Positron emission tomographic assessment of cerebral hemocirculation and glucose metabolism in malignant glioma following treatment with intracarotid recombinant human tumor necrosis

Journal of Neuro-Oncology 23: 67-73, 1995. �9 1995 Kluwer Academic Publishers. Printed in the Netherlands.

ClinicalStudy

Positron emission tomographic assessment of cerebral hemocirculation and glucose metabolism in malignant glioma following treatment with intracarotid recombinant human tumor necrosis factor-~

Toshio Sasajima, 1 Katsuyoshi Mineura, 1 Junkoh Sasaki, 1 Masayoshi Kowada, 1 Noriaki Tomura, 2 Jun Hatazawa, 3 Toshihide Ogawa 3 and Kazuo Uemura 3 1 Neurosurgical Service and 2 Department of Radiology, Akita University Hospital, and 3 Department of Radiology and Nuclear Medicine, Research Institute for Brain and Blood Vessels-Akita, Akita, Japan

Key words: malignant glioma, recombinant human tumor necrosis factor-c~, positron emission tomography, hemocirculation, glucose metabolism

Summary

Cerebral hemocirculation and glucose metabolism in a malignant astrocytoma were repeatedly quantified before and after intracarotid injection of recombinant human tumor necrosis factor-a (rH-TNF) using pos- itron emission tomography (PET). The patient received an intracarotid injection of a 3 x 104 U/m 2 dose of rH-TNF three times over a two week period. PET was performed prior to and 24 hr after the first injection, and two weeks after the third injection. Prior to the first rH-TNF treatment, two lesions demonstrating high perfusion and hypermetabolism of glucose were noted in the right frontal and temporal regions. The frontal hypermetabolic lesion showed decreases in hemocirculation and metabolism 24 hr after the first injection and then increases beyond the pre-treatment level two weeks after the third treatment, whereas the temporal lesion remained unchanged during the follow-up period. No appreciable changes were noted in the adjacent cortex where rH-TNF was perfused, with the exception of a transient decrease in regional blood volume. Magnetic resonance images of the tumor showed no changes as a result of treatment with intracarotid rH- TNK Intracarotid rH-TNF preferentially affects tumor tissue as opposed to normal cortex.

Introduction

Recombinant human tumor necrosis factor-o~ (rH- TNF), a non-glycosylated 17 kD protein, was suc- cessfully cloned using a recombinant DNA tech- nique [1]. Bulk production of purified rH-TNF facil- itates the elucidation of its biological effects in ex- perimental and clinical studies, rH:TNF has been characterized as a cytokine showing a variety of bi- ological activities including direct and indirect anti- tumor effects, rH-TNF exhibits direct cytolytic and cytostatic effects on tumor cell lines [2]. It also in- duces damage to the endothelial cells of tumor ves- sels [3], disrupts the permeability of tumor vessels

[4, 5], and mediates activated macrophage cytotox- icity [6]. The indirect action of rH-TNF on tumor vessels results in coagulation necrosis and leads to tumor shrinkage in transplanted tumors that lack sensitivity to rH-TNF in vitro [7, 8]. rH-TNF inhib- ition of tumor growth has proven to be dose-de- pendent in both in vivo and in vitro experiments [2, 9].

Clinical application of intravenous high dose rH- TNF demonstrates limited efficacy in patients with advanced cancers [10], and causes adverse effects such as hypotension and hepatotoxity [11]. Intra-ar- terial administration of rH-TNF is a pertinent al- ternative to lessen systemic toxicity and to maxi-

68

mize local drug delivery to the tumor bed. Recently, Yoshida et al. [12] reported the clinical efficacy of intracarotid rH-TNF in recurrent and progressive malignant gliomas. Tumors responded well to the treatment in three out of 15 patients. Response to rH-TNF on computed tomography (CT) and mag- netic resonance (MR) images was observed for at least four weeks after treatment. Neurological symptoms improved in four cases without objective responses on either CT or MR images. It is impor- tant in the management of patients treated with in- tracarotid rH-TNF to understand the effects of rH- TNF on tumor response and neurological symp- toms.

