Use of Magnetic Resonance Imaging to Assess Blood-Brain/Blood-Glioma Barrier Opening During Conformal Radiotherapy
http://www.100md.com
《临床肿瘤学》
the Departments of Radiation Oncology and Radiology, University of Michigan, Ann Arbor, MI
ABSTRACT
PURPOSE: For chemotherapy to act synergistically and safely with radiation against high-grade gliomas, drugs must pass the endothelial junctions of the blood-tumor barrier (BTB) to reach all tumor cells, and should not pass the blood-brain barrier (BBB) to cause toxicity to normal brain. The objective of this study was to assess BBB/BTB status using magnetic resonance imaging (MRI) during a course of radiotherapy of high-grade gliomas.
PATIENTS AND METHODS: Sixteen patients with grade 3 or 4 supratentorial malignant glioma receiving conformal radiotherapy (RT) underwent contrast-enhanced MRI before, during, and after completion of RT. A gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA) uptake index was analyzed with respect to the tumor and RT dose received.
RESULTS: In the nonenhanced tumor region, contrast uptake increased significantly after the receipt of approximately 10 Gy (P < .01), and reached a maximum after the receipt of approximately 30 Gy. In the initially contrast-enhanced tumor region, contrast uptake decreased over the course of RT and became significant after completion of RT in patients without progressive disease. The healthy brain showed only nonsignificant changes during and after irradiation.
CONCLUSION: Contrast MRI reveals increases in Gd-DTPA uptake in the initially nonenhanced tumor region but not in the remaining brain during the course of RT, suggesting opening of the BTB. This finding suggests that the effect of conformal radiation is more selective on the BTB than the BBB, and there may be a window extending from 1 week after the initiation of radiotherapy to 1 month after the completion of treatment during which a pharmaceutical agent has maximum access to high-grade gliomas.
INTRODUCTION
The median survival for patients with glioblastoma remains less than 1 year.1 Adjuvant radiotherapy (RT) has been shown to increase median survival,2 but local tumor progression is the predominant pattern of failure.3 Clinical trials based on intensifying local RT treatment have met with limited success,4,5 which has given rise to attempts to combine chemotherapy with radiation. Unfortunately, sequential radiotherapy and chemotherapy have not consistently prolonged survival.6 One potential reason for the ineffectiveness of chemotherapy is that few drugs can pass the endothelial junctions that make up the blood-brain barrier (BBB) and blood-tumor barrier (BTB) to reach tumor cells.
Radiation could act to increase vascular permeability—causing loosening of the endothelial tight junctions, vascular leakage, or endothelial cell death,7,8 and thus, potentially increasing drug transfer capability through the BBB to tumor cells.9 Selective disruption of the barrier between the vascular supply of the tumor and tumor itself, compared with the surrounding healthy tissue receiving less radiation, could provide an optimal time window for chemotherapy. Few studies have investigated this possibility in gliomas. A study of 14 patients with brain tumors who underwent RT and were assessed by 99m Tc-glucoheptonate (99mTc-GH) imaging demonstrated an average signal enhancement increase of 25% in normal tissue and of 22% in tumor after 30 to 40 Gy compared to baseline, suggesting BBB and BTB opening.10 However, there is little information about the effect of RT on the BBB or BTB using modern techniques with high spatial resolution during the course of treatment in high-grade gliomas.
We attempted to assess the effect of RT on the BBB and BTB, with the aim of evaluating the potential for RT to improve the delivery of chemotherapeutic agents to primary gliomas by disrupting the BBB. Gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA) has a molecular size in the range of many chemotherapeutic agents (approximately 470 Da) and does not pass through the healthy BBB. When Gd-DTPA passes through the barrier and reaches brain tissue, signal intensities in postcontrast T1-weighted magnetic resonance imaging (MRI) increase. Therefore, we decided to study changes in Gd-DTPA uptake in tumor, tumor periphery, and healthy brain tissue in high-grade gliomas over the course of RT and within the first 6 months of the completion of RT using MRI.
PATIENTS AND METHODS
Between December 2000 and February 2004, 16 patients with newly diagnosed high-grade gliomas underwent three-dimensional conformal radiotherapy with a median dose of 70 Gy (range, 52 to 72 Gy in 1.8-Gy [1 patient], 2.2-Gy [1 patient], 2.4-Gy [1 patient], combination of 2-Gy and 3-Gy [1 patient] or 2-Gy [12 patients] fractions), and participated in a prospective, institutional review board–approved, clinical MRI study (Table 1). All patients met the inclusion criterion of a minimum assessable tumor volume of 4 mL after resection (ie, no patient had undergone a gross total resection). Three patients received concurrent drug treatment (Poly-ICLC, an experimental proposed immune modulator11 [patient 12] and temozolomide [patients 13 and 15]). Their results did not differ substantially from the rest of the patients (data not shown), and have therefore been analyzed together. All patients received corticosteroids during RT. As it is possible that changes in corticosteroid use during treatment could affect imaging independently of radiation, the two patients (patients 9 and 14) who had a dose reduction during the period between the pretreatment and week 3 MRI were analyzed separately.
