- 124 -

RADIOISOTOPIC STUDIES OP BRAIN UPTAKE

. WOldendorH . f Vfedsworth Veterans Administration Hospital Los Angeles, California, U.S.A.

ABSTRACT

Measurement uptake th f radioactivf o eo s e substancee th n si brain tissues after their administratio y injectionb inhalar o n - tion provid n atraumatia e c approac stude th f bloo o yo t h d flow d metabolian c processe braine th n i .s This paper reviewe th s anatomical,physiologica d physicaan l l problems arisine th n gi measuremen f radioactivito t braine th e factor n Th .i y s governing the passage of various classes of substances through the brain and their transport through the brain tissues are first considered. The physical problems arising in the measure- men radioactivitf o t braie th e the nar n i ny discussed maie Th .n difficulties in such measurements is shown to arise from the con- tributio e observeth o t n d counting rate from radioactivite th n i y scal d skullan p . This contributiof o e minimizee b us e n th ca n y b d special collimators designed to view only a part of the brain but to include in their field of view a minimum of non-neural tissue. A further possibility arises with radioisotopes such as In which emit characteristi X radiatioc y radiatios wel na s a l n since the contribution of the former to the total observed counting rate is almost entirely due to radioactivity in the superficial tissues whereas that of the latter is due to radioactivity in the super- ficial e braintissueth y recordind B .an s e countingth g raten i s appropriate channel photoe th f o ns spectru s thui t si m possiblo t e correct the results for radioactivity in the scalp and skull.- With radioisotope c whicS s suc r morho s a hemieo tw photont 7"5 n i s cascade, coincidence counting techniques offer still a further possibilit minimizo t y e contributioth e n from radioactivite th n i y superficial tissues. Various potential applications of these techniques are described. - 5 12 -

Although most solutes exchange slowly "between "blood and brain, "brain uptake measurement varieta f o s substancef o y s revea wida l e rang f freedoo e m of exchange. Some generalizations relating chemical composition to the rate of exchange can be made.

blooe . 1 th Wate dd gaseran s exchange freely between plasm braind an a .

2. There is a restriction of all of the common electrolytes found in plasma.

3. Most small molecule plasman i s , suc ures ha d creatinine aan , showa restricted passage related in some degree to molecular weight.

4. Large molecules, such as the plasma proteins, show an extreme re- strictio passagf o n e into brain healthn I .nearls i t ,i y impossibl mako t e a e meaningful measurement of large molecular uptake because this uptake is so low. The exchang largf o e e molecules between plasm braid an amors i n e tha n ordea n r of magnitude less than that of small molecules such as sucrose or mannitol. This extreme restrictio largf o n e molecules result considerabla n si e dependence f capillaro y permeabilit smalo t y l molecules upon their completenesf o s bindin plasmo gt a proteins A substanc. e which might otherwis e exchangeb e d freely wil e markedlb l y restricte s firmli t i y f bouni d plasmo t d a proteins.

5- Lipid-soluble substance plasmn si a generally exchange freely with the brain. Non-electrolytes penetrate into brain tissue with variable ease and this variability usually is related, in addition to molecular size, to their lipid-solubility. This probably pertains to the solubility of the substanc plasme lipie th th f ao dn i emembran e capillarth f o es i y t i cell f I . solublplasme th n ai e membran n readile cellca th t f ,i o e y pass inte th o interio e capillarth f o r yexternae th cell t ou ,l plasma membrane d throug,an h the basement membrane. This free penetratio lipid-solublf o n e substances ordinarily doe expost sno e brain cell harmfuo t s l substances because there ar e essentailly no natural, lipid-soluble noxious substances present in plasma which are not bound firmly to plasma proteins. Thus, bilirubin which could be seriously neurotoxic, particularly when blood levels are elevated, achieves very little penetration of the brain presumably because of its protein binding. Substances suc antipyrines ha , barbiturates, ethyl alcohol, anesthetic agents and many central nervous system drugs are lipid-soluble and thus freely penetrate from blood into brain. - 126 -

6. In addition to these relatively non-specific exchanges "between "blood and brain, certain metabolites suc somd glucoss ha an e ) amin(l e o} 3 acid , (2 s apparently pass throug capillare th h y wall n specifio s c carrier systemse Th . entry of both glucose and amino acids are quite efficient at low plasma concentrations. This transpor glucosf o t d aminan e o acid s saturablsi d an e simplo t e edu thu diffusiont sno . This transpor s specifii t r thesfo c e metabolites with restrictio f otheo n r solutes, even thoug f vero h y simila, - r molecular structure.

