- Email: [email protected]
Single photon emission computed tomography imaging of brain tumors
Single Photon Emission Computed Tomography Imaging Of Brain Tumors H.J. Biersack, F. Grenwald, and J. Kropp Five radiotracers may be used for single-photon emission computed tomography (SPECT) imaging of brain tumors, namely technetium 99m pertechnetate, iodine123 amphetamine derivatives, "mTc-hexamethyl propylene amine oxime (HMPAO), thallium 201, and 1231 alpha methyl tyrosine. Of these, pertechnetate may be considered as an "historical" procedure in brain t u m o r s , However, there may be some equivocal cases in computed tomography or magnetic resonance imaging, where this procedure may still be used. In 1981, 1231 isopropyl amphetamine was first used in brain tumors. Further studies showed, however, that IMP is not a usedful tool for brain imaging in tumorous lesions. In 1986, ggmTcHMPAO appeared on the European market
RAIN SCINTIGRAPHY with technetium
B 99m pertechnetate had been the only imaging procedure for brain tumors until the mid-1970s when computed tomography (CT) appeared on the market. Serial scintigraphy, including the perfusion phase as well as early and delayed images, allowed a differential diagnosis of brain tumors. 1 The serial sequential scintigraphic method enhanced the CT rate of correctly identified tumor types in meningioma and glioma as well as metastases. 2'3 Although iodine 123-labeled amphetamine derivates led to an increase in brain investigations after the "CT depression," many years passed until nuclear medicine could offer new procedures for the detection of tumorous brain lesions. The 99mTcHMPAO perfusion marker was evaluated by several groups as a tumor imaging agent? In 1989, ot-L-3-(123I) iodo-a-methyl tyrosine (IMT), a new radiopharmaceutical, was introduced for the evaluation of its incorporation into protein of brain tumors? This article summarizes the results of brain single-photon emission computed tomography (SPECT) in tumors and also attempts to analyze the potential use of this imaging procedure. From the Department of Nuclear Medicine, University of Bonn, Germany. Address reprint requests to H.J. Biersack, MD, Department of Nuclear Medicine, University of Bonn, Sigmund-FreudStrasse 25, D-5300 Bonn L Germany. Copyright 9 1991 by W.B. Saunders Company 0001-2998/91/2101-0001505. 00/0
2
as a new tumor imaging agent. Some useful clinical results were obtained in patients before and after chemotherapy or radiotherapy. Thallium-201 was incidentally noted to accumulate in tumors. Using a threshold index, this agent can be used to distinguish lowversus high-grade lesions. The most promising agent for brain tumor SPECT is 1231-~methyl tyrosine, which showes potential to evaluate therapeutic procedures in brain tumors and may improve the differentiation between abcess and glioblastoma. The most promising aspect is the differentiation of tumor recurences and scar tissue after brain surgery. Copyright 9 1991 by W.B. Saunders Company
99mTc PERTECHNETATE
Radionuclide brain imaging, performed as serial scintigraphy, proved to be effective in establishing type-specific tumor patterns. 1-3 Cerebral serial scintigraphy includes both radionuclide angiography as well as early and late static images.2 R6sler et al, 1 as well as Buell et al,2'3 presented different criteria for the differential diagnosis of brain tumors, including the perfusion ratio as well as increasing or decreasing contrast ratios. While the dynamic perfusion phase cannot be used for SPECT investigations, early and delayed activity accumulation may be evaluated by tomographic techniques (Fig 1). Our group presented the first results in 1981 describing brain SPECT using a rotating gamma camera. 6 In five of 23 patients, tumor diagnosis was only possible by SPECT. The combination of SPECT and planar imaging resulted in a specificity of 98% (412 of 422 patients). 6 Several results in a considerably smaller group of patients had already been reported by Ell et al. 7 However, SPECT with 99mTcpertechnetate is currently considered as a "historical" procedure, because CT and magnetic resonance imaging (MRI) are the tools of choice for the evaluation of brain tumors. 1231AMPHETAMINES
Soon after the first description of N-isopropylp-(l:3I)iodoamphetamine (IMP) by Winchell et al in 1980,8 LaFrance et al9 reported on the decreased accumulation of IMP in brain tumors
Seminars in Nuclear Medicine, Vol XXI, No 1 (January), 1991 : pp 2-10
BRAIN SPECT IN TUMORS
3
Fig 1. (A) Metastasis of the right cerebellum, partly superimposed by vascular activity (sinus) in the planar study, (B) Brain SPECT with "mTc allows a superior visualization of the mestastesis.
(Fig 2). However, Ell et al 1~ described some cases in which IMP concentrated actively in neoplasia of the brain. This uptake seemed not to be related to the grade of malignancy. Schober et al H reported a comparison of IMP-SPECT and amino acid-PET in brain tumors. None of the malignant brain lesions accumulated IMP, whereas "C-L-methionine uptake was present.
It was speculated that the tumors did not concentrate IMP, most probably due to the lack of appropriate receptor sites or metabolic pathways and decreased perfusion." Moretti et a112 investigated 27 patients with different brain tumors, including astrocytoma, glioblastoma, oligodendroglioma, meningioma, and metastasis, and demonstrated a decreased uptake in
4
BIERSACK, GRONWALD, AND KROPP
Fig 2. (A) Oligodendroglioma, right frontal view by CT. (B) Brain SPECT with 1231butyl amphetamine shows reduced uptake in the tumor region.