Patients underwent MRI 1 to 2 weeks before RT
at weeks 1 to 2 and 3 to 4 during RT
at 1 month, 3 months, and 6 months after the completion of RT. When the MRI was performed at weeks 1 to 2 during RT, the patients had received a median dose of 10 Gy (range, 10 to 16 Gy), and at weeks 3 to 4, a median dose of 30 Gy (range, 23.4 to 37 Gy). For the six patients who had partial tumor resection, the pre-RT MRI was performed 15 to 26 days after surgery. MRI included T1-weighted imaging, T2-weighted imaging, fluid-attenuated inversion recovery imaging (FLAIR), and postintravenous administration of a single dose of Gd-DTPA T1-weighted imaging. Gd-DTPA uptake was assessed using a pair of T1-weighted images acquired pre- and postcontrast injection. Pre- and postcontrast T1-weighted images were registered to the radiation treatment-planning computed tomography scan, and the total accumulated planned radiation dose was registered with the diagnostic MRI using a mutual information algorithm. Changes in Gd-DTPA uptake over the course of RT and within the first 6 months of the completion of RT were assessed using a pair of T1-weigted images acquired pre- and post–Gd-DTPA injection. Signal intensities of pre- and post–Gd-DTPA T1–weighted images are expressed as
(1)
and
where A is a receiver gain that might vary from scan to scan, So is a magnitude of magnetization, TR is a repetition time, R1o is a spin-lattice relaxation rate without contrast, R1 is a change in R1 due to contrast uptake in tissue, and j represents the jth scan (or time when the scan was performed). The magnitude of magnetization and the spin-lattice relaxation rate without contrast may change over time due to factors such as tumor growth, response to RT, or both in heterogeneous tumors. An increase in R1 (an increase in Gd-DTPA uptake) over the course of RT and following completion of RT would be an indicator of an opening of the barrier between the blood supply and the tissue. To reduce the influence of changes in So and R1o, a natural logarithm of a ratio of post- to precontrast T1-weighted images was used as a matrix (a Gd-DTPA uptake index) for assessment of the change in Gd-DTPA uptake,
(2)
The latter two terms reasonably cancel out the effect of T1 enhancement in the Gd-DTPA uptake index map given by equation 2 (Fig 1). In order to remove the scan-dependent receiver gain (the first term in equation 2), all precontrast T1-weighted images and postcontrast images were normalized to their corresponding ones acquired before RT by assuming that signal intensities in the brain region (excluding large vessels) that received less than 10 Gy were unchanged. Thus, the first term in Eq (2) became a constant, ln[Apost(preRT)/Apre(preRT)] that is independent of a scan or a subject. Also, to reduce edge effects as a result of imperfect image registration, tumor volume change, or brain volume change, a 7 x 7 Gaussian filter was used to smooth images before the computation given in equation 2. The contrast uptake index given in equation 2 is a small number, and in order to display it in an image, we scaled it by a factor of 100. This index is a measurement of permeability and not simple diffusion, due to the fact that the Gd-DTPA molecules (approximately 470 Da in size) can diffuse only approximately 0.2 mm in 5 minutes in brain tissue, with a typical time from contrast injection to imaging of 1 to 3 minutes.
Changes in the Gd-DTPA uptake index maps were analyzed in regions of interest with respect to the distribution of the total accumulated planned dose and the gross tumor volume (GTV). GTV was defined on the postoperative, pre-RT, postcontrast T1-weighted MRI to include abnormal hyperintensity and the surgical cavity that would be treated by radiation (Figs 1 and 2), by a radiation oncologist in conjunction with a radiologist, both of whom were unaware of the image analysis plan. Within the GTV, voxels were classified into two groups as contrast enhanced and nonenhanced, by automatically thresholding signal intensities in the Gd-DTPA uptake index maps observed before RT. The voxels having indices one standard deviation (SD) above the mean (established from the brain region receiving < 10 Gy of the total accumulated planned dose) were defined as enhanced, and the remaining voxels in the GTV defined as nonenhanced. Furthermore, the resection cavity and necrosis were excluded from the nonenhanced GTV by rejecting voxels that had cerebral blood volumes (CBVs) two SDs below the mean of CBV defined in the contralateral healthy white matter region. The region adjacent to the GTV by 5 mm wide was targeted to received the full dose of radiation delivered to the GTV, and was classified as a periphery region of the GTV. The remainder of the brain was defined with respect to the distribution of the total accumulated planned dose as the regions receiving more than 68 Gy (excluding the GTV and the periphery of the GTV), between 60 and 68 Gy, and between 10-Gy intervals to a minimum of less than 10 Gy.
