Differential Suppression of Spontaneous and Noxious-evoked Somatosensory Cortical Activity by Isoflurane in the Neonatal Rat

AN optimal level of neonatal anesthesia achieves both hypnosis and antinociception while maintaining physiologic stability and minimizing potential neurotoxicity.^1,2  As both anesthesia and uncontrolled pain may alter cortical activity and impair neurodevelopmental outcomes,^3–5  the impact of anesthetic agents on both spontaneous and noxious-evoked neural activity in the developing brain requires further evaluation. An important aspect of neonatal anesthesia research is the effect of nociceptive sensory input on activity within cortical sensory circuits and the degree to which central nociceptive activity is modulated by anesthesia and analgesia. Both animal and clinical evidence point to long-term consequences of early life procedural and surgical tissue injury on somatosensory and nociceptive systems,^6,7  highlighting the need to consider the impact of postnatal age on changes in both spontaneous and noxious-evoked cortical activity during surgery and anesthesia.

Extracellular field recording, including electroencephalogram, electrocorticogram, and local field potentials (intracortical activity) are commonly used to monitor ongoing spontaneous brain activity and levels of anesthesia in human and rodent neonates but are also used to record specific potentials evoked by a sensory stimulus. Somatosensory potentials evoked by experimental noxious cutaneous stimulation^8–11  are commonly used to measure pain activity in the adult human and rodent brain.¹²  Specific nociceptive potentials are also evoked by single, clinically required skin breaking procedures in the human infant brain^13,14  and have been used in this age group to measure the postnatal development of cortical pain processing.¹⁰  In adult humans and rodents, nociceptive-evoked potential amplitudes decrease with increasing concentrations of isoflurane,^9,15  but, to date, the sensitivity of nociceptive potentials to anesthesia in infants has not been studied. In a study of the whisker barrel cortex of neonatal rats, sensory potentials evoked by whisker deflection persisted at surgical isoflurane levels (1.5 to 2%) that completely suppressed the electroencephalogram and silenced spontaneous neuronal firing.¹⁶  This suggests that noxious-evoked potentials and spontaneous intracortical activity might be differentially sensitive to anesthesia in infants. Furthermore, because nociceptive potential amplitudes can be increased by peripheral C-fiber sensitization in both humans and rat cortex,^8,11  surgical injury may enhance evoked activity and add to this differential response to anesthesia in infancy.

The primary aim of this experimental laboratory study was to test the impact of increasing isoflurane concentration on the pattern of spontaneous and evoked nociceptive activity in the somatosensory cortex of infant rats undergoing hind paw incision. In addition, we compared the effects of isoflurane on cortical activity at different postnatal ages. We hypothesized that, in infant rats, noxious-evoked brain activity is more resistant than spontaneous brain activity to isoflurane anesthesia and that this difference declines with postnatal age.

All experiments were performed under personal and project licenses approved by the Home Office, London, United Kingdom, under regulations of the UK Animal (Scientific Procedures) Act, 1986. Male Sprague-Dawley rats aged postnatal day (P) 7, P14, P21, and P30 were obtained from the Biological Services Unit, University College London, London, United Kingdom. All animals were from the same colony, bred and maintained in-house, and exposed to the same caging, diet, and handling throughout development. Rats were housed in cages of six age-matched animals (P30) or with the dam and littermates (P7, P14, and P21) under controlled environmental conditions (24° to 25°C; 50 to 60% humidity; 12-h light/dark cycle) with free access to food and water. Animals were randomly picked from litters by hand for recording and alternately assigned to an incision or no incision group. Treatment groups were distributed across multiple litters and/or adult cage groups. At the end of an experiment, the isoflurane was increased to 5% until there was no heart beat and the neck dislocated.