88%, although two lesions were missed. It was postulated that the removal of the monoamine IMP from circulation is achieved through the podocytes into glial cells. However, astrocytes are known to have a high affinity for noradrenaline and possess -/-aminobutyric acid (GABA) binding sites. 12To answer the question whether IMP accumulates in normal astrocytes, Moretti et a112performed investigations in astroglial cell cultures. The results showed the presence of a saturable uptake system in normal astrocytes. Tumorous astrocytes seem to loose this ability,
for example, when the GABA pathway is not functional. This may explain the lack of accumulation of IMP in astrocytomas. The results obtained with IMP in brain tumors do not provide any evidence that IMP is a useful tool for brain imaging in tumorous lesions. 99mTCHMPAO
Soon after the availability of 99mTCHMPAO on the European market, the first paper on the results of cerebral uptake of this new radiophar-
BRAIN SPECT IN TUMORS
maceutical in patients with brain tumors was published. 4 Lindegaard et al4 presented their data in 12 patients with cerebral glioma scheduled for intracarotid chemotherapy. Diagnosis had been established by surgical or stereotactic biopsies. Perfusion indexes derived from selected regions of interest (ROIs) were calculated by dividing the activity uptake in the tumor region through a similarly sized area in the corresponding region in the contralateral hemisphere. The HMPAO uptake in the tumor region was significantly lower than in the corresponding region of the contralateral, presumed normal, cerebral hemisphere in 10 of 12 patients. In one patient with a significantly higher uptake in the tumor region, it was speculated that this was caused by an earlier brain infarction in the contralateral hemisphere. No significant differences between astrocytomas and glioblastomas were found. These data indicate that blood flow in gliomas seems to be variable, but is generally lower than in a normal brain. One of the drawbacks of perfusion estimation is that a differentiation between viable tumor tissue and variable amounts of necrotic tissue or edema is not possible. These results are in so far controversial, because 133Xe injection studies have shown high local blood flow in hypervascular gliomas. This difference may be due to the binding mechanism of HMPAO. Another study on the use of HMPAO in brain tumor patients was presented by Babich et a113 before, during, and after radiotherapy. Untreated brain tumors were found to exhibit a range of 99mTCHMPAO uptake, varying from areas of markedly increased isotope activity to photopenic areas, when compared with normal brain tissue (Fig 3). As in the studies reported by Lindegaard et al, 4 a ratio of HMPAO tumor uptake to contralateral normal tissue uptake was calculated prior to and during radiotherapy. Two patients with glioma showed increased uptake; three cases with glioma and two with metastases presented with normal uptake. In five patients, HMPAO uptake was decreased. Tumor to nontumor ratios were calculated for seven patients (two with increased and five with decreased uptake in the tumor) prior to and during radiotherapy. This ratio tended to return toward unity in lesions responding to therapy. These preliminery results indicate that HMPAO
5
Fig 3. (A) Oligodendroglioma, parietal right view by CT. (B) Brain SPECT with ~rc-HMPAO shows slightly increased activity uptake in the region of the tumor.
6
imaging may be useful for providing an indirect measure of tumor therapy response, especially in metastases. In 1987, Hoshi et a114reported on a mismatch between ~23I IMP and 99mTc HMPAO brain perfusion imaging in a patient with meningioma. In their case report, the tumor showed a high blood flow as evident by the 133Xe inhalation method. In contrast, IMP imaging showed no activity uptake in the tumor. Imaging with HMPAO showed--in concordance with the xenon method--high tumor activity. It was speculated that this disparate behavior of tumor uptake is most probably due to the lack of binding sites for amphetamines. This study was later extended to seven patients with meningiomas and published by Nakano et al. ~s Brain tissue ratios calculated from the tumor ROI and the ROI of contralateral homologous area were lower for HMPAO than those of the IMPSPECT study during the first 10 minutes after intravenous injection of the tracer. It must be mentioned that Nakano et a115used a dynamic SPECT technique (ring-type gamma camera) for their investigations. These data make evident that accurate assessment of vascularity of meningiomas is not possible, and are supported by the studies of Lindegaard et al4 in which the two hypervascular tumors showed HMPAO uptake similar to tumors without this feature. In summary, HMPAO brain SPECT in tumorous lesions seems to be only of value in the follow-up of patients under treatment. In another series Langen et a116studied cerebral uptake of 99mTc-HMPAO in 66 patients with various types of brain tumors. A quantitative evaluation of HMPAO brain SPECT was performed using tumor to cerebellar ratios. Again, the uptake of HMPAO by gliomas and meningiomas showed a wide range of values with no significant differences between malignant and benign gliomas. However, meningiomas exhibited a significantly higher uptake of HMPAO. Malignant gliomas presented with an uptake ratio of 0.75, whereas meningiomas showed values of 1.14. Two patients were studied after intraarterial ACNU (nitrosurea derivative) chemotherapy. No significant changes of the uptake pattern were observed. One patient with a malignant glioma had HMPAO-SPECT before and two times during the course of
BIERSACK, GRONWALD, AND KROPP
radiotherapy. In this case, a decrease of the tumor to cerebellar ratio could be observed. THALLIUM 201 Thallium 201, first used for myocardial imaging, was incidentally noted to accumulate in tumors. Ancri et a117']8presented data in primary and metastatic brain tumors in 1978 and 1980. These studies were further extended by Kaplan et al~9 in 1987, who compared the respective findings with the results of the histopathologic examination. In a more extensive study, Kim et al2~ published a paper on 2~ SPECT in 40 patients with low-grade and 11 patients with high-grade lesions. All patients had biopsy or autopsy. ROIs were created for the tumor lesion, and ROIs for the nonaffected tissue were determined by creating a mirror image of the initially defined region on the contralateral side (Fig 4). Kim et al2~found a strong statistical difference between the 2~ uptake indexes in low-grade versus high-grade lesions. In 14 patients with biopsy- or autopsy-documented lowgrade astrocytoma, the mean 2~ index was 1.27 -+ 0.4 compared with a mean index of 2.4 _+ 0.6 for all patients with high-grade astrocytomas. Using a threshold index of 1.5 to distinguish low- versus high-grade lesions, Kim et al2~ could predict the grade with an accuracy of 89%. It was concluded that 2~ is a useful radiopharmaceutical for preoperative evaluation of brain tumors. The routine use of attenuation correction was recommended because this method was most useful in low-grade lesions in the deep parenchyma. Thallium brain SPECT may be useful for differentiation of tumor recurrence from scar tissue and to predict low- or high-grade lesions. 123[ IMT
Compared with normal brain tissue, many tumors have increased protein synthesis rates and, consequently, an increased uptake of amino acids that can be measured quantitatively by positron emission tomography (PET). 21 These data make evident that 123I-labeled amino acid analogues might be used for SPECT. It had been previously reported that radioiodinated IMT is a radiopharmaceutical with high pancreatic specificity in mice22and was thus applied for imaging of the pancreas of patients. 23It had also been shown in rodents that ~31IIMT exhibited a
BRAIN SPECT IN TUMORS
Fig 4. (A) Z~ brain SPECT in a patient with glioblastoma grade IV. There is increased uptake (uptake ratio, 1.72), upper parietal left view. (B) The corresponding ROIs. (Courtesy of H. Wieler, MD, Military Hospital, Koblenz, Germany.)
