Alternative technique or mitigating strategy for sevoflurane-induced neurodegeneration: a randomized controlled dose-escalation study of dexmedetomidine in neonatal rats

Abstract

Background: Brain injury in newborn animals from prolonged anaesthetic exposure has raised concerns for millions of children undergoing anaesthesia every yr. Alternative anaesthetic techniques or mitigating strategies are urgently needed to ameliorate potentially harmful effects. We tested dexmedetomidine, both as a single agent alternative technique and as a mitigating adjuvant for sevoflurane anaesthesia.

Methods: Neonatal rats were randomized to three injections of dexmedetomidine (5, 25, 50, or 100 µg kg -1 every 2 h), or 6 h of 2.5% sevoflurane as a single agent without or with dexmedetomidine (1, 5, 10, or 20 µg kg -1 every 2 h). Heart rate, oxygen saturation, level of consciousness, and response to pain were assessed. Cell death was quantified in several brain regions.

Results: Dexmedetomidine provided lower levels of sedation and pain control than sevoflurane. Exposure to either sevoflurane or dexmedetomidine alone did not cause mortality, but the combination of 2.5% sevoflurane and dexmedetomidine in doses exceeding 1 µg kg -1 did. Sevoflurane increased apoptosis in all brain regions; supplementation with dexmedetomidine exacerbated neuronal injury, potentially as a result of ventilatory or haemodynamic compromise. Dexmedetomidine by itself increased apoptosis only in CA2/3 and the ventral posterior nucleus, but not in prefrontal cortex, retrosplenial cortex, somatosensory cortex, subiculum, lateral dorsal thalamic nucleaus, or hippocampal CA1.

Conclusions: We confirm previous findings of sevoflurane-induced neuronal injury. Dexmedetomidine, even in the highest dose, did not cause similar injury, but provided lesser degrees of anaesthesia and pain control. No mitigation of sevoflurane-induced injury was observed with dexmedetomidine supplementation, suggesting that future studies should focus on anaesthetic-sparing effects of dexmedetomidine, rather than injury-preventing effects.

Keywords: anaesthetics; apoptosis; brain injury; dexmedetomidine; inhalation; neuroprotection; safety; sevoflurane; toxicity.

Fig 1

Schematic of method for quantification of apoptotic neurones in representative brain regions. (A) Similar sections were coronally cut from each animal to analyse tissue samples from prefrontal cortex (PFC), hippocampal cornu ammonis CA1 and CA2/3, thalamic laterodorsal nucleus (LDN), thalamic ventroposterior nucleus (VPN), somatosensory cortex (SSC), subiculum (S), and retrosplenial cortex (RSC), as outlined in red. (B) Quantification of degenerating neurones that were colabelled with the apoptotic executioner caspase 3, the neuronal marker NeuN, and the nuclear marker DAPI performed by unbiased stereology using the optical disector method after confocal microscopy imaging of z-stacks throughout entire tissue sections. Cells were excluded from counts if they transected the three exclusion planes (top, back, and right side, represented by "x" in schematic) and were included if located within the counting frame or transecting the other three sides of the frame (labelled with check marks in schematic). Density counts were calculated using tissue volume determined in Neurolucida using the counting dimensions and width of the tissue, and were averaged for each experimental group by brain region.

Fig 2

Survival during sevoflurane exposure was affected by dexmedetomidine injections. Mortalities occurred after the second and third injections of 10 or 20 µg kg⁻¹ dexmedetomidine (Sevo 2.5 + Dex 10 or Sevo 2.5 + Dex 20) at 120 and 240 min, or after the third injection of 5 µg kg⁻¹ dexmedetomidine (Sevo 2.5 + Dex 5) at 240 min, indicating a dose-related effect. Graphs indicate the fraction of animals surviving during a six-h exposure to 2.5% sevoflurane with or without dexmedetomidine injections. No mortalities occurred in animals exposed to sevoflurane as a single agent (Sevo 2.5) or during co-administration of 1 µg kg⁻¹ dexmedetomidine every two h (Sevo 2.5 + Dex 1). All animals exposed to dexmedetomidine alone, injected three times every 120 min with doses of 5 to 100 µg kg⁻¹, or fasted in room air for six h survived (data not shown).

Fig 3

Heart rate and peripheral oxygen saturation measured during anaesthetic exposure in seven day-old rats. Spot checks were performed every 30 min during a 360-min exposure to sevoflurane 2.5% without (Sevo 2.5, n=10) or with dexmedetomidine at 1 µg kg⁻¹ (Sevo 2.5 + Dex 1, n=10), 5 µg kg⁻¹ (Sevo 2.5 + Dex 5, n=9), 10 µg kg⁻¹ (Sevo 2.5 + Dex 10, n=7), or 20 µg kg⁻¹ (Sevo 2.5 + Dex 20, n=6), or three intraperitoneal injections of increasing doses of dexmedetomidine every two h, Dex 5 µg kg⁻¹ (n=10), Dex 25 (n=10), Dex 50 (n=11), or Dex 100 (n=10). Control animals (n=10), were injected three times with equal volumes of normal saline every two h. Heart rate (A) decreased significantly at dexmedetomidine>25 µg kg⁻¹, but not sevoflurane 2.5%. Addition of dexmedetomidine≥5 µg kg⁻¹ to sevoflurane decreased heart rates, compared with sevoflurane alone. Similarly, addition of dexmedetomidine>1 µg kg⁻¹ diminished oxygen saturation (B) compared with sevoflurane alone, while dexmedetomidine of any dose or sevoflurane by itself did not diminish oxygen saturation compared with controls. Data are represented as mean (SEM) to facilitate comparison; however, non-parametric statistics were used for group comparisons because several data cells violated normality criteria.

