Pharmacological blood-brain barrier modification for selective drug delivery

Vasoactive agents have been identified through studies of peritumoral edema and effects on systemic capillaries. Abnormal blood-brain barrier or blood-tumor barrier can develop transient increases in permeability with the intraarterial delivery of vasoactive agents. Normal blood-brain barrier resists the effects of these compounds because of a biochemical barrier that may inactivate or become inert to vasoactive agents. Vasoactive compounds, including leukotrienes, bradykinin, and histamine appear to selectively increase permeability in abnormal brain capillaries. Intracarotid infusion of leukotrienes, bradykinin, and other vasoactive agents can increase drug delivery to diseased tissue.

High expression of the Glut1 glucose transporter in human brain hemangioblastoma endothelium

The principal glucose transporter at the blood-brain barrier is Glut1, and GLUT1 expression is downregulated in high grade gliomas. In the present study, glucose transporter expression was studied in surgically resected hemangioblastoma tissue. Light microscopic immunochemistry indicated the high expression of the Glut1 glucose transporter isoform throughout the central vascular endothelium of this tissue. Glial fibrillary acidic protein (GFAP) was observed only at the tumor border, with no GFAP immunoreactivity in stromal cells, pericytes or endothelia in the central tumor regions. It is generally believed that more Glut1 is found in erythrocytes than any other cell, but quantitative electron microscopic immunogold analyses of Glut1-immunoreactive sites per micron of capillary membrane showed the Glut1 density in tumor endothelial membranes glucose transporter was 2-3-fold higher than in human red cells. In the same tissue samples, qualitative immunogold electron microscopy of human serum albumin indicated that this protein (MW 65,000) moved freely from the vascular space into pericapillary regions, confirming the leaky barrier characteristics of the hemangioblastoma. These studies show that Glut1 expression may be high in endothelia that are highly permeable and devoid of astroglial contacts. Thus, human cerebral hemangioblastomas may provide a novel system for studying the induction of Glut1 in the blood-brain barrier.

Hyperacute thermal lesions: MR imaging evaluation of development in the brain

PURPOSE

To determine the natural time course of development of hyperacute thermal lesions in the brain.

MATERIALS AND METHODS

Ten interstitial lesions were created in five rabbit brains with a radio-frequency probe; an electrode-tip temperature of 80 degrees C was maintained for 60 seconds. Continuous fast spin-echo magnetic resonance (MR) imaging was used to follow lesion development for a minimum of 30 minutes. Temporal variations in lesion size and signal intensity were examined. Findings in final images were correlated with histologic findings.

RESULTS

Images demonstrated a focal hyperintense zone, which developed into an expanding ring of edema surrounding a necrotic center in about 10 minutes. Quantitative analysis revealed a 23% +/- 6 (standard deviation) increase in average signal intensity of the edema layer and a 152% +/- 41 increase in overall lesion size.

CONCLUSION

Full development of a thermal lesion is delayed for a period of minutes. Clinical implications of this effect should be considered when MR imaging-guided thermal ablation is performed.

Adenovirus-mediated p53 gene delivery inhibits 9L glioma growth in rats

Adenoviral vectors have recently been shown to effectively deliver genes into a variety of tissues. Since these vectors have some advantages over the more extensively investigated retroviruses, we studied the effect of two replication-defective adenovectors bearing human wild type tumor suppressor gene p53 (Adp53) and Escherichia coli beta-galactosidase gene (AdLacZ) on 9L glioma cells. Successful in vitro gene transfer was shown by DNA polymerase chain reaction (PCR), and expression was confirmed by reverse transcriptase RNA PCR and Western blot analyses. Transduction of 9L cells with the Adp53 inhibited cell growth and induced phenotypic changes consistent with cell death at low titers, while AdLacZ caused cytopathic changes only at high titers. Stereotactic injection of AdLacZ (10(7) plaque forming units) into tumor bed stained 25 to 30% of tumor cells at the site of vector delivery. Injection of Adp53 (10(7) plaque forming units), but not AdLacZ (controls), into established 4-day old 9L glioma brain tumors decreased tumor volume by 40% after 14 days. As a step toward gene therapy of brain tumors using replication-defective adenoviruses, these data support the use of tumor suppressor gene transfer for in vivo treatment of whole animal brain tumor models.

Preliminary experience with MR-guided thermal ablation of brain tumors

PURPOSE

To evaluate the feasibility of a technique of MR-guided stereotactic radio frequency ablation, which was developed as a minimally invasive treatment for brain tumors, and to determine MR characteristics and sequential evolution of radio frequency lesions created to ablate brain tumors.

METHODS

Fourteen lesions in 12 patients with primary and metastatic brain tumors were treated with this technique and followed for up to 10 months. The stereotactic coordinates of the tumor and the angle of the radio frequency probe were calculated on MR imaging. The radio frequency lesion was generated in the awake patient by increasing the temperature to 80 degrees C within the tumor for 1 minute. This was repeated until the entire tumor volume was destroyed. MR imaging was performed before, during, and immediately after the radio frequency procedure, and sequential MR was obtained during clinical follow-up.

RESULTS

MR imaging clearly showed well-defined radio frequency lesions and provided feedback for treatment planning. The radio frequency lesion boundary was well identified as a dark signal rim on T2-weighted images and showed ring enhancement on contrast-enhanced T1-weighted images. The sequential MR imaging showed the radio frequency lesions decreased in volume in all cases, suggesting focal control.

CONCLUSION

Stereotactic MR-guided radio frequency brain tumor ablation is a feasible and promising technique that can be an attractive brain tumor treatment alternative. MR provided not only accurate tumor location but also visualization of feedback of thermal tissue changes that reflected therapeutic effect.

Intracarotid infusion of RMP-7, a bradykinin analog: a method for selective drug delivery to brain tumors

The bradykinin analog, RMP-7, was investigated for its ability to selectively increase uptake of molecular tracers in RG2 glial tumors. When infused in low doses (0.1 microgram/kg/min) through the intracarotid artery ipsilateral to RG2 gliomas in rats, RMP-7 significantly increased the permeability of tumor capillaries to methotrexate and to four other tracers of varying molecular weights, compared to intracarotid infusion of vehicle alone. Tracers used to examine permeability included radiolabeled alpha-aminoisobutyric acid (M(r) 103 D), sucrose (M(r) 342 D), methotrexate (M(r) 454.5 D), inulin (M(r) 5000 D), and dextran (M(r) 70,000 D). Permeability was expressed as the unidirectional transfer constant, Ki (microliters/gm/min). The permeability (Ki) of tumors in the RMP-7 group compared to the vehicle control group was as follows: alpha-aminoisobutyric acid, 35.3 +/- 9.11 versus 12.7 +/- 4.56 (p < 0.001); sucrose, 16.5 +/- 3.83 versus 9.28 +/- 3.12 (p < 0.05); methotrexate, 26.3 +/- 10.3 versus 8.98 +/- 6.78 (p < 0.005); inulin, 13.5 +/- 3.23 versus 6.55 +/- 4.32 (p < 0.005); dextran, 15.2 +/- 3.42 versus 1.47 +/- 1.24 (p < 0.001). The permeability of RG2 gliomas to high-molecular-weight dextran (70,000 D) was 10.3-fold higher in the RMP-7 group than in the vehicle control group. Intracarotid infusion of RMP-7 did not significantly increase the blood volume in tumor or brain tissue. The permeability of normal brain capillaries was unaffected by intracarotid infusion of 0.1 microgram/kg/min RMP-7 relative to that achieved in tumor. These data support the idea that intracarotid infusion of RMP-7 will be a useful technique for selective delivery of antitumor compounds to brain tumors.