Rats were anaesthetized with 4% isoflurane (Abbot, AbbVie Ltd., United Kingdom) in 100% oxygen through a nose cone and after insertion of a tracheal cannula were mechanically ventilated (Small Animal Ventilator, Bioscience, United Kingdom) with 100% oxygen and isoflurane delivered from a recently calibrated vaporizer (Harvard Apparatus, USA). The adjustments for ventilation of rats across the ages were based on the breath rates and tidal volume. The intermittent positive pressure ventilation was achieved by using a T-type system in conjunction with a small animal lung ventilator. This system affords control of inspired gas mixture (isoflurane concentration and oxygen) inspiratory flow rate, respiratory rate, and peak inspiratory pressure. A simple water manometer placed in the inspiratory limb provided a monitoring and pressure limiting device with visual display. Using an inspiratory flow rate between 1 and 1.5 l/min and adjusting the peak inspiratory pressure between 12 and 15 cm H₂O¹⁷  allowed the tidal volume to be adjusted according to the size of animals. This mode was confirmed to be adequate in our laboratory by using a transcutaneous combination probe to monitor transcutaneous oxygen and carbon dioxide. Body temperature was maintained with a thermostatically controlled heated blanket, and the electrocardiogram was monitored throughout (NeuroLog, Digitimer, United Kingdom).

During anesthesia with 2.5% isoflurane, animals were placed in a stereotaxic frame, and a craniotomy was performed to expose the surface of the cerebral cortex. A recording electrode (stainless steel, E363/1, Plastic One Inc., USA) was inserted into the primary somatosensory cortex in the somatotopic region for the hind paw. Coordinates for P7 and P14 rats were lateral 2.0 mm from midline and posterior 0.5 mm from the bregma; and coordinates for P21 and P30 rats were lateral 2.5 mm from midline and posterior 1 mm from the bregma.^18,19  The reference electrode was placed subcutaneously on the surface of the skull anterior to the bregma, and the ground electrode was placed subcutaneously in the back. The inspired isoflurane concentration was then reduced to 1.5% and allowed to equilibrate for at least 40 min.

Continuous intracortical activity recordings were performed using a NeuroLog NL100 headstage connected to a NL104 amplifier and a NL125 filter (100 Hz). The signal was digitized at 16K Hz using Axon Instruments (Digidata 1400A, Molecular Devices, USA). Data were acquired and stored using a Windows PC–based program, WinEDR v3.3.6 (John Dempster, University of Strathclyde, United Kingdom) for later analysis. The depth of the recording electrode was adjusted to optimize the somatosensory-evoked potential amplitude. At the end of an experiment, the brain was removed and immediately immersed in 4% paraformaldehyde for over 24 h and then transferred to 30% sucrose postfixation solution. Brain sections (40-μm thick thickness) were cut using a micotome (Leica SM2000R, Leica Microsystems (UK) Ltd., United Kingdom) and stained with cresyl violet for histologic location of the electrode track. This procedure verified that recordings were in layer 5 to 6 of the somatosensory cortex (fig. 1). All the data recorded were included in this study, and no animals were excluded.

Two stainless steel pin electrodes were placed subcutaneously 3 to 5 mm apart on the plantar hind paw, contralateral to the cortical recording site. A train of 10 stimuli of 3.2 mA, 500 μs were applied at 10-s interstimulus intervals, using a constant current stimulator (NeuroLog, Digitimer). These stimulation parameters are sufficient to recruit both A- and C-fibers²⁰  (here called “C-fiber” stimulation) and were established in pilot experiments to evoke clear potentials restricted to the somatosensory cortex. At all ages, electrical hind paw stimulation failed to evoke visible hindlimb reflex responses at 1.5% or higher inspired isoflurane concentrations. In a separate group of P7 animals, a train of lower intensity stimuli of 0.32 mA for 50 μs was also applied, sufficient to recruit only Aβ-fibers²⁰  (here called “A-fiber” stimulation), for comparison.

Plantar hind paw incision was performed in P7 and P30 animals (n = 6 per age group) after equilibration at 1.5% isoflurane in oxygen. A midline longitudinal incision through the skin and fascia extended from the midpoint of the heel to the proximal border of the first footpad to incise a similar relative length of paw at different ages, and the underlying plantar muscle was elevated and incised.^21,22  Skin edges were closed with 5-0 nylon suture (Ethicon, United Kingdom), and the procedure took 2.5 to 3 min. The initial skin incision evoked a brief visible muscle reflex in young animals (P7) but not in older animals (P30).

The experimental timeline is illustrated in figure 1. After surgical preparation and equilibration at 1.5% inspired isoflurane concentration, spontaneous intracortical activity was recorded for 100 s, followed by evoked activity during hind paw electrical stimulation in P7, P14, P21, and P30 rats (fig. 1, Group 1). In pilot studies, and subsequent experimental recordings, it was clear that increasing the inspired isoflurane concentration rapidly altered spontaneous activity. Therefore, the inspired isoflurane concentration was increased at 5-min intervals (step 1 = 2%, step 2 = 3%, step 3 = 4%, and step 4 = 5%), and spontaneous and evoked potentials were measured in the 3- to 5-min period after each step. The step terminology was adopted because this frequency of change would not have allowed sufficient time for equilibration at each inspired concentration. However, this protocol allowed us to assess the response to increasing isoflurane exposure without the cardiovascular compromise associated with more prolonged exposure to high concentrations.