Statistical analysis was performed by using the analysis of variance. To adjust for multiple comparisons, only P values of < .01 were considered as significant. The correlation between the change in contrast uptake in the initially enhanced or nonenhanced GTV and overall survival was tested by linear regression.
RESULTS
We needed to validate several technical aspects of image analysis before we could conduct this study. First, it was important to generate the appropriate Gd-DTPA uptake index maps. We needed to focus on changes in contrast enhanced hyperintensity rather than on changes in T1 hyperintensity, whereas both types of hyperintensities were shown on post Gd-DTPA T1-weighted images. By using equation 2, we were able to eliminate T1 hyperintensity from the contrast uptake images (Fig 1). Second, we needed to define the regions of interest. Volumes of interest used in this analysis include the GTV that was drawn on the post-Gd-DTPA T1 weighted images, the contrast-enhanced GTV, the nonenhanced GTV, the periphery of the GTV, and the additional regions defined based on accumulated planned dose. Figure 2 shows the GTV (red inner contours pointed by red arrows), and isodose contours of 68 (yellow), 60 (dark pink), 50 (cyan), 40 (blue), 30 (green) and 20 Gy (outer red) overlaid on the postcontrast T1 weighted images (top row). In images on the bottom row of Figure 2, blue color marks the initially nonenhanced GTV but with CBV in the normal range compared to normal white matter while green marks the nonenhanced GTV with CBV 2 SDs lower than the mean of normal white matter. The regions enclosed by the cyan contours (cyan arrows) represent the periphery of the GTV. The unmarked regions pointed to by the red arrows represent the initially enhanced GTV. Before RT, the mean enhanced GTV occupied approximately 60% of the whole volume in the group. The median dose to the GTV was 68.4 Gy, with a range of 71.9 to 52.7 Gy (Table 1). The dose received by the GTV periphery was, as anticipated, similar to that of the GTV, and the difference between the two regions was no more than 0.5 Gy. Third, we needed to exclude the surgical cavity and necrosis in patients who underwent subtotal resection (Fig 2). As a result, volumes of the nonenhanced GTV ranged from 8.4 to 27.4 mL (median volume, 15.1 mL), which were adequate for analysis.
In addition to generating uptake maps and quantifying volumes, it was important to validate our image registration approach. Figure 3 shows two typical examples of postcontrast T1 weighted images acquired at week 3 of RT registered to those obtained before RT. Using mutual information and simplex optimization, we were able to produce excellent-quality image registration across different time points, which corrected for changes in head position and image slice orientation.
Before proceeding with a full analysis, we reviewed the clinical course of these patients. We noted that two patients (patients 5 and 6) had large tumor volume increases 3 months after completion of RT. For patient 5, the GTV increased by 50% at week 1 and by three-fold at 3 months after completion of RT compared to one observed before RT. For patient 6, the GTV increased by 50% at 1 month after completion of RT and by almost four-fold at 3 months after completion of RT. For the other 14 patients, changes in the GTV were much less (Table 1). Changes in the longest dimension of tumor at week 3 of RT versus pre-RT varied from 22% (patient 5) to –15% (patient 14). Given the large clinical difference observed, we elected to perform separate analyses of the 14 patients with relatively stable or responding tumors and the two patients whose tumor demonstrated rapid progression (patients 5 and 6). Also, for the patients who had decreases in tumor volumes during RT (eg, patients 7, 9, and 14), we manually readjusted the volume of the nonenhanced GTV region to ensure that the region did not overlap with the enhanced rim at a later time point due to a decrease in the tumor volume (Fig 4, top row). Similar adjustments were made to the enhanced volumes.
Having addressed these technical issues, we could then assess the effect of radiation on the permeability of different regions of the tumor and surrounding normal brain. Radiation affected the initially contrast-enhanced GTV differently from the other regions. Before RT, the mean Gd-DTPA uptake index in the enhanced GTV (28.6 ± 2.2) was significantly greater than in the region receiving less than 10 Gy (0.3 ± 0.1) and in the nonenhanced GTV (–0.8 ± 0.6) (P < .0005). The mean index in the initially enhanced GTV showed a nonsignificant decrease over the course of radiation treatment, and continued decreasing after completion of RT (Fig 5). The decrease became significant three months after completion of RT (P < .01).