The assessment of hemocirculation and glucose metabolism using positron emission tomography (PET) has been reported to be valuable in the de- termination of histological malignancy [13, 14], monitoring of therapeutic effects [15-17] and pre- diction of survival [18] in glioma patients. The most pronounced changes in tumor glucose metabolism were observed five or six days after the first intra- arterial nimustine hydrochloride (ACNU) injec- tion [15]. Rozental et al. [17] demonstrated that tu- mor glucose metabolism increased 24 hr after eight- drugs-in-one-day chemotherapy and subsequently decreased until 28 days after treatment. Under- standing of the effects on blood flow and metabo- lism has been scarce. We applied PET in the eluci- dation of the effects of intracarotid rH-TNF, and monitored blood flow and metabolism before and after rH-TNF-treatment of malignant glioma.

Materials and methods

P a ~ e n t

A 48-year-old male with a one-month history of mild left hemiparesis was admitted. MR images re- vealed gadolinium-enhancing lesions with sur- rounding high T2-weighted signal in the right front- al and temporal regions (Fig. 1). In cerebral angio- grams, early venous filling and tumor staining were mainly fed by the right operuculofrontal artery (Fig. 2, left). The patient underwent an uneventful

Fig. 1. MR images revealed solid and cystic tumor components with extensive perifocal edema consistent with tumor invasion. (from top to bottom row: Tl-weighted (490/10) images, T2- weighted (3000/90) images, postcontrast Tl-weighted (490/10) images.)

exploratory biopsy followed by external decom- pression. Specimens were histologically diagnosed as malignant astrocytoma (Grade III).

Six weeks after the surgery, treatment with rH- TNF (Asahi Chemical Industry Co., Ltd., Tokyo) was initiated, at an injection rate of 1 ml/min (a total dose of 3 x 104 U/m2/day in 15 ml) through the right internal carotid artery. Intracarotid rH-TNF was in- jected three times over a two week period. The pa- tient remained asymptomatic except for a mild chill during the injection. The right internal carotid and middle cerebral arteries were diffusely narrowed only at the first injection (Fig. 2, right). The patient received no other treatment before or during intra- carotid rH-TNF.

69

Fig. 2. Right carotid angiogram demonstrated tumor staining fed by the operuculofrontal artery (left). Immediately after the injection, angiogram revealed diffuse narrowing of internal carotid and middle cerebral arteries, but no change in tumor staining (right).

P E T m e a s u r e m e n t Results

PET studies were performed on the day before the first injection, 24 hr after the first injection, and two weeks after the third injection. The Headtome IV apparatus employed for this study is a high resolu- tion PET scanner. The patient inhaled 1850 MBq of C150 with room air. After a four minute equilibrium period, radioactivity of hemoglobin-bound C150 in the vessels was measured to determine regional blood volume (rCBV) [19]. Regional cerebral blood flow (rCBF) was measured by administering 1110 MBq of H2150 intravenously according to the autoradiographic method [20]. Between 290 and 370 MBq of 18F-fluorodeoxyglucose (FDG) was then administered intravenously, and the distribu- tion of FDG in the brain was measured 60 minutes later. Periodical arterial blood sampling was per- formed to measure plasma FDG radioactivity. Us- ing FDG radioactivity of brain and plasma, regional cerebral metabolic rate of glucose (rCMRG1) was calculated by the Phelps' equation, with kinetic rate constants for normal brain tissue [21] and a lumped constant of 0.52 [22].

Prior to the rH-TNF treatment, two FDG-hyper- metabolic lesions as well as high hemodynamic val- ues were noted in the frontal and temporal regions. Values for the frontal and temporal portions were as follows: rCBF: 49.5ml/100ml/min, 43.6ml/ 100 ml/min; rCBV: 5.6 ml/100 ml, 6.1 ml/100 ml; and rCMRG1; 6.4 rag/100 ml/min, 7.7 mg/100 ml/min, respectively. These values were higher than those for the contralateral cortex. Hemocirculation and metabolism were reduced in the perifocal edema- tous portion (rCBF: 25.0ml/100ml/min; rCBV: 3.0 ml/100 ml; rCMRGl: 3.9 mg/100 ml/min) (Figs 3-5, top row; Fig. 6).