Capillary permeability

The mechanisms of the restricted permeability of brain capillaries are not completely understood t certai,bu n aspect f thio s s blood brain barrier (BBB) have been clarified. The capillary wall of most other tissues exhibits the

following characteristics* t 1. Water and blood gases are freely exchanged.

2. Lipid-soluble substances are very rapidly exchanged; their rates of exchange approaching that of water.

3. Molecules less than about 20,00 40,00o 0t 0 molecular weight equrli- brate quite rapidly between plasma and (4)» exhibiting an equilibration half-time less than one minute. This is considerably less rapid thar lipid-solublfo n e substances.

4. Substances with a molecular weight over about 40,000 exhibit a passage through the capillary wall -more than an order of magnitude less efficiently thar small molecules. Even very large macromolecules, however, continu o past e s through capillary walls wit rata h passagf o e e tha roughls i t y independenf o t their molecular size (4).

Ther e probablar e y three route passagf o s e throug capillare th h y walf o l most tissues (Fig.l.).

. 1 Directl ye plasm wallo througth tw f ao e s membranth h capillare th f o e j cell with passage through the car-illary cellular cytoplasm.

2. Passage through clefts (about 90 A width) between the capillary cells whic e attachear h d onl point a y t junctions. Mucopolysaccharide e cleftth n i s may assis filterinn i t largt gou ebrain i molecule t nbu thi ) unlikels i s(5 s y since intraventricular horseradish peroxidase (M.W. 42,000) passeS sEC froe th m of brai ne tigh levee intth clefe th tf th o o lo intercellula t r junction (6). - 127 -

3. Pinocytosis (fluid phagocytosis) of small portions of " which are transported through the cytoplasm, isolated from the cytoplasm "by the fragemen plasmf o t a membrane surroundin pinocytotie gth c vesicle. They are exuded into the extracellular space (ECS) at the basement membrane surroundi the capillary.

These speculation e compatiblsar e witconcepe th h t that water, blood gases, and lipid-soluble substances pass directly throug membrane walle th th hf o sf o e the capillary cell. The intercellular clefts are the probable major route for the less efficiently exchanged substances suc s plasmha a electrolytes, urea, glucose, amino acids etc. The relatively inefficient transport of macromolecule independent of their molecular size could be attributed to pinocytosis (?)• Sinc pinocytotie th e c vesicl substantialls i e y largemacroe th f ro - thay an n molecules in blood, this mode of transport would be independent of molecular l plasmal siz d aan e solutes coul e expectedb o traverst d e capillarth e y celli at similar rates.

There probably is no specific, carrier-mediated capillary transport in tissues other than nervous system. Such transport would not be required because intercellular non-specific passage woul e adequatb d meeo t e t most metabolic exchange requirements. If present, it would be inconspicuous against a back- ground of efficient non-specific transport.

It is possible that a system of very small pores (less than 10 A) in the plasma membran f generao e l capillary cellmajoa s i spassage r th rout r eefo of some ions and very small molecules such as water and urea. Although these postulated pores havt beeeno n visualize y electrob d n microscopy s stili t i ,l possible that transcellular transport of certain small molecules takes place by this route. The issue of how much capillary transport takes place through the the cells and how much between capillary cells is still open. The extensive recent literature on membrane transport has been reviewed (8).

Brain transport mechanisms

In the brain there probably is passage directly through the capillary cell of water, blood gases and lipid-soluble substances. The junctions between brain capillary cells have been show y electronb n microscop e differenb o t y t from capillary cells elsewhere in that they constitute a continuous strip of fusion between adjacent plasma membranes (9) welding the entire capillary into what - 128 -

is, in effect, a continuous plasma membrane. This eliminates the possibility of intercellular cleft passage in brain of molecules and ions which elsewhere equilibrate with extracellular fluid by this route. It has been Bhown in recent studies (6) of the electron microscopic distribution of horse- radish peroxidase (42,000 molecular weight) show this indicator in plasma is markedly restricte e entranc e intercellulath th t a df o e rwale cleff th o l n i t brain capillarie e fuse th e sit df th o e junctiont a s s between adjacent cells wherea clefte th s s pasperoxidase th s n othei e r tissues. Peroxidas t intpu eo the ventricular fluid freely communicates throug extracellulae hth r fluif o d brain up to the outside of this same fusion point (6).