Gene expression of GLUT3 and GLUT1 glucose transporters in human brain tumors

GLUT3 glucose transporter gene expression is confined to neurons, while GLUT1 gene expression is limited to endothelial cells in normal brain. Thus far, neither of the GLUT genes has been shown to be consistently expressed in glial cells in adult brain in vivo under normal conditions. However, GLUT gene expression may be aberrant in human brain glial tumors. The present investigation shows that the GLUT1 and GLUT3 transcripts are differentially expressed in a series of 20 human brain tumors. The GLUT1/actin mRNA ratio increased in parallel to the astrocytoma grade, compared to a control human brain cortex, although no change in this ratio was seen in 5 meningiomas. Immunoreactive GLUT1 protein was not detectable in human brain tumors, including high-grade gliomas. Both 4.2 or 2.7 kb GLUT3/actin mRNA ratios showed a linear correlation with the glioma grade (P < 0.025), and the GLUT3-immunoreactive protein was also expressed in high grade gliomas. These studies provide evidence for induction of GLUT1 and GLUT3 gene expression in malignant glial cells, and the mRNA levels correlate with the biologic aggressiveness of the tumor. The detection of immunoreactive GLUT3, but not GLUT1, in the high grade gliomas suggest the GLUT3 isoform may be the predominant glucose transporter in highly malignant glial cells of human brain.

Stereotactic delivery of a recombinant adenovirus into a C6 glioma cell line in a rat brain tumor model

The dismal results of conventional therapy for primary malignant brain tumors has justified exploring gene therapy approaches for this disease. Transduction of animal brain tumor models in vivo has been reported previously with retroviruses and herpes viruses. Because adenoviruses have the advantage of transducing quiescent and actively dividing tumor cells, they may prove to be more effective in such therapy. We used a replication-deficient recombinant adenovirus bearing the Escherichia coli beta-galactosidase gene in a rat C6 glioma tumor model. Transduced cells were detected by X-5-bromo-4-chloro-3-indolyl beta-D-galactoside staining to reveal beta-galactosidase activity. Initial experiments in vitro showed 50% and 90% transduction at vector titers of approximately 10(7) and 10(8) plaque-forming units/ml, respectively. Although no cytopathic effects were seen at 10(7) plaque-forming units/ml, more than 50% reduction in tumor cell growth was noted at 10(8) plaque-forming units/ml both in vitro and in vivo. Stereotactic delivery of the recombinant adenovirus into the frontal lobe of normal rat brains resulted in intense staining of all cell types, that is, neurons, astrocytes, and ependymal cells. Stereotactic injection into C6 glioma brain tumors in rats stained 25 to 30% of the tumor cells. We conclude that adenovirus vectors can be used to transfer genes to central nervous system tumors in vivo. Using stereotactic delivery, adenovirus vectors can transfer genes into the central nervous system intended for tumor therapy.

Enzymatic barrier protects brain capillaries from leukotriene C4

Leukotriene C4 (LTC4) increases vascular permeability in systemic, brain tumor, and ischemic brain capillaries, but not in normal brain capillaries. This study examines whether the abundance of gamma-glutamyl transpeptidase (gamma-GTP) in normal brain capillaries might act as an enzymatic barrier to vasoactive leukotrienes in the brain. Blood-brain barrier (BBB) permeability was determined by quantitative autoradiography using 14C-aminoisobutyric acid. Ischemia was produced by occluding the middle cerebral artery. Seventy-two hours after occlusion, gamma-GTP activity in ischemic brain disappeared, and LTC4 (4-micrograms total dose), which was infused into the carotid artery ipsilateral to the occlusion, selectively increased permeability, Ki, approximately twofold within core ischemic tissue and adjacent tissue, compared to vehicle alone in seven brains (15.53 +/- 6.03 vs. 7.29 +/- 3.36, p < 0.05, and 8.76 +/- 4.02 vs. 4.32 +/- 2.65, p < 0.05, respectively). No effect on BBB was seen in nonischemic brain tissue. Twenty-four hours postocclusion, gamma-GTP activity was still present, and LTC4 infusion did not increase permeability within ischemic tissue. However, inhibition of gamma-GTP with acivicin allowed LTC4 to increase permeability even 24 hours after occlusion in ischemic core and adjacent tissue compared to vehicle alone in seven brains (17.21 +/- 16.32 vs. 8.23 +/- 6.58, p < 0.05, and 11.78 +/- 7.96 vs. 4.56 +/- 1.93, p < 0.01, respectively). Acivicin almost completely blocked both the histochemical activity of gamma-GTP in brain capillaries and the metabolism of LTC4 in isolated bovine capillaries. These findings suggest that gamma-GTP may help normal brain capillaries resist the vasoactive effects of LTC4. In contrast, gamma-GTP is lost in injured brain capillaries, which allows LTC4 (in combination with other factors) to increase vascular permeability in ischemic brain and brain tumors.

Intracarotid infusion of bradykinin selectively increases blood-tumor permeability in 9L and C6 brain tumors

This study investigated the effects of bradykinin on blood-tumor barrier (BTB) permeability in transplanted 9L gliosarcomas (9L) and C6 gliomas (C6) in rats. Permeability, expressed as the unidirectional transfer constant, Ki (microliter/g/min), was measured by quantitative autoradiography. Tracers used to examined permeability included radiolabeled alpha-aminoisobutyric acid ([14C]AIB), sucrose ([14C]sucrose) and dextran ([14C]dextran). Intracarotid infusion of bradykinin (10 mg/kg/min) significantly increased the BTB permeability in both 9L and C6 tumors to [14C]AIB and [14C]sucrose, but did not increase permeability to [14C]dextran. Blood-brain barrier (BBB) permeability in normal (non-tumor) brain was not significantly increased to any of the tracers by intracarotid bradykinin infusion. Ki values for [14C]AIB, [14C]sucrose and [14C]dextran of 9L tumors in the bradykinin group versus control group were 41.6 +/- 12.6 vs. 24.8 +/- 6.30 (P < 0.02), 17.5 +/- 9.34 vs. 9.05 +/- 4.36 (P < 0.05), and 3.90 +/- 2.59 vs. 2.42 +/- 1.76, respectively (mean +/- S.D.). Ki values to [14C]AIB, [14C]sucrose and [14C]dextran of C6 tumors in the bradykinin group versus control group were 41.4 +/- 19.0 vs. 19.5 +/- 11.4 (P < 0.01), 18.0 +/- 8.88 vs. 7.06 +/- 3.05 (P < 0.01), and 4.07 +/- 1.45 vs. 2.27 +/- 1.26, respectively (mean +/- S.D.). Intracarotid infusion of bradykinin did not significantly increase the blood volume in tumor or brain tissue despite its known vasodilative effect. Intracarotid infusion of bradykinin may be a useful technique for selective delivery of compounds to brain tumors.