In additional groups at P7 or P30, a plantar hind paw incision was performed after equilibration at 1.5% isoflurane (fig. 1, Group 2). In these experiments, spontaneous activity was measured (sampled from a 100-s epoch) in the 2-min period before, during, and 2 min after plantar hind paw incision with inspired isoflurane maintained at 1.5%. Ten minutes after plantar hind paw incision, recordings were performed during step increases in inspired isoflurane concentration as described for group 1.

Electrocardiography confirmed that the heart rate remained stable throughout the recording periods at 1.5% isoflurane in both groups. At all ages, the heart rate was reduced by step 4. When expressed as a percentage change from baseline (as heart rate differs with postnatal age in rodents), the degree of change in heart rate was not significantly different at P7 and P30 (P7: 81 ± 3% vs. P30: 88 ± 3%, P = 0.157, Student’s t test; data not shown).

Sample sizes (n = 7 animals for each age) were based on previously published group effects of volatile anesthesia on cortical activity. Statistically significant dose-dependent effects of isoflurane on evoked responses have been reported in the somatosensory cortex of adult rats (n = 5)⁹  and in the barrel cortex of neonatal (P2 to P7) rats (n = 6).¹⁶ 

Ongoing spontaneous intracortical activity was analyzed in 100-s epochs at each isoflurane step and before, during, and after plantar hind paw incision. Frequency domain analysis was performed by fast Fourier transformation (spectrum type: one sided) using the “Welch Window.”²³  Intracortical responses were converted from the time domain to the frequency domain, and the Fourier transform components amplitude (millivolts per Hertz) was computed using OriginPro 9 (OriginLab Corporation, USA). Brain waves have been categorized into five frequency bands: delta (δ = 2 to 4 Hz), theta (θ = 4 to 8 Hz), alpha (α = 8 to 12 Hz), beta (β = 13 to 30 Hz), and gamma (γ more than 30 Hz).²⁴  Network oscillations with a range of frequencies are thought to control the flow of information through anatomical pathways and communicate among local networks. The changing patterns of network oscillation are tightly correlated with behavior or features of sensory stimuli.²⁵ 

Individual noxious-evoked potentials were averaged (10 stimuli per animal), and peak amplitudes were computed for each animal. Data are given as grand mean ± SEM (number of stimuli × number of animals). In addition, evoked potentials at steps 2 to 5 were normalized to baseline (1.5% of inspired isoflurane) and presented as % of baseline. Statistical analysis of peak amplitudes was performed using one-way repeated measures ANOVA (RM ANOVA) followed by Dunnett post hoc comparison tests.

Time–frequency (TF) analysis was also used to detect the energy of the oscillations in the brain evoked by the noxious electrical stimulus. TF analysis has been extensively used in functional studies of brain activity in humans and animal models.^10,26,27  First, continuous intracortical activity was high-pass filtered at 0.5 Hz with a zero-phase second-order Butterworth filter. TF analysis was then performed using a complex Morse wavelet transform.^26,28  This allowed us to calculate a complex TF spectral estimate W(a,b) of the intracortical activity at each point (a,b) of the TF plane from 3 s before the stimulus to 3 s after stimulus in the time domain and between 0.5 and 100 Hz (in logarithmic steps) in the frequency domain. The changes in TF spectral energy (i.e., modulus square) in the intracortical activity in response to the stimulation were estimated by normalizing to the mean energy content of the baseline period (a period of quiescence of 1 s) at each frequency.¹⁰  The TF energy, time locked to the stimulation in each group (anesthetic step, age, incision), was presented as a group median. The stimulus-induced energy changes, time locked to each stimulus, were estimated separately at each postnatal age and anesthetic step.

