• Ei tuloksia

LFPI model produced under isoflurane anesthesia or under pentobarbital-based (TOPI) anesthesia result in similar injury phenotype: No differences in lesion size, location or growth rate

N/A
N/A
Info
Lataa
Protected

Academic year: 2022

Jaa "LFPI model produced under isoflurane anesthesia or under pentobarbital-based (TOPI) anesthesia result in similar injury phenotype: No differences in lesion size, location or growth rate"

Copied!
25
0
0

Kokoteksti

(1)

LFPI model produced under isoflurane

anesthesia or under pentobarbital-based (TOPI) anesthesia result in similar injury phenotype:

No differences in lesion size, location or growth rate

UNIVERSITY OF EASTERN FINLAND, A.I.V INSTITUTE, EPILEPSY RESEARCH LABORATORY. ASLA PITKÄNEN GROUP,

SUPERVISORS: RIIKKA IMMONEN, XAVIER EKOLLE NDODE- EKANE

WALTTERI KOSKELO

(2)

2

Table of contents

Abstract ... 3

Introduction ... 3

Animal models and MRI as research tool ... 4

Effect of Anesthesia ... 4

Objectives... 5

Hypothesis ... 5

Methods ... 6

Animal model ... 6

MRI ... 6

Data analysis by unfolded maps ... 7

Statistics ... 10

Results ... 10

Lesion location ... 10

Areas damaged in LFPI model produced under isoflurane anesthetic – final lesion coverage 5 months after injury ... 11

Comparison of cortical areas damaged by LFPI operation between anesthetics – are there signs of selective vulnerability? ... 11

Comparison of lesion dynamics and growth rate between anesthetics ... 12

Lesion size did not differ between the anesthetic cohorts ... 15

Lesion growth from 9 days to 30 days did not differ between the anesthetic cohorts ... 15

Edema size and early lesion size correlated with the final lesion size 5-6 months later ... 16

How is the acute edema size and location associated with the final lesion size and location ... 16

How is the size of early lesion (day 9) associated with the final lesion size and location ... 17

Discussion ... 18

Conclusion ... 22

References ... 22

Supplementary data ... 23

Comparison of lesion growth rate and size in each time point between anesthetics ... 25

(3)

3

Abstract

Animal models are paramount for the research on complex brain diseases such as traumatic brain injury (TBI). The surgical anesthetics influence the animal model produced and are to be accounted for when comparing studies. The present study tested if lateral fluid percussion injury (LFPI) under isoflurane anesthesia caused brain injury to develop and progress the same way as in previous studies where LFPI was induced under an anesthetic cocktail called TOPI. The lesion formation and progression were monitored by T2 weighted magnetic resonance imaging (MRI) at 2 days, 9 days, 30 days and 5 months post-TBI. Lesion areas, anatomical location and growth dynamics were analyzed using unfolded maps and compared to corresponding data of a previous cohort to observe differences between differently anesthetized cohorts. Results show no differences in lesion locations, sizes or progression pattern between differently anesthetized cohorts. This supports the notion that the LFPI model produced under isoflurane is comparable with that produced under TOPI-anesthetic, and thus the research data obtained are comparable. Secondly, the analysis highlights similar prevalence of two lesion phenotypes within both cohorts: 30% of animals develop stable lesion phenotype while the lesion in 70% of animals keeps on expanding until 5-6months post injury.

For this project work, the unfolded maps of isoflurane –cohort were generated based on pre-existing MRI data (EpiBios 2017-2018), and results of spatial and temporal evolvements of lesion morphometry were compared to those obtained in a parallel study with TOPI-anesthetics (TRAUMA1 series 2003-2005 [2]).

Introduction

Traumatic brain injury (TBI) is one of the main causes of disability and mortality in young persons.

It is possible for secondary injuries to appear months or even years after primary impact. These secondary injuries can manifest as devastating symptoms but usually the level of disability among patients varies [1]. To examine these pathological changes, non-invasive imaging is an important tool.

Animal models are usually used since imaging data acquired can be directly correlated with histopathological outcomes and changes in brain function. [1]. These secondary injuries are poorly understood, and they comprise of a combination cellular, molecular and metabolic alterations in the brain causing post-injury disabilities. One of these delayed pathological processes is post traumatic epilepsy (PTE). [2]. A 30-year cumulative research indicates that incidence of epilepsy is 16.7% for severe, 4.2% for moderate and 2.1% for mild traumatic brain injury. With penetrating head injury like bullet wounds, the incidence of PTE goes up to 53%. TBI causes altogether approximately 10-

(4)

4

20% cases of acquired epilepsy. The experimental studies have shown that only a subgroup of animal models have different types of TBI develop epilepsy during monitoring period. [3].

One third of epilepsy patients suffers from drug-refractory seizures and most patients suffers adverse effects from treatments even though there are dozens antiepileptic drugs available [3].

Thus, preventing the development of epilepsy after a TBI could raise patients’ long-term quality of life which makes it a major priority in epilepsy research. Since many of these delayed pathological changes could potentially be reversed, information and deeper understanding of these processes could provide targets for developing recovery enchanting and antiepileptogenic treatments [2].

Animal models and MRI as research tool

In approximately 80% of all PTE cases, epilepsy occurs during the first 2-years after TBI. Because of long follow-up period, studies widely use animal models to investigate the mechanics of TBI and post traumatic epilepsy. As presented earlier, most animal models with different severities of brain injury show similar incidence of TBI develop epilepsy during the follow-up as statistics show for humans. [3][8]. The lateral fluid-percussion injury (LFPI) model is one of the most widely used experimental model in studying the dynamics of the post-TBI cortical lesion progression and the incidence of epilepsy and a clinically relevant rat model of closed head TBI in humans [2] [4]. In LFPI surgery the skull is exposed and circular craniectomy is performed leaving the dura mater intact.

Brain injury is induced with a brief pressure pulse impact against the exposed dura done by fluid- percussion device. LFPI is induced to rats during anesthesia, and the depth of the anesthesia as well as the selection of anesthetics have a major impact on the tissue vulnerability and response to the injury.