The capillary cells of brain are remarkably free of pinocytotic vesicles. Thi uniqus i s e relativ o othet e r tissues where such vesicle e commosar d an n may constitute a substantial proportion of the entire cytoplasm of the capillary cells (10).

e basiTh f thesso e unique propertie e braith n i s capillary cels i l presently quite unknown. It has been speculated that the sucker-foot of the astrocyte, which largely cover capillare th s braine y th wal y secretn i ,lma e some substance which would affec capillare tth y cell plasma membrane causing fuso it t e witadjacene th h to stot cell pd pinocytosinan s g (ll).

It is conceivable that the brain capillary represents a continuous plasma membran capillare th f o e y endothelial cel lcharacteristice (12th d )an e th f o s B actuallBB y represen permeabilite th t generae th f o yl tissue endothelial cell plasma membran without bu e t pinocytosis. Suc ha generalizatio n would remove e requirementh muc f o hhighla r fo ty specialized modificatio braie th nf o n capillary sinc woult i e d the e specializeb n d only insofa madt i es ra tight , circumferential junctions with adjacent cell d stoppean s d pinocytosingf I . this js a correct analysis, substances which pass into brain freely are substanc which would pass directly through general capillary cells.

e capillarieTh brain i s n apparently hav uniqua e e mechanis y whicb m h selected entry of certain metabolites into brain is allowed (l, 3). To explain this rate-limited specific transport we must postulate specific carrier sites capillare ith n y cell plasma membranes which would allo passage th w f theso e e solutes throug capillare th h limitea t ya celt d bu lrate . Whethe t thesno r ero carrier e energy-consuminsar unsettleds gi possibls i t I . e that they simply constitute facilitated diffusion which would cause the carriers to behave as though they were creating selective, limited solubilities of these substances in the plasma membrane capillarf o s y cell brainn si metabolitee Th . s would then - 9 12 -

diffuse passively throug capillare th h y cells, muc e sam s th hlipid-solubla e e substances, dow concentrationa n gradient from plasma into brain extracellular fluid.

e carrier-mediateTh d metabolite transport presen t eviden n braino i t t nbu t in general tissues migh e presenb t obscuret tbu othen i d r tissue y efficiensb t intercellular diffusion into the ECS which would parallel the carrier-mediated transcellular route f thiI . s carrier-mediated transport coul e establisheb d d in general tissues,it would make brain capillaries less unique woult I . d seem a complex proces somehoo t s w provide brain capillary cells with "carriers". It woul e simpleb d thesf ri e carriers were presenl capillarieal n t i t no t bu s eviden n othei t r tissues, becausefficiene th f o e t non-specific intercellular diffusion of small molecules. These carriers in brain capillaries could be evident solely because the usually predominant non-specific intercellular ' pathwa s eliminatedi y .

Tt has also been postulated that the membranous layer created by the expande sucker-foot of the astrocyte which largely surrounds the capillary (Fig.l) constitutes a barrier between the capillary lumen and the brain extracellular fluid (13). Although it is possible that this plasma membrane of the astrocyte surroundin capillare gth o somt es i ydegre e effectiv barriera s a e woult ,i t no d explai e absencth n pinocytosif o e e capillarth n i s e braie yth th celr o n i l between capillary cells. In a species with a prominant peri- vascular space inside the glial membrane, peroxidase does not penetrate beyond the capillary junction (14)•

Animal brain studies

There are several methods of studying the rates of penetration into brains animalf o s where remove b braie n th er analysis ca nfo d e mosTh .t definitive approac measuro t s e e substanci "spaceh th th e r "fo brain i e n relativ o plasmat e , This "space" is a numerical concept and does not imply confinement in a specific anatomic compartment. Such a quantitative approach is desirable since valid comparison made b en betweeca s n different substances e sinTh .k actio f cerebro n c spinal fluid complicate measuremente th s substancef ,o s entering brain slowly and which remain in the brain ECS. This sink action removes the-indicator

substances frobraine th m rmakin braie gth n space seem unrealisticall w (15)lo y » Other common approaches utilize autoradiography (l6 d fluorescenc)an e micro- scopy (17). All of these regional distribution methods can be misleading since they tell little about the absolute concentrations achieved. - 130 -

Human brain studies

If comparable studies are to be carried out in humans, we must use atraumatic technique whicn i s h both administratio d measuremenan n f tracero t s are totally safe. For human use this implies inhalation or intravenous injection and external counting.