We first recorded intracortical spontaneous activity in the primary somatosensory cortex of P7 rats after 40-min equilibration at 1.5% isoflurane and through subsequent step increases in inspired concentration (n = 7 animals). Figure 2A shows typical traces of S1 intracortical activity at each inspired isoflurane concentration. No spontaneous activity was recorded in any frequency band in the P7 primary somatosensory cortex at 1.5% inspired isoflurane or at higher concentrations (fig. 2B)

Despite the lack of spontaneous activity, hind paw electrical stimulation evoked clear low amplitude–evoked potentials, with simple positive–negative voltage waveforms, in the P7 somatosensory cortex (fig. 3). A significant decrease in the mean evoked potential amplitude did not occur until step 3, where it dropped to half the amplitude recorded at 1.5% inspired isoflurane (50 ± 9%; fig. 3, A and B). At the maximum inspired isoflurane concentration (step 4), evoked potentials were greatly diminished (6 ± 1%) but importantly were still clearly detectable in five of the seven animals studied. A TF analysis of this evoked activity is shown in figure 3C to demonstrate the mean energy across different frequencies (1 to 100 Hz) evoked in 3 s after noxious electrical stimulation. At 1.5% inspired isoflurane, a strong increase in energy at all frequencies is observed in the first 500 ms after stimulus, but this declines with increasing levels of inspired isoflurane, notably at steps 3 and 4 (fig. 3C). At step 4, the oscillation energy is greatly reduced, but still present.

We next tested whether hind paw incision influenced the sensitivity of plantar C-fiber–evoked potentials to step increases in anesthesia. The immediate effect of the plantar incision is shown in figure 4. In P7 rats, spontaneous activity was increased during, and immediately after, the incision (δ band activity, 0.02 ± 0.01 to 0.10 ± 0.02 μV/Hz; RM ANOVA F_{(3, 15)} = 13.5, P = 0.0002; post hoc Dunnett test: P < 0.001, and β band activity (0.01 ± 0.00 to 0.02 ± 0.01 μV/Hz, RM ANOVA, F_{(3, 15)} = 6.1, P = 0.0062; post hoc Dunnett test: P < 0.05), which returned to baseline at 10 min (fig. 4, A and B). In addition figure 3 shows that hind paw incision had a more persistent effect on the response to subsequent noxious electrical stimuli. In animals with previous hind paw incision, the mean peak amplitude of the electrical-evoked potential was not significantly altered by step increase in isoflurane level, except at step 4 (RM ANOVA F _{(4, 20)} = 3.7, P = 0.0210; post hoc Dunnett test: P < 0.01 between baseline 1.5% and step 4), and even at the maximal inspired isoflurane concentration, the mean response remained at 67 ± 16% of the peak amplitude evoked at the initial 1.5% isoflurane concentration (fig. 3, D and E). This is further confirmed by the TF analysis (fig. 3F). At baseline 1.5% isoflurane and at step 1, the strong increase in energy at all frequencies in the first 500 ms after stimulus does not differ in animals with and without skin incision, but this is not so at higher levels of isoflurane. Thus, after hind paw incision, the energy of cortical activity does not diminish with increasing inspired concentration of isoflurane. The comparison of oscillation energies with and without plantar hind paw incision at step 4, shown in figure 3, shows that the presence of surgical injury increases the resistance of infant S1 cortex nociceptive activity to isoflurane (fig. 3, C and F, bottom panels).

To test whether this effect of skin incision was nociceptive specific, we compared the effects on potentials evoked by low-intensity innocuous A-fiber stimulation versus those evoked by high-intensity noxious C-fiber stimulation in the same animal. Figure 5, A and B show that, while in intact skin A-fiber stimulation evokes distinct potentials that slowly diminish with the increasing levels of isoflurane, after skin incision, the A-fiber–evoked response is completely absent at all isoflurane levels (fig. 5, C and D). This is in marked contrast to the C-fiber–evoked potentials that persist unchanged after skin incision at all but the highest step level of isoflurane (fig. 5, E and F).

We next examined the effect of postnatal age on the relationship between primary somatosensory cortical spontaneous and evoked activity at increasing inspired isoflurane concentrations. To do this, the same experiment was performed at P14, P21, and P30, and the results were compared with those obtained at P7.

