MATERIAL PROPERTIES OF THE DEVELOPING PORCINE BRAIN

Transcription

1 MATERIAL PROPERTIES OF THE DEVELOPING PORCINE BRAIN Kirk L Thibault and Susan S Margulies Dept of Bioengineering, University of PA, Philadelphia, PA USA ABSTRACT The objetive of this projet is to quantify the age-dependent material properties of porine brain tissue and to orrelate them with strutural alterations assoiated with growth and development Samples of frontal ortex from neonatal (2-3 days) and adult pigs were harvested and tested within 3 hours post-martern The omplex shear modulus of the samples was measured in a ustom-designed osillatory shear testing devie Samples were tested at a shear strain amplitude of 25% from 2-2 Hz at 25 C and 1% humidity The elasti and visous omponents of the omplex shear modulus hange signifiantly with the development of the ortial region of the brain These hanges in material properties orrelate with inreases in myelin, brain mass, total ell number, and a derease in water ontent This projet is the first step in developing head injury tolerane riteria for the infant and young hild SEVERE HEAD INJURY is the most frequent ause of death in the pediatri population (Weston 1968, Kraus et al 1987), omprising 5% to 8% of all trauma-related deaths eah year Annual pediatri head injury statistis desribe an enormous emotional and finanial burden of 3, to 4, hospitalizations, 6, to 7, deaths, and an estimated $1 billion in osts for hildren under 14 years of age (Waller et al 1989) lmportantly, the majority of traumati head injuries to hildren less than 6 years old result from motor vehile aidents and fa lls (Pasui 1988, Brue 199) Trauma in the under four years-of-age group is partiularly important beause it is a period of aelerated growth and development (Calder et al 1984) and it is a time assoiated with a relatively high frequeny of nonaidental injury aused by ustodial abuse (Brue 199) Approximately 1 % of all traumati injuries in hildren will have a non-aidental etiology (MClelland et al 198), and in urban populations, 25 hildren per 1, people will be vitims of abuse annually (Helfer and Pollak 1968) The frequent inidene of morbidity and mortality assoiated with abuse emphasizes the linial and soio-eonomi importane for a greater

2 understanding of the biomehanis and linial outome of pediatri head injury Although investigations studying the biomehanis of pediatri head injury have been performed (Dejeammes et al 1984, Duhaime et al 1987, Stürtz 198, Mohan 1979), their results and their interpretation have relied on saling applied loads based on brain mass to infer the effet of external loads on the neural and vasular elements of the pediatri brain (Ommaya et al 1967) The differenes between adult head injury and pediatri head injury are intrinsially linked to both the anatomy and mehanial properties of the developing skull/brain struture Currently, brain mass is the only physial harateristi used to distinguish the mehanial response of the pediatri head from the adult head when developing separate head injury riteria, beause there are no data available that haraterize the properties of brain and skull in hildren The fous of this ommuniation is on the mehanial properties of the developing brain The first four to five years of life is a period of rapid growth and development in the human This growth spurt is haraterized by inreases in neural and glial ell number, rapid myelination, dereases in water ontent, inreases in dendriti branhing and synapti onnetions In this study we report mehanial properties of tissue obtained from the pig Dobbings and olleagues (Dikerson and Dobbing 1966, Dobbing) ompared human entral nervous system (CNS) development with several other speies to develop a means for interspeies saling of CNS development Using the guideline proposed by Dobbings, one an extrapolate the stage of CNS development between mammalian speies if one mathes the timing of omponent growth, and assumes that interspeies differenes in the fundamental units, like myelin omposition, are minor Using this approah Dobbings and oworkers found that during the first deade of human life, months of life in a human were roughly omparable to weeks in a pig Thus, a 1 year-old pig would orrespond to a >4 year-old human hild with a fully developed CNS, and a 2-3 day old pig would orrelate roughly with a <1-month old human newborn In this study, to examine the mehanial properties in the developing brain, we ompare properties measured in tissue from 1 year-old pigs with those from young (2-3 day old) pigs This study reports, for the first time, the properties of developing brain tissue Samples of ortial tissue were tested sinusoidally in simple shear over a broad frequeny range (2-2 Hz) and the omplex shear modulus was determined This information is essential for prediting the unique kinematis assoiated with brain injury in the pediatri population and, thus, is a first-step in identifying brain injury thresholds for hildren METHODS Fresh porine brain tissue was obtained from 2-3 day old (N=6) and 1 year old (N=6) domesti pigs in ompliane with the Animal Care and Use Committee of the University of Pennsylvania's and the United States Department of Agriulture's regulations regarding remnant tissue