Bradykinin selectively opens blood-tumor barrier in experimental brain tumors

Bradykinin, infused in low doses (10 micrograms/kg/min) through the carotid artery ipsilateral to RG2 glioma in rats, significantly increased the permeability in tumor capillaries to six different tracers of varying molecular weights compared with intracarotid infusion of saline alone. Permeability in normal brain capillaries was not significantly increased by intracarotid bradykinin infusion. Tracers used to examined permeability included radiolabeled alpha-aminoisobutyric acid (AIB; MW 103), sucrose (MW 342.3), inulin (MW 5000), and dextran (MW 70,000), horseradish peroxidase (HRP) and Evans blue (EB). Permeability was expressed as the unidirectional transfer constant K(i) (microliter/g/min). The permeabilities (K(i)) of tumors in the bradykinin group versus the control saline group for AIB, sucrose, inulin, and dextran were 25.91 +/- 6.78 vs. 13.95 +/- 4.29 (p < 0.01), 17.90 +/- 2.65 vs. 10.75 +/- 4.55 (p < 0.01), 23.92 +/- 6.99 vs. 6.20 +/- 4.37 (p < 0.01), and 17.84 +/- 1.00 vs. 1.47 +/- 1.24 (p < 0.001), respectively (mean +/- SD). Permeability of RG2 gliomas to high molecular weight dextran (70,000) was 12-fold higher in the bradykinin group than in the saline infusion group. Intracarotid infusion of bradykinin did not significantly increase the blood volume in tumor or brain tissue despite its known vasodilative effect. The permeability of normal brain capillaries was unaffected by intracarotid bradykinin infusion. The increased permeability was reversed 20 min after stopping the intracarotid infusion. Electron microscopic and gross qualitative analysis was performed using HRP and EB. Intracarotid bradykinin infusion increased HRP and EB within tumor tissue but not normal tissue.(ABSTRACT TRUNCATED AT 250 WORDS)

Trans retinoic acid inhibits in vivo tumour growth of C6 glioma in rats: effect negatively influenced by nerve growth factor

Retinoic acid (RA) and nerve growth factor (NGF) are both differentiation factors for central nervous system tumours. Mouse-derived NGF inhibits proliferation of C6 glioma cells in vivo in the absence of serum. Retinoic acid inhibits in vivo growth of C6 gliomas in the subcutaneous tissue of rats. This study evaluated the response of C6 cells implanted in the rat cortex to NGF, RA, or a combination of the two in 89 rats. Tumour size, cellular density and morphology were analysed using light microscopy. Treatment with RA alone resulted in tumour volumes that were 38% of control and 48% of NGF-treated groups. There was no significant difference in the tumour volumes or in cell morphology in C6 cells treated with NGF alone compared to controls. Tumours treated with a combination of RA and NGF were larger however, than tumours treated with RA alone. This suggests that despite the growth inhibitory effects of NGF in vitro, NGF acts to prevent the growth inhibitory effect of RA in vivo.

Thallium-201 SPECT and positron emission tomography equal predictors of glioma grade and recurrence

Quantitative indexes based on regions of interest ratios from preoperative thallium-201 SPECT and F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) scans were compared for their ability to predict the histological grade of cerebral gliomas. Of the five different ratios used to determine the index for each PET scan, the ratio of tumour versus ipsilateral hemisphere in the same plane as the tumour showed the most significant correlation with glioma grade (p < 0.001). A large study of 62 thallium-201 SPECT scans showed equally good correlation with glioma grade (p < 0.0005). The use of thallium-201 SPECT for detection of tumour recurrence had 100% sensitivity in this series, while PET scans had 90.9% sensitivity. Patients with a rate of change in serial thallium indexes of 0.475 +/- 0.278 per month survived 5.93 +/- 2.25 months from onset of index increase, while those with little change in the index (0.069 +/- 0.063 per month) survived 17.93 +/- 5.25 months from onset of index increase. Results indicate thallium-201 SPECT correlates well with tumour grade and is equally efficient as PET in screening for glioma recurrence.

Intracarotid histamine infusion increases blood tumour permeability in RG2 glioma

Histamine will alter blood flow and permeability in systemic and cerebral vessels. We reported that intracarotid infusion of histamine selectively increased the blood flow in experimental brain tumours and caused extravassation of Evans blue within tumours. In this study, the effects of histamine on tumour and brain capillary permeability were quantified using autoradiography. RG2 glioma cells were implanted in female Wistar rats. Seven days after implantation, either low doses of histamine (1 or 10 micrograms kg-1 min-1) or saline as a control was infused through the carotid artery of rats. Regional permeability was measured by autoradiography using [14C] aminoisobutyric acid, and the unidirectional transfer constant, Ki (microliters g-1 min-1), was calculated. Intracarotid infusion of 10 micrograms kg-1 min-1 histamine resulted in significant increase in brain tumour permeability, compared to controls. The permeability, Ki, for the 10 micrograms kg-1 min-1 histamine group, the 1 micrograms kg-1 min-1 histamine group, and the control group was 18.8 +/- 4.6 (p < 0.05), 14.9 +/- 5.2, 13.9 +/- 3.7 microliters g-1 min-1, respectively. There was no significant change in blood brain permeability in other brain regions. The effect of increased permeability by 10 micrograms kg-1 min-1 histamine was suppressed by the H2-blocker, cimetidine. This suggests that the effect of histamine on tumour capillaries is mediated by H2-receptors. Intracarotid histamine infusion selectively increases permeability in brain tumours.

Peripheral benzodiazepine stimulates secretion of growth hormone and mitochondrial proliferation in pituitary tumour GH3 cells

High affinity receptors for peripheral benzodiazepines (PBD) are present within the pituitary gland. Whether these receptors influence cell growth, mitochondrial morphology and secretion of growth hormone was examined in pituitary tumour GH3 cells. Cells were incubated in medium with either 10% calf serum, or serum free medium, or serum free medium plus 10 nM of selective PBD ligands (PK11195 or Ro5-4864) or a central benzodiazepine ligand (clonazepam). Peripheral or central ligands had no effect on cell growth. Quantitative electron micrography, however, revealed that cells treated with PBDs had a 2.5-fold increase in the total mitochondrial area within cells, a 1.6-fold increase in the number of mitochondria, and a 2.0 to 2.7-fold increase on the number of dividing mitochondria compared to serum free controls. The activity of DNA polymerase gamma, which replicates mitochondrial DNA, was increased approximately 3 times the initial value after 6 h exposure to PBDs. The number of nucleolar organizer regions (NORs), which relate to the ultimate synthesis of cellular protein, showed a 1.4-fold increase in PBDs treated cells. Secretion of growth hormone was stimulated (240% of serum free control) by exposure to PBDs for 12 h. The central benzodiazepine ligand, clonazepam, had little effect on mitochondrial morphology or secretion of growth hormone. These findings suggest that PBD receptors may be involved in mitochondrial proliferation and may affect the secretion of pituitary hormones.