high accumulation in melanomas.24Bockslaff et a124'2sreported the uptake of 1231IMT in patients with ocular melanoma. From these findings it was speculated that IMT might also be a suitable agent for SPECT imaging of brain tumors. In 1989, Biersack et als reported results in 10 patients with tumorous brain lesions, including glioblastoma, oligodendroglioma, lymphoma, and metastasis (Fig 5). Radiolabeling of IMT was performed by direct electrophilic radioiodination based on previous procedures. 22'23All 10 patients had early (10 minutes) and delayed (60 minutes) brain SPECT using a rotating gamma camera equipped with high-resolution, lowenergy collimator. Using a ROI technique, tu-
7
mor to brain tissue ratios were calculated for early and delayed images. Diagnosis of tumor was established by CT and (in four cases with primary brain tumor) by neurosurgical procedures. Above that, all patients had planar sequential scintigraphy to evaluate the kinetics of IMT within the first 5 minutes after bolus injection. The time-activity curves of brain and tumor in all 10 patients showed an initial perfusion-related peak followed by a rapid decrease, and after 30 to 60 seconds by a plateau phase. All six patients with glioblastoma or oligodendroglioma and three patients with brain metastases of bronchogenic carcinoma had increased IMT uptake in the tumor. The tumor to brain tissue ratio showed values between 1.4 and 2.6 and did not change with time. Only one patient with a small lymphoma was classed as a false negative. These data make evident that IMT is accumulated in brain tumors like 11C-labeled amino acids being used for PET studies. The incorporation rates of 1231 IMT in normal tissue are much lower compared with those obtained in rats with L-(14C)-leucine and L(11C)methionine 40 minutes postinjection.26However, the tumor uptake of the radioiodinated amino acids is comparable to that of their 11C analogues. Strong arguments against a bloodbrain barrier-mediated uptake (like with pertechnetate) are the rapid uptake of the radiopharmaceutical in the tumors as well as the constant tumor to brain tissue ratio with time. L-a-methyl-tyrosine is an enzyme inhibitor; therefore, it seems possible that there is also a "trapping mechanism" for the iodinated analogue via enzyme inhibition. UV-absorbance and radioactivity distribution after molecular mass separation of cerebral proteins containing 123I IMT by discontinuous SDS gel electrophoresis gave evidence that most of the radioactivity was contained in an apparently monomolecular fraction, indicating no incorporation into protein. 27From these data it may be concluded that IMT can be used to measure amino acid transport, as had already been suggested by Kawai et al. 28Langen et a127examined the kinetic behavior of 124I IMT in brain and plasma in two patients with glioblastoma using dynamic PET. 1241 IMT accumulated in brain and tumor tissue reaching a maximum after 15 minutes, with a washout of 22% to 35% after 1 hour. The 123I IMT-SPECT gave nearly
8
BIERSACK, GRUNWALD, AND KROPP
C F ' S / T Ir'lE< " i:, ,,,
A
H
TUMF, . . . . . . .
120-
~1. (~
....
-255. 0
Fig 5. Glioblastoma of the left hemisphere. (A) The time activity curve after IV bolus injection of lz31-ct-methyl tyrosine shows a peak after approximately 20 seconds, reaching a plateau phase after approximately 30 seconds. (B) CT shows a lesion with surrounding edema, temporoparietal left, corresponding to a lesion with increased IMT uptake in (C) brain SPECT. (Reprinted with permission of the Society of Nuclear Medicine from Biersack HJ, Coenen HH, St6cklin G, et al: Imaging of brain tumors with L-3-(ml)lodo-alpha-methyl tyrosine and SPECT. J Nucl Med 30:110-112, 1989. 5)
the same result as the 124I I M T - P E T scan. These data illustrate that the results of P E T can be transferred to SPECT using IMT. In the same study, Langen et a127 investigated 32 patients with different types of brain tumors, 26 of whom showed increased uptake of IMT. No significant differences in the degree of IMT-uptake between grade IV, grade III, or grade II gliomas could be identified in this group of patients. The radiation dose to the whole body was estimated at 0.025 rad/mCi for 123I IMT. 27 The generally relatively low amino acid uptake by the brain interferes with the evaluation of SPECT scans, but may be overcome by injecting tracer doses from 5 to 10 mCi. The latter dose appears acceptable because of the relatively low radiation dose and the rapid urinary elimination of 1231IMT. 27
Guth-Tougelides et a129 used I M T in patients with brain tumor recurrences. In an open prospective study 11 patients had IMT brain SPECT after surgery because of malignant brain tumors. In nine cases the patients had a surgically proven recurrence or were inoperable and under radiation or chemotherapy. Four patients did not show evidence of recurrent tumor. In seven of nine studies the tumor could be identified immediately on the IMT scan as an area of intense focal uptake. All patients without tumor had a negative IMT scan. From these studies it was concluded that I M T - S P E C T is a reliable tool for the differentiation of space-occupying lesions after tumor surgery. It allows a differentiation of tumor recurrence from scar tissue. These initial clinical data give evidence that IMT shows potential to evaluate therapeutic