Magnetic resonance imaging (MRI) is the most widely used tool in describing the TBI alterations in central nervous system. Common MRI findings after traumatic brain injury are lesion, edema and intracerebral hemorrhage. During the acute phase edema is the leading factor, contrary to the subacute phase where edema has dissolved and lesion has started to grow. [2]. Since the severity of TBI is a major risk factor for PTE, MRI is used as a tool to describe the size and location of the lesion to predict possible PTE cases [3].

Effect of Anesthesia

In earlier studies rats were anesthetized with an anesthetic cocktail called “TOPI-anesthetic”

containing sodium pentobarbital (58 mg/kg), chloral hydrate (60 mg/kg), magnesium sulfate (127.2 mg/kg), propylene glycol (42.8%), and absolute ethanol (11.6%) [2]. Isoflurane is commonly used

(5)

5

on humans during anesthesia and therefore the use of isoflurane promotes the translational aspect of our studies. Isoflurane has the largest circulatory margin of safety of all potent halogenated agents.

Like other inhalation anesthetics, isoflurane has various pharmacodynamics and its effect on human body cannot be explained with any uniform theory. Although inhalation anesthetics exact mechanics is yet unknown, sedative effect seems to be caused by hyperpolarization of neurons in thalamus and alterations on synapses in central nervous system. [5]. Anesthetics also have an effect on the blood pressure and blood-brain-barrier permeability, and that may affect the severity of postinjury hemorrhage and cumulative effects of that. Furthermore, some substances (e.g. magnesium in TOPI cocktail) are neuroprotective putatively alleviating the neuronal damage after the injury. [2].

Since the latest laboratory animal licenses and guidelines require the TOPI to be replaced by isoflurane, our recent study designs (EpiBios-project) have changed TOPI-anesthetic to isoflurane. Now, we should evaluate if the injury in previous model and in current model are comparable or different in some aspect.

Objectives

Our objective was to compare effects of TOPI-anesthetic and isoflurane on traumatic brain injury lesion location and dynamics of lesion progression. The aim of his study was to find which cortical areas were atrophied after LFPI under isoflurane and does the size or location of the lesion differ between isoflurane –anesthetized and TOPI-anesthetized cohorts. We also intended to compare the topographic progression over time 2 days, 9 days, 30 days and 5 months after injury and find differs between anesthetics.

Hypothesis

In the EpiBios-project LFPI was induced to rats under isoflurane while previous studies were done with TOPI-anesthetics. We hypothesized that using isoflurane as anesthetic during LFPI will cause lesion to act same way as TOPI-anesthetized cohorts. To test this hypothesis, we collected MRI T2 data from animals imaged at time points of EpiBios 2 days, 9 days, 30 days and 5 months and we compared them to MRI time points 3 days, 9 days, 23 days, and 6 months of the TOPI-cohort.

(6)

6

Methods

Animal model

In the EpiBios-project lateral fluid percussion injury was induced to 15 rats under isoflurane anesthesia and executed with the technique described by Kharatishvili and her colleagues [4]. The skull was exposed and a 5 mm diameter circular craniectomy was performed leaving the dura mater intact. Location of craniectomy was standardized as described earlier (at anterior edge 2,0 mm posterior to the bregma and the lateral edge adjacent to the lateral ridge). A brief (21-23 ms) transient fluid pressure pulse impact against the exposed dura was done by fluid-percussion device with adjusted force of 2.6-3.3 atm inducing the LFPI. During this procedure rats were under isoflurane- anesthesia. Isoflurane was chosen to investigate if the histopathological outcomes and changes in brain function between different anesthetics differ during follow-up. All animal procedures were conducted in accordance with the guidelines of animal ethics committee of the Provincial Government of Southern Finland and fulfilled the guidelines of the European Community Council Directives (86/609/EEC).

MRI

At time-points of 2 days, 9 days, 30 days and 5 months MRI data was acquired at 7T magnet with actively shielded gradients using volume coil as a transmitter and quadrature surface RF coil as a receiver. For imaging rats were placed into stereotactic holder and kept under 1.7% isoflurane anesthesia (with carrier gas 70% N2O / 30% O), physiological temperature was maintained by heating bed and rectal temperature monitoring. T2-weighted images for optimal contusion and lesion detection were obtained by fast spin echo sequence (TurboRARE) with 3400 ms repetition time, 45 ms echo time, RARE factor 8, 3 averages, and scan time of 5min26sec. Total 23 coronal slices were taken with 0.80 mm slice thickness, field of view (FOV) 30 x 30 mm covered by 256 x 256 matrix resulting in images with 0.117 x 0.117 mm in-plane resolution (Figure 1A-B). The T2 weighted images of TOPI-cohort from TRAUMA1-project (Nov2003-June2004) were acquired with 4.7 T Varian magnet, and actively decoupled volume transmitter and quadrature surface receiver with similar fast spin echo technique and comparable resolution. Time-points in TOPI-cohort are 3 days, 9 days, 23 days (compared to 1 month), and 6 months (compared to 5 months).

(7)

7

Figure 1. (A) In T2-weighted MRI image (coronal plane) of a representative rat 2 days postinjury the contusion complex shows as hyperintensity with some hypointense hemorrhages and cortical swelling (Arrow). (B) In-plane resolution of 117x117micrometers is achieved with FOV 30x30mm covered by 256x256 matrix, shown is a portion of cortex including the lesion. Scale bar equals 1mm.

Data analysis by unfolded maps

MRI images were used to generate 2D unfolded maps of the cortical lesion according to the technique described by Ndode-Ekane and his colleagues [6] using the rat brain atlas for reference points [7]. In this study we used ImageJ-application to measure the size of lesion (Figure 2) from multi slice MRI images. Measurement was done by drawing a line with segmented line tool along the surface of brain.

Line was divided into three segments (Line 1, Line 2 and Line 3) and started from corpus gallosum to rhinal fissure. Each line was measured and length of line 2 mirrors the size of the lesion in one slice. This procedure was repeated to each (coronal) MRI image plane, covering the entire lesion depth. The measurements were downloaded to unfolded map application created by Ndode-Ekane and colleagues to generate 2D map presenting which cortical areas were atrophied by lesion (Figure 3). The application imports and superimposes simple user-made measurements from histologic sections or MRI images on a template. The application then quantifies the total lesion area and the area of damage in each cortical cytoarchitectonic region.