Figure 2 shows the intracortical spontaneous activity in the primary somatosensory cortex of the rat, at the four postnatal ages after equilibration at 1.5% isoflurane and through subsequent step increases in inspired isoflurane concentration (n = 7 animals at each age). In contrast to the lack of spontaneous activity at P7, figure 2A shows that at P14, P21, and P30, an inspired isoflurane concentration of 1.5% induced burst suppression activity in the somatosensory cortex, characterized by intermittent, highly synchronized neuronal discharges (bursts) separated by silent periods (suppression), as described elsewhere.²⁹  The burst suppression activity disappears as inspired isoflurane concentration increases. Figure 2B shows the distribution of frequency components of spontaneous somatosensory cortical activity at each age. At P14 to P30, activity across all frequencies decreases steadily as the inspired isoflurane concentration increases, and there is no spontaneous activity and an isoelectric intracortical activity by step 4.

We next compared the primary somatosensory (S1) cortical potentials evoked by noxious, C-fiber intensity electrical stimuli in the contralateral hind paw (10 stimuli/step/animal, n = 7 animals) in P7, P14, P21, and P30 rats during increasing inspired isoflurane concentrations (fig. 6). Sample traces are shown in figure 6A, and the mean peak amplitude of the evoked potential is plotted in figure 6B, as the percentage decrease with each step increase in isoflurane, normalized to the initial values recorded after equilibration at 1.5% isoflurane. The relative decrease in evoked potential peak amplitude with each step increase in isoflurane is more marked in older animals (P14, P21, and P30) than at P7. At step 3, the mean peak amplitude has reduced to half in P7 animals (50 ± 9% of initial 1.5% isoflurane values), in contrast to P21 and P30 where it drops below 10% (P21: 6 ± 1%; P30: 4 ± 1%). At Step 4, no discernible evoked potentials were recorded from any P21 or P30 animals but were recorded from three of seven P14 animals and five of seven P7 animals.

Figure 7 shows TF analysis of evoked activity at each age. The mean energy of cortical activity across different frequencies (1 to 100 Hz) evoked in the 3 s after noxious electrical stimulation is shown at the four steps of increasing inspired isoflurane (10 stimuli/step/animal, n = 7 animals for each age group). At all ages, the evoked response energy decreases with increasing inspired isoflurane. Although noxious-evoked cortical activity is relatively resistant to increasing isoflurane concentration at P7, there is a gradual increase in the degree of suppression by isoflurane at older ages (fig. 7).

As surgical incision had a profound effect on the subsequent cortical response to noxious electrical stimulation at P7, we tested whether this also occurred at an older age. Figure 4 shows that there was no immediate effect of the plantar incision on spontaneous activity during, and immediately after, the incision at P30 (fig. 4, C and D). Figure 8 illustrates the effect of plantar incision on nociceptive-evoked potential at each anesthetic step in P30 rats. Electrical-evoked nociceptive potentials are progressively suppressed by increasing inspired isoflurane concentration in the absence (fig. 8, A and B) and presence (fig. 8, C and D) of skin incision. The reduction in peak evoked potential amplitude produced by step increases in inspired isoflurane (fig. 8B) was not altered after hind paw incision (fig. 8D). There was a steady fall in mean evoked potential amplitude, expressed as a percentage change from baseline (1.5% isoflurane) in both nonincised (step 1, 88 ± 14; step 2, 62 ± 8; step 3, 4 ± 1; step 4, 1.4 ± 0.5, F _{(4, 20)} = 49, P < 0.0001) and incised groups (step 1, 62 ± 9; step 2, 50 ± 19; step 3, 3.5 ± 1; step 4, 0.13 ± 0.1, F _{(4, 20)} = 25, P < 0.0001) (fig. 8, B and D). TF analysis (fig. 8, E and F) shows that skin incision at 1.5% isoflurane causes an increased energy and duration of evoked cortical oscillations at P30, not seen at P7, suggesting some underlying sensitization of cortical pain circuits. Although clear at an inspired isoflurane concentration of 1.5%, this increased cortical response is highly sensitive to the subsequent increases in isoflurane, and the difference in response between P30 animals with (fig. 8F) and without (fig. 8E) hind paw incision is largely lost at step 1, and evoked responses in both groups are significantly diminished at steps 2 and 3 and effectively gone at step 4. These results differ significantly from those obtained at P7 (fig. 3), where cortical-evoked potentials and oscillatory activity persisted and were resistant to increasing inspired isoflurane concentration.