Twenty-four hours after the first injection, rCBF, rCBV and rCMRG1 in the frontal FDG-hyperme- tabolic lesion decreased to 85%, 55% and 55% of pre-treatment levels, respectively. In contrast, val- ues in the temporal FDG-hypermetabolic lesion re- mained unchanged, rCBV fell to 80% of pre-treat- ment values in the ipsilateral cortex corresponding to the territory of the right internal carotid artery, although rCBF and rCMRG1 remained almost con-

70

Fig. 3. The first PET study (top row) revealed two hyperperfu- sion areas in the right frontal and temporal lobes. In the frontal lesion, rCBF temporarily decreased 24 hr after the first rH-TNF (middle row), but increased markedly beyond the pre-treatment levels two weeks after the third treatment (bottom row). In con- trast, rCBF remained unchanged in the temporal lesion.

stant. All values in the contralateral cortex showed no significant changes (Figs 3-5, middle row; Fig. 6).

A follow-up PET obtained two weeks after the third t reatment demonstra ted a marked increase in rCBF and rCMRG1 in the frontal lesion (rCBF: 96.4 ml/100 ml/min; rCMRGI: 7.6 mg/100 ml/min), although these values remained unchanged in the temporal lesion (rCBF: 44.9 ml/100 ml/min; rCMRGI: 6.8 mg/100 ml/min) (Figs 3-5, bo t tom row; Fig. 6). Neither ipsilateral nor contralateral non-involved cortex demonstra ted obvious chang- es in hemodynamic or metabolic values. These dra- matic changes did not appear on the concurrent M R

images. During the four months following the ther- apy, the patient has been free of symptoms, showing no evidence of tumor regrowth on M R images.

Fig. 4. Changes in rCBV at the tumor site were similar to those in rCBF. In the normal cortex corresponding to the territory of right internal carotid artery, rCBV fell abruptly after the first in- jection (middle row), but returned to the pre-treatment level af- ter the third treatment (bottom row).

Discussion

In the present study, we measured changes in hemo- circulation and glucose metabol ism in tumor fol- lowing intracarotid rH-TNF, even when concurrent M R images showed no morphological changes. Both rCBF and rCBV in the frontal hypermetabol- ic lesion decreased 24 hr after the treatment, rH- TNF caused a greater reduction in rCBV than in rCBF, which may reflect a decrease in tumor vascu- larity, although angiograms revealed no changes in tumor staining immediately after the treatment. In murine Meth A f ibrosarcoma treated with rH-TNF, structural vascular volume does not decrease mark- edly until 24 hours later, In contrast, functional vas- cular volume, which correlates well to blood flow, is reduced within four hours, and then returns to its

Fig. 5. FDG-PET images showed two hypermetabolic lesions in the frontal and temporal lobes (top row). The frontal hyperme- tabolic lesion temporarily disappeared after the first injection (middle row), but appeared after the third one (bottom row). In contrast, the temporal lesion remained unchanged.

previous level 24 hours after treatment with a lower dose of rH-TNF [23]. Nawroth et al. [24] suggest that intravascular thrombi contribute greatly to the reduction in tumor perfusion, because fibrin depo- sition is present 30 min after rH-TNF treatment and occlusive thrombi and formed two hours after the treatment. Our observation of the decrease in he- mocirculation agrees with these experimental re- sults. The reduction in tumor rCBF and rCBV in the present study likely corresponds to the above-men- tioned experimental results.