We have pursued direct extracranial countin meana s measurinf ga o s g human brain uptake (l8, 19)- When first considered, the external measurement braif o n uptake seem preseno t s n ideaa t l geometry (Fig.2) e braiTh . n occupies most of the upper half of the head well away from other organs allowing it to be surrounded by external detectors which can be shielded from other tissues. A numbe f problemo r s arise, however complicato t , e this seemingly ideal situation.

Althoug l organal h s viewed frooutside th m e mus e seetb n through skid nan other overlying tissues, the brain presents a particular problem because the isotope levels achieve nervoun i d s tissue usuall e extremelyar y low n tissue.I s suc s liverha , kidne r thyroido y , where high concentration f tracere o sb n sca achieved e overlyinth , g skin isotope conten s insignificanti t , even though f skio abous ni thS 0 ieEC 4 °t thus allowin e accumulatiogth largf o n e amounts of plasma solutes.

Scalp and Skull

The human cerebral hemispheres constitute a mass of about 1100 ml of tissue and are surrounded by approximately 250 ml of bone and, by our radio- graphic measurements, a slightly smaller volume of scalp (20). The scalp is a particularly difficult problem because skin contains suc hlarga e ECS. Within minutew fe a injectionf o s , pertechnetate n examplea s ,a , achieves about fifteen times the concentration in scalp that it does in brain. Most of the material that appeare blooth f e n dbraii n i i sth pood s f i nan o l correction for blood pool is made the scalp to brain ratio becomes about fifty.

The extracellular space of skull is about one-half of that of the scalp and, in addition, the blood pool volume in skull is very large. Our measurement of scalp and skull blood volume in rabbits, if extrapolated to Humans, indicate that these tissuebloodf o l m s.0 6 contai Thio t perhaps si 0 5 n s one-halh t f fo normal blood pool braine volumth t mus recognizedf I e o .e "b t , however, that i t probably is not valid to make human estimates from rabbit skull since the bone structure is these two species is quite different. It is clear, nevertheless, from anatomic studie humaf so n skull wher externae th e l tabl bees ha en removed - 131 -

(2l) that the diploic spaces of the skull constitute a large "blood pool. Our studies of the turnover of blood in the skull of the rabbit indicate that although the pool volume is large, the turnover rate is quite prolonged (22). The skull also actshiela s internallr sa fo d y originating gamma rays, particularly thos lowef o e r energy.

r severaFo l year pursuee sw n externaa d l detector geometry which attempt to see all cranial tissues rostral to the floor of the cranial cavity with a counting efficienc unifors a y possibls l regiona m al d r depthssan fo e . When we studie uptake th d f labelleo e d albumi n humani n s wit varieta h f braio y n pathology and corrected for the blood content in the field we found no correla- tion between brain disease and the cranial uptake of the albumin (23). Subsequent studies of the albumin content of scalp, skull and brain made it clear tha were w t e seeing predominantly non-neural tissues e (20)on f .I administers albumin and measures the upper half of the head, virtually all the counts will originate from the blood pools of brain, skull and scalp, and the extra- and intracellular compartments of the scalp and skull. Essentially none will, in the normal, originate in the extracellular spaces of the brain.

Rather than try to see the entire cranium, one can considerably improve on the proportions of non-neural to neural tissue by restricted collimation which will view this surface shell of high-activity tissue at right angles none (anse ed tangentially) wit divergina h g internal con f acceptanco e o t e maximize the volume of brain (Pig.2). A focusing collimator is of no advantage since it still sees a diverging cone of tissue on the superficial side of the focal point.

