Cerebrospinal Fluid Circulation: Evaluation by Single

557
Cerebrospinal Fluid Circulation: Evaluation by
Single-Photon and Positron Emission
Tomography
Gustaf Bergstrand,' Stig Larsson, 2 Mats Bergstrbm, 3 Lars Eriksson, 4 and Gbran Edner 5
Gamma camera cisternography after intrathecal administration of "'In-DTPA is a widely used and well established technique
for studies of cerebrospinal fluid (CSF) circulation. The radiation
dose from'''ln, however, is a limiting factor that sometimes
seriously affects the image quality and, hence, correct interpretation . This investigation was designed to assess the value of
tomographic techniques for studies of CSF flow. Single-photon
emission computed tomography (SPECT) was performed on 15
patients by means of the Karolinska Hospital rotating gamma
camera system after injection of ''' In-DTPA or 99mTc_DTPA. The
combination of SPECT and 99mTc_DTPA proved to be a valuable
technique. It permits a clear evaluation of the location of the
isotope intracranially and its precise relation to the ventricles.
Positron emission tomography (PET) applied after intrathecal
administration of 68Ga-EDT A demonstrates that small amounts
of radioactivity can be registered and that PET is a promising
technique for the evaluation of CSF dynamics.
Indium-"'-DTPA was introduced in 1972 as a suitable radiopharmaceutical for studies of the cerebrospinal fluid (CSF) circulation [1]. Since then "' In has been the most commonly used
radionuc li de for such studies. Its physica l half-life of 2.81 days is
we ll su ited for studie s of CSF flow over a period of up to 72 hr.
The disadvantages are th at only low amounts of radioactivity ca n
be ad ministered due to high rad iation doses to the spinal co rd and
that th e photon energies (1 71 keV and 245 keV) are not optimal for
gamma camera scintigraphy . Therefore, the spatial resolution is
poor, and accurate interpretation may be impossible, whether or
not persistent intraventric ular uptake is present . This is essential
for correct evaluation of CSF flow, for example in patients with
communicating hydrocep halus.
In order to improve spatial resolution and to reduce noise, ,,, InDTPA was replaced by 99mTc_DTPA, which has been used in conventional gamma cisternography [2].
Single-photon em ission tomography (SPECT) has been used in
the Karolinska Hospital sin ce 1978 [3 ]. Thi s technique seemed
suitable for stud ies of th e CSF c ircu lation and , therefore , was
applied in this investigation.
Positron emission tomography (PET) offers certain important
advantages over SPECT, such as higher spatial resolution and more
' Departm ent
' Department
JDepartm ent
' Departm ent
SDepartm ent
of
of
of
of
of
accurate quantitation . Th e device is rou tin ely used in Karo lin ska
Hospital for localization of CSF fi stula s [4] and it s value for CSF
studi es has now been evalu ated .
Materials and Methods
Conventional gamma c istern og raphy and SPECT were performed
in seven cases with 18.5 MBq (0 .5 mCi) "' In-DTPA and in eight
cases with 100 MBq (2 .7 mCi) 99mTc_DTPA , using th e rotatin g
gamma camera system [3]. Th e radiopharmace utica ls were injected
intrathecall y by the lumbar route.
In ord er to fac ilitate adequate interpretation of reg ional CSF fl ow
from SPECT stud ies, co mputed tomographi c (CT) studi es were
inc luded in the examination proced ure. Both th e CT and th e SPECT
studies were perform ed with th e patient' s head fi xed to the examinati on table using an extern al fi xati on by a plasti c helmet system of
the sa me type as th e one used during stereotaxic CT procedures
[5]. A spec iall y designed loca lizing plastic box was co nnec ted to
th e sa me base plate as the helmet (fig . 1). In th e lateral wall th e box
ca rri es an aluminum rod th at is tilt ed 45 ° relative to the base plate.
Thi s rod can be loca li zed in th e CT im age, and by means o f the
cursor system of th e display unit th e actual level of each slice ca ll
be determined . Th e wall s of th e bo x are further supplied wit h a tube
system with one tube tilted in th e sa me way as th e aluminum rod .
For visua lization in th e SPECT stud y th e tube system is fill ed with
99mTc . The level of the individu al SPECT slice as we ll as the
coo rdinates within th e slice ca n su bseq uently be ca lc ulated from
th e location of the tubes . This device all ows a corre lati on between
di stribution of th e radi onuc li de in th e head using SPECT and th e
morphologic details of th e CSF space from CT .
Th e PET scanner of th e Karolin ska Hospital [6] was used for
localization of CSF fistul as in 14 patients. Th e intrath ecally ad ministered radioactivity was 30 MBq (0 .8 mCi) sBGa-EDTA.
CT and PET slices of th e sa me level ca n be superimposed usin g
a specially designed software , which includes th e transform ation of
the CT image to the matri x of th e PET scanner system . Th e accuracy
is very high due to th e extern al helmet fi xation . Th e possibilities of
further analyzing the radionuclid e distribution during CSF fl ow
studies and of performing quantitative measurements were evaluated in thi s investigation .
Neuroradiolog y, Karolin ska Hospital, S-1 04 01 Stockholm , Sweden. Address reprint requests to G. Bergstrand .