An opposite temporal trend was observed in the regions of the GTV that were not enhanced initially (Fig 5). Before RT, the mean Gd-DTPA uptake index in the nonenhanced GTV region did not differ from the region receiving less than 10 Gy (P > .1). The mean index in the nonenhanced GTV increased significantly by week 1 of RT (P < .005), and reached a peak or plateau between week 3 during RT (P < .002) and 1 month after completion of RT (P < .01). Although the nonenhanced GTV region showed a significant uptake during the course of RT, the contrast uptake index remained significantly lower than in the initially contrast-enhanced GTV region during the course of treatment (P < .001). However, the difference between the two regions diminished over time, and was not significant by three months after completion of RT (P > .05). Both the increase in contrast uptake in the initially nonenhanced GTV and the decrease in contrast uptake in the initially enhanced GTV at week 3 of RT were visible on the Gd-DTPA uptake index maps (see Fig 4). The exclusion of the two patients in whom the corticosteroid dose decreased between pretreatment and week-3 MRI in the analysis did not alter the significant contrast uptake increase in the initially nonenhanced GTV (Fig 6). Furthermore, there were no significant differences in the contrast enhancement indices or changes in the indices between patients who had partial tumor resection and biopsy alone (P > .2).
In contrast to the effect of radiation on tumor enhancement, radiation had little effect on the periphery of the GTV. The Gd-DTPA uptake index in the periphery of the GTV did not differ from that of the region receiving less than 10 Gy (P > .1), and also no time-dependent change was detected (P > .4). Similarly, no time-dependent change in the contrast uptake was detected in the remaining regions of the brain, regardless of dose.
As it seemed possible that an increase in GD-DPTA uptake in the previously un-enhanced region could be a reflection of tumor progression rather than radiation-induced permeability changes, we performed two types of analyses to assess the relationship between uptake and progression. First, we analyzed separately the patients who showed clinical progression during or immediately after treatment and those who did not. Temporal changes in the contrast uptake of the 2 patients who showed rapid tumor progression during treatment differed from the group (Fig 7). For patient 6, an initial decrease in the pre-RT enhanced GTV was observed at week 1 during RT, similar to the results from the group. However, there was a large increase (two-fold greater than the group mean) 1 month after completion of RT. Furthermore, the increase in the Gd-DTPA uptake index in the initially nonenhanced GTV was greater than the group mean at week 1 during RT, and the extent of the increase was four-fold greater than the group mean 1 month after completion of RT. Patient 5 showed similar results: a marked increase of the contrast uptake in both initially enhanced and nonenhanced GTV. Therefore, although progression was associated with an increase in Gd-DPTA uptake in these two patients, this increase was easily distinguishable from the increase in permeability in the other 14 patients.
Our second method of assessing the relationship between Gd-DPTA uptake and progression was to attempt to correlate uptake changes with survival. If increases in Gd-DPTA were simply a reflection of progression, then patients showing the largest changes would be anticipated to have worse survival. In fact, the changes in Gd-DTPA uptake of the GTV, observed at week 3 of RT or 1 month after completion of RT, were not correlated to survival, especially in the initially nonenhanced GTV. The cross correlation coefficients between survival and the changes of Gd-DTPA uptake in the nonenhanced GTV at week 3 of RT or 1 month after completion of RT versus pre RT was –0.01 (P > .9), or –0.43 (P > .1). Similarly, no significant correlation was found between changes in contrast uptake in the enhanced GTV at week 3 of RT or 1 month after completion of RT versus pre-RT (cross correlation coefficients –0.14 and P > .5, and –0.53 and P > .05, respectively). Taken together, these findings suggest that although progression is associated with increased Gd-DPTA uptake, the specific changes that we describe in this study in the initially nonenhanced region reflect increased permeability.
DISCUSSION
To our knowledge, this is the first study is to use quantitative and high-resolution imaging to assess the effect of radiation on the permeability of tumor and normal brain to a molecule in the size range of chemotherapeutic agents in high-grade gliomas. We found that permeability in the regions of the GTV that initially showed little enhancement increases significantly during a course of RT. The onset of the increase in Gd-DTPA uptake occurs after approximately 1 week (10 Gy), reaches its maximum or plateau after 3 weeks (30 Gy), and subsides 1 month after completion of RT. In contrast, although Gd-DTPA uptake in the initial enhanced GTV remains high compared to the nonenhanced regions during treatment, uptake decreases over the course of RT and after completion of RT. In fact, 3 months after the completion of RT, Gd-DTPA uptake does not differ significantly between the initially enhanced and nonenhanced GTV regions.