BRAIN SPECT IN TUMORS
9
procedures in brain tumors and might thus have widespread clinical application as a SPECT radiopharmaceutical. Furthermore, IMTSPECT may improve diagnostic accuracy in patients with controversial CT or MRI findings. The differentiation between abscess and glioblastoma seems also to be possible. CONCLUSION
A variety of new radioactive tracers are now available for brain SPECT in patients with tumorous brain lesions. However, radiolabeled
amphetamines as well as HMPAO seem to be of limited use. The new brain imaging agent 123Ilabeled a-methyl tyrosine offers a useful tool for the evaluation of brain tumors, rendering possible the diffential diagnosis between abscess and tumor, the differentiation of tumor recurrence and scar tissue after brain surgery, and the evaluation of therapeutic procedures such as chemotherapy or radiotherapy. IMT-SPECT is another example that PET results can be transferred to SPECT and thus widen the spectrum of routine nuclear medicine procedures~
REFERENCES 1. R6sler H, Huber P: Die cerebrale Serienszintigraphie. Ergebnisse bei 208 Patienten. Fortschr R6ntgenstr 111:467480, 1969 2. Buell U, Niendorf HP, Kazner E, et al: Computerized transaxial tomography and cerebral serial scintigraphy in intracranial tumors--Rate of detection and tumor-type identification: Concise communication. J Nucl Med 19:476479, 1978 3. Buell U, Lanksch W: Tumor-specific pattern in brain scintigraphy, in Magistretti PL (ed): Functional Radionuclide Imaging of the Brain, Serono Symposia Publications, vol 5. New York, NY, Raven, 1983, pp 61-71 4. Lindegaard MW, Skretting A, Hager B, et al: Cerebral and cerebellar uptake of 99mTe-(d,1)-hexamethyl-propyleneamine oxime (HM-PAO) in patients with brain tumor studied by single photon emission computerized tomography. Eur J Nucl Med 12:417-420, 1986 5. Biersack HJ, Coenen HH, St6cklin G, et al: Imaging of brain tumors with L-3-(123I)Iodo-alpha-methyl tyrosine and SPECT. J Nucl Med 30:110-112, 1989 6. Biersack H J, Knopp R, Wappenschmidt J, et al: Single photon emission computed tomography of the brain with a rotating gamma camera--Results in 471 patients. NucCompact 12:130-134, 1981 7. Ell P J, Deacon JM, Dueassou D, et al: Emission and transmission brain tomography. Br Med J 280:438-440, 1980 8. Winchell HS, Horst WD, Braun L, et al: N-Isopropyl (iodine 123)p-iodo amphetamine: Single pass brain uptake and washout: Binding to brain synaptosomes and localization in dog and monkey brain. J Nucl Med 21:947-952, 1980 9. LaFranee ND, Wagner HN, Whitehouse R, et al: Decreased accumulation of isopropyl-iodo amphetamine (lz3I) in brain tumors. J Nucl Med 22:1081-1083, 1981 10. Ell PJ, Jarritt H, Costa C, et at: Functional imaging of the brain. Semin Nuel Med 17:214-229, 1987 11. Sehober O, Meyer GJ, Creuzig H, et al: IMP-SPECT and amino acid-PET in brain tumors, in Biersack HJ, Winkler C (eds): Amphetamines and pH-Shift Agents for Brain Imaging; Basic Research and Clinical Results. New York, NY, de Gruyter, 1986, pp 157-164 12. Moretti JL, Askienazy S, Raynaud C, et al: 123-I-piodo-isopropyl amphetamine for brain tumor diagnosis, in Biersack HJ, Winkler C (eds): Amphetamines and pH-Shift Agents for Brain Imaging; Basic Research and Clinical Results. New York, NY, de Gruyter, 1986, pp 167-170
13. Babich JW, Keeling F, Flower /VIA, et al: Initial experience with Tc-99m-HM-PAO in the study of brain tumors. Eur J Nucl Med 14:39-44, 1988 14. Hoshi H, Jinnouchi S, Watanabe K, et al: Mismatch between iodine-123 IMP and technetium-99m HM-PAO brain perfusion imaging in a patient with meningioma. Clin Nucl Med 12:737-740, 1987 15. Nakano S, Kinoshita K, Jinnouchi S, et al: Dynamic SPECT with technetium-99m HM-PAO in meningiomas--A comparison with iodine-123 IMP. J Nucl Med 30:1101-1105, 1989 16. Langen KJ, Roosen N, Herzog H, et al: Investigations of brain tumours with 99m-Te-HM-PAO SPECT. Nucl Med Comm 10:325-334, 1989 17. Ancri D, Bassett JY, Lonchampt MF, et al: Diagnosis of cerebral lesions by thallium-201. Nucl Med 28:417-422, 1978 18. Aneri D, Bassett JY: Diagnosis of cerebral metastases by thallium-201. Br J Radio153:443-445, 1980 19. Kaplan WD, Takronan T, Morris A, et al: Thallium201 brain tumor imaging: A comparative study with pathologic correlation. J Nucl Med 28:47-52, 1987 20. Kim T, Black KL, Marciano D, et al: Thallium-201 SPECT imaging of brain tumors: Methods and results. J Nuel Med 31:965-969, 1990 21. Bolster JM, Vaalburg W, Paans MJ, et al: Carbon labeled tyrosin to study tumor metabolism by positron emission tomography (PET). Eur J Nucl Med 12:321-324, 1986 22. Tisljar U, Kloster G, Ritzl F, et al: Accumulation of radioiodinated L-a-methyltyrosine in pancreas of mice: Concise communication. J Nucl Med 20:973-976, 1979 23. Kloster G, Coenen HH, Szabo Z, et al: Radiohalogenated 1-u-methyltyrosine as potential pancreas imaging agents for PET and SPECT, in Cox PH (ed): Progress in Radiopharmacology, vol 3. Den Haag, The Netherlands, Nijhoff, 1982, pp 97-102 24. Bockslaff H, Kloster G, StOcklin G, et al: Studies on L-3-(123I)iodo-methyl tyrosine: A new potential melanoma seeking compound. Nucl Med 17:179-182, 1980 (suppl) 25. Bockslaff H, Kloster G, Dausch D, et al: First clinical results using L-3-123I-c~-methyl tyrosine for non-invasive detection of intraoceular melanoma. Nucl Med 18:840-844, 1981 (suppl) 26. Lauenstein L, Meyer GJ, Sewing KF, et al: Uptake