(8)

8

Figure 2. ImageJ is used to measure the size of the lesion after 30 days from LFPI-operation

Figure 3. Unfolded map of rat brain with lesion marked as blue area. Tablet on right shows which cortical areas are atrophied and amount of damage in percent.

Lesion location and size was measured from MRI images by using ImageJ and each coronal plate was uploaded into ImageJ as .tiff files. The bregma scale was set based on the atlas [7]

to identify in which depth of the brain each plate was representing (Figure 5B-C). Twenty-three coronal planes located between the anteroposterior (AP) coordinates +3.0 to -9.24 from bregma were included in the template at 0.80 mm intervals. The surface length of each lesion was measured by using the segmented line tool in ImageJ (Figure 5A). Since the size of lesion differs in different depths of cortex, the measurement points of lesions’ length were standardized. The measurement depth for lesion was set to the layer of large pyramidal cells of the cortex. First line was drawn from medial reference point which was set to be where cortex contracted the corpus callosum to the point on

(9)

9

surface where lesion started to show (See figure 4Aa). The second line was drawn from the medial starting point of the lesion to the most lateral point of the lesion (See figure 4Ab). The third line was drawn from the most lateral part of the lesion on the surface to the rhinal fissure which was set to be the lateral reference point (See figure 4Ac). The procedure was repeated to each coronal plate and lines were measured. The length of line two represented the size of the lesion and lines one and three were used to locate the lesion on each plate. Each measurement was collected to a excel-file and images with drawn lines were saved for later inspections.

Figure 4. Measuring the size of the lesion. Example images from rat 1008 at time point of 5 months shows the size of the lesion measured and the bregma scale identified. (A) The MRI was used to measure the size of the lesion by drawing three lines (a-c) in ImageJ and measuring the length of each line. (B-C) The location of the plate between the anteroposterior (AP) coordinates was identified by using atlas [7] and setting bregma scale for each plate. As displayed, images B and C can be identified to be from the same depth of the brain. Note that image A is from different depth than B and C.

The measurements from ImageJ and estimated bregma scales were uploaded to unfolded map application created by Ndode-Ekane and colleagues from excel-file to generate two- dimensional unfolded cortical maps presenting which cortical areas were influenced by lesion (Figure 3). The application then quantifies the total lesion area and the area of damage in each cortical cytoarchitectonic region from measured lines in each bregma point. The procedure was repeated to each time point for every animal. The measurements were extracted from unfolded map application and saved to excel-file. Results are reported as square millimeters. The sizes, location and progressions of lesions of each individual rat were studied and compared between the two anesthetic- cohorts.

A B

C

a b

c

(10)

10 Statistics

To elucidate the dynamics of the lesion progression, lesion areas over time are presented for each individual animal. Data is compared to that of animals with the “TOPI” protocol at the individual level and at the group level. Division of the animals into different endophenotypes, and whether the lesion dynamics differ between anesthetics is descriptively analyzed. Size of the lesion, progression speed of the lesion and which cortical areas are atrophied are compared between cohorts.

Group differences are analyzed with Student’s t-test, and difference over time by related samples t- test. Association between acute edema (at day 2) and final lesion size were analyzed using linear regression and Pearson correlation.

Results

Lesion location

Figure 5 is a collage of multi-slice T2 weighted MRI showing a representative injury in a LFPI rat 30 days after the operation. Cortical lesion appears bright and is visually discernable.

Figure 5. Example mosaic of coronal slices of a rat brain at time point of 30 d after lateral FPE operation. Lesion can be seen as hyperintensity with some hypointense hemorrhages and cortical swelling.

(11)

11

Areas damaged in LFPI model produced under isoflurane anesthetic – final lesion coverage 5 months after injury

The total lesion area and the area of damage in each cortical cytoarchitechtonic region were quantified. Damage percentages in each region are reported. The areas damaged at end-stage (time point of 5 months) are displayed in Supplementary Table 1 for isoflurane-cohort, and in Supplementary Table 2 for TOPI-cohort (6 months), and compared in Figure 6. In isoflurane cohort, areas that had averagely more than 50% damaged by lesion are Au1 (primary auditory area), AuD (secondary auditory cortex, dorsal area), S1 (Primary somatosensory cortex), PtPR (rostral posterior parietal cortex) and V2L (secondary visual cortex, lateral part). The most damaged area in this study was S1 with average of 92.6% of area destroyed by lesion caused by LFPI. Areas that had averagely 20-50% damaged by lesion are TeA (temporal cortex, association area), AuV (secondary auditory cortex, ventral area), S1BF (somatosensory 1, barrel field), PtPD (dorsal posterior parietal cortex), V1B (primary visual cortex, binocular region) and V2ML (visual cortex 2, mediolateral part). Other areas shown in figure 6 had damage average less than 20%. Cortical areas that are not listed in figure 6 were not affected by LFPI-operation.

Comparison of cortical areas damaged by LFPI operation between anesthetics – are there signs of selective vulnerability?

One of the main purposes of this study was also to compare the lesion location and damage done to different cortical areas in percentages after LFPI operation between isoflurane- anesthetized and TOPI-anesthetized cohorts. The areas damaged at end-stage (time point of 6 months) for TRAUMA1 series 2003-2005 (LFPI operated under TOPI anesthetics) are displayed in Supplementary Table 2. Similar table for EpiBios C1 rats was presented in Supplementary Table 1.

When comparing average damages for each individual cortical area there appears some variations between cohorts. Bar chart in figure 6 elucidates the differences.

(12)

12

Figure 6 – Each damaged cortical area is presented individually by their abbreviation. Columns are presenting damage done to cortical area in percentage for differently anesthetized cohorts. EpiBios C1 was anesthetized with isoflurane and TRAUMA1 series 2003-2005 was anesthetized with TOPI-anesthetic. AVERAGE of cohort or MAX damage of cohorts. Abbreviations: Primary auditory area; Au1, secondary auditory cortex, dorsal area; AuD, secondary auditory cortex, ventral area; AuV, dysgranular insular cortex; DI, ectorhinal cortex; Ect, granular insular cortex; GI, lateral parietal association cortex; LPtA, perirhinal cortex; PRh, dorsal posterior parietal cortex; PtPD, rostral posterior parietal cortex; PtPR, primary somatosensory cortex; S1, somatosensory 1, dysgranular zone; S1DZ, somatosensory 1, barrel field; S1BF, upper lip of the primary somatosensory cortex; S1ULp, second somatosensory cortex; S2, temporal cortex, association area; TeA, primary visual cortex; V1, primary visual cortex, binocular region; V1B, primary visual cortex, monocular region; V1M, secondary visual cortex, lateral part; V2L, visual cortex 2, mediolateral part; V2ML, visual cortex 2, mediomedial part; V2MM.