There are several limiting factors in our study. First, only a relatively small number of patients have been studied. Secondly, given the nature of clinical high-grade gliomas, tumor volumes can change during and immediately following the course of RT, and in some cases demonstrate rapid tumor growth. Rapid tumor growth is often associated with angiogenesis, which could confound our results. We have tried to address this weakness by excluding the two patients in whom the tumor volumes showed a rapid increase during our evaluation period. Also, we have tested the correlation between the contrast-uptake change in the initially nonenhanced tumor region and survival, which shows no significant correlation. Third, the MR images used in the current study were acquired via a standard clinical protocol. The contrast uptake index is calculated from the clinical pre- and postcontrast T1 weighted images. This index might not have the sensitivity to detect the subtle BBB opening in the normal brain tissue regions that received high radiation dose.
Our findings may have important implications for how to combine chemotherapy and radiation therapy. For chemotherapy to be effective, both alone and as synergistic with radiation, it must gain access to tumor. One reason for the overall modest efficacy of chemotherapy may be that the blood-tumor barrier prevents delivery of these agents to all tumor cells. In most high-grade gliomas, the blood-brain/tumor barrier is partially disrupted, as evidenced by the post Gd-DTPA T1-weighted images and histopathology.12-14 In the postcontrast images, there is an enhanced tumor region, usually (but not always) located in the periphery of tumor, indicating the existence of highly permeable vessels (Figs 1 and 2). However, the contrast enhancement in high-grade gliomas is not uniform, and in our study, the nonenhanced region occupied approximately 40% of the GTV. This suggests that even in highly malignant gliomas there are regions that are not accessible to drug. We would hypothesize that chemotherapeutic agents may not be able to pass through the blood-brain/tumor barrier to reach these tumor cells.
Our analysis also reveals that the effect of radiation on the blood-brain barrier and blood-tumor barrier is selective. In the periphery of the GTV, we did not observe a significant change in the contrast agent uptake, even though this region received the same dose as the GTV. The median volume of the regions that received more than 68 Gy but excluded the GTV was 89 mL, range from 26 to 330 mL, which was twice as great as the median volume of the GTV. Likewise, the healthy brain more distant from the high-dose region also showed no effect. Although the underlying mechanism of our finding remains undetermined, our observation implies that the radiation-induced permeability change is more selective for tumor than normal brain vasculature. This finding is encouraging, in that it offers hope that the radiation-induced increase in blood-brain barrier disruption may improve tumor selectivity of chemotherapy.
Our findings are overall consistent with an earlier report using a radioisotope.10 In that study, patients received whole-brain radiation of 2 Gy per fraction via parallel opposed and paired portals. Before radiation, 99mTc-GH counts in the tumor regions were 30% greater than in the normal tissue region. After patients received 30 to 40 Gy, 99mTc-GH counts increased 25% and 22% in the healthy tissue and tumor regions, respectively, compared to before radiation. The use of only a single time point during treatment and the low spatial resolution of this technique did not permit the types of quantitative measurements used in the current study. Experiments in animal models also support our findings. In a study of radiation-induced blood-brain barrier permeability changes in a rat glioma model, MRI-derived Gd-DTPA transport coefficient increased approximately 2.5-fold after receiving 25 Gy whole brain irradiation compared with before radiation.15 Another study assessed the dynamic change of the blood-brain barriar permeability produced by radiation using fluorescein isothiocyanate (FITC)-dextran molecules with molecule sizes 4.4, 10, 38.2, 70, and 150 kDa in the healthy rat.16 A single dose of 20 Gy significantly increased the permeability of the blood-brain barrier to FITC-dextran in the pial vessels, an effect that peaked 24 hours after radiation and depended on the molecule size. Note, however, that, in contrast to these studies, we did not find an increase in permeability of the normal brain. It is possible that the single large fraction used in the animal studies does not produce the same effect on the blood-brain barrier as fractionated low doses of radiation used in the clinic.
Our study reveals that the permeability of the blood-vessel barrier in the high-grade gliomas increases in response to radiation. This finding potentially supports the survival advantage of concomitant and adjuvant temozolomide and RT versus RT alone in glioblastoma multiform patients as reported from a randomized phase III trial by the European Organisation for Research and Treatment of Cancer (EORTC) Brain & RT Groups and National Cancer Institute of Canada (NCIC) Clinical Trials Group.17 However, many questions still remain. It would be interesting to determine if these results will apply to low-grade gliomas, especially grade 2 tumors, in which the use of chemotherapy is now being extensively explored, as well as to assess whether radiation affects the permeability of the blood-tumor barrier in brain metastases. In addition, new studies using dynamic contrast enhanced imaging could complement the current study by providing a more quantitative assessment of permeability and offer the possibility of pharmacokinetic modeling (at the cost of lower spatial resolution and signal-to-noise ratio).18-23 Furthermore, the use of contrast agents with different molecule sizes may allow us to estimate the maximum size of therapeutic agents that is able to cross the opening barrier, and the time that it takes for the agent to across the barrier. We feel these types of studies have high potential to help us design more rationale and, we hope, effective schedules combining chemotherapy and radiation therapy for the treatment of high-grade brain tumors.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Honoraria: Theodore S. Lawrence, Astro, Center for Biomedical Communications, Elsevier, National Institutes of Health, NMCR, S.G. Madison, University of Pennsylvania, William Vlasek. Research Funding: Theodore S. Lawrence, AstraZeneca, Eli Lilly, MedImmune. For a detailed description of these categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and Disclosures of Potential Conflicts of Interest found in Information for Contributors in the front of each issue.