3 prourement (IACUC protool #1) Time of death and brain weight were reorded for eah animal immediately post-martern The erebrum and thalamus of eah pig was removed from the skull en blo, stored in separate airtight ontainers filled with hilled mok erebrospinal fluid, and refrigerated until preparation for testing Samples were tested within 1-5 hours post martern The fous of our study was to determine the age-related behavior of brain tissue, independent of its inhomogeneous and anisotropi properties Therefore, regardless of age, all speimens were removed from the same loation in the frontal ortex and tested in the same orientation Eah brain tissue speimen was prepared by first hemiseting the erebrum along the longitudinal fissure A ylindrial sample of ortial tissue was removed from the erebrum by plunging a ylindrial ork barer laterally into the medial surfae of the ingulate gyrus, just superior to the genu of the orpus allosum The ylindrial sample was then slied perpendiular to its lang axis to remove a dis-shaped tissue speimen approximately 1 mm thik and 1-12 mm in diameter Gare was taken to ensure that the dis-shaped speimen was omposed entirely of ortial tissue and was free of any penetrating suli One sample per animal was tested Samples were randomized with respet to right or left hemisphere but the loation and orientation of eah sample remained onstant aross all studies The stressfree sample diameter and thikness (h) were measured while the sample was floating in mok CSF solution Cross-setional area (A) was omputed Mehanial properties of the porine brainstem were determined in shear beause brain tissue has a high bulk modulus but a very low shear modulus, and is therefore most likely to fail in shear (MElhaney et al (1976), Holburn 1943) Shear properties were determined using a ustom-designed devie apable of performing both osillatory (2-2Hz) and stress relaxation tests (Figure 1 ) Figure 1 - Brain Tissue Testing Devie shear fore transduer linear atuator sample humidity hamber linear bearing LVDT -71 -

4 Frequeny and strain amplitude were independently ontrolled in the osillatory tests reported in this ommuniation The devie used a horizontal parallel plate onfiguration with a variable gap distane, delivered a displaement to the bottom fae of the test sample, and measured the shear fore transmitted through the sample at its top fae The bottom plate was mounted on a linear bearing for support and ontrol, and was onneted to a LVDT (Trans-Tek ln, Ellington, CT) to measure the amplitude of the input displaement An isometri fore transduer (Kulite Semiondutor Produts ln, Leonia, NJ) was onneted to the top plate (a glass over slip) to measure the shear fore transmitted through the speimen Fore and displaement transduers were alibrated before every experiment Tissue samples were mounted in the testing apparatus and the shear sandwih assembly was enlosed within a humidity hamber to prevent dehydration during testing One in plae, the sample was allowed to equilibrate to the temperature and humidity onditions within the hamber (approximately 25 C and 1% humidity) for five minutes Mirometers supported the top plate/fore transduer assembly and were used to adjust the gap between the plates to the thikness (h) of the test speimen, and eliminated any pre-stress on the sample To drive the system for osillatory tests, a funtion generator in series with a power operational amplifier was used to exite a linear voie oil atuator (BEI Motion Systems, San Maro, CA) By varying the generator signal, the bottom plate of the osillating assembly was displaed sinusoidally with a speified amplitude and frequeny Eah sample was subjeted to a sequene of sinusoidal shear strains with an amplitude of 25% over the frequeny range of 2 to 2 Hz in 1 Hz inrements, randomized to an inreasing or dereasing frequeny protool The displaement of the bottom plate and the resulting fore at the top plate were measured Shear prestress was eliminated at eah frequeny by adjusting the plate alignment until the DC offset of the fore signal was nulled The voltages orresponding to displaement and fore amplitudes and the phase differene between the two signals were measured at eah frequeny with a digitizing osillosope (Hewlett Pakard, Love land, CO) at 1 ksamples/se Several preautions were take to ensure that these measurements represent the true response of the material Eletronially, the frequeny response of both the displaement and fore transduer are adequate over the frequeny range, and the two transduers were onditioned identially to remove alterations in the data due to a phase shift from the filtering omponents Mehanially, the inertial effets of the system were minimized by reduing the mass of moving portion of the devie Likewise, the inertial effets assoiated with the aeleration of the material itself were redued by limiting the size and therefore, the mass of the sample The measurements of the devie were validated for a visoelasti silione gel with mehanial properties similar to brain (Arbogast et al, in review) The assumption of a non-slip surfae at the tissue-plate interfae was onfirmed previously in osillatory shear tests between 2 and 2 Hz