Fig 4

Level of anaesthesia and response to painful stimulation measured during anaesthetic exposure in seven day-old rats. Spot checks were performed every 30 min during a 360-min exposure to sevoflurane 2.5% without (Sevo 2.5, n=10) or with dexmedetomidine at 1 µg kg⁻¹ (Sevo 2.5 + Dex 1, n=10), 5 µg kg⁻¹ (Sevo 2.5 + Dex 5, n=9), 10 µg kg⁻¹ (Sevo 2.5 + Dex 10, n=7), or 20 µg kg⁻¹ (Sevo 2.5 + Dex 20, n=6), or three intraperitoneal injections of increasing doses of dexmedetomidine every two h, Dex 5 µg kg⁻¹ (n=10), Dex 25 (n=10), Dex 50 (n=11), or Dex 100 (n=10). Control animals (n=10) were injected three times with equal volumes of normal saline every 2 h. The level of consciousness (A), measured by loss of righting reflex (LORR), was not decreased by dexmedetomidine 5 µg kg⁻¹, compared with controls, but was diminished for dexmedetomidine≥5 µg kg⁻¹ and for all sevoflurane animals, with or without dexmedetomidine. Sevoflurane, even by itself, led to deeper levels of anaesthesia than even the highest dexmedetomidine dose. Response to painful stimulation was diminished for all dexmedetomidine doses; however, sevoflurane by itself provided pain relief superior to even the highest dexmedetomidine dose. Data are represented as mean (SEM) to facilitate comparison, however, non-parametric statistics were used for group comparisons because several data cells violated normality criteria. LORR assessed animals' responses to being turned on their back from 0 to 4: 0 – no response to being placed in supine position, 1 – delayed attempt to right itself, but failing, 2 – delayed, uncoordinated return to upright position, 3 – sedated, but righting themselves in a coordinated fashion, but eventually successful, or 4 – awake, not remaining supine. Pain responses were assessed by applying a mechanical force to the hind paws and grading the withdrawal response from 0 to 4; 0 – no response, 1 – single limb movement to stimulus, 2 – delayed response, but generalized movement, 3 – delayed motor response and subsequent increase in motor activity, or 4 – immediate, vigorous response with vocalization.

Fig 5

Prolonged sevoflurane exposure qualitatively increases the number of neurones undergoing apoptotic cell death, as labelled with the apoptotic marker activated caspase 3. Representative photomicrographs of sections of retrosplenial cortex (left column, RSC) and subiculum (right column, SUB) at 20x magnification from seven day-old rat pups after three injections of normal saline (control), 50 µg kg⁻¹ (Dex 50) or 100 µg kg⁻¹ (Dex 100) dexmedetomidine every two h, or six h of sevoflurane 2.5% without (Sevo 2.5) or with three injections of 5 µg kg⁻¹ (Sevo 2.5 + Dex 5) or 20 µg kg⁻¹ (Sevo 2.5 + Dex 20) dexmedetomidine every two h. Other doses and brain regions not shown. Green puncta represent caspase 3-positive neurons (arrows), which qualitatively increase in number with increased anaesthetic doses. Scale bar equals 100 µm.

Fig 6

High-magnification photomicrograph from a seven day-old rat pup after a six-h exposure to 2.5% sevoflurane demonstrating neurones undergoing programmed cell death. Tissue sections were stained for the postmitotic neuronal marker neuronal nuclei (NeuN, blue, A), the cellular marker propidium iodide (red, B), the apoptotic marker activated caspase 3 (green, C); D represents a merged image of all three stains. Arrows represent neurones at various states of neurodegeneration, expressing activated caspase 3; note dendritic beading as a sign for cellular disintegration (arrowheads). Degenerating neurones are surrounded by immediately adjacent, seemingly viable neurones (*). Scale bar equals 25 µm.

Fig 7

Prolonged anaesthetic exposure increased density of apoptotic neurones in all examined brain regions. Bar graphs represent density counts of dying neurones as assessed by expression of the apoptotic marker activated caspase 3 in representative cortical, thalamic, and hippocampal brain regions in seven day-old rats after a six-h exposure to sevoflurane 2.5% without (Sevo 2.5, n=10) or with dexmedetomidine at 1 µg kg⁻¹ (Sevo 2.5 + Dex 1, n=10), 5 µg kg⁻¹ (Sevo 2.5 + Dex 5, n=9), 10 µg kg⁻¹ (Sevo 2.5 + Dex 10, n=7), or 20 µg kg⁻¹ (Sevo 2.5 + Dex 20, n=6), or three intraperitoneal injections of dexmedetomidine every two h, Dex 5 (n=10), Dex 25 (n=10), Dex 50 (n=11), or Dex 100 (n=10). Dex 0 represents the control group (n=10), which was injected three times with normal saline every two h. Data are represented as bar graphs to facilitate comparison, however, non-parametric statistics were used for group comparisons, because several data cells violated normality criteria.