10
kinetics of 14 C-L-leucine and 14 C-L- and 14 C-Dmethionine in rat brain and incorporation into protein. Neurosurg Rev 10:147-150, 1987 27. Langen KJ, Coenen HH, Roosen N, et al: SPECT studies of brain tumors with L-3-(123I)iodo-ct-methyl tyrosine: Comparison with PET, 124IMT and first clinical results. J Nucl Med 31:281-286, 1990
BIERSACK, GRONWALD, AND KROPP
28. Kawai K, Fujibayashi Y, Saji H, et al: New radioiodinated radiopharmaceutical for cerebral amino acid transport studies: 3-1odo-ct-methyl-L-tyrosine. J Nucl Med 29: 778, 1988 (abstr) 29. Guth-Tougelidis B, M/iller SP, Mehdorn HM, et al: 1-123-a-methyl-tyrosine in brain tumor recurrences. J Nucl Med 31:766, 1990 (abstr)
RAIN SCINTIGRAPHY with technetium
B 99m pertechnetate had been the only imaging procedure for brain tumors until the mid-1970s when computed tomography (CT) appeared on the market. Serial scintigraphy, including the perfusion phase as well as early and delayed images, allowed a differential diagnosis of brain tumors. 1 The serial sequential scintigraphic method enhanced the CT rate of correctly identified tumor types in meningioma and glioma as well as metastases. 2'3 Although iodine 123-labeled amphetamine derivates led to an increase in brain investigations after the "CT depression," many years passed until nuclear medicine could offer new procedures for the detection of tumorous brain lesions. The 99mTcHMPAO perfusion marker was evaluated by several groups as a tumor imaging agent? In 1989, ot-L-3-(123I) iodo-a-methyl tyrosine (IMT), a new radiopharmaceutical, was introduced for the evaluation of its incorporation into protein of brain tumors? This article summarizes the results of brain single-photon emission computed tomography (SPECT) in tumors and also attempts to analyze the potential use of this imaging procedure. From the Department of Nuclear Medicine, University of Bonn, Germany. Address reprint requests to H.J. Biersack, MD, Department of Nuclear Medicine, University of Bonn, Sigmund-FreudStrasse 25, D-5300 Bonn L Germany. Copyright 9 1991 by W.B. Saunders Company 0001-2998/91/2101-0001505. 00/0
2
as a new tumor imaging agent. Some useful clinical results were obtained in patients before and after chemotherapy or radiotherapy. Thallium-201 was incidentally noted to accumulate in tumors. Using a threshold index, this agent can be used to distinguish lowversus high-grade lesions. The most promising agent for brain tumor SPECT is 1231-~methyl tyrosine, which showes potential to evaluate therapeutic procedures in brain tumors and may improve the differentiation between abcess and glioblastoma. The most promising aspect is the differentiation of tumor recurences and scar tissue after brain surgery. Copyright 9 1991 by W.B. Saunders Company
99mTc PERTECHNETATE
Radionuclide brain imaging, performed as serial scintigraphy, proved to be effective in establishing type-specific tumor patterns. 1-3 Cerebral serial scintigraphy includes both radionuclide angiography as well as early and late static images.2 R6sler et al, 1 as well as Buell et al,2'3 presented different criteria for the differential diagnosis of brain tumors, including the perfusion ratio as well as increasing or decreasing contrast ratios. While the dynamic perfusion phase cannot be used for SPECT investigations, early and delayed activity accumulation may be evaluated by tomographic techniques (Fig 1). Our group presented the first results in 1981 describing brain SPECT using a rotating gamma camera. 6 In five of 23 patients, tumor diagnosis was only possible by SPECT. The combination of SPECT and planar imaging resulted in a specificity of 98% (412 of 422 patients). 6 Several results in a considerably smaller group of patients had already been reported by Ell et al. 7 However, SPECT with 99mTcpertechnetate is currently considered as a "historical" procedure, because CT and magnetic resonance imaging (MRI) are the tools of choice for the evaluation of brain tumors. 1231AMPHETAMINES
Soon after the first description of N-isopropylp-(l:3I)iodoamphetamine (IMP) by Winchell et al in 1980,8 LaFrance et al9 reported on the decreased accumulation of IMP in brain tumors
Seminars in Nuclear Medicine, Vol XXI, No 1 (January), 1991 : pp 2-10
BRAIN SPECT IN TUMORS
3
Fig 1. (A) Metastasis of the right cerebellum, partly superimposed by vascular activity (sinus) in the planar study, (B) Brain SPECT with "mTc allows a superior visualization of the mestastesis.
(Fig 2). However, Ell et al 1~ described some cases in which IMP concentrated actively in neoplasia of the brain. This uptake seemed not to be related to the grade of malignancy. Schober et al H reported a comparison of IMP-SPECT and amino acid-PET in brain tumors. None of the malignant brain lesions accumulated IMP, whereas "C-L-methionine uptake was present.
It was speculated that the tumors did not concentrate IMP, most probably due to the lack of appropriate receptor sites or metabolic pathways and decreased perfusion." Moretti et a112 investigated 27 patients with different brain tumors, including astrocytoma, glioblastoma, oligodendroglioma, meningioma, and metastasis, and demonstrated a decreased uptake in
4
BIERSACK, GRONWALD, AND KROPP
Fig 2. (A) Oligodendroglioma, right frontal view by CT. (B) Brain SPECT with 1231butyl amphetamine shows reduced uptake in the tumor region.