Comparing the %-damage in each area between anesthetic cohorts show similar overall coverage while some variations can be seen. Both anesthetics cause the lesion to develop in the same areas. That is, generally the same functional regions are affected, but to a different extent. DI and GI are spared in isoflurane-cohort but affected only in 2 animals of TOPI-cohort. In some cortical areas difference between anesthetics is diminutive (e.g. secondary auditory cortex dorsal area, primary somatosensory cortex, and visual cortex 2 mediolateral part) contrary to some areas (e.g.

somatosensory 1 dysgranular zone, primary visual cortex monocular region and lateral parietal association cortex) where percentage of damaged area differs considerably between anesthetics.

Overall, the inter-animal variation within each cohort is greater than the variation between the cohorts.

This suggests that the LFPI model generated under isoflurane anesthesia results in comparable cortical injury topography than in LFPI model generated under TOPI anesthetics.

Comparison of lesion dynamics and growth rate between anesthetics

Generally, at the time point of 2 days after LFPI operation MRI shows the size of formed edema. Thereafter the edema dissolves and the lesion become observable. [1]. To inspect the progression of the lesion MRIs were taken at time points of 2 days, 9 days, 30 days and 5 months.

Figure 7 and 8 show the topography of how the large edematic area at day 2 is partly resolved by day 9 leaving the atrophic lesion core. Thereafter the atrophy and necrosis progress and the lesion cavity expand. Cavity is filled with cerebro spinal fluid (CSF).

0 20 40 60 80 100

Damaged area (%)

Cortical areas damaged by lesion on different anesthetics

EpiBios C1 Trauma1 series 2003-2005

(13)

13

Figure 7. Progression of the lesion during the 5 months follow-up in a representative rat.

Figure 8. Unfolded maps at different time points. (A) Unfolded map of the edema from rat 1008 at time point of 2 days. (B)Time point of 9 days where lesion has started to show after edema has dissolved. (C) Unfolded map from time point of 30 days showing how the lesion has started to expand. (D) Final time point, 5 months, where the final size of the lesion is shown.

MRI images were used to generate 2D unfolded maps of the cortical lesion according to the technique described by Ndode-Ekane and his colleagues [6]. Cortical lesion areas were measured and lesion areas over time are presented in a bar plot for each individual animal (Figure 9).

In bar plot we can see that usually the size of the lesion is higher at day 2 than day 9. This is because of the swelling of brain tissue after LFPI causing an edema. The actual lesion is visible after edema dissolves which have happened at time point of 9 days. Therefore, actual lesion size and its progression is sensible to be observed from time pints of 9 days, 30 days and 5 months. Day 2 measures can be used for example to measure diagnostic power of the edema volume on the size of the lesion to develop.

(14)

14

Figure 9 demonstrates that not every progression is like each other. Size of damaged area is highest at day 2 after LFPI operation because of the edema. In every case edema dissolves and underlying lesion is smaller than edema (seen at time-point of 9 days). Size of the lesion grows bigger then original edema in some follow-ups but always this is not the case. In some cases, lesion doesn’t grow at all or very marginally. The growth pattern suggests that there are two subgroups:

Ones whose lesion is stable and doesn’t grow over time and others whose lesion is expanding over the observed time points. This can be observed from plot chart (Figure 9). Rats 1019, 1031, 1036,

0 5 10 15 20 25 30 35 40 45 50

1008 1010 1012 1018 1019 1024 1027 1028 1029 1031 1036 1038 1043 1044 1046

Lesion area (mm2 )

Rats

A Lesion size over time in isoflurane-cohort

2d 9d 30d 5mo

0 10 20 30 40 50 60 70 80

2 3 4 5 6 10 13 14 15 16 20

Lesion area (mm2 )

Rats

B Lesion size over time in TOPI-cohort

3d 9d 23d 6mo

Figure 1. Lesion growth over time in individual animals in both anesthesia cohorts. (A) Isoflurane-cohort. Lesion size progression on different Isoflurane-cohort rats over time shows that there are two subgroups. On the first subgroup the lesion is stable and doesn’t grow over time (Rats 1019, 1031, 1036, 1038). On the second subgroup the lesion is expanding over time, from 9 days to 5 months (Rats 1008, 1010, 1012, 1018, 1024, 1028, 1029, 1043, 1044, 1046). First column (Lightest blue) is the size of the lesion 2-3 days after LFPI operation and is highest because of the edema size. In every case edema dissolves and underlying lesion is smaller than edema. Size of the lesion grows bigger then original edema in some animals. (B) TOPI-cohort. Lesion growth over time in each rat shows that after the edema has resolved the lesion size remains stable in rats 14, 15 and 20, while keeps expanding in 2,3,4,5,6,10,13 and 16.

(15)

15

1038 from isoflurane-cohort represents the first subgroup whose lesion stays stable and the size of the lesion doesn’t grow during follow-ups. Rats 1008, 1010, 1012, 1018, 1024, 1028, 1029, 1043, 1044, 1046 from isoflurane-cohort represents the subgroups whose lesion is expanding over the chosen time points. Data from rat 1027 couldn’t be collected after time point 9 days so its lesion progression stays unknown. That is, 4/14 (29%) have stable, and 10/14 (71%) have expanding phenotype in the cohort operated under isoflurane. Similarly, in the cohort operated under TOPI- anesthestetic, 3/11 (27%) have stable, and 8/11 (73%) have expanding phenotype (Figure 9B). This indicates that LEPI operation under either anesthetic have similar division of stable vs expanding cases.