The primary aim of this study was to test the impact of increasing isoflurane concentration on spontaneous and evoked nociceptive activity in the somatosensory cortex of infant rats undergoing hind paw incision. We hypothesized that, in infant rats, noxious-evoked brain activity is more resistant than spontaneous brain activity to isoflurane anesthesia. The results, obtained by recording intracortical neuronal activity from layer 5 to 6 of the somatosensory cortex in P7 rats during increasing inspired concentrations of isoflurane, support this hypothesis. The data show that isoflurane influences spontaneous activity and evoked activity in the infant rat somatosensory cortex quite differently.

At P7, all concentrations of inspired isoflurane (1.5 to 5%) silenced cortical neurons and suppressed spontaneous bursts and oscillations. This effect is consistent with a previous study in the neonatal somatosensory cortex, where cortical activity was completely suppressed in P7 animals by 1.5 to 2% isoflurane.¹⁶  This is in contrast to the spontaneous cortical activity observed in awake or very lightly anesthetized animals at this age, which is characterized by intermittent bursts characterized by α–β (spindle bursts) and γ frequency bands (early γ oscillations).^16,30–33 

In contrast to the absence of spontaneous activity, noxious somatosensory–evoked potentials and evoked oscillatory activity persisted, even at high inspired concentrations of isoflurane. This result highlights the differences in the neural mechanisms generating spontaneous activity and evoked activity in the immature somatosensory cortex. Activity in the developing brain changes with age, and early cortical development is marked by unique patterns of activity as functional circuits mature.^34,35  Early oscillatory bursts are generated in thalamocortical circuits^32,36  but can also be triggered by sensory inputs and are likely to be generated by glutamatergic α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid and N-methyl-d-aspartate receptors, whereas GABAergic (γ-aminobutyric acid releasing) interneurons compartmentalize the activated areas via surround inhibition.^37,38  The transition from predominantly nonspecific neuronal bursts to specific cortical sensory evoked potentials seems to be triggered by increasing sensory input³¹  in the neonatal rodent somatosensory cortex.¹⁶  A similar transition begins in the preterm human infant brain at around 35 weeks postconceptional age.^26,39  Thus, the nociceptive-evoked activity in the newborn infant cortex recorded at P7 is still relatively immature and may be less tightly coupled within specific networks, rendering it less susceptible to suppression by volatile anesthetics than at older ages.

A further finding here is that the relative insensitivity of noxious-evoked activity to isoflurane was enhanced after plantar hind paw incision. This well-established model of surgical incision pain,²¹  adapted for younger rat pups,^40,41  rendered infant rat nociceptive-evoked cortical potentials and oscillatory network activity totally resistant to the maximum inspired isoflurane tested (5%). The mechanism for this is not clear but may reflect widespread depolarization of central neurons by the incoming nociceptive barrage. It is also notable that the incision itself increased cortical activity in young animals, consistent with the reported spike activity and sensitization of receptive fields in infant dorsal horn cells after hind paw incision.⁴²  Importantly, the enhanced isoflurane resistance after skin incision at P7 was selective for C-fiber nociceptive-evoked potentials. In contrast, A-fiber innocuous evoked potentials were completely abolished after skin incision at all isoflurane levels, reflecting the different sensitivities of immature A- and C-fiber–evoked activity to skin incision.⁴³ 

A secondary outcome of this study was the significant age dependence of the isoflurane effects on spontaneous and noxious-evoked activity in the somatosensory cortex. At P14 and older ages, 1.5% isoflurane produced a typical pattern of burst suppression with high-amplitude low-frequency activity and a progressive decrease in spontaneous activity with step increases in inspired isoflurane concentration, as reported elsewhere.^29,44,45  This contrasted with the suppression of spontaneous spindle bursts and γ oscillations even at the initial inspired concentration of 1.5% isoflurane at P7. More important was the greater sensitivity of the noxious evoked potentials to isoflurane in older animals. As reported in adult rats⁹  noxious-evoked potentials at P14 to P30 decreased markedly with increasing concentrations of inspired isoflurane and at P21 and P30 were completely absent at 5% isoflurane. Differences were especially clear using TF analysis, which reveals the energy and power of oscillating signals in the cortex, and thus the changes within different frequency bands that are linked to specific sensory and motor functions.²⁵  This form of analysis has been used during isoflurane anesthesia in adult rats to characterize concentration-dependent changes in both low (30 to 50 Hz) and high (70 to 140 Hz) frequency γ power in different brain regions.⁴⁴ 

A further secondary outcome was the failure of surgical incision to affect noxious-evoked potential sensitivity to isoflurane at the older age of P30. Unlike the P7 rat, surgical incision itself at P30 caused no immediate cortical activity but did have a sensitizing effect on noxious cortical–evoked activity at 1.5% isoflurane. This is consistent with previous reports of increased C-fiber input to the adult rat somatosensory cortex in a UV-B irradiation model of hyperalgesia.⁸  However, in contrast to P7 rats, the reduction of evoked potentials and oscillatory activity by isoflurane were not altered by surgical incision in rats at P30. These data suggest that the powerful effect at P7 is not due to the effects of central sensitization, but a different yet unknown effect.

