain and blood of rats exposed to high concentrations of carbon dioxide. Am J Physiol 1958;192:91-4. [14] Woodhury DM. Karler R. The role of carbon dioxide in the nervous system. Anesthesiology 1960;21:686-703. [15] Goddard J, Lewis RM, Armstrong DL, Zeller RS. Moderate. rapidly induced hypertension as a cause of intraventricular hemorrhage in the newborn beagle model. J Pediatr 1980;96: 1057-60. [16] Yoshioka H, Goma H, Nioka S. et al. Bilateral carotid artery occlusion causes periventricular leukomalacia in neonatal dogs, Dev Brain Res 1994;78:273-8. [17] Goddard J, Lewis RM, Zeller RS. Intraventricular hemorrhage-An animal model. Biol Neonate 1980;37:39-52.