Lesion size did not differ between the anesthetic cohorts

The lesion sizes in each time point was compared between cohorts, and they did not differ. Average edema size at day 2 post-injury were 32.2 ±10.1 mm2 in isoflurane cohort, and 38.9

± 14.8 mm2 at day 3 in TOPI-cohort (p>0.05). Average lesion size at day 9 post-injury were 16.4 ± 8.3 mm2 and 18.1 ± 14.5 mm2, respectively. At day 30 post-injury the average lesion size was 22.8

±10.7 mm2 in isoflurane cohort, and at 23 days post-injury 22.3 ± 17.6 mm2 in TOPI-cohort. Final lesion sizes were 26.2 ± 12.4 mm2 and 30.6 ± 20.8 mm2, respectively. Variation in lesion sizes were large in both cohorts.

Lesion growth from 9 days to 30 days did not differ between the anesthetic cohorts

Lesion growth rates as the relative change of size and absolute change of size from 9d to 30 days / 23 days post-injury for both anesthetics were calculated (Supplementary Tables 5-6). The absolute increase in lesion size from day 9 to day 30 days was 5.9 ± 5.5 mm2 in isoflurane cohort and 4.2 ± 4.3 mm2 from day 9 to day 23 in TOPI-cohort (p=0.410). The relative increase in lesion size from day 9 days to day 30 was 36 ± 37% in isoflurane cohort and 29 ± 43% from day 9 to day 23 in TOPI-cohort (p=0,655). Thus, the cohorts did not differ in their growth. The lesion growth in each animal are also demonstrated in scatter diagram (figure 13) which shows that there are no observable clusters but the distributions are overlapping.

(16)

16

Figure 13 – Lesion growth is measured for each animal and presented in scatter diagram dark dots represents TOPI-anesthetized animals and light dots represents isoflurane anesthetized animals. Since there are no observable clusters diagram indicates that cohorts do not differ in their growth.

Edema size and early lesion size correlated with the final lesion size 5-6 months later

As seen in figure 9, the standardized LFPI operation causes different types of lesion progressions. In addition, the final lesion sizes vary between animals. Observing the plot chart (figure 9) the end-stage lesion (5 months after LFPI) seems to correlate with the size of the original edema.

The bigger is the edema at time point of 2 days, larger is the lesion after 5 months after LFPI operation.

Also, in some cases it seems like that rats with larger lesion at time point of 9 days after LFPI ends up with larger lesion than those with small lesion at time point of 9 days. To test these hypotheses 2- tailed Pearson correlation was used to evaluate whether there is correlation between early and final lesion sizes.

How is the acute edema size and location associated with the final lesion size and location

The edema size correlated positively with the end-stage lesion size (r=0.608, p=0.028, n=13, Pearson), Figure 10. Thus, bigger edema after TBI predicts larger lesion to be developed.

Interestingly however, the unfolded maps show that the area of the early edema is not an exact match of the final atrophy area (shown in Figure 11). That is, the cell death and tissue degradation occur beyond the regions pinpointed by the early edema.

-5,0 0,0 5,0 10,0 15,0 20,0

Lesion growth (mm2)

Growth from 9d to 23/30d

TOPI ISOFLURANE

(17)

17

Figure 2 – Correlation between the size of edema at time-point of 2 days and the size of the lesion after 5 months. r=0.608, p=0.028, n=13, Pearson. Measurements used is presented in Figure 7.

Figure 11 – Comparing the area of the early edema (Time point of 2 days) and the final atrophy area (Time point of 5 months) we can observe that the final stage lesion is not an exact match with the edema formed after LFPI operation. Images are from EpiBios C1 rat 1024.

How is the size of early lesion (day 9) associated with the final lesion size and location

To investigate whether the size of the lesion at day 9 (i.e. irreversibly damaged cortical tissue) would predict the final lesion size better than the size of the transient edema, the correlation between 9d and 5mo lesion sizes were also examined. Result is shown in Fig 12 (figure 12). Trend of positive correlation can be seen (r=0.545, p=0.054, n=13), but the Pearson correlation p-value was

>0.05, i.e. not quite statistically significant. Between day 9 and 5 months the lesion grew in all directions

0 10 20 30 40 50

15 20 25 30 35 40 45

Lesion area after 5 months (mm2)

Edema area (mm2)

Correlation

(18)

18

Figure 3 - Correlation between the size of the lesion at time-point of 9 days and the size of the lesion after 6 months. r=0.545, p=0.054, n=13, Pearson. Measurements used is presented in Figure 6.

Discussion

In this study we focused on lesion development and procession on isoflurane-anesthetized rats (EpiBios C1) and compared the results to previous data collected from TOPI-anesthetized rats from TRAUMA 1 series 2003-2005. To understand the dynamics of brain damage after TBI,

a lateral fluid-percussion model of TBI in rat and quantitative MRI were performed. In our 5 months follow-up study we found 2 major findings. First, brain damage after LFPI is more common to grow continually over the follow-up period than remain stable. In 10 of 14 (71%) cases the lesion continued to grow. The same result was found in a study by Immonen et al, 2009 [2]. Second, we found that bigger edema after TBI predicts larger lesion to be developed during follow-up. P-value for correlation between size of edema and end-stage lesion was 0.028 indicating statistical significance.

Differently, correlation between lesion sizes at time points of 9 days and 5 months were not statistically significant (p>0.05).

Studies suggest that isoflurane directly affects brain vascular endothelial cells to increase blood-brain barrier (BBB) permeability which promotes plasma leaks from blood vessels to brain tissue [9]. After brain injury edema is developed around damaged brain tissue. Increased BBB permeability causes inflation on plasma leakage which might boost the expansion of the edema. In addition, increased permeability on BBB and promoted plasma leaks increases the chance of bleeding from damaged blood vessels. This might cause additional intracerebral hemorrhages after brain injury and infiltration of blood-derived immune cells causing inflammation. During inflammation immune cells excrete inflammatory factors which damages brain tissue further. Since LFPI was executed under isoflurane-anesthesia, changes in BBB permeability might accelerate the progression and

y = 1,059x + 7,8086 R² = 0,297

0 5 10 15 20 25 30 35 40 45 50

0 5 10 15 20 25 30

Lesion area after 5 months (mm2)

Lesion area after 9days (mm2)

Correlation

(19)

19

enhance the size of the lesion after brain injury. Therefore, end-stage sizes of lesions caused by LFPI under isoflurane anesthesia might differ from brain injuries occurred without isoflurane anesthesia.