88%, although two lesions were missed. It was postulated that the removal of the monoamine IMP from circulation is achieved through the podocytes into glial cells. However, astrocytes are known to have a high affinity for noradrenaline and possess -/-aminobutyric acid (GABA) binding sites. 12To answer the question whether IMP accumulates in normal astrocytes, Moretti et a112performed investigations in astroglial cell cultures. The results showed the presence of a saturable uptake system in normal astrocytes. Tumorous astrocytes seem to loose this ability,
for example, when the GABA pathway is not functional. This may explain the lack of accumulation of IMP in astrocytomas. The results obtained with IMP in brain tumors do not provide any evidence that IMP is a useful tool for brain imaging in tumorous lesions. 99mTCHMPAO
Soon after the availability of 99mTCHMPAO on the European market, the first paper on the results of cerebral uptake of this new radiophar-
BRAIN SPECT IN TUMORS
maceutical in patients with brain tumors was published. 4 Lindegaard et al4 presented their data in 12 patients with cerebral glioma scheduled for intracarotid chemotherapy. Diagnosis had been established by surgical or stereotactic biopsies. Perfusion indexes derived from selected regions of interest (ROIs) were calculated by dividing the activity uptake in the tumor region through a similarly sized area in the corresponding region in the contralateral hemisphere. The HMPAO uptake in the tumor region was significantly lower than in the corresponding region of the contralateral, presumed normal, cerebral hemisphere in 10 of 12 patients. In one patient with a significantly higher uptake in the tumor region, it was speculated that this was caused by an earlier brain infarction in the contralateral hemisphere. No significant differences between astrocytomas and glioblastomas were found. These data indicate that blood flow in gliomas seems to be variable, but is generally lower than in a normal brain. One of the drawbacks of perfusion estimation is that a differentiation between viable tumor tissue and variable amounts of necrotic tissue or edema is not possible. These results are in so far controversial, because 133Xe injection studies have shown high local blood flow in hypervascular gliomas. This difference may be due to the binding mechanism of HMPAO. Another study on the use of HMPAO in brain tumor patients was presented by Babich et a113 before, during, and after radiotherapy. Untreated brain tumors were found to exhibit a range of 99mTCHMPAO uptake, varying from areas of markedly increased isotope activity to photopenic areas, when compared with normal brain tissue (Fig 3). As in the studies reported by Lindegaard et al, 4 a ratio of HMPAO tumor uptake to contralateral normal tissue uptake was calculated prior to and during radiotherapy. Two patients with glioma showed increased uptake; three cases with glioma and two with metastases presented with normal uptake. In five patients, HMPAO uptake was decreased. Tumor to nontumor ratios were calculated for seven patients (two with increased and five with decreased uptake in the tumor) prior to and during radiotherapy. This ratio tended to return toward unity in lesions responding to therapy. These preliminery results indicate that HMPAO
5
Fig 3. (A) Oligodendroglioma, parietal right view by CT. (B) Brain SPECT with ~rc-HMPAO shows slightly increased activity uptake in the region of the tumor.
6
imaging may be useful for providing an indirect measure of tumor therapy response, especially in metastases. In 1987, Hoshi et a114reported on a mismatch between ~23I IMP and 99mTc HMPAO brain perfusion imaging in a patient with meningioma. In their case report, the tumor showed a high blood flow as evident by the 133Xe inhalation method. In contrast, IMP imaging showed no activity uptake in the tumor. Imaging with HMPAO showed--in concordance with the xenon method--high tumor activity. It was speculated that this disparate behavior of tumor uptake is most probably due to the lack of binding sites for amphetamines. This study was later extended to seven patients with meningiomas and published by Nakano et al. ~s Brain tissue ratios calculated from the tumor ROI and the ROI of contralateral homologous area were lower for HMPAO than those of the IMPSPECT study during the first 10 minutes after intravenous injection of the tracer. It must be mentioned that Nakano et a115used a dynamic SPECT technique (ring-type gamma camera) for their investigations. These data make evident that accurate assessment of vascularity of meningiomas is not possible, and are supported by the studies of Lindegaard et al4 in which the two hypervascular tumors showed HMPAO uptake similar to tumors without this feature. In summary, HMPAO brain SPECT in tumorous lesions seems to be only of value in the follow-up of patients under treatment. In another series Langen et a116studied cerebral uptake of 99mTc-HMPAO in 66 patients with various types of brain tumors. A quantitative evaluation of HMPAO brain SPECT was performed using tumor to cerebellar ratios. Again, the uptake of HMPAO by gliomas and meningiomas showed a wide range of values with no significant differences between malignant and benign gliomas. However, meningiomas exhibited a significantly higher uptake of HMPAO. Malignant gliomas presented with an uptake ratio of 0.75, whereas meningiomas showed values of 1.14. Two patients were studied after intraarterial ACNU (nitrosurea derivative) chemotherapy. No significant changes of the uptake pattern were observed. One patient with a malignant glioma had HMPAO-SPECT before and two times during the course of
BIERSACK, GRONWALD, AND KROPP
radiotherapy. In this case, a decrease of the tumor to cerebellar ratio could be observed. THALLIUM 201 Thallium 201, first used for myocardial imaging, was incidentally noted to accumulate in tumors. Ancri et a117']8presented data in primary and metastatic brain tumors in 1978 and 1980. These studies were further extended by Kaplan et al~9 in 1987, who compared the respective findings with the results of the histopathologic examination. In a more extensive study, Kim et al2~ published a paper on 2~ SPECT in 40 patients with low-grade and 11 patients with high-grade lesions. All patients had biopsy or autopsy. ROIs were created for the tumor lesion, and ROIs for the nonaffected tissue were determined by creating a mirror image of the initially defined region on the contralateral side (Fig 4). Kim et al2~found a strong statistical difference between the 2~ uptake indexes in low-grade versus high-grade lesions. In 14 patients with biopsy- or autopsy-documented lowgrade astrocytoma, the mean 2~ index was 1.27 -+ 0.4 compared with a mean index of 2.4 _+ 0.6 for all patients with high-grade astrocytomas. Using a threshold index of 1.5 to distinguish low- versus high-grade lesions, Kim et al2~ could predict the grade with an accuracy of 89%. It was concluded that 2~ is a useful radiopharmaceutical for preoperative evaluation of brain tumors. The routine use of attenuation correction was recommended because this method was most useful in low-grade lesions in the deep parenchyma. Thallium brain SPECT may be useful for differentiation of tumor recurrence from scar tissue and to predict low- or high-grade lesions. 123[ IMT
Compared with normal brain tissue, many tumors have increased protein synthesis rates and, consequently, an increased uptake of amino acids that can be measured quantitatively by positron emission tomography (PET). 21 These data make evident that 123I-labeled amino acid analogues might be used for SPECT. It had been previously reported that radioiodinated IMT is a radiopharmaceutical with high pancreatic specificity in mice22and was thus applied for imaging of the pancreas of patients. 23It had also been shown in rodents that ~31IIMT exhibited a
BRAIN SPECT IN TUMORS
Fig 4. (A) Z~ brain SPECT in a patient with glioblastoma grade IV. There is increased uptake (uptake ratio, 1.72), upper parietal left view. (B) The corresponding ROIs. (Courtesy of H. Wieler, MD, Military Hospital, Koblenz, Germany.)