Glucose metabolism in the frontal hypermeta- bolic lesion was also suppressed after the first treat- ment. Recently, the acute effect of rH-TNF on tu- mor blood flow and high energy-phosphate metab- olism has been noted in experimental studies using 31P-nuclear magnetic resonance spectroscopy

71

(MRS). Intravenous rH-TNF induces a progressive depletion of adenosine triphosphate (ATP) togeth- er with an increase in inorganic phosphate in me- thylcholanthrene-induced sarcomas. Karczmar et

al. [25] measured tumor blood flow by MRS to de- termine whether changes in energy metabolism re- flect the direct metabolic effects of rH-TNF or sec- ondary effects due to vascular damage and subse- quent ischemia. These metabolic changes in high- energy phosphate may be due to a decrease in tu- mor blood flow, since the decrease in ATP generally follows a significant reduction in D20 uptake, which correlates well with tumor perfusion. In addi- tion, acute ischemia in the tumor produces similar metabolic changes [26]. These notions are support- ed by the fact that tumor blood flow is reduced with- in the first hour following treatment, after which time tumor energy metabolism is suppressed in a dose-dependent manner [27]. Effects of rH-TNF on metabolism may be secondary to its suppression of tumor perfusion.

Two weeks after the third treatment, tumor he- mocirculation and metabolism increased beyond the pre-treatment level. This augmentation was in- terpreted as having resulted from a release from the effect of rH-TNF on hemocirculation, since we have not previously observed such a high rCBF in gliomas, unlike meningiomas [28]. Another possi- bility is that these changes indicate tumor progres- sion. Low dose rH-TNF increases tumor blood flow and tumor growth in moderately rH-TNF-sensitive transplanted tumors [29]. An increase in glucose metabolism after chemotherapy predicts tumor progression [15, 17].

Another hypermetabolic lesion with a different response to intracarotid rH-TNF was noted in the lateral part of the tumor. It is still not understood why only part of the tumor demonstrated these changes. There is some evidence, however, for het- erogeneous response of blood flow in experimental tumors after rH-TNF administration. Preferential reduction of blood flow is noted in the central areas of experimental murine tumors [23]. rH-TNF im- pairs peripheral tumor blood flow prior to global reduction in rodent tumors [27]. Hemocirculation and glucose metabolism in gliomas may change with site-specificity after intracarotid injection.

72

ml/lOOml/min

100

80

60

40

rCBF

.•• 6

4

- - - - ~ 2 ~ I * " - - - - " - - *

ml/lOOml

8

rCBV rCMRGI mg/lOOrnl/min

8

4 ,_____~,______,

2O

I " , , ' 0 ' ' ' 0 " [ ' ' ' ' Pre Post 1 st 3rd Pre Post 1 st 3rd Pre Post 1 st 3rd

r frontal hypermetabolic lesion temporal hypermetabolic lesion

* perifocal edema �9 right parietal cortex

/~ left parietal cortex �9 right occipital cortex 0 left occipital cortex

Fig. 6. Changes in hemocirculation and glucose metabolism following intracarotid rH-TNE In the frontal lesion (open star), all values fell abruptly after the first treatment, and later rose beyond pre-treatment levels, while values for the temporal lesion (closed star) remained unchanged. In the perifocal edema (asterisk), all values were suppressed to a lower level as compared with normal cortex. In the cortex corresponding to the territory of the right carotid artery (closed triangle), there were no significant changes in hemocirculation or metabolism, with the exception of a transient decrease in rCBV. In the normal cortex without infusion of rH-TNF (open triangle, circle),

all values were constant.

Further study in a large series will make clear the effects of rH-TNF on hemocirculation and metabo- lism in gliomas.

Current histological studies have suggested an acceptable level of toxicity of rH-TNF to normal brain structure [4]. To our knowledge, the effect of rH-TNF on hemocirculation and metabolism in normal cortex has not yet been reported. In the ad- jacent cortex where rH-TNF was perfused, rCBV fell 24 hr after the first injection, but returned to its pre-treatment level two weeks after the third injec- tion. Additionally, rCBF and rCMRG1 values re- mained almost constant during the follow-up peri- od. Hemocirculation and metabolism remained un- changed in the contralateral normal cortex where rH-TNF was not perfused, rH-TNF has only a tran- sient effect on hemocirculation and metabolism in the normal cortex. These preliminary results sug- gest that intracarotid rH-TNF preferentially influ- ences hemocirculation and metabolism in tumor as compared to normal cortex.