In turn, TOPI-anesthetic contains chloral hydrate which is considered as neuroprotective substance. Chloral hydrate upregulates anti-inflammatory mechanisms which reduces brain edema after injury and attenuate neuronal death. [10]. This could suggest that TOPI- anesthesia reduces the size and inhibits the progression of the lesion when used during LFPI operation. Therefore, brain injuries under TOPI-anesthetic could cause lesser damage when compared to brain injuries occurred under isoflurane-anesthesia. In addition, isoflurane could enhance the size of end-stage lesion which causes more definite differences to occur when comparing the anesthetics.

During the comparison between TOPI-anesthetized and isoflurane-anesthetized cohorts, 3 major findings were found. First, the size and procession of the lesion doesn’t seem to differ between anesthetics. The division between stable and grooving lesion types were nearly the same (Isoflurane cohort had 4/14 (29%) stable cases and TOPI cohort had 3/11 (27%) stable cases).

Second, in those cases when lesion processed to grow continually over follow-up period there were no difference in growth rate or end-stage sizes between cohorts. When T-test were executed the growth rate as the absolute change of size from 9d to 30d between groups resulted p=0,410 (>0,05) indicating that there is no statistical significance. In addition, t-test for growth rate as the relative change of size from 9d to 30d between groups gave p-value p=0,655 (>0,05) which indicates that cohorts do not differ. This study indicates that LFPI operation under either anesthetic have similar result in its entirety. Third, the location of the lesion occurs in the same cortical areas with both anesthetics. In both cases areas with most damage by the lesion in percentages were in this study secondary auditory cortex dorsal area (AuD), primary somatosensory cortex (S1), rostral posterior parietal cortex (PtPR) and secondary visual cortex, lateral part (V2L). The most damaged area in this study was S1 with average of 92.6% of area destroyed by lesion. This might suggest that LFPI operation most likely influences brain ability to process information from sensory receptor cells such as mechanoreceptors, chemoreceptors, and nociceptors causing problems with sense of touch, proprioception, and haptic perception.

In summary, this study indicates that both anesthetics causes similar lesion profile when used during LFPI operation although isoflurane was hypothesized to enhance the size of end-stage lesion when comparing the anesthetics. Only minor variations can be observed between anesthetics which are not statistically significant. This indicates that data collected previously using TOPI- anesthetic is comparable with the data from current model done with isoflurane-anesthesia. Because LFPI is not controlled without anesthetics, it is difficult to estimate the effect of anesthetics on the

(20)

20

size of the injury when comparing to brain injuries occurring outside of anesthesia. Since anesthesia impacts various organ functions (e.g. suppresses central nervous system and cardiovascular system) it may be assumed that anesthetics might influence the dynamics of the lesions as well.

Anesthetics has vast effects on various organs and therefore might influence the progression and end-stage size of the lesions. In addition for increasing BBB permeability as discussed earlier, isoflurane reduces cardiac output and dilates blood vessels causing mean arterial pressure (MAP) to decrease during anesthesia [11]. Lower blood pressure causes ruptures on arteries to bleed less to brain tissue as the cerebral perfusion pressure (CCP) decreases and blood flows slower from ruptures and coagulates faster. This might suggest that isoflurane could have a protective property for internal hemorrhages during LFPI operation. After injury infiltration of blood-derived immune cells occurs causing inflammation to brain tissue. Immune cells release inflammatory factors that damages neurons and increase edema formation in injured brain tissue. Edema in turn causes pressure to surrounding brain tissue causing ischemia and additional neuronal death. Anesthetics (e.g.

isoflurane) directly suppresses inflammatory response by suppressing cellular and neurohumoral immunity through inflammatory gene expression and inflammatory factor secretion [12]. This anti- inflammatory effect of anesthetics may slow down the expansion of the lesion over time and reduce the end-stage lesion size by attenuating neuronal death. Therefore, it is possible that traumatic brain injuries occurring without anesthesia could cause slightly larger lesions than ones caused by LFPI operation.

Anesthetics also dilates cerebral blood vessels thus increasing cerebral blood volume, which increases intracranial pressure (ICP). Increased ICP reduces cerebral flow and perfusion causing brain to be more susceptible to injuries since less nutrients circulates to brain tissue. In addition, during post traumatic inflammatory energy consumption is even higher promoting additional neuronal death through ischemia. This suggest that anesthetics could also have a negative effect on lesion progression causing additional neuronal damage. In turn, increased ICP and reduced perfusion diminish plasma leakage to brain tissue after injury which might reduce the expansion of the edema and lessen the end-stage lesion size. Therefore, it is complex to predict whenever anesthetics have a more positive or more negative affect when it comes to lesion progression and lesion size after LFPI operation. When excessive amounts on anesthetics is used ICP could approach the level of MAP, cerebral perfusion might stop all together. This causes body to raise systemic blood pressure and dilate cerebral blood vessels to prevent ischemia. This results in increased cerebral blood volume, which increases ICP lowering CPP further and could cause a vicious cycle leading to

(21)

21

swelling of brain and eventually to brain death. Although when correctly used anesthetics might have a reducing effect on lesion after TBI.

In Finland the most used anesthetic substance on humans is propofol which is used for induction and maintenance of general anesthesia intravenously. It is the latest iv-anesthetic brought into wide use. The only gas-anesthetic currently used is nitrous oxide also known as laughing gas and is widely used as an analgesic during minor operations such as giving birth or dental procedures.

Isoflurane, which is the anesthetic used in this study, is a halogenated ether and is one of the inhaled anesthetics used on humans since 1980s. Compared to nitrous oxide isoflurane is more potent anesthetic and therefore can be used for induction and maintenance of general anesthesia. Like other inhalation anesthetics, isoflurane has various pharmacodynamics and its effect on human body cannot be explained with any uniform theory. Although it’s specific pharmacodynamic mechanisms are yet unknown it is suspected to affect GABAA -receptors as propofol does which explains the similar anesthesia profiles. [11]. Similar effects and mechanisms support the hypothesis that anesthetics used in this study has similar effects on brains during TBI and progression of lesions as propofol has. This suggests that studies performed under isoflurane anesthesia corresponds to clinical cases under propofol anesthesia and our results may potentially be utilized for clinical studies. Also, isoflurane is commonly used on humans during anesthesia and therefore the use of isoflurane promotes the translational aspect of our studies.