Differential tissue expression of immunoreactive dehydropeptidase I, a peptidyl leukotriene metabolizing enzyme

We previously reported that intracarotid infusion of leukotriene C4 (LTC4) causes a selective increase in vascular permeability within brain tumor capillaries in experimental rat brain tumor. Normal brain capillaries are rich in gamma-glutamyl transpeptidase (gamma-GTP), an enzyme which converts LTC4 to leukotriene D4 (LTD4), and acts as an 'enzymatic barrier' to the vasoactive effects of LTC4. Metabolism of LTD4 in brain capillaries is, however, not known. In this study, rat renal dipeptidase (dehydropeptidase-I, microsomal dipeptidase; EC 3.4.13.11), which converts LTD4 to leukotriene E4 (LTE4) in kidney, was purified from rat kidney and the distribution of immunoreactive dipeptidase in multiple rat organs was determined. Immunocytochemical multi-organ analysis in the rat, which included brain, lung, heart, liver, spleen, small intestine, and testis, was performed. The antigen corresponding to renal dipeptidase was recognized in lung, liver, and testis. There was no antigen in the brain, heart, spleen, and small intestine. In order to confirm the absence of dipeptidase activity in brain capillaries, the metabolism of LTD4 by isolated brain capillaries were examined by reversed phase high performance liquid chromatography. When LTD4 was incubated with the isolated rat brain capillary, no measurable conversion of [3H] LTD4 to LTE4 and leukotriene F4 (LTF4) by brain capillaries was observed with 30 min of incubation. These findings suggest that although gamma-GTP acts as an enzymatic barrier and inactivates LTC4, brain capillaries do not have metabolic activity against LTD4.

The petrosal approach with hearing preservation

Twenty-six patients with petroclival lesions were operated on via a petrosal approach designed to preserve hearing. The surgical pathology included 14 meningiomas, three chordomas, three epidermoid cysts, four vertebrobasilar aneurysms, and two pontine cavernous malformations. The approach allowed complete resection of 14 of 20 tumors and definitive treatment of all six vascular lesions. Complications included cerebrospinal fluid leakage in three patients, high-frequency sensorineural hearing loss in three, meningitis in one, and cranial nerve palsies (which were usually transient). This approach allows a wide exposure of the petroclival region with decreased operating distance. Cerebellar and temporal lobe retraction are minimized, dural sinus patency is maintained, and the inner ear structures are not sacrificed. The approach is suitable for neoplastic or vascular lesions involving the petroclival region, the ventral pons, or the basilar artery trunk. The surgical technique, indications, and neuro-otological considerations are discussed.

Clinical trial of Serratia marcescens extract and radiation therapy in patients with malignant astrocytoma

PURPOSE

A clinical trial was undertaken to determine the safety and efficacy of combining a biologic response modifier derived from the bacterium Serratia marcescens (ImuVert) and radiation therapy (RT) in patients with newly diagnosed anaplastic astrocytoma (AA) or glioblastoma multiforme (GBM).

PATIENTS AND METHODS

Fifteen patients who had undergone either a gross total resection, a partial resection, or a biopsy were treated concurrently with ImuVert and RT. Safety and tolerance were examined by assessment of symptomatic reactions recorded at each ImuVert treatment. Efficacy of treatment was examined in terms of time to progression of tumor and survival.

RESULTS

All patients experienced local reactions at the injection sites that consisted of erythema and induration. The majority of patients experienced flu-like symptoms. Hypotension was responsible for the most significant morbidity (which required fluid resuscitation and extended observation) and dose deescalation. No patients were removed from the study because of toxicity. There were no on-study deaths related to ImuVert treatment. Median time to progression was 33.4 weeks, and median survival was 78 weeks.

CONCLUSION

These results compare favorably with those of recent studies in patients with malignant astrocytomas who received multimodality therapy.

Increased opening of blood-tumour barrier by leukotriene C4 is dependent on size of molecules

Intracarotid injections of leukotriene C4 (LTC4) will increase blood-tumour barrier (BTB) permeability for 14C aminoisobutyric acid (AIB) in RG-2 tumours in rats. In this study, rats with RG-2 tumours were given 15 minute intracarotid infusions of LTC4 (5 micrograms total dose). Blood-tumour and BBB permeability were determined by quantitative autoradiography using different sized 14C tracers; AIB (103.1 daltons, radius = 2.8 A), Sucrose (342 daltons, radius = 5 A), Inulin (5000 daltons, radius = 15 A) and Dextran (70,000 daltons, radius = 60 A), respectively (n = 7 in each group). Permeability (Ki) within tumours for AIB, Sucrose and Inulin increased significantly after LTC4 infusion (92.7 +/- 15.0 vs 39.4 +/- 6.5, 47.8 +/- 5.2 vs 29.1 +/- 8.8, and 46.4 +/- 8.4 vs 26.4 +/- 4.8 microliters/gm/min, respectively, p < 0.01) (mean +/- SD). No significant change in Ki for 14C Dextran was observed. Increased permeability within tumours after LTC4 infusion is dependent on molecular size. This suggests that LTC4 opens tight junctions rather than increases vesicular transport through endothelial cells.

Differential expression of arachidonate 5-lipoxygenase transcripts in human brain tumors: evidence for the expression of a multitranscript family

In addition to the important role of leukotrienes as mediators in allergy and inflammation, these compounds are also linked to pathophysiological events in the brain including cerebral ischemia, brain edema, and increased permeability of the blood-brain barrier in brain tumors. Although brain tumors have been shown to secrete leukotrienes, no studies to date have provided evidence for the tumor expression of genes encoding enzymes involved in leukotriene production. Therefore, the present study determined the abundance of the mRNA for arachidonate 5-lipoxygenase (5-LO; arachidonate:oxygen 5-oxidoreductase, EC 1.13.11.34), which is the rate-limiting enzyme in leukotriene synthesis, in a series of human brain tumors. Macrophage/monocyte infiltration of the tumor was estimated by measuring the abundance of the transcript for the 91-kDa glycoprotein phagocyte-specific oxidase (gp91-phox), which is the phagocyte-specific cytochrome b heavy chain. The present study shows that (i) the 5-LO transcript is expressed in normal bovine brain and in human brain tumors; (ii) the 5-LO gene in human brain tumors and in the dimethyl sulfoxide-induced promyelocytic human leukemic HL-60 cells is expressed as a multitranscript family (2.7, 3.1, 4.8, 6.4, 8.6 kilobases); and (iii) the abundance of 5-LO transcripts, the expression of the larger transcripts, and the 5-LO/gp91-phox ratio correlate with the tumor malignancy. Overall, the present study supports the hypothesis that the 5-LO gene product may play a role in human tumor-induced brain edemas and provides evidence for tumor-associated expression of high molecular weight 5-LO transcripts in human brain tumors.

The effect of 5-lipoxygenase inhibition on blood-brain barrier permeability in experimental brain tumors

To determine if leukotrienes are important mediators of vascular permeability in brain tumors, the effect of 5-lipoxygenase inhibitors on blood-tumor barrier permeability in rats harboring HK Walker 256 brain tumors was examined using quantitative autoradiography with alpha-14C-aminoisobutyric acid. The 5-lipoxygenase enzyme converts arachidonic acid to leukotrienes. Three 5-lipoxygenase inhibitors were utilized: BW755C, nordihydroguaiaretic acid, and AA-861. All three 5-lipoxygenase inhibitors significantly decreased vascular permeability both within the tumors and in brain adjacent to tumor. This suggests that capillary permeability in and adjacent to tumors is influenced by endogenous leukotrienes and that leukotrienes play an important role in brain tumor edema.