Characteristic X-rays

We have attempted to correct for extracranial isotope by utilizing the characteristic X-ray emissio certaif commoe o n th f no n radioisotopes (24). Nuclides with an atomic number of about 45 emit characteristic X-rays with variable efficiencies at energies approximating 25 keV. This will be too low n energa penetrato t y electron-dense th e e skull t wil,bu l escape froe scalpth m . A very suitable isotope for this purpose is In. This nuclide emits approximatel characteristie on y X-raV everr ke fo y4 V gamma-ray 2 cy ke fou0 39 r s (25). - 132 -

If the side of the head is examined with In "both in brain and scalp, the characteristic X-ray seen by the external detector will originate essential] completely from scalp whereas the gamma rays seen will originate from both scalf and brain (Fig.3). Since the number of gamma rays seen per X-ray seen can be determined, and will, with In, be approximately 4*1, a number of gamma rays correspondin X-raye th o sgt countee assume b e originatin b n o ca dt d g superficial e skulltth o . e subtracteb Thes n ca e d frototae th m l numbe f gammro a rays obtained from the entire field.

Metabolic studies

We are now pursuing a metabolic application of external counting of 75Se-selenomethionine have W e . determine n humani d d ratan ss that seleno- methionine enters the brain freely and, in rats, apparently traverses the brain I amine th of capillar o aci y dwa carriey b y r system. n keepini Thi s i s g wite th h generally accepted equivalenc f selenomethionino e d methioninean e have W .e also determined that phenylalanin particularla s i e y effective competitive inhibitor braie oth f n uptak selenomethioninf o e ratn i e s (Fig.4) n thiI . s capacits a y competitive inhibitor it is more effective than the same molar concentration of natural methionine. Phenylalanine, however, does not measurably interfere with pancreati r liveo c r uptak f selenomethionineo e , eve highet a n r dosage levels thae requirear n essentiallo t d y completely inhibi e braith t n uptak f selenoo e - methionine .

Our clinical interest in Se—selenomethionine uptake in brain is in 75 the stud phenylketonuriaf o y e gammTh .e make aS ray sf o shuma n studies feasible even though 75Se is not particularly desirable for human use. It has been postulated that the pathophysiology of phenylketonuria is related to saturatio e braith nf o ncapillar y amino acid transport hige systeth h y bloob m d levels of phenylalanine (26). This speculation is compatible with the observa- tion that feeding a phenylalanine-deficient diet in early infancy results in a considerable improvement in intellectual development (271- This suggest that tï high blood phenylalanine level has some toxic influence on brain. This would be compatible with its impairment of the uptake of other amino acids. We are pursuing 75^Se-selenomethionine wite intentioth h f correlatino n humae th g n brair uptake of selenomethionine with blood phenylalanine level in phenylketonurics. Thi y possiblsma usefue b emeana s screeo a lt s n patient r metabolifo s c defects whic affecy hma t amino acid transpor centran i t l nervous system. - 133 -

75 Coincidence counting of Se

The essentiall gamma-rayV ke 0 y27 simultaneoud san V ke 0 13 s "7 c emissions of Se (28) are a theoretical possibility of considerable interest in brain monitoring. Isolated depth localisation shoul e possiblb do tw f i e focal collimators are arranged to meet the following specifications (Fig.5)*

1. Positioned with their focal centers at the same point. 2. Their axes are at right angles. 3. Each detector is provided with a pulse height analyzer window V gammke a 0 seaccep o t 27 rayst r o t. 0 eithe13 e th r

4. The two window outputs are fed to a coincidence circuit.

With this arrangement the counting efficiency of any point in the tissue will be the product of the counting efficiencies of the two detectors foi that e outputpointth f I botf .o s h detector e summear s d (not coincidenc' e counting) the counting efficiency of the two detectors for any point will be the sum of the efficiencies of both for that point.

Coincidence counting in this situation should allow extreme isolation near the overlap of the focal points of the two detectors. At all other loca- tions one of the detectors will be very inefficient because of the elliptical shape of the acceptance pattern of a converging multi-hole collimator. Such an arrangement might allow countin a stereotacticall f go y located point sucs a h globus pallidu a tumou r o m r with rejectio counf o n t originating superficiao t l the poin f interesto t e countinTh . g efficienc f suco y a hsyste mw woullo e b d since the absolute counting efficiency of even the central convergence point will be the product of the absolute efficiencies of the two detectors. Each of the detectors might be expected to have an absolute efficiency at the focal o thas o tf e orde n thousanpoin1 onlth te f yf ro n o t i abou e d on tevent s would be counted. Nevertheless, prolonged counting periods would make this possibly useful since the coincidence background count should be essentially zero. This same technique should be useful with many other nuclides which emit nearly simultaneous, isotropically propagated high energy photons originating either nucleue ith n electror o s n shell structure. They need only occur sufficiently close togethe electronice th e see b n tim y ri nb e o event t eon s sa . - 134 -