Acknowledgment
We thank Diana Gomez-Hassan, MD, PhD, for advice on radiological evaluation.
NOTES
Supported in part by NIH grants 2 PO1 CA59827 and PO1 CA85878, and the Walther Foundation.
Presented at the 46th Annual Meeting of the American Society for Therapeutic Radiation and Oncology (ASTRO), Atlanta, GA, October 3-7, 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Curran WJ Jr, Scott CB, Horton J, et al: Recursive partitioning analysis of prognostic factors in three Radiation Therapy Oncology Group malignant glioma trials. J Natl Cancer Inst 85:704-710, 1993
Walker MD, Green SB, Byar DP, et al: Randomized comparisons of radiotherapy and nitrosoureas for the treatment of malignant glioma after surgery. N Engl J Med 303:1323-1329, 1980
Wallner KE, Galicich JH, Krol G, et al: Patterns of failure following treatment for glioblastoma multiforme and anaplastic astrocytoma. Int J Radiat Oncol Biol Phys 16:1405-1409, 1989
Loeffler JS, Alexander E III, Shea WM, et al: Radiosurgery as part of the initial management of patients with malignant gliomas. J Clin Oncol 10:1379-1385, 1992
Werner-Wasik M, Scott CB, Nelson DF, et al: Final report of a phase I/II trial of hyperfractionated and accelerated hyperfractionated radiation therapy with carmustine for adults with supratentorial malignant gliomas: Radiation Therapy Oncology Group Study 83-02. Cancer 77:1535-1543, 1996
Randomized trial of procarbazine, lomustine, and vincristine in the adjuvant treatment of high-grade astrocytoma: A Medical Research Council Trial. J Clin Oncol 19:509-518, 2001
van der Kogel AJ: Radiation-induced damage in the central nervous system: An interpretation of target cell responses. Br J Cancer Suppl 7:207-217, 1986
Rubin P, Gash DM, Hansen JT, et al: Disruption of the blood-brain barrier as the primary effect of CNS irradiation. Radiother Oncol 31:51-60, 1994
Schultheiss TE, Kun LE, Ang KK, et al: Radiation response of the central nervous system. Int J Radiat Oncol Biol Phys 31:1093-1112, 1995
Qin DX, Zheng R, Tang J, et al: Influence of radiation on the blood-brain barrier and optimum time of chemotherapy. Int J Radiat Oncol Biol Phys 19:1507-1510, 1990
Salazar AM, Levy HB, Ondra S, et al: Long-term treatment of malignant gliomas with intramuscularly administered polyinosinic-polycytidylic acid stabilized with polylysine and carboxymethylcellulose: An open pilot study. Neurosurgery 38:1096-1103, 1996
Weindel K, Moringlane JR, Marme D, et al: Detection and quantification of vascular endothelial growth factor/vascular permeability factor in brain tumor tissue and cyst fluid: The key to angiogenesis Neurosurgery 35:439-448, 1994
Takano S, Yoshii Y, Kondo S, et al: Concentration of vascular endothelial growth factor in the serum and tumor tissue of brain tumor patients. Cancer Res 56:2185-2190, 1996
Chaudhry IH, O'Donovan DG, Brenchley PE, et al: Vascular endothelial growth factor expression correlates with tumour grade and vascularity in gliomas. Histopathology 39:409-415, 2001
Krueck WG, Schmiedl UP, Maravilla KR, et al: MR assessment of radiation-induced blood-brain barrier permeability changes in rat glioma model. AJNR Am J Neuroradiol 15:625-632, 1994
Yuan H, Gaber MW, McColgan T, et al: Radiation-induced permeability and leukocyte adhesion in the rat blood-brain barrier: Modulation with anti-ICAM-1 antibodies. Brain Res 969:59-69, 2003
Stupp R, Mason WP, Van den Bent MJ, et al: Concomitant and adjuvant temozolomide (TMZ) and radiotherapy (RT) for newly diagnosed glioblastoma multiform (GBM): Conclusive results of a randomized phase III trial by the EORTC brain & RT groups and NCIC clinical trials group. American Society of Clincial Oncology, New Orleans, LA
p. 1, 2004
Tofts PS, Brix G, Buckley DL, et al: Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: Standardized quantities and symbols. J Magn Reson Imaging 10:223-232, 1999
Tofts PS: Modeling tracer kinetics in dynamic Gd-DTPA MR imaging. J Magn Reson Imaging 7:91-101, 1997
Larsson HB, Tofts PS: Measurement of blood-brain barrier permeability using dynamic Gd-DTPA scanning—A comparison of methods. Magn Reson Med 24:174-176, 1992