Selective blood-tumor barrier disruption by leukotrienes

The authors have previously reported that intracarotid infusion of 5 micrograms leukotriene C4 (LTC4) selectively increases blood-tumor barrier permeability in rat RG-2 tumors. In this study, rats harboring RG-2 tumors were given 15-minute intracarotid infusions of LTC4 at concentrations ranging from 0.5 microgram to 50.0 micrograms (seven rats in each dose group). Blood-tumor and blood-brain barrier permeability were determined by quantitative autoradiography using 14C aminoisobutyric acid. The transfer constant for permeability (Ki) within the tumors was increased twofold by LTC4 doses of 2.5, 5.0, and 50.0 micrograms compared to vehicle alone (90.00 +/- 21.14, 92.68 +/- 15.04, and 80.17 +/- 16.15 vs. 39.37 +/- 6.45 microliters/gm/min, respectively; mean +/- standard deviation; p less than 0.01). No significant change in Ki within the tumors was observed at the 0.5-microgram LTC4 dose. Blood-brain barrier permeability was selectively increased within the tumors. At no dose in this study did leukotrienes increase permeability within normal brain. To determine the duration of increased opening of the blood-tumor barrier by LTC4 administration, Ki was measured at 15, 30, and 60 minutes after termination of a 15-minute LTC4 infusion (seven rats at each time point). The mean Ki value was still high at 15 minutes (92.68 +/- 15.04 microliters/gm/min), but declined at 30 minutes (56.58 +/- 12.50 microliters/gm/min) and 60 minutes (55.40 +/- 8.10 microliters/gm/min) after the end of LTC4 infusion. Sulfidopeptide leukotrienes LTC4, LTD4, LTE4 and LTF4 were infused to compare their potency in opening the blood-tumor barrier. The mean leukotriene E4 was the most potent, increasing the permeability value 3 1/2-fold compared with vehicle alone (139.86 +/- 23.95 vs. 39.37 +/- 6.45 microliters/gm/min).

Inflammatory leukocytes associated with increased immunosuppression by glioblastoma

In order to determine the in vivo immune response in glioblastoma, monoclonal and polyclonal antibodies specific for inflammatory leukocytes and immunoregulatory products were utilized to stain tissue from four surgical specimens. The more activated the inflammatory cells, the more activated the tumors appeared to be. In the tumor with the largest infiltration (Case 3), inflammatory cells were stained for interferon-gamma, interleukin-2, interleukin-1 beta, lymphotoxin, tumor necrosis factor-alpha, and transforming growth factor-beta. The tumor cells also expressed interleukin-1 beta, interleukin-6, transforming growth factor-beta, tumor necrosis factor-alpha, and prostaglandin E. In contrast, in the tumor with the least inflammatory response (Case 1), the tumor cells did not express any cytokines. Expression of cytokines by glioma cells was modest in the two cases with modest inflammatory responses. Cellular inflammation, primarily consisting of T cells and macrophages with few or no B cells or natural killer cells, was two- to 15-fold greater outside the tumor than within. In contrast to leukocytes outside the tumor, which were activated and expressing class II major histocompatibility antigens, leukocytes within the tumor parenchyma or at the tumor's edge were negative for these antigens. In the four specimens studied here, the tumor cells themselves were also negative for class II major histocompatibility antigens. These findings, although preliminary, suggest that inflammatory cells within gliomas are inactivated and that glioma cells may increase the expression of immunosuppressive cytokines in response to an increased lymphocyte infiltrate. This observation, if corroborated by more extensive studies, may help to explain the failure of immune treatments in glioblastoma multiforme.

Inflammatory cell infiltrates vary in experimental primary and metastatic brain tumors

We have studied the cellular immune response that accompanies primary and metastatic brain cancers induced experimentally in rats by inoculation of RG-2 glioma and Walker 256 (W256) carcinoma cells, respectively. The inflammatory cell infiltrates were characterized with lectin histochemistry to visualize microglial cells and macrophages and with immunohistochemistry, using a panel of monoclonal antibodies, to detect major histocompatibility complex (MHC), lymphocytic, and macrophage antigens. The metastatic tumor was composed of a loose stroma with multiple, often large, necrotic areas, whereas the RG-2 glioma was composed of a dense collection of tumor cells showing only rare necrotic foci. Both tumor types were heavily infiltrated with microglia and/or macrophages, and these were positive for MHC Class II (Ia) antigens. Expression of MHC Class I antigens was absent from RG-2 glioma cells, but it was present in W256 metastatic carcinoma cells. The metastatic tumor was also characterized by a much heavier infiltrate of lymphocytes, as shown by the presence of cells positive for CD4, CD8, and leukocyte common antigens. These lymphocytic markers were absent from reactive microglia in the W256 carcinoma, whereas they were present in the RG-2 glioma. Polymorphonuclear leukocytes were seen only in the metastatic tumor. Our study delineates differences between the inflammatory cell infiltrates found in metastatic brain tumors and those found in primary brain tumors. The differences in cell composition and immunophenotype may indicate a more effective antitumor response in the metastatic tumor that could account for the observed tissue destruction.

Immunophenotypic analysis of infiltrating leukocytes and microglia in an experimental rat glioma

The appearance and cellular distribution of major histocompatibility complex (MHC), as well as lymphocytic and macrophage antigens has been studied in a fully developed experimental rat forebrain glioma. Activated microglial cells and microglia-derived macrophages expressing CR3 complement receptor molecules and MHC class II (Ia) antigen were found throughout the tumor, and with increased density along the tumor's periphery. MHC class I antigen expression was entirely absent from tumor cells, and found only occasionally on microglia. The expression of leukocyte common antigen, and CD4 and CD8 antigens was conspicuous throughout the tumor, and associated with lymphocytes, perivascular cells, and microglia. Cells expressing the ED2 macrophage epitope were almost exclusively of the perivascular type and revealed a distribution dissimilar to that of cells positive for Ia antigen. The ED2 epitope was found sporadically on ramified microglial cells. The results show that despite heavy infiltration with blood mononuclear and CNS microglial cells, the tumor showed no evidence of destruction caused by inflammatory cells. Possible mechanisms of tumor immunosuppressive activity preventing the full immunological activation of microglia and blood mononuclear cells are discussed.

Blood-brain barrier and new approaches to brain drug delivery

Morbidity caused by brain dysfunction affects more than 50 million persons in the United States. Although new neuropharmaceuticals have the potential for treating specific brain diseases, they may not effectively enter brain from blood. Safe strategies are needed for drug delivery through the brain capillary wall, which makes up the blood-brain barrier in vivo. Two of these strategies are reviewed, as are related new developments in the molecular and cell biology of the brain capillary endothelium. The production of chimeric peptides represents a physiologic-based strategy for drug delivery. It entails the covalent coupling of the neuropharmaceutical to a brain transport vector, allowing transportation through the blood-brain barrier. Another strategy is biochemical opening of the blood-brain barrier: intracarotid leukotriene infusion is a method for selectively increasing blood-brain barrier permeability in brain tumors without affecting barrier permeability in normal brain tissue.

Topographic mapping of somatosensory evoked potentials helps identify motor cortex more quickly in the operating room

Median nerve somatosensory evoked potentials were recorded from exposed cerebral cortex during craniotomies. This technique is valuable when knowledge of the motor cortex location can influence surgical decisions about resection limits or biopsy sites. Two different recording techniques were compared: strips of electrodes and arrays of electrodes. The arrays recorded electrical potentials suitable for topographic mapping. We found that motor cortex could be identified more quickly when using the topographic mapping of SEPs from arrays. We conclude that topographic mapping of SEP from sensorimotor regions during craniotomies works well in general and can be done more quickly than the traditional electrode strip technique.