Tumour detection An approach being pursued by us is the definition of various parameters permeabilite th f o capillarief o y brain si n tumours versus normal brai assio nt s in pre-operative tumour classification. It is our plan to utilize our paired focal collimation system to view a known tumour in one cerebral hemisphere with e detecto on looo t kd witran othee th h r detecto controa t ra l are braif ao n in the other hemisphere as removed as possible from the tumour site (Fig.6). We would then study the relative uptake of various tracers. By subtracting the control region counts from the tumour region counts, a correction for blood, scalp and skull should be possible.

Smal d large-moleculaan l r permeabilities wil e assesseb l y chelatedb d smale th ls a molecul n I protein-bound an e large th d es acidifiea n I d molecule t woulI . dinteresf o als e ob o stud t e selenomethionin th y e uptake characteristics of tumour versus non-tumour control brain. '

Since some tumours exhibit considerable arteriovenous shunting evident by angiograph determinatioe th y f arteriovenouo n s shunting coul e determinedb d by noting the passage of an intravenous bolus of labelled iodoantipyrine, since the difference between passage of this substance through arteriovenous malforma- tion d normasan l brai quits ni e easily detected after rapid intravenous injectior *p 3 i (29). The availability of I-iodoantipyrine would greatly facilitate this study by allowing much larger doses with rapid disappearance of the nuclide.

Another interesting approach would be to study the changes in pool volun of the tumour versus the control area of brain as a function of systole. The chang bloon i e d volume smalo would requirto an le db rapio to e a ddefinitio n i n measurable b tim o t e e with attainable count re.tes r methoOu . d uses protein- bound In with measurement of changes in the pool volume in the tumour and e controith n l brai s indicatea n e cross-correlatioth y b d n between cound an t the F-r-wave of the EGG. By utilizing about 1,000 heart actions a very large numbe f e countobtainero b n ca s d resultin n abougi thirty-fol. ts d reduction i n statistical noise. This would produc a noise e level that woul equivalene b d c t t increasing the isotope dose by a fa.ctor of 1,000. This might allow the defini- tio f changeo n bloon i s d pool e tumouvolumth f ro e havin a gfixe d relatioo t n heart action. It is anticipated that tumours of differing vascularity will exhibit different changes in blood volume with systole. - 135 -

Regional cerebral blood flow

The future availability of some of the newer radioisotopes should be 123 considerablf o e interes brain i t n blood flow studies a f I I-iodoantipyrin. e could be obtained sufficiently cheaply it would make possible relative regional distribution of blood flow measureable by the gamma camera. A rapid intravenous injection of this material would allow about a one minute view of the side of the head with a distribution of isotope approximately in proportion to regional blood flow.

Positron emission

If carbon technology can be advanced to the point where it can usefull e incorporateb y d into glucos d aminan e o acids ,larga e numbe humaf ro n brain metabolic studies would become feasible.. Such positron emitter e idear s a r craniafo l counting because annihilation gamm at emphasiz no ray o d s e counth e t from overlying cranial tissues as occurs with single gamma emissions. - 136 -

FIG 1 .(overleaf )

Showing diagrammatically some basic probable differences between general capillarie braid san n capillaries generae th n I l. capillar dominane th ) (a yt route passagf e extracellulao s th o t e r space from plasma are via the intercellular cleft and, to a lesser extent, pinocytotic vesicles. d—solublLi dp e materia n alsca lo pass directly through the capillary cell at sites other than the inter- cellular cleft. Very small molecule iond san s probabl n alsca yo pas a thisvi s same route traversin plasme th g a membrane through extremely small pores. Carrier-mediated transport througe th h capilla.ry wall has not been demonstrated in general capillaries. braie Ith n n (b) capillare ,th modifies i y d insofae interth s ra - cellular cleft is sealed and pinocytosis is essentially absent. Lipid-soluble materials still pass freely through the brain capil- lary cell and carrier mediated transport of amino acids and glucose is demonstrable. Pore filtration probably occurn generai s sa l capillaries. The brain capillary is nearly completely surrounded by the expanded astrocytic processes. Their relationship to the restricted permeability of brain capillaries is unclear. Although the directio f passago n e her indicates i e d outward froe capilth m - lary lumen becaus basie th e c consideratio escape n th her f s o ei e radioactive labels frocapillarye th m , these transfer processes probably are bidirectional. LIPID SOLUBLE CARRIER MEDIATED CARRIER MEDIATED TRAMS-CELLULAR TRANS-CELLULAR PASSAGE FILTRA PASSAGE PORE TION FILTRA- INTER-CELLULAR, TION CLEFT PASSAGE