5 DATA ANALYSIS Shear stress, -r, and Lagrangian shear strain, ro were omputed from the amplitude of the sinusoidal fore (F) and displaement () signals at eah test frequeny using the following relationships: 'l' F =- and A where A is the shear area of the sample and h is the thikness of the sample A visoelasti material subjeted to a periodi shear strain, y(t) = yos(wt), will respond with a periodi shear stress that leads the strain by a phase angle, o The omplex shear modulus, G* =G1+ ig2, of a visoelasti material may be determined from periodi stress and strain measurements using the following equations: (;:) (;:) os(8) G, = G, = s i n (8) where G 1 and G2 represent the elasti and visous omponents of the omplex shear modulus, respetively, -r is the amplitude of the shear stress, y is the amplitude of the shear strain, and o is the phase angle between the signals This formulation assumes linearity of the material and neglets inertial effets of aelerating the mass of the sample The measured fore and displaement signals were sinusoidal, permitting the use of these equations to alulate the omplex modulus (G*, G1 and G2 ) at eah frequeny The signifiane of the material properties' (G*, G1 and G2 ) dependene on age and frequeny were evaluated using analysis of variane (ANOVA p<1 ) RESULTS The frequeny response of the omplex modulus, G*, and the elasti and visous omponents G 1 and G2 are shown in Figures 2 and 3a and 3b, respetively (mean±se) Eah of the material property parameters were larger in the adult tissue than the pediatri tissue, and values inreased with inreasing frequeny Analysis of variane revealed that G*, G1 and G2 eah depend signifiantly (p<1) on age and frequeny I n addition, analysis of the ross-orrelation (age x frequeny) indiated that the slopes of G* and G2-79 -

6 Figure 2 - G* vs Frequeny for Porine Brain Tissue 3 adult 25 'f 2 "' 15 >- ::e (, Frequeny (Hz) 15 2 Figure 3a - G1 vs Frequeny for Porine Brain Tissue 3 25 NE 2 15 >- ::! C> 1 adult Frequeny (Hz) 15 2 Figure 3b - G 2 vs Frequeny for Porine Brain Tissue 3 25 adult E "' 2 15 >- ::! 1 C> Frequeny (Hz)

7 are affeted signifiantly by age, whereas age resulted in only a parallel shift in G 1 with no signifiant hange in slope The omposition (perent total sample area that was white or gray matter) of eah sample was determined using a mirosope-based video image analysis system The average (mean±sd) white matter surfae area fration was 4±1 perent in the pediatri brain samples and 81 ±12 perent in the adult tissue samples This differene is attributable to keeping the diameter of the ylindrial ork borer tool onstant, and using the same test site and orientation for both age groups Beause the test site was entered about a white matter ore, idential-sized samples from adult ortex ontained less gray matter than pediatri animals beause the white matter nearly filled the entire diameter in that site Therefore, the adult samples ontained a proportionally!arger fration of white matter than the pediatri samples lmportantly, the data within eah group was quite onsistent, and we ould find no onsistent relationship between variation of gray/white matter ontent and properties within eah group DISCUSSION Our study has two major limitations: that porine tissue was tested rather than human tissue, and that properties were determined only in one loation and one orientation There are two aspets we will address regarding porine tissue First, that 2-3 day old pig brain tissues are similar to those of human infants, and similarly that 1-year old pigs brain tissue is representative of adult tissue Seond, that pig brain tissue at the same phase of CNS development resembles human tissue The first four to five years of life is a period of rapid growth and development in the human This growth spurt is haraterized by inreases in neural and glial ell number, rapid myelination, dereases in water ontent, inreases in dendriti branhing and synapti onnetions Dobbings and olleagues (Dikerson and Dobbing 1966, Dobbing) measured the magnitude and rate of hange of brain mass, holesterol ontent (indiating lipid ontent, whih is a major omponent in myelin), water ontent, and DNA-P ontent (an indiator of total ellularity) in the erebrum For the purposes of diret omparison between pig and man, we normalized eah parameter by the maximum value and replotted the results in Figure 4, 5, 6, 7, respetively The x-axis in all plots is time, in units of weeks for pigs, and months for humans, with zero indiating birth, and the y- axis is perent maximum value While not a perfet orrelation for all four aspets of development, when omparing the Dobbings et al (Dikerson and Dobbing 1966, Dobbing) CNS developmental data for humans and pigs, we onur with Dobbings et al proposal that during the first deade of human life, months of life in a human are roughly omparable to weeks in a pig Thus, assuming that ellularity, brain mass, water ontent and myelin ontent are important ontributors to material properties, we believe that measurements of brain tissue from 1 year-old pigs orrespond to those in a >4 year-old human hild with a fully developed CNS, and that our measurements of tissue from 2-3 day old pigs orrelate roughly with those from a <1-month old human newborn