Response of microglial cells to experimental rat glioma

The response of indigenous CNS microglia to an experimentally induced glioma has been studied in rat brain using lectin histochemistry with the Griffonia simplicifolia B4-isolectin. The study was undertaken 2 weeks after tumor cell injection when tumor size was near maximal. Reactive microglial cells formed a dense band that surrounded most of the well-circumscribed tumor mass, and extended along the corpus callosum into the contralateral cerebral hemisphere. From the periphery inward, reactive microglia extended into the tumor tissue, where large numbers of them were found to be present as microglia-derived macrophages. The lectin stain, which also labels endothelial cells, revealed a highly vascularized tumor with ongoing neovascularization apparent as vascular sprouts. Moderate numbers of lectin-stained blood monocytes were localized primarily inside the vessel lumina. Our results show that microglial cells react to brain tumors; however, it remains to be determined whether the microglial response represents an active antitumor defense mechanism that could be manipulated during immunotherapeutic approaches.

Peripheral benzodiazepine induces morphological changes and proliferation of mitochondria in glioma cells

Peripheral benzodiazepine (PBD) receptors are localized on the mitochondrial membrane and are highly expressed in brain tumors compared to normal brain. To elucidate the biological role of the PBD receptor on mitochondria, we examined the effect of PBDs on mitochondrial morphology in C6 and T98G glioma cells using rhodamine 123 and quantitative electron microscopy. In cells incubated in serum-free medium alone, mitochondria were distributed in a filamentous pattern throughout the cytoplasm. By contrast, the mitochondria aggregated in the perinuclear region in PK11195 or Ro5-4864 (10 nM) treated cells. Quantitative electron micrography revealed a 250% increased in the number of mitochondria with elongated cristae and a fivefold increase in dividing mitochondria in PK11195-treated cells compared with cells incubated in serum-free medium alone. PBD treatment also resulted in vacuolation within the matrix and mitochondrial swelling. These data suggest that PBDs influence mitochondrial morphology and induce mitochondrial replication in cultured glioma cells.

Intracarotid infusion of leukotriene C4 selectively increases blood-brain barrier permeability after focal ischemia in rats

Intracarotid infusions of leukotriene C4 (LTC4) were used to open selectively the blood-brain barrier (BBB) in ischemic tissue after middle cerebral artery (MCA) occlusion in rats. BBB permeability was determined by quantitative autoradiography using [14C]aminoisobutyric acid. Seventy-two hours after MCA occlusion, LTC4 (4 micrograms total dose) infused into the carotid artery ipsilateral to the MCA occlusion selectively increased the unidirectional transfer constant for permeability Ki approximately threefold within core ischemic tissue and tissue adjacent ot the ischemic core. No effect on BBB permeability was seen within nonischemic brain tissue or in ischemic tissue after only 24 h after MCA occlusion. gamma-Glutamyl transpeptidase (gamma-GTP) activity was decreased in capillaries in ischemic tissue at 48 and 72 h after infarction, compared to high gamma-GTP in normal brain capillaries and moderate gamma-GTP in capillaries in the ischemic tissue at 24 h after infarction. These findings suggest that normal brain capillaries resist the vasogenic effects of LTC4. In contrast, LTC4 increases permeability in capillaries of ischemic tissue, where gamma-GTP is decreased. gamma-Glutamyl transpeptidase, an enzyme that inactivates LTC4 to LTD4 and LTE4 to LTF4, may act as an "enzymatic barrier" in normal brain capillaries to leukotrienes.

Interstitial chemotherapy with drug polymer implants for the treatment of recurrent gliomas

Malignant gliomas have been difficult to treat with chemotherapy. The most effective agent, BCNU (carmustine), has considerable systemic toxicity and a short half-life in serum. To obviate these problems, a method has been developed for the local sustained release of chemotherapeutic agents by their incorporation into biodegradable polymers. Implantation of the drug-impregnated polymer at the tumor site allows prolonged local exposure with minimal systemic exposure. In this Phase I-II study, 21 patients with recurrent malignant glioma were treated with BCNU released interstitially by means of a polyanhydride biodegradable polymer implant. Up to eight polymer wafers were placed in the resection cavity intraoperatively, upon completion of tumor debulking. The polymer releases the therapeutic drug for approximately 3 weeks. Three increasing concentrations of BCNU were studied; the treatment was well tolerated at all three levels. There were no adverse reactions to the BCNU wafer treatment itself. The average survival period after reoperation was 65 weeks for the first dose group, 64 weeks for the second dose group, and 32 weeks for the highest dose group. The overall mean survival time was 48 weeks from reoperation and 94 weeks from the original operation. The overall median survival times were 46 weeks postimplant and 87 weeks from initial surgery. Eighteen (86%) of 21 patients lived more than 1 year from the time of their initial diagnosis and eight (38%) of 21 patients lived more than 1 year after intracranial implantation of the polymer. Frequent hematology, blood chemistry, and urinalysis tests did not reveal any systemic effect from this interstitial chemotherapy. Since the therapy is well tolerated and safe, a placebo-controlled clinical trial has been started. The trial will measure the effect of the second treatment dose on survival of patients with recurrent malignant glioma.

Brain tumors

Recent advances in experimental tumor biology are being applied to critical clinical problems of primary brain tumors. The expression of peripheral benzodiazepine receptors, which are sparse in normal brain, is increased as much as 20-fold in brain tumors. Experimental studies show promise in using labeled ligands to these receptors to identify the outer margins of malignant brain tumors. Whereas positron emission tomography has improved the dynamic understanding of tumors, the labeled selective tumor receptors with positron emitters will enhance the ability to specifically diagnose and greatly aid in the pretreatment planning for tumors. Modulation of these receptors will also affect tumor growth and metabolism. Novel methods to deliver antitumor agents to the brain and new approaches using biologic response modifiers also hold promise to further improve the management of brain tumors.

Tumor-associated neurological dysfunction prevented by lazaroids in rats

The efficacy of U-74006F and U-78517F in the treatment of blood-tumor barrier permeability and tumor-associated neurological dysfunction was evaluated in a brain-tumor model in rats. U-74006F is a 21-aminosteroid and U-78517F is a 2-methylamino chroman. Rats with stereotactically implanted Walker 256 tumors were treated with methylprednisolone, U-74006F, U-78517F, or vehicle (0.05 N HCl) on Days 6 through 10 following implantation. Neurological function and vascular permeability were assessed on Day 10. Methylprednisolone and U-74006F were equally effective at preventing neurological dysfunction compared to the control group (p less than 0.01); U-78517F was slightly less effective than U-74006F and methylprednisolone but was significantly better than vehicle in preventing neurological dysfunction. Delivery of methylprednisolone resulted in a significant decrease in tumor vascular permeability (p less than 0.006) while U-74006F and U-78517F had no effect on permeability. This suggests that U-74006F and U-78517F prevented tumor-associated neurological dysfunction by a mechanism other than decreasing permeability in tumor capillaries, and that U-74006F or U-78517F could prove useful in the treatment of brain tumors.