high accumulation in melanomas.24Bockslaff et a124'2sreported the uptake of 1231IMT in patients with ocular melanoma. From these findings it was speculated that IMT might also be a suitable agent for SPECT imaging of brain tumors. In 1989, Biersack et als reported results in 10 patients with tumorous brain lesions, including glioblastoma, oligodendroglioma, lymphoma, and metastasis (Fig 5). Radiolabeling of IMT was performed by direct electrophilic radioiodination based on previous procedures. 22'23All 10 patients had early (10 minutes) and delayed (60 minutes) brain SPECT using a rotating gamma camera equipped with high-resolution, lowenergy collimator. Using a ROI technique, tu-
7
mor to brain tissue ratios were calculated for early and delayed images. Diagnosis of tumor was established by CT and (in four cases with primary brain tumor) by neurosurgical procedures. Above that, all patients had planar sequential scintigraphy to evaluate the kinetics of IMT within the first 5 minutes after bolus injection. The time-activity curves of brain and tumor in all 10 patients showed an initial perfusion-related peak followed by a rapid decrease, and after 30 to 60 seconds by a plateau phase. All six patients with glioblastoma or oligodendroglioma and three patients with brain metastases of bronchogenic carcinoma had increased IMT uptake in the tumor. The tumor to brain tissue ratio showed values between 1.4 and 2.6 and did not change with time. Only one patient with a small lymphoma was classed as a false negative. These data make evident that IMT is accumulated in brain tumors like 11C-labeled amino acids being used for PET studies. The incorporation rates of 1231 IMT in normal tissue are much lower compared with those obtained in rats with L-(14C)-leucine and L(11C)methionine 40 minutes postinjection.26However, the tumor uptake of the radioiodinated amino acids is comparable to that of their 11C analogues. Strong arguments against a bloodbrain barrier-mediated uptake (like with pertechnetate) are the rapid uptake of the radiopharmaceutical in the tumors as well as the constant tumor to brain tissue ratio with time. L-a-methyl-tyrosine is an enzyme inhibitor; therefore, it seems possible that there is also a "trapping mechanism" for the iodinated analogue via enzyme inhibition. UV-absorbance and radioactivity distribution after molecular mass separation of cerebral proteins containing 123I IMT by discontinuous SDS gel electrophoresis gave evidence that most of the radioactivity was contained in an apparently monomolecular fraction, indicating no incorporation into protein. 27From these data it may be concluded that IMT can be used to measure amino acid transport, as had already been suggested by Kawai et al. 28Langen et a127examined the kinetic behavior of 124I IMT in brain and plasma in two patients with glioblastoma using dynamic PET. 1241 IMT accumulated in brain and tumor tissue reaching a maximum after 15 minutes, with a washout of 22% to 35% after 1 hour. The 123I IMT-SPECT gave nearly
8
BIERSACK, GRUNWALD, AND KROPP
C F ' S / T Ir'lE< " i:, ,,,
A
H
TUMF, . . . . . . .
120-
~1. (~
....
-255. 0
Fig 5. Glioblastoma of the left hemisphere. (A) The time activity curve after IV bolus injection of lz31-ct-methyl tyrosine shows a peak after approximately 20 seconds, reaching a plateau phase after approximately 30 seconds. (B) CT shows a lesion with surrounding edema, temporoparietal left, corresponding to a lesion with increased IMT uptake in (C) brain SPECT. (Reprinted with permission of the Society of Nuclear Medicine from Biersack HJ, Coenen HH, St6cklin G, et al: Imaging of brain tumors with L-3-(ml)lodo-alpha-methyl tyrosine and SPECT. J Nucl Med 30:110-112, 1989. 5)
the same result as the 124I I M T - P E T scan. These data illustrate that the results of P E T can be transferred to SPECT using IMT. In the same study, Langen et a127 investigated 32 patients with different types of brain tumors, 26 of whom showed increased uptake of IMT. No significant differences in the degree of IMT-uptake between grade IV, grade III, or grade II gliomas could be identified in this group of patients. The radiation dose to the whole body was estimated at 0.025 rad/mCi for 123I IMT. 27 The generally relatively low amino acid uptake by the brain interferes with the evaluation of SPECT scans, but may be overcome by injecting tracer doses from 5 to 10 mCi. The latter dose appears acceptable because of the relatively low radiation dose and the rapid urinary elimination of 1231IMT. 27
Guth-Tougelides et a129 used I M T in patients with brain tumor recurrences. In an open prospective study 11 patients had IMT brain SPECT after surgery because of malignant brain tumors. In nine cases the patients had a surgically proven recurrence or were inoperable and under radiation or chemotherapy. Four patients did not show evidence of recurrent tumor. In seven of nine studies the tumor could be identified immediately on the IMT scan as an area of intense focal uptake. All patients without tumor had a negative IMT scan. From these studies it was concluded that I M T - S P E C T is a reliable tool for the differentiation of space-occupying lesions after tumor surgery. It allows a differentiation of tumor recurrence from scar tissue. These initial clinical data give evidence that IMT shows potential to evaluate therapeutic
BRAIN SPECT IN TUMORS
9
procedures in brain tumors and might thus have widespread clinical application as a SPECT radiopharmaceutical. Furthermore, IMTSPECT may improve diagnostic accuracy in patients with controversial CT or MRI findings. The differentiation between abscess and glioblastoma seems also to be possible. CONCLUSION