Larsson HB, Stubgaard M, Frederiksen JL, et al: Quantitation of blood-brain barrier defect by magnetic resonance imaging and gadolinium-DTPA in patients with multiple sclerosis and brain tumors. Magn Reson Med 16:117-131, 1990
Brix G, Semmler W, Port R, et al: Pharmacokinetic parameters in CNS Gd-DTPA enhanced MR imaging. J Comput Assist Tomogr 15:621-628, 1991
Ewing JR, Knight RA, Nagaraja TN, et al: Patlak plots of Gd-DTPA MRI data yield blood-brain transfer constants concordant with those of 14C-sucrose in areas of blood-brain opening. Magn Reson in Med 50:283-292, 2003(Yue Cao, Christina I. Tsi)
ABSTRACT
PURPOSE: For chemotherapy to act synergistically and safely with radiation against high-grade gliomas, drugs must pass the endothelial junctions of the blood-tumor barrier (BTB) to reach all tumor cells, and should not pass the blood-brain barrier (BBB) to cause toxicity to normal brain. The objective of this study was to assess BBB/BTB status using magnetic resonance imaging (MRI) during a course of radiotherapy of high-grade gliomas.
PATIENTS AND METHODS: Sixteen patients with grade 3 or 4 supratentorial malignant glioma receiving conformal radiotherapy (RT) underwent contrast-enhanced MRI before, during, and after completion of RT. A gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA) uptake index was analyzed with respect to the tumor and RT dose received.
RESULTS: In the nonenhanced tumor region, contrast uptake increased significantly after the receipt of approximately 10 Gy (P < .01), and reached a maximum after the receipt of approximately 30 Gy. In the initially contrast-enhanced tumor region, contrast uptake decreased over the course of RT and became significant after completion of RT in patients without progressive disease. The healthy brain showed only nonsignificant changes during and after irradiation.
CONCLUSION: Contrast MRI reveals increases in Gd-DTPA uptake in the initially nonenhanced tumor region but not in the remaining brain during the course of RT, suggesting opening of the BTB. This finding suggests that the effect of conformal radiation is more selective on the BTB than the BBB, and there may be a window extending from 1 week after the initiation of radiotherapy to 1 month after the completion of treatment during which a pharmaceutical agent has maximum access to high-grade gliomas.
INTRODUCTION
The median survival for patients with glioblastoma remains less than 1 year.1 Adjuvant radiotherapy (RT) has been shown to increase median survival,2 but local tumor progression is the predominant pattern of failure.3 Clinical trials based on intensifying local RT treatment have met with limited success,4,5 which has given rise to attempts to combine chemotherapy with radiation. Unfortunately, sequential radiotherapy and chemotherapy have not consistently prolonged survival.6 One potential reason for the ineffectiveness of chemotherapy is that few drugs can pass the endothelial junctions that make up the blood-brain barrier (BBB) and blood-tumor barrier (BTB) to reach tumor cells.
Radiation could act to increase vascular permeability—causing loosening of the endothelial tight junctions, vascular leakage, or endothelial cell death,7,8 and thus, potentially increasing drug transfer capability through the BBB to tumor cells.9 Selective disruption of the barrier between the vascular supply of the tumor and tumor itself, compared with the surrounding healthy tissue receiving less radiation, could provide an optimal time window for chemotherapy. Few studies have investigated this possibility in gliomas. A study of 14 patients with brain tumors who underwent RT and were assessed by 99m Tc-glucoheptonate (99mTc-GH) imaging demonstrated an average signal enhancement increase of 25% in normal tissue and of 22% in tumor after 30 to 40 Gy compared to baseline, suggesting BBB and BTB opening.10 However, there is little information about the effect of RT on the BBB or BTB using modern techniques with high spatial resolution during the course of treatment in high-grade gliomas.