Imaging peripheral benzodiazepine receptors in brain tumors in rats: in vitro binding characteristics

Peripheral benzodiazepine binding constants for transplanted RG-2 gliomas and HD and LK Walker 256 tumors (metastatic breast carcinoma) were determined in Wistar rats using autoradiography. In addition, Kd and Bmax parameters for peripheral benzodiazepine receptors on RG-2 tumors were directly visualized using digital image analysis of autoradiograms. High specific binding of [3H]PK11195, a selective peripheral benzodiazepine ligand, had excellent topographical correlation to areas of histologically verified tumor. Scatchard analysis suggested a single class of peripheral binding sites with similar binding affinities in RG-2 and LK Walker 256 tumors and normal cortex. Bmax was 20-fold greater in glial tumors and 11.6- and 10.6-fold greater in LK and HK Walker 256 tumors, respectively, compared to normal cortex. The location of metastatic tumors, either intracerebrally or subcutaneously, did not effect their Kd or Bmax values. Kd and Bmax values for RG-2 tumors were similar whether determined densitometrically or by direct visualization with image analysis. Binding parameters within normal brain were difficult to visualize by image analysis due to the low level of specific binding. The ability to label specifically intracerebral tumor cells and to characterize the binding parameters shown in this study suggest that peripheral benzodiazepine receptor ligands could be utilized by PET to analyze directly a variety of tumors in humans.

Three-dimensional comparison of peripheral benzodiazepine binding and histological findings in rat brain tumor

Experiments were undertaken to determine the in vivo utility of the mixed benzodiazepine ligand [3H]flunitrazepam and the selective peripheral benzodiazepine ligand [3H]PK 11195 [1-(2-chlorophenyl)-N-methyl-N-(1-methylpropyl)-3-isoquinoline carboxamide] to outline the borders of rat C6 glial tumors in three dimensions. Intravenous injection of [3H]flunitrazepam resulted in a tumor/cortex ratio of radioactive densities between 2.7 and 1.5 within the first 60 minutes after injection. [3H]PK 11195 demonstrated a higher tumor/cortex ratio (5.3) than [3H]flunitrazepam. For three-dimensional studies, images were generated from thionin-stained histological sections and autoradiograms. The mixed type benzodiazepine ligand [3H]flunitrazepam was superior in showing some of the normal anatomical structures surrounding the tumor, whereas [3H]PK 11195, a specific peripheral ligand, demonstrated higher tumor/brain contrast and superior topographical correlation between histological and autoradiographic images. Implications of peripheral benzodiazepine receptor ligands for positron emission tomography are discussed.

Thallium-201 SPECT imaging of brain tumors: methods and results

Recent studies suggest that thallium-201 (201Tl) planar scans of brain tumors more accurately reflect viable tumor burden than CT, MRI, or radionuclide studies with other single-photon emitting compounds. We have previously reported the utility of 201Tl SPECT index in distinguishing low- from high-grade gliomas elsewhere. Here we describe the technical considerations of deriving a simple 201Tl index, based on uptake in the tumor normalized to homologous contralateral tissue, from SPECT images of brain tumors. We evaluated the importance of consistently correcting for tissue attenuation, as it may achieve better lesion discrimination on qualitative inspection, and the methodologic limitations imposed by partial volume effects at the limits of resolution.

Selective opening of the blood-tumor barrier by intracarotid infusion of leukotriene C4

Intracarotid infusions of leukotriene C4 (LTC4) were used to selectively open the blood-tumor barrier in rats with RG-2 gliomas. Blood-brain and blood-tumor permeability was determined by quantitative autoradiography using 14C aminoisobutyric acid. Leukotriene C4 (4 micrograms total dose) infused into the carotid artery ipsilateral to the tumor increased twofold the unidirectional transfer constant for permeability within the tumor while no effect on permeability was seen in normal brain. No gamma glutamyl transpeptidase (gamma-GTP) activity was seen in tumor capillaries in contrast to high gamma-GTP in normal brain capillaries. These findings suggest that normal brain capillaries may resist the vasogenic effects of LTC4, while LTC4 will increase permeability in tumor capillaries. This could relate to the ability of gamma-GTP to act as an enzymatic barrier and inactivate leukotrienes in normal brain capillaries. Intracarotid LTC4 infusion may be a useful tool to selectively open the blood-tumor barrier for delivery of antineoplastic compounds.

Stimulation of cell growth and DNA synthesis by peripheral benzodiazepine

The effects of peripheral benzodiazepine receptor ligands on cell proliferation were evaluated. PK11195 increased the growth rate of C6 glioma cells by 20-30% in the nanomolar range in serum free medium. [3H]thymidine incorporation into C6 glioma cells also were increased 22% and 25% after treatment by PK11195 and Ro5-4864, respectively. The effect of PK11195 as a mitogenic agent was estimated by mitogenic agent was estimated by [3H]thymidine incorporation using Swiss 3T3 cells. PK11195 increased DNA synthesis 170% over control at 10 nM. Higher concentrations of benzodiazepines showed inhibition of the DNA synthesis. Peripheral benzodiazepine binding sites underwent downregulation after exposure to serum free medium or to 10 nM PK11195. These findings suggest that peripheral benzodiazepines may be involved in the regulation of cell proliferation as a growth factor in lower concentration and as a antiproliferative agent in higher concentration.

Arachidonic acid metabolite production following focal cerebral ischemia: time course and effect of meclofenamate

Arachidonic acid metabolites have been implicated in the development of cerebral edema following ischemia. To define the time course of metabolite production, subtemporal craniectomies were performed on 60 male Sprague-Dawley rats (350-400 g). Thirty rats underwent middle cerebral artery occlusion while 30 rats underwent craniectomy alone. Five rats in each of two groups (middle cerebral artery occlusion and sham) were sacrificed at 15 minutes, 1 hour, 4 hours, 1 day, 3 days, and 6 days. The cerebral hemispheres were removed and divided in the midsagittal plane. Each hemisphere was immediately frozen in isopentane cooled in dry ice and stored at -70 degrees C. Tissue prostaglandins E2 and 6-keto F1 alpha, and leukotrienes (LT) B4 and C4 were measured by radioimmunoassay. Prostaglandin E2 and 6-keto prostaglandin F1 alpha were significantly elevated at 15 minutes in the middle cerebral artery occlusion hemispheres (p less than 0.05). Prostaglandins were not significantly elevated after 15 minutes. LT B4 and C4 were never significantly elevated. Meclofenamate, a nonsteroidal anti-inflammatory agent, was administered to 21 additional rats. Seven controls underwent middle cerebral artery occlusion alone, 7 were given intraperitoneal meclofenamate (20 mg/kg) 30 minutes prior to middle cerebral artery occlusion, and 7 underwent middle cerebral artery occlusion followed immediately by intraperitoneal meclofenamate (20 mg/kg). The animals were sacrificed at 15 minutes and similarly studied. There was a significant reduction of prostaglandin E2 and 6-keto prostaglandin F1 alpha following pretreatment with meclofenamate (p less than 0.01 and p less than 0.05). In pretreated rats, leukotrienes were not affected by meclofenamate. Similarly, prostaglandins and leukotrienes did not change when meclofenamate was administered after middle cerebral artery occlusion. We conclude that cyclo-oxygenase metabolite production begins within 15 minutes of middle cerebral artery occlusion. Treatment with meclofenamate prior to middle cerebral artery occlusion significantly reduced cyclooxygenase metabolite production, suggesting a protective effect of meclofenamate against ischemia-induced elevations of vasoactive prostaglandins implicated in the development of cerebral edema. Lipoxygenase metabolite production was not affected by middle cerebral artery occlusion or pharmacological intervention.