The LFPI reproduces many of the features of human PTE and researches done previously (D’Ambrosio et al, 2004) suggests an epileptogenic role for the neocortex at the site of injury [13]. However, some earlier studies indicate that PTE may be independent when observing connection between PTE and major structural, functional, and behavioral changes induced by TBI [14]. Thus, the model and data collected from this study can be used in future studies for identifying specific epileptogenic mechanisms and discovering antiepileptogenic agents. In clinical use these results could potentially shed light to identifying which traumatic brain injuries will most likely cause PTE to develop. Identifying those cases and discovering specific epileptogenic mechanisms could ease the prevention of PTE, enable prophylaxis medication to be initiated after injury on apparent cases and possibly lead to new drugs to be develop for controlling epileptic seizures. These are potentially flourishing field of research since epileptic seizures are poorly controlled by currently available drugs [15].

(22)

22

Conclusion

This study indicates that LFPI operation under either anesthetic have similar result in its entirety. In both cases, two subgroups in lesion progression emerge: Ones whose lesion is stable and doesn’t grow over time and others whose lesion is expanding over the observed time points. Anesthesia cohorts have similar portion/division of stable vs expanding cases. There was also no difference between anesthetics when comparing the size or the location of the lesion in any of the corresponding time points. Hence, cohorts done with isoflurane-anesthetized and TOPI-anesthetized rats are comparable with each other.

References

[1] Immonen, R. J., Kharatishvili, I., Gröhn, H., Pitkänen, A., & Gröhn, O. H. J. 2009. Quantitative MRI predicts long-term structural and functional outcome after experimental traumatic brain injury.

NeuroImage, 45(1), 1–9.

[2] Immonen, R. J., Kharatishvili, I., Gröhn H., Niskanen J., Pitkänen, A., & Gröhn, O. H. J. 2009. Distinct MRI pattern in lesional and perilesional area after traumatic brain injury in rat — 11 months follow-up.

Experimental Neurology, 215, 29-40.

[3] Immonen, R. J., Kharatishvili, I., Gröhn, O., Pitkänen, A. 2013. MRI Biomarkers for Post-Traumatic Epileptogenesis. Journal of Neurotrauma, 30, 1305-1309.

[4] Kharatishvili, I., Nissinen, J.P., McIntosh, T.K., Pitkanen, A. 2006. A model of posttraumatic epilepsy induced by lateral fluid-percussion brain injury in rats. Neuroscience 140, 685–697.

[5] EI Eger 2nd, The pharmacology of isoflurane, British journal of anaesthesia, 1984

[6] Ndode-Ekane, X. E., Kharatishvili, I., Pitkänen, A. 2017. Unfolded maps for Quantitative Analysis of Cortical Lesion Location and Extent after Traumatic Brain Injury. Journal of Neurotrauma, 34, 459-474.

[7] Franklin, K., Watson, C. 2007. The rat brain in Stereotaxic Coordinates Sixth Edition Elsevier Inc.:

London.

[8] Annegers JF, Rocca WA, Hauser WA. Causes of epilepsy: contributions of the Rochester epidemiology project. Mayo Clin Proc 1996; 71: 570–5.

(23)

23

[9] Nimish K. Acharya, Eric L. Goldwaser, Martin M. Forsberg, George A. Godsey, Cristina A. Johnson, Sevoflurane and Isoflurane induce structural changes in brain vascular endothelial cells and increase blood−brain barrier permeability: Possible link to postoperative delirium and cognitive decline. Brain Research, Volume 1620, 2015, Pages 29-41,

[10] Liu, J., Feng, D., Zhang, Y., Shang, Y., Wu, Y., Li, X. and Pei, L. (2015), Chloral Hydrate Preconditioning Protects Against Ischemic Stroke via Upregulating Annexin A1. CNS Neurosci Ther, 21: 718- 726

[11] Farmakologia ja toksikologia, Medicina, Volume 6, Pages 329-342

[12] Kurosawa, S. & Kato, M. J Anesth, Anesthetics, immune cells, and immune responses, Journal of anesthesia (2008) 22: 263.

[13] Raimondo D’Ambrosio, Jared P. Fairbanks, Jason S. Fender, Donald E. Born, Dana L. Doyle, John W. Miller; Post‐traumatic epilepsy following fluid percussion injury in the rat, Brain, Volume 127, Issue 2, 1 February 2004, Pages 304–314

[14] Sandy R. Shultz et al, Epilepsia,Volume54, Issue7 July 2013 Pages 1240-1250

[15] Löscher W, Schmidt D. New horizons in the development of antiepileptic drugs. Epilepsy Res 2002, 50, 3–16.

.

Supplementary data

Supplementary Table 1 – Cortical areas damaged on EpiBios C1 rats: LFPI done under isoflurane anesthetics. In this table the damage on different cortical areas are displayed as percentages at time point of 5 months. Damage for each individual rat is displayed separately, and their reference number is shown in first row. Abbreviations: Primary auditory area; Au1, secondary auditory cortex, dorsal area; AuD, secondary auditory cortex, ventral area; AuV, dysgranular insular cortex; DI, ectorhinal cortex;

Ect, granular insular cortex; GI, lateral parietal association cortex; LPtA, perirhinal cortex; PRh, dorsal posterior parietal cortex;

PtPD, rostral posterior parietal cortex; PtPR, primary somatosensory cortex; S1, somatosensory 1, dysgranular zone; S1DZ, somatosensory 1, barrel field; S1BF, upper lip of the primary somatosensory cortex; S1ULp, second somatosensory cortex; S2, temporal cortex, association area; TeA, primary visual cortex; V1, primary visual cortex, binocular region; V1B, primary visual cortex, monocular region; V1M, secondary visual cortex, lateral part; V2L, visual cortex 2, mediolateral part; V2ML, visual cortex 2, mediomedial part; V2MM. Average damage for each cortical area was calculated and is displayed at the last column (highlighted in green ).