Hospital Physics , Karoli nska Hospital, S-104 01 Stockholm , Sweden .
Radiation Physics , Karolinska Institute, S-1 04 0 1, Stockholm , Sweden .
Physics, University o f Stockh olm , Stockholm, Sweden .
Neurosurg ery, Karolinska Hospital , S- 104 0 1 Stock holm, Sweden .
AJNR 4:557-559, May / June 19830195 - 6 108 / 83 / 0403-0557 $00 .00 © Ameri ca n Roentgen Ray Society
PET AND RADIONUCLIDE STUDIES
558
Results
Du e to the poor spatial reso lution using "'In-DTPA, it was impossible to obtain SPECT images of the CSF space of sufficient
clinica l value. SPECT using 99mTc_DTPA on the other hand demon-
AJNR:4, May / June 1983
strated additional detail , especially in cases where it was difficult to
decide whether intraventricular radioactivity was present or not (fig .
2).
When conventional views demonstrated clear-cut normal or
pathologic (fig . 3) conditions, SPECT did not add further morphologic information. Tomography, however, is probably necessary for
accurate quantification of CSF flow .
St~d i es done with PET and 68Ga-EDTA c learly show that th e
anatomy of CSF spaces can be accurately stud ied with positronem itting isotopes (fig. 4). However, the short half-life (68 min) of
68Ga-EDTA makes it unsuitable as a tracer for periods longer than
2 - 3 hr.
Discussion
Fig . 1.-Plasti c helmet and localizing box appli ed to head and exam ination
table. Tube system for sli ce level determination (arrows).
A
c
A rad ioactivity of 100 MBq (2.7 mCi) of 99mTc_DTPA will be
sufficient for SPECT studies and for conventiona l gamma camera
scintigraphy of the CSF space within about 24 hr after lumbar
injection. This period of time is, according to our experience,
sufficient for a correct interpretation of CSF flow. Static views have
been taken at 2, 4, 6, 9, and 24 hr after administration, but
registrations after 4, 6, and 24 hr are probably sufficient for evaluation of flow dynamics.
Although 100 MBq of 99mTc_DTPA is insufficient for tomographi c
registration after 24 hr , the presence of an intraventricular uptake
B
Fig. 2.-Frontal views 4 (A) and 12 (B) hr after
intrathecal injection of 100 MBq of 99mTc_DTPA. C,
Radioactivity in lateral ventricles (arrows) cannot be
excl ud ed in this patient with aq ueduct stenosis. D,
SPECT slice at level of lateral ventri c les demonstrates radio activity in sylvian fi ssures (arrows) and
quadrigeminal cis tern but no radioactivity within ventricle system. Peripheral rad ioactivity represents
cross section s of tube system.
o
AJ NR :4 . May / June 1983
A
PET AND RADI ON UCLIDE STUDIE S
559
B
Fig . 3. -Patient with commu nica ting hydrocephalus. Gamma c istern og raphy . co nvenlio nal fro ntal
view (A) and transverse axial SPECT slice (B) at level of late ral ventricles. Radioac tivity w ithin ventric ul ar
system (arrowheads ) is well demo nstrated with bot h techniq ues.
can usually be inferred by comparin g th e stati c view w ith th e
tomog raphi c im ag es obtain ed at 6-9 hr.
For accurate quantificati on and c learance c urves of th e radi oacti vity from th e CSF spaces repeat SPECT studi es are needed .
Thi s is presentl y under investi gati on. PET offers th ese opti ons bu t
new positron-emitting radioph arm aceuticals have to be prepared
and tested .
REFERENCES
1. Hosain F. Phil D. Som P. Ch elated "' In : an ideal radi oph armaceutica l for c istern ograph y . Br J Radio/1972 ;45 : 677 - 6 7 9
2. Som p . Hosain F. Wag ner HN .·Scheffel U. Ci stern og raph y with
Fig . 4 .- Superimposed CT and PET slices at
same level demonstrate activity in ventric ular system
(arrowheads ) and CSF c istern s and relatio n to sku ll
bo ne .
c helated comp lex of 99mTc . J Nucl Med 1972;13 : 551 -553
3 . Larsson SA. Gamm a camera emiss ion tomograp hy . Ac ta Radiol
[ Suppl] (Stock h) 1981 ;363
4 . Bergstrand G. Bergstrom M. Eriksson L. Edner G. Wi den L.
Pos itro n emi ssion tomog raph y with GBGa-EDT A in th e diag nosis
and loca lizati o n of CSF fi stul as. J Comput Assist Tomogr
1982; 6: 320-324
5 . Berg strom M . Boethiu s J . Eriksson L. Greitz T. Ribbe T. Wi den
L. Head fi xati on device for reproduc ible position ali gnmen t in
transmi ssi on CT and positron emi ssion tomog raphy . J Comput
Assist Tomogr 1981 ;5 : 136-1 41
6 . Eriksson L. Bergstrom M . Bohm C. et al. A four ri ng posi tro n
ca mera system for emi ssion tomog raph y of th e brain . IEEE
Tra ns Nuc l Sci 1982;1 : 539-545

Download Report