ASTROCYTIC PROCESS

GENERAL BRAIN CAPILLARY CAPILLARY

Fig. 1 - 138 -

FIG(overleaf2 . )

Indicatin e shel gth higf o l h activity tissues which surround the low activity brain tissue. External measurements on the brain must be made through this shell of high activity. If the entire craniu crystale viewes i mth y db , considerable scalp and skul e seear ln tangentially with relatively little contribu- tion from brain minimizatioA . scalf no skuld pan l e counb n tca obtained by restricting the collimation as shown on the right so that the scalp and skull are seen largely at right angles. Focal collimation such as shown at the top is of limited value because it sees a diverging cone of superficial tissues. Even though the scalp and skull are seen at reduced efficiency a large area of these tissues are seen. For a unit volume of tissue the focal point is heavily weighted but is of relatively small volume. The larger superficiae volumth f o e l cone still result significana n si t scal skuld pan l count even thoug absolute hth e counting efficiencs i y lower than at the focal point. XTAL

LEAD SHIELD

PHOTO- XTAL MULT.

FIG.2. - 140 -

FIG (overleaf3 . )

A correction can "be made for radioactivity in the large extracellular spac scalf e o utilizatioy p"b e characterth f no - istic X-rays. Nuclides such as In produce characteristic X-ray sofo spenetrato to t tha e tar skulle eth note s A . n di the text ,correctioa superficiallr nfo y originating gamma externalle radiatio th made f "b i e n nca y detected character- istic X-ray countede sar . CHARACTERISTIC X-RAY

SCALP SKULL

FIG.3. - 142 -

FIG. 4 (overleaf)

Showing the "brain uptake of selenomethionine alone and when a pre-load of 1-methionine and 1-phenylalanine are administered. The pancreati lived can r uptak selenomethioninf eo suppresses ei d by pre-loading with 1-methionine "but unaffecte 1-phenylalaniney db . Becaus pancreatie th e c capillary wall pases selenomethionine, methionine and phenylalanine by non-specific diffusion the block of tissue uptake is presumed to be at the parenchymal cell wall. The brain uptake of selenomethionine is more suppressed by 1-phenylalanine tha 1-methioniney nb . Studies introducing 1-phenylalanin- 1 d ean methionine directly into the extracellular space of rabbit brain by means of a ventriculocisternal perfusion indicate that the phenyl- alanine bloc braif ko n uptak selenomethioninf eo e from blood probably occurs at the capillary wall and the brain uptake block by 1-methionine probably parenchymae occurth t sa l cell wall. 2200. PANCREAS PLASMA 2000

1800.

1600

1400

I20C 1000

800 • L-METHIONINE PRE-LOAD L-PHENYLALANINF PRE-LOAD 600, S eMETHIONIN- E ALONE 40Q

10 24 8 16 MIN. 32 I 2 4 MIN. 32

LIVER

70C 80. 600 70.

500 60.

50. 40C

40. 300 30. 200 20.

100 10.

32 4 2 I 16 MIN. 32 2 4 16 MIN. - 144

FIG (overleaf5 . )

Showing diagrammaticall possible th y e utilizatio cascadf no e obtaio gammt e aS n raypoin f so t localizatio75 n with nearly complete rejection of overlying superficial radioactivity. As noted in the focao texttw le ,th collimator arrangee sar d with their focal points coinciding and their axes of viewing at right angles. When the two detectors are connected to count only in coincidence, the counting efficienc y poin produce an absolute spacn th tr i th s fo yi f e to e counting efficiencies of the two detectors. If connected so that the two collimators count in summation, the counting efficiency for individuae y pointh an f to becomem lsu collimatoe th s r counting efficiencies. The elliptical acceptance pattern of the focal collima- tor with good rejectio laterallf no y originating gamma rays should allow good isolation of rays originating at the overlapping focal point absolute Th . e counting efficienc thif yo s coincidence arrange- menhowever, is t , very low. SUMMED 'COUNT COINCIDENCE COUNT\ XTAL