A variety of new radioactive tracers are now available for brain SPECT in patients with tumorous brain lesions. However, radiolabeled
amphetamines as well as HMPAO seem to be of limited use. The new brain imaging agent 123Ilabeled a-methyl tyrosine offers a useful tool for the evaluation of brain tumors, rendering possible the diffential diagnosis between abscess and tumor, the differentiation of tumor recurrence and scar tissue after brain surgery, and the evaluation of therapeutic procedures such as chemotherapy or radiotherapy. IMT-SPECT is another example that PET results can be transferred to SPECT and thus widen the spectrum of routine nuclear medicine procedures~
REFERENCES 1. R6sler H, Huber P: Die cerebrale Serienszintigraphie. Ergebnisse bei 208 Patienten. Fortschr R6ntgenstr 111:467480, 1969 2. Buell U, Niendorf HP, Kazner E, et al: Computerized transaxial tomography and cerebral serial scintigraphy in intracranial tumors--Rate of detection and tumor-type identification: Concise communication. J Nucl Med 19:476479, 1978 3. Buell U, Lanksch W: Tumor-specific pattern in brain scintigraphy, in Magistretti PL (ed): Functional Radionuclide Imaging of the Brain, Serono Symposia Publications, vol 5. New York, NY, Raven, 1983, pp 61-71 4. Lindegaard MW, Skretting A, Hager B, et al: Cerebral and cerebellar uptake of 99mTe-(d,1)-hexamethyl-propyleneamine oxime (HM-PAO) in patients with brain tumor studied by single photon emission computerized tomography. Eur J Nucl Med 12:417-420, 1986 5. Biersack HJ, Coenen HH, St6cklin G, et al: Imaging of brain tumors with L-3-(123I)Iodo-alpha-methyl tyrosine and SPECT. J Nucl Med 30:110-112, 1989 6. Biersack H J, Knopp R, Wappenschmidt J, et al: Single photon emission computed tomography of the brain with a rotating gamma camera--Results in 471 patients. NucCompact 12:130-134, 1981 7. Ell P J, Deacon JM, Dueassou D, et al: Emission and transmission brain tomography. Br Med J 280:438-440, 1980 8. Winchell HS, Horst WD, Braun L, et al: N-Isopropyl (iodine 123)p-iodo amphetamine: Single pass brain uptake and washout: Binding to brain synaptosomes and localization in dog and monkey brain. J Nucl Med 21:947-952, 1980 9. LaFranee ND, Wagner HN, Whitehouse R, et al: Decreased accumulation of isopropyl-iodo amphetamine (lz3I) in brain tumors. J Nucl Med 22:1081-1083, 1981 10. Ell PJ, Jarritt H, Costa C, et at: Functional imaging of the brain. Semin Nuel Med 17:214-229, 1987 11. Sehober O, Meyer GJ, Creuzig H, et al: IMP-SPECT and amino acid-PET in brain tumors, in Biersack HJ, Winkler C (eds): Amphetamines and pH-Shift Agents for Brain Imaging; Basic Research and Clinical Results. New York, NY, de Gruyter, 1986, pp 157-164 12. Moretti JL, Askienazy S, Raynaud C, et al: 123-I-piodo-isopropyl amphetamine for brain tumor diagnosis, in Biersack HJ, Winkler C (eds): Amphetamines and pH-Shift Agents for Brain Imaging; Basic Research and Clinical Results. New York, NY, de Gruyter, 1986, pp 167-170
13. Babich JW, Keeling F, Flower /VIA, et al: Initial experience with Tc-99m-HM-PAO in the study of brain tumors. Eur J Nucl Med 14:39-44, 1988 14. Hoshi H, Jinnouchi S, Watanabe K, et al: Mismatch between iodine-123 IMP and technetium-99m HM-PAO brain perfusion imaging in a patient with meningioma. Clin Nucl Med 12:737-740, 1987 15. Nakano S, Kinoshita K, Jinnouchi S, et al: Dynamic SPECT with technetium-99m HM-PAO in meningiomas--A comparison with iodine-123 IMP. J Nucl Med 30:1101-1105, 1989 16. Langen KJ, Roosen N, Herzog H, et al: Investigations of brain tumours with 99m-Te-HM-PAO SPECT. Nucl Med Comm 10:325-334, 1989 17. Ancri D, Bassett JY, Lonchampt MF, et al: Diagnosis of cerebral lesions by thallium-201. Nucl Med 28:417-422, 1978 18. Aneri D, Bassett JY: Diagnosis of cerebral metastases by thallium-201. Br J Radio153:443-445, 1980 19. Kaplan WD, Takronan T, Morris A, et al: Thallium201 brain tumor imaging: A comparative study with pathologic correlation. J Nucl Med 28:47-52, 1987 20. Kim T, Black KL, Marciano D, et al: Thallium-201 SPECT imaging of brain tumors: Methods and results. J Nuel Med 31:965-969, 1990 21. Bolster JM, Vaalburg W, Paans MJ, et al: Carbon labeled tyrosin to study tumor metabolism by positron emission tomography (PET). Eur J Nucl Med 12:321-324, 1986 22. Tisljar U, Kloster G, Ritzl F, et al: Accumulation of radioiodinated L-a-methyltyrosine in pancreas of mice: Concise communication. J Nucl Med 20:973-976, 1979 23. Kloster G, Coenen HH, Szabo Z, et al: Radiohalogenated 1-u-methyltyrosine as potential pancreas imaging agents for PET and SPECT, in Cox PH (ed): Progress in Radiopharmacology, vol 3. Den Haag, The Netherlands, Nijhoff, 1982, pp 97-102 24. Bockslaff H, Kloster G, StOcklin G, et al: Studies on L-3-(123I)iodo-methyl tyrosine: A new potential melanoma seeking compound. Nucl Med 17:179-182, 1980 (suppl) 25. Bockslaff H, Kloster G, Dausch D, et al: First clinical results using L-3-123I-c~-methyl tyrosine for non-invasive detection of intraoceular melanoma. Nucl Med 18:840-844, 1981 (suppl) 26. Lauenstein L, Meyer GJ, Sewing KF, et al: Uptake
10
kinetics of 14 C-L-leucine and 14 C-L- and 14 C-Dmethionine in rat brain and incorporation into protein. Neurosurg Rev 10:147-150, 1987 27. Langen KJ, Coenen HH, Roosen N, et al: SPECT studies of brain tumors with L-3-(123I)iodo-ct-methyl tyrosine: Comparison with PET, 124IMT and first clinical results. J Nucl Med 31:281-286, 1990
BIERSACK, GRONWALD, AND KROPP
28. Kawai K, Fujibayashi Y, Saji H, et al: New radioiodinated radiopharmaceutical for cerebral amino acid transport studies: 3-1odo-ct-methyl-L-tyrosine. J Nucl Med 29: 778, 1988 (abstr) 29. Guth-Tougelidis B, M/iller SP, Mehdorn HM, et al: 1-123-a-methyl-tyrosine in brain tumor recurrences. J Nucl Med 31:766, 1990 (abstr)