(24)

24

Supplementary Table 2 – Cortical areas damaged by lesion of TRAUMA1 series 2003-2005: LFPI done under TOPI anesthetics. In this table the damage on different cortical areas are displayed as percentages at time point of 6 months. Damage for each individual rat is displayed separately, and their reference number is shown in the first row. Abbreviations of cortical areas are displayed at the left side of table. Abbreviations: Primary auditory area; Au1, secondary auditory cortex, dorsal area; AuD, secondary auditory cortex, ventral area; AuV, dysgranular insular cortex; DI, ectorhinal cortex; Ect, granular insular cortex; GI, lateral parietal association cortex; LPtA, perirhinal cortex; PRh, dorsal posterior parietal cortex; PtPD, rostral posterior parietal cortex; PtPR, primary somatosensory cortex; S1, somatosensory 1, dysgranular zone; S1DZ, somatosensory 1, barrel field; S1BF, upper lip of the primary somatosensory cortex; S1ULp, second somatosensory cortex; S2, temporal cortex, association area; TeA, primary visual cortex; V1, primary visual cortex, binocular region; V1B, primary visual cortex, monocular region; V1M, secondary visual cortex, lateral part; V2L, visual cortex 2, mediolateral part; V2ML, visual cortex 2, mediomedial part; V2MM. Also, average damage for each cortical area was calculated and is displayed at the right side of table (highlighted in green).

(25)

25

Comparison of lesion growth rate and size in each time point between anesthetics

The aim of this study was to compare the lesion sizes and dynamics after LFPI operation between isoflurane-anesthetized and TOPI-anesthetized cohorts. The lesion sizes (mm2) for each individual rat over time are presented in Supplementary Tables 4-5. The calculated average lesion size and the standard deviation (STD) are shown in last two columns.

Rats 1008 1010 1012 1018 1019 1024 1027 1028 1029 1031 1036 1038 1043 1044 1046 AVERAGE STD

2d

31,4 35,4

36,7 30,4

20,2 27,5

28,0 32,1

29,3 23,8

32,7 41,0

32,8 39,6

42,6 32,2

10,1

9d

19,3 10,7

14,1 11,4

7,2 15,6

9,9 17,5

11,1 10,9

19,7 25,4

23,3 28,2

22,4 16,4

8,3

30d

34,1 19,4

32,1 12,1

7,4

24,1 - 20,7

11,3 13,9

24,9 27,3

33,6 34,0

24,8 22,8

10,7

5mo

46,9 - 44,3

17,9 4,3

26,6 - 28,2

18,2 11,1

21,3 20,7

32,7 34,7

34,2 26,2

12,4

Supplementary Table 4- Lesion sizes (mm2) of EpiBios C1 rats: LFPI operated under ISOFLURANE. Rats 1019, 1031, 1036, 1038 represents the first subgroup whose lesion stays stable. Rats 1008, 1010, 1012, 1018, 1024, 1027, 1028, 1029, 1043, 1044, 1046 represents the other subgroups whose lesion is expanding over the chosen time points.

Rats

2

3 4 5 6 10 13 14 15 16

20 AVERAGE STD

3d 36,9

30,0 33,6

32,2 40,1

45,6 36,7

19,5 59,8

69,7

23,7 38,9 14,8 9d 5,5

17,0 16,5

12,8 33,7

18,1

10,5 9,5

9,5 55,8

10,5 18,1 14,5 23d 13,6

24,6

22,2 9,2

40,2 23,2

12,7 10,0

11,8 67,5

11,0 22,3 17,6 6mo 29,8

42,0 39,1

14,5 52,3

36,5

16,2 8,8

11,1 74,7

11,3 30,6 20,8

Supplementary Table 5- Lesion sizes (mm2) of TRAUMA1 series 2003-2005: LFPI operated under TOPI anesthetics Rats 14, 15 ,20 represents the first subgroup whose lesion stays stable. Rats 2, 3, 5, 6, 10, 13, 16 represents the other subgroups whose lesion is expanding over the chosen time points.

Growth rate as the relative change of size from 9d to 30d (i.e., lesion grows X% from 9d to 30d)

AVERAGE STD

isoflurane 77,1 80,8 128,3 5,8 2,1 54,7 NA 18,1 1,4 28,3 26,2 7,6 43,9 20,6 11,0 36,1 37,2

TOPI 144,7 44,4 34,7 -28,6 19,5 28,1 21,3 4,8 23,9 21,0 4,0 28,9 42,9

Supplementary Table 6 – Lesion growth rate of each rat as the relative change of size from 9d to 30d Growth rate as the absolute change of size from 9d to 30d (i.e.,

lesion grows mm2 from 9d to 30d)

AVERAGE STD

isoflurane 14,9 8,6 18,1 0,7 0,2 8,5 3,2 0,2 3,1 5,2 1,9 10,2 5,8 2,5 5,9 5,5

TOPI 8,0 7,6 5,7 -3,7 6,6 5,1 2,2 0,5 2,3 11,7 0,4 4,2 4,3

Supplementary Table 7 - Growth rate as the absolute change of size from 9d to 30d

Viittaukset

LIITTYVÄT TIEDOSTOT

We studied whether duration deviant sounds elicit mismatch responses in the auditory cortex and CA1 of hippocampus in urethane-anesthetized rats in oddball condition, and whether

Retinal and LGN cells are only a few synapses away from the photorecep- tors, so it is comparatively easy to model their function. In fact, within limits, these cells can be modeled

The development of neocortical cytoarchitecture in visual cortex is thought to be dependent on subplate dependent neuronal activity, experimental destruction of which in

G>NKHMKHIABGF>=B:M>=:<MBO:MBHGH<<NKLPBMABGFBGNM>L MA>MK:GL:<MBO:MBHGIKH<>LLBLLEHP>K:G=.

Several interconnected brain regions, such as the amygdala, anterior cingular cortex, prefrontal and orbitofrontal cortex, and insula, are involved in the regulation of

Regional differences in the regulation of dopamine and noradrenaline release in medial frontal cortex, nucleus accumbens and caudate-putamen: a microdialysis study in the rat..

The maximal osteonal diameter and mean wall width were the highest in the external cortex of the iliac crest (p b 0.05), and the mean cortical width, osteon number per cortical

Toi min ta ym - päristön muutos sekä johtamiseen ja vallanja- koon liittyvä tyytymättömyys ovat läsnä niin ai- emmassa tutkimuksessa (esim. Svara & Watson 2010)