Our aim of this study was not to compare “hypnotic potency” at the different ages, but rather to evaluate within age group changes in cortical activity during isoflurane anesthesia. Equipotent concentrations of volatile anesthetics have been traditionally based on the minimum alveolar concentration (MAC) that prevents movement to a standardized noxious stimuli. However, these values do not reflect hypnotic potency in the brain.⁴⁶  The MAC of isoflurane is higher in P7 to P9 rodents,^47,48  as spinal reflex excitability is greater at this age, but a clear stimulus–response relationship that is sensitive to injury, analgesia, and inspired volatile agent concentration is evident across all ages.^22,49  Although isoflurane actions in the spinal cord may reduce ascending somatosensory information and indirectly alter cortical activity,⁵⁰  the level of cortical-evoked activity cannot be inferred from the presence or absence of a visible reflex response. Here, hind paw incision (but not electrical stimulation) produced a brief visible reflex response in P7 but not older animals, consistent with the reported greater MAC, but in older animals noxious stimuli produced clear cortical-evoked responses at 1.5% isoflurane, despite a lack of reflex response. The age dependence of cortical-evoked response sensitivity to isoflurane may have been influenced by alterations in physiologic parameters. Heart rate was monitored throughout the experiments and even in the more prolonged protocols including hind paw incision; decreases were only seen at the highest level of inspired isoflurane and to the same degree (approximately 20% reduction) in P7 and P30 groups. Animals were mechanically ventilated in isoflurane and oxygen with age-adjusted settings, and although we cannot confirm that partial pressures of carbon dioxide (Pco₂) were the same across the different ages, changes with increasing isoflurane concentration would not have been influenced by respiratory parameters.

These rat data obtained have considerable implications for clinical practice. The timing and sequence of key events in brain development are exceptionally similar across mammals, and a recent neuroinformatic analysis shows that sensorimotor events in the cortex of the P7 rat translate to 1 to 2 months in the human.⁵¹  This study clearly demonstrates that spontaneous intracortical network activity is more effectively silenced by isoflurane in the neonatal brain, but that isoflurane has less effect on cortical activity evoked by peripheral noxious sensory inputs. As the developing central nervous system is vulnerable to changes in neural activity, maintaining appropriate anesthesia in infants is likely to require avoidance of both excessive reductions in cortical activity that may enhance neuronal apoptosis and excessive increases in activity due to uncontrolled noxious inputs.^1–4  Anesthesia-induced developmental neurotoxicity has been clearly demonstrated in neonatal animals,³  and the associated long-lasting cognitive impairment has raised concern for young children undergoing anesthesia.⁴  The data here highlight the complexity of using intracortical activity parameters to define or measure the level of anesthesia required to produce “hypnosis” in neonates and infants. The persistence of noxious sensory–evoked responses despite increasing isoflurane concentration in young rats emphasizes the critical need to provide analgesia in neonates and infants. Surgery and tissue injury in neonatal rodents can produce long-term changes in sensory processing,⁵²  but this can be modified by morphine^53,54  or local anesthetic blockade.^40,55  Increased understanding of age- and anesthesia-dependent changes in intracortical activity may improve algorithms for evaluating depth of anesthesia in neonates and infants, and comparative studies of the ability of different types and doses of analgesic to minimize noxious-evoked responses are likely to improve acute clinical care and long-term neurodevelopmental outcome after neonatal surgery.

The authors thank Tom Carson, B.Sc., Department of Neuroscience, Physiology and Pharmacology, University College London, London, United Kingdom, for his technical support.

This study was supported by the Medical Research Council, London, United Kingdom, grants G0901269 (to Dr. Fitzgerald) and MR/K022636/1 (to Dr. Walker).

The authors declare no competing interests.