8 Figure 5 - Normalized Cholesterol Figure 4 - Brain Growth Spurt 1 Growth Rate 6 of brain in % of adult brain weight per time 8 () (;) 4 6 () "O - (ij E z birth I2:> 9 : :, <, hol, plg Time (pig: week, human: month) 8 6 z Cl "O (ij E z 4 & water, pig 2 water, human 8 r Time (pig: week, human: month) } : t :i i\ :* A ", - 85 () hol,human "O - (ij E z j 1 1 & Figure 7 - Normalized DNA-P Figure 6 - Normalized Water Content u Time Sale = 2 = 95, 4 pigs weeks human months & 2 DNA-P, plg DNA-P, human 4 Time (pig: week, human: month)

9 The fous of this ommuniation is on the mehanial properties of the developing brain A review of the mehanial properties of adult human and animal brain (Ommaya 1968, Fallenstien et al 1969, Shuk and Advani 1972, Metz et al 197, Koeneman 1966, MElhaney et al 1976) reveals that the reported properties of brain tissue vary by as muh as an order of magnitude, as well as speies of donor There is no published data regarding porine and human tissue mehanial tests with mathed experimental onditions Shuk and Advani (1972) measured omplex shear modulus in human brain tissue over a range of frequenies that inludes 2-2 Hz They tested adult human tissue undergoing sinusoidal torsion at small amplitudes and omputed omplex shear moduli The frequeny-dependent behavior they report for "unyielded white and gray" adult human ortex bears remarkable resemblane to our findings, inluding a!arger slope in G2 than G1 that results in a "ross-over" between 1 and 15 Hz However, the magnitude they report for G1 and G2 are on the order of a magnitude higher than our values for adult porine tissue The reasons for this differene may be attributed to differenes in experimental protool or speies Their protool inluded testing tissue a few hours after autopsy (whih might have ourred hours or days after death) and it is unlear the effet post mortem time may have had on their results Regardless, it is important to note that the frequeny dependent behavior of adult porine tissue was onfirmed by previous findings in the literature for human tissue, and that the frequeny dependene of the visous omponent of pediatri porine tissue differed onsiderably from adult tissue The range in reported tissue properties may also be related to the anisotropi and inhomogeneous nature of brain tissue Only Shuk and Advani (1972) inluded brain's anisotropy and inhomogeneity in their study design They reported that gray (thalami nulei) matter displayed some anisotropy in its storage modulus, and that multi-oriented white matter from the orona radiata was nearly isotropi Their data also shows overall differenes in the two types of brain tissue, indiating heterogeneity in the brain On average, they report that the storage modulus for these two regions are similar, and that the loss modulus for gray matter is approximately half that for randomly-oriented white matter However, Shuk and Advani's important and enouraging data on gray and white matter differenes were obtained over a frequeny range of 2-1 Hz, onsiderably lower than frequenies assoiated with traumati head injury loads In this ommuniation we sought to measure the age-related properties of brain tissue independent of the anisotropi and inhomogeneous nature of brain tissue, and therefore tested samples oriented in the same diretion and harvested from the same loation Future studies may be onduted to investigate the inhomogeneity and anisotropi properties of brain tissue in more detail, and their hanges with development The response of the head to traumati injury is influened by the mass, shape, dimensions, and material properties of the skull and the brain Previous investigations have determined the material properties of mature human skull and brain to predit the mehanial response of the adult head to traumati loads and to determine thresholds for brain injury However, the