A c k n o w l e d g e m e n t s

We arc greatly indebted to the staff of the Research Institute for Brain and Blood Vessels-Akita for their cooperation and assistance.

R e f e r e n c e s

1. Shirai T, Yamaguchi H, Ito H, Todd CW, Wallace RB: Clon- ing and expression in Escherichia coli of the gene for human tumour necrosis factor. Nature 313: 803-806,1985

2. Sugarman B J, Aggarwal BB, Hass PE, Figari IS, Palladino MA Jr, Shepard HM: Recombinant human tumor necrosis factor-c~: Effects on proliferation of normal and transformed cells in vitro. Science 230: 943-945,1985

3. Watanabe N, Niitsu Y, Umeno H, Kuriyama H, Neda H, Ya- mauchi N, Maeda M, Urushizaki I: Toxic effect of tumor ne- crosis factor on tumor vasculature in mice. Cancer Res 48: 2179-2183, 1988

4. Kido G, Wright JL, Merchant RE: Acute effects of human recombinant tumor necrosis factor-~ on the cerebral vascu- lature of the rat in both normal brain and in an experimental glioma model. J Neuroonco110: 95-109, 1991

5. Mullin JM, Snock KV: Effect of tumor necrosis factor on epithelial tight junctions and transepithelial permeability. Cancer Res 50: 2172-2176, 1990

6. Le J, gil~ek J: Tumor necrosis factor and interleukin 1: Cyto-

kines with multiple overlapping biological activities. Lab In- vest 56: 234-248, 1987

7. Liu SKM, Jakowatz JG, Pollack RB, Ceraldi C, Yamamoto R, Dett C, Lopez E Camacho C, Carson WE, Sentovich SM, Jacques D, Granger GA: Effects of intracarotid and intrave- nous infusion of human TNF and LT on established intrace- rebral rat gliomas. Lymphokine Cytokine Res 10: 189-194, 1991

8. Shimomura K, Manda T, Mukumoto S, Kobayashi K, Naka- no K, Mori J: Recombinant human tumor necrosis factor-co Thrombus formation is a cause of anti-tumor activity. Int J Cancer 41: 243-247, 1988

9. Sohmura Y, Nakata K, Yoshida H, Kashimoto S, Matsui Y, Furuichi H: Recombinant human tumor necrosis factor. II. Antitumor effect on routine and human tumors transplant- ed in mice. Int J Immunopharmacol 8: 357-368, 1986

10. Kemeny N, Childs B, Larchian W, Rosado K, Kelsen D: A phase II trial of recombinant tumor necrosis factor in pa- tients with advanced colorectal carcinoma. Cancer 66: 659- 663, 1990

11. Creaven PJ, Brenner DE, Cowens JW, Huben RP, Wolf RM, Takita H, Arbuck SG, Razack MS, Proefrock AD: A phase I clinical trial of recombinant human tumor necrosis factor given daily for five days. Cancer Chemother Pharmaco123: 186-191, 1989

12. Yoshida J, Wakabayashi T, Mizuno M, Sugita K, Yoshida T, Hori S, Mori T, Sato T, Karashima A, Kurisu K, Kiya K, Uo- zumi T: Clinical effect of intra-arterial tumor necrosis fac- tor-c~ for malignant glioma. J Neurosurg 77: 78-83, 1992

13. Di Chiro G: Positron emission tomography using [18F]fluo- rodeoxyglucose in brain tumors: A powerful diagnostic and prognostic tool. Invest Radiol 22: 360-371, 1987

14. Mineura K, Yasuda T, Kowada M, Shishido E Ogawa T, Uemura K: Positron emission tomographic evaluation of histological malignancy in gliomas using oxygen-15 and fluo- rine-18-fluorodeoxyglucose. Neurol Res 8: 164-168, 1986

15. Langen KJ, Roosen N, Kuwert T, Herzog H, Kiwit JCW, Kops ER, Muzik O, Bock WJ, Feinendegen LE: Early ef- fects of intra-arterial chemotherapy in patients with brain tumours studied with PET: Preliminary results. Nuc Med Comm 10: 779-790, 1989