DIRECTION FOCAF O L COLLIMATORS

XTAL

COINCIDENCE CIRCUIT

PIG.5, - 146 -

(overleaf6 FIG. )

assessmene th r Fo uptakf to focaf eo l lesions, sucs ha tumors, correctio "bloodr nfo , scalp d skul,an l radioactivity can be realised "by simultaneous comparison with a control region of normal "brain as far removed as possible from the lesione th sit f o e. FIG. 6 . - 148 -

REFERENCES

(1) CRONE, C., J. Physiol. (Lond.) 181 (1965) 103. (2) BLASBERG, R., LAJTHA, A., Brain Res. !_ (1966) 86. (3) UDENFRIEND, S., Amer. J. Nutr. 12. (1963) 28?. ) (4 RENKIN, E.M., Physiologis (1964_ ] t . 13 ) (5) LUFT, J.H., In The Inflammatory Process (ZWEIFACH, B., Ed.) Academic Press Yorw kNe , (1965). (6) BRIGHTMAN, M.W., REESE, T.S., J. cell. Biol. £0 (1969) 648. (7) PALADE, G.E., Anat. Rec. 136 (i960) 254. (8) DIAMOND, J.M., WRIGHT, E.M., In Annual Review of Physiology (HALL, V.E. GIESE, A.C., Eds.). Annual Reviews Inc., Palo Alto, Calif. (1969). (9) REESE, T.S., KARNOVSKY, M.J., J. cell. Biol. 3& (196?) 20?. (10) PALADE, G.E., J. appl. Physiol. 24 (1953) 1424. (11) DAVSON , OLDENDORFH. , , W.H., Proc. roy. Soc. Med 0 (196?.6 ) 326. (12) KROGH , ProcA. , . roy. Soc. Med 3 (1946.13 ) 140. (13) DEMPSEY, E.W., WISLOCKI, G.B., J. biophys. biochem. Cytol. 1. (1955) 245. (14) BODENHEIMER, T.S., BRIGHTMAN, M.tf., Amer . J Anat. 2 (196812 . ) 299. (15) OLDENDORF, W.H., DAVSON , ArchH. , . Neurol. (Chic. . (196717 ) ) 197. (16) BAKAY , Blood-BraiL. , n Barrier, C.C. Thomas, Springfield (1956). (17) KLATZO, I., WISNIEWSKI, H., SMITH, D.E., In Biology of Neuroglia, Progres n Braii s n Researc e ROBERTIS(d ? hl] , E.D.P., Ed.) Elsevier, Amsterdam (1965). (18) OLDENDORF, W.H., KITANO, M., Arch. Neurol. (Chic.) 2 (1963) 574. (19) OLDENDORF, W.H., KITANO , ArchM. , . Neurol. (Chic. , (196513 ) ) 533- (20) OLDENDORF. W.H., IISAKA . J nucl , Y. ,. Med 0 (1969.1 ) 177, 184. (21) PENDERGRASS, E.P., SCHAEFFER, J.P., HODES, P.J., The Head and Neck n Roentgei n Diagnosis, C.C. Thomas, Springfield (1956). (22) OLDENDORF, tf.H., Proc. Int . e ClinicaSympth n o . l Applicationf o s Isotope Clearance Measurement of Cerebral Blood Flow, Mainz, Sed. Re'p f Germano . y (1969). (23) OLDENDORF. W.H., Unpublished data. (24) OLDENDORF, W.H. . J nucl, . Med 0 (1969.1 ) 740. (25) DILLMAN, L.T., J. nucl. Med. 10 Suppl. 2 (1969) 5. (26) UDENFRIEND, S., Amer. J. clin. nutr. £ (l96l) 691. (27) SUTHERLAND , UNBAHGERB. , , BERRYB. , , H.K., Amer . J Dis. . Child 3. (1966) 505. (28) LEDERER, C.M., HOLLANDER, J.M., PERLMAN, I., Table of Isotopes, 6th edn. Wiley, New York (1967). (29) OLDENDORF, W.H., Neurology (Minneap. 4 (19641 ) ) 1078.