Selective opening of the blood-tumour barrier by intracarotid infusion of leukotriene C4

Intracarotid infusions of leukotriene C4 (LTC4) were used to selectively open the blood-tumour barrier in rats with RG-2 gliomas. Blood-brain and blood-tumour barrier permeability was determined by quantitative autoradiography using 14C aminoisobutyric acid. LTC4 (4 micrograms total dose) infused into the carotid artery ipsilateral to the tumour increased the unidirectional transfer constant for permeability, Ki, two-fold within the tumour while no effect on permeability was seen in the normal brain. No gamma glutamyl transpeptidase (gamma-GTP) activity was seen in tumour capillaries in contrast to high gamma-GTP in normal brain capillaries. These findings suggest that normal brain capillaries may resist the vasogenic effects of LTC4 while LTC4 will increase permeability in tumour capillaries. This could relate to the ability of gamma-GTP to act as an enzymatic barrier and inactivate leukotrienes in normal brain capillaries. Intracarotid LTC4 infusion may be a useful tool to selectively open the blood-tumour barrier for delivery of antineoplastic compounds.

Novel 21-aminosteroids prevent tumour associated neurological dysfunction

The efficacy of the potent lipid peroxidation inhibitors U-74006F and U-78517F in the treatment of blood-tumour barrier permeability and tumour associated neurological dysfunction was evaluated. Rats with stereotactically implanted Walker 256 tumours were treated with methylprednisolone (MP), U-74006F, U-78517F, or vehicle. (0.05 N HCl) on days 6 through 10 following implantation. Neurologic function and vascular permeability was assessed on day 10. U-74006F and MP were equally effective at preventing neurologic dysfunction compared to control (p less than 0.01). U-78517F was slightly less effective than U-74006F and MP but was significantly better than vehicle in preventing neurological dysfunction. MP resulted in a significant decrease in tumour vascular permeability (p less than 0.006) while the lipid peroxidation inhibitors had no effect on permeability.

Specific high-affinity binding of peripheral benzodiazepine receptor ligands to brain tumors in rat and man

Two types of benzodiazepine receptors have been identified in mammalian tissues: a central type which is localized to neuronal elements in the brain, and a peripheral type which is present on glial cells and in tissues outside the central nervous system such as kidney. The authors report an increase in specific binding of peripheral benzodiazepine receptor ligands in certain human brain tumors using computer assisted quantitative image analysis of autoradiograms. Higher densities of binding sites to a 3H-labeled selective peripheral benzodiazepine ligand, PK11195 [1-(2-chlorophenyl-N-methyl-N-(1-methylpropyl)-3-isoquinoline carboxamide] were observed in human gliomas as the malignancy of these tumors increased. Specific binding was also present in some non-glial tumors but little binding was demonstrated in necrotic tissue or normal brain. In in vitro binding studies in rats, there was a significant increase in Bmax (1089.3 +/- 232.2 fmol/mg tissue) in C6 glial tumors and LK Walker 256 metastatic tumors (924.2 +/- 183.7) compared with normal brain (62.1 +/- 12.8 fmol/mg tissue). Binding affinities were, however, similar (Kd = 2.09, 2.17, and 2.04 nmol/l, respectively). These findings suggest that the number of peripheral benzodiazepine receptors are increased in brain tumors. These receptors could be utilized in positron emission tomography to image brain tumors.

Cerebral edema induced by arachidonic acid: role of leukocytes and 5-lipoxygenase products

Arachidonic acid is released from cellular phospholipid membranes after brain injury associated with vasogenic edema. Intracerebral injection of arachidonic acid results in rapid breakdown of the blood-brain barrier, followed by an increase in brain water and sodium content. This effect is diminished by a 5-lipoxygenase inhibitor, but is unaffected by indomethacin, an inhibitor of cyclooxygenase. Leukocytes are rich in 5-lipoxygenase and mediate posttraumatic extracellular edema in other tissues. We sought to determine whether leukocytes are necessary for arachidonic acid-induced vasogenic edema and whether they are the primary source of 5-lipoxygenase activity. Intracerebral injection of arachidonic acid (10 micrograms) was performed in 21 rats divided into three groups. One hour after injection, the area of Evans blue stain extravasation on the corona slice through the needle tract was quantitated by polar planimetry and taken as a measure of blood-brain barrier permeability. Control animals (n = 7) had a 3.44 +/- 0.19 mm2 area of Evans blue cortical stain. Rats (n = 7) pretreated with a lipoxygenase inhibitor (BW755C) had a significant decrease (P less than 0.05) in the area of Evans blue extravasation was not significantly different from that seen in the control animals. Intracerebral injection of saline or eicosapentaenoic acid showed only minimal staining along the needle tract (0.14 +/- 0.08 mm2). We have confirmed the role of the 5-lipoxygenase products in arachidonic acid-induced vasogenic edema. The primary source of cerebral 5-lipoxygenase activity does not appear to be in leukocytes and is most likely within the brain parenchyma.

Use of thallium-201 SPECT to quantitate malignancy grade of gliomas

A quantitative preoperative technique using thallium-201 single-photon emission computerized tomography is described which predicts whether specific gliomas are of high- or low-grade malignancy. An index, based on the ratio of thallium uptake in the tumor versus the homologous contralateral brain, was calculated and compared with tumor histology. The index in 14 patients with low-grade malignant gliomas was 1.27 +/- 0.40 in contrast to an index of 2.40 +/- 0.61 in 11 patients with high-grade malignant gliomas (p less than 0.0005). Whether gliomas were of low- or high-grade malignancy could be predicted with 89% accuracy using a threshold of 1.5. Low-grade gliomas with an index higher than 1.5 acted biologically more like high-grade tumors, and no tumor histologically classified as being of high-grade malignancy had an index lower than 1.7. This technique could help to reduce unrecognized sampling errors during needle biopsies of brain tumors, particularly of high-grade lesions classified in error as low-grade tumors due to inadequate biopsy material.

Imaging of brain tumors using peripheral benzodiazepine receptor ligands

Peripheral benzodiazepine receptor ligands were utilized to selectively image intracerebrally implanted C6 gliomas, RG-2 gliomas, and Walker 256 metastatic tumors by means of quantitative autoradiography. Intravenous injections of 3H-PK11195 (1-(2-chlorophenyl)-N-methyl-N-(1-methylpropyl)-3-isoquinoline carboxamide) or 3H-flunitrazepam in combination with clonazepam revealed high densities of peripheral benzodiazepine binding in glial tumors, with less binding in metastatic tumors. Peripheral binding was displaced by preadministration of excess PK11195. Topographical correlation was excellent between areas of histologically verified tumor and high densities of peripheral benzodiazepine binding. The choroid plexus, ependyma, and pineal gland also showed a moderate level of binding, but there was little binding in other normal brain structures or necrotic tissue. Binding densities were three- to fivefold higher in C6 glial tumors compared to normal cortex. Injection of 3H-flunitrazepam alone, which binds to both central and peripheral receptors, had the advantage of showing normal anatomic structures in addition to a clear definition of tumor topography. The potential value of peripheral benzodiazepine ligands in selectively imaging brain tumors in man with positron emission tomography is discussed.

Seizures and hemiparesis in a young woman 24 years after treatment of astrocytoma

An edited transcript of Neurology Grand Rounds held at the University of California, Los Angeles, Medical Center on January 27, 1988. John Mazziotta, MD, PhD, Professor of Neurology and Radiology, is the coordinator of these conferences. This conference was edited by Harry V. Vinters, MD.