16. Mineura K, Yasuda T, Kowada M, Ogawa T, Shishido E Uemura K: Positron emission tomographic evaluation of ra- diochemotherapeutic effect on regional cerebral hemocir- culation and metabolism in patients with gliomas. J Neu- rooncol 5: 277-285, 1987

17. Rozental JM, Levine RL, Nickles RJ, Dobkin JA: Glucose uptake by gliomas after treatment: A positron emission to- mographic study, Arch Neuro146: 1302-1307, 1989

18. Patronas NJ, Di Chiro G, Kufta C, Bairamian D, Kornblith PL, Simon R, Larson SM: Prediction of survival in glioma

73

patients by means of positron emission tomography. J Neu- rosurg 62: 816-822, 1985

19. Phelps ME, Huang SC, Hoffman EJ, Kuhl DE: Validation of tomographic measurement of cerebral blood volume with C-11-1abeled carboxyhemoglobin. J Nucl Med 20: 328-334, 1979

20. Kanno I, Iida H, Miura S, Murakami M: Optimal scan time of oxygen-15-1abeled water injection method for measure- ment of cerebral blood flow. J Nucl Med 32: 1931-1934, 1991

21. Phelps ME, Huang SC, Hoffman E J, Selin C, Sokoloff L, Kuhl DE: Tomographic measurement of local cerebral glu- cose metabolic rate in humans with (F-18)2-fluoro-2-deoxy- D-glucose: Validation of method. Ann Neurol 6: 371-388, 1979

22. Reivich M, Alavi A, Wolf A, Fowler J, Russell J, Arnett C, MacGregor RR, Shiue CY, Atkins H, Anand A, Dann R, Greenberg JH: Glucose metabolic rate kinetic model pa- rameter determination in humans: The lumped constants and rate constants for [18F]fluorodeoxyglucose and [11C]de- oxyglucose: J Cereb Blood Flow Metab 5: 179-192, 1985

23. Van de Wiel PA, Bouma GJ, Van der Pijl A, Weitenberg ES, Lam AW, Bloksma N: Effect of tumour necrosis factor and lipid A on functional and structural vascular volume in solid nmrine tumours. Br J Cancer 62: 718-723, 1990

24. Nawroth R Handley D, Matsueda G, Waal RD, Gerlach H, Blohm D, Stern D: Tumor necrosis factor/cachectin-induced intravascular fibrin formation in Meth A fibrosarcomas. J Exp Med 168: 637-647, 1988

25. Karczmar GS, Meyerhoff D J, Speder A, Valone F, Wilkin- son M, Shine N, Boska MD, Weiner MW: Response of tu- mors to therapy studied by 31p magnetic resonance spectros- copy. Invest Radiol 24: 1020-1023, 1989

26. Shine N, Palladino MA Jr, Patton JS, Deisseroth A, Karcz- mar GS, Matson GB, Weiner MW: Early metabolic response to tumor necrosis factor in mouse sarcoma: A phospho- rus-31 nuclear magnetic resonance study. Cancer Res 49: 2123-2127, 1989

27. Kluge M, Elger B, Engel T, Schaefer C, Seega J, Vaupel P: Acute effects of tumor necrosis factor a or lymphotoxin on global blood flow, laser doppler flux, and bioenergetic status of subcutaneous rodent tumors. Cancer Res 52: 2167-2173, 1992

28. Mineura K, Sasajima T, Kowada M, Shishido F, Uemura K: Positron emission tomography (PET) study in patients with meningiomas. No To Shinkei 42:145-151,1990 (in Japanese)

29. Kallinowski F, Schaefer C, Tyler G, Vaupel P: In vivo targets of recombinant human tumour necrosis factor-co blood flow, oxygen consumption and growth of isotransplanted rat tumours. Br J Cancer 60: 555-560, 1989

Address for offprints: T. Sasajima, Neurosurgical Service, Akita University Hospital, 1-1-1 Hondo, Akita 010, Japan