10 brain, ranial bone, and sutures undergo rapid and dramati developmental hanges following birth These hanges affet the material and strutural properties of the individual tissues, and substantially affet the response of the head to trauma during hildhood Presently, little is known about how the material and strutural properties of the brain hange with age We measured the age-dependent material properties of young (2-3 day old) and adult (>1 year old) porine brain tissue during sinusoidal tests (2-2 Hz) at a strain amplitude of 25 perent, using a unique shear test apparatus Age signifiantly affeted the magnitude of both the storage and loss omponents of the brain tissue properties, and the frequeny-dependene of the visous omponent Charaterizing the developmental hanges of the brain's material properties is an essential omponent to determine the unique response of the pediatri brain to traumati loading onditions As suh, this projet is the first step in developing injury tolerane riteria for the infant and young hild, and eventually to more effetive preventive, therapeuti, and rehabilitative strategies for the pediatri ommunity REFERENCES Arbogast KB, Thibault KL, Pinheiro BS, Thibault LE, Winey KI, Margulies SS A high-frequeny shear devie for testing soft biologial tissues J Biomeh (in review) Brue, D A (199) Head injuries in the pediatri population Current problems in pediatris Chiago, IL, Year Book Medial Publishers Calder, 1 M 1 Hili, et al (1984) "Primary Brain Trauma in Non-Aidental lnjury" J Clin Pathol 37: Dejeammes, M C Tarriere, et al (1984) "Exploration of Biomehanial Data Towards a Setter Evaluation of Tolerane for Children lnvolved in Automotive Aidents" SAE 8453 : Dobbing, J The Later Development of the Brain and its Vulnerability In Sientifi Foundations of Paediatris, eds JA Davis and J Dobbings Heinemann Medial, London Dikerson JWT and Dobbing J (1966) Prenatal and Postnatal Growth and Development of the Central Nervous System of the Pig Pro Roya l So of London, Series B 166: Duhaime, A C T A Gennarelli, et al (1987) "The shaken baby syndrome--a linial pathologial, and biomehanial study" J Neurosurgery 66: Fallenstein, G T Hule, VD Melvin, JW (1969) "Dynami Mehanial Properties of Human Brain Tissue" J Biomeh 2: Helfer, R E and C H Pollak (1968) ''The Battered Child Syndrome" Adv Pediatr 15: 9-27 Holburn, A H S (1943) "Mehanis of Head lnjury" Lanet 2: Koeneman, J B (1966) Visoelasti Properties of Brain Tissue Case Institute of Tehnology

11 Kraus, J F, D Fife, et al (1987) "Pediatri Brain lnjuries: The Nature, Clinial Course, and Early Outomes in a Defined United States' Population" Pediatris 79(4): MClelland, C Q, H Rekate, et al (198) "Cerebra! lnjury in Child Abuse: A Changing Profile" Child's Brain 7: MElhaney JH, Roberts VL and Hilyard JF (1976) Handbook of Human Tolerane Jap Auto Res Inst Metz, H, J MElhaney, et al (197) "A Comparison of the Elastiity of Live, Dead, and Fixed Brain Tissue" J Biomeh 3: Mohan D Bowman BM, Snyder RG, Foust DR A Biomehanial Analysis of Head Impat lnjuries to Children Trans ASME 25: Ommaya, A K (1968) "Mehanial Properties of Tissues of the Nervous System" JBiomeh 1: Ommaya, A K, P Yarnall, et al (1967) Saling of Experimental Data on Cerebra! Conussion in Sub-Human Primates to Conussive Thresholds for Man 1 1th Stapp Car Crash Conferene, Soiety of Automotive Engineers Pasui, R C (1988) "Head trauma in the hild" Intensive Care Med 14: Shuk, L Z and S H Advani (1972) "Rheologial Response of Human Brain Tissue in Shear" ASME J of Basi Eng'g De: Stürtz, G (198) Biomehanial Data of Children - SAE th Stapp Car Crash Conferene, Troy, MI, Soiety of Automotive Engineers Waller, A E, S P Baker, et al (1989) "Chilhood injury deaths, national analysis and geographial variations" Am J Publi Health 79: Weston, J T (1968) The Pathology of Child Abuse The Battered Child Chiago, The University of Chiago Press