We attempted to assess the effect of RT on the BBB and BTB, with the aim of evaluating the potential for RT to improve the delivery of chemotherapeutic agents to primary gliomas by disrupting the BBB. Gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA) has a molecular size in the range of many chemotherapeutic agents (approximately 470 Da) and does not pass through the healthy BBB. When Gd-DTPA passes through the barrier and reaches brain tissue, signal intensities in postcontrast T1-weighted magnetic resonance imaging (MRI) increase. Therefore, we decided to study changes in Gd-DTPA uptake in tumor, tumor periphery, and healthy brain tissue in high-grade gliomas over the course of RT and within the first 6 months of the completion of RT using MRI.
PATIENTS AND METHODS
Between December 2000 and February 2004, 16 patients with newly diagnosed high-grade gliomas underwent three-dimensional conformal radiotherapy with a median dose of 70 Gy (range, 52 to 72 Gy in 1.8-Gy [1 patient], 2.2-Gy [1 patient], 2.4-Gy [1 patient], combination of 2-Gy and 3-Gy [1 patient] or 2-Gy [12 patients] fractions), and participated in a prospective, institutional review board–approved, clinical MRI study (Table 1). All patients met the inclusion criterion of a minimum assessable tumor volume of 4 mL after resection (ie, no patient had undergone a gross total resection). Three patients received concurrent drug treatment (Poly-ICLC, an experimental proposed immune modulator11 [patient 12] and temozolomide [patients 13 and 15]). Their results did not differ substantially from the rest of the patients (data not shown), and have therefore been analyzed together. All patients received corticosteroids during RT. As it is possible that changes in corticosteroid use during treatment could affect imaging independently of radiation, the two patients (patients 9 and 14) who had a dose reduction during the period between the pretreatment and week 3 MRI were analyzed separately.
Patients underwent MRI 1 to 2 weeks before RT
at weeks 1 to 2 and 3 to 4 during RT
at 1 month, 3 months, and 6 months after the completion of RT. When the MRI was performed at weeks 1 to 2 during RT, the patients had received a median dose of 10 Gy (range, 10 to 16 Gy), and at weeks 3 to 4, a median dose of 30 Gy (range, 23.4 to 37 Gy). For the six patients who had partial tumor resection, the pre-RT MRI was performed 15 to 26 days after surgery. MRI included T1-weighted imaging, T2-weighted imaging, fluid-attenuated inversion recovery imaging (FLAIR), and postintravenous administration of a single dose of Gd-DTPA T1-weighted imaging. Gd-DTPA uptake was assessed using a pair of T1-weighted images acquired pre- and postcontrast injection. Pre- and postcontrast T1-weighted images were registered to the radiation treatment-planning computed tomography scan, and the total accumulated planned radiation dose was registered with the diagnostic MRI using a mutual information algorithm. Changes in Gd-DTPA uptake over the course of RT and within the first 6 months of the completion of RT were assessed using a pair of T1-weigted images acquired pre- and post–Gd-DTPA injection. Signal intensities of pre- and post–Gd-DTPA T1–weighted images are expressed as
(1)
and
where A is a receiver gain that might vary from scan to scan, So is a magnitude of magnetization, TR is a repetition time, R1o is a spin-lattice relaxation rate without contrast, R1 is a change in R1 due to contrast uptake in tissue, and j represents the jth scan (or time when the scan was performed). The magnitude of magnetization and the spin-lattice relaxation rate without contrast may change over time due to factors such as tumor growth, response to RT, or both in heterogeneous tumors. An increase in R1 (an increase in Gd-DTPA uptake) over the course of RT and following completion of RT would be an indicator of an opening of the barrier between the blood supply and the tissue. To reduce the influence of changes in So and R1o, a natural logarithm of a ratio of post- to precontrast T1-weighted images was used as a matrix (a Gd-DTPA uptake index) for assessment of the change in Gd-DTPA uptake,
(2)
The latter two terms reasonably cancel out the effect of T1 enhancement in the Gd-DTPA uptake index map given by equation 2 (Fig 1). In order to remove the scan-dependent receiver gain (the first term in equation 2), all precontrast T1-weighted images and postcontrast images were normalized to their corresponding ones acquired before RT by assuming that signal intensities in the brain region (excluding large vessels) that received less than 10 Gy were unchanged. Thus, the first term in Eq (2) became a constant, ln[Apost(preRT)/Apre(preRT)] that is independent of a scan or a subject. Also, to reduce edge effects as a result of imperfect image registration, tumor volume change, or brain volume change, a 7 x 7 Gaussian filter was used to smooth images before the computation given in equation 2. The contrast uptake index given in equation 2 is a small number, and in order to display it in an image, we scaled it by a factor of 100. This index is a measurement of permeability and not simple diffusion, due to the fact that the Gd-DTPA molecules (approximately 470 Da in size) can diffuse only approximately 0.2 mm in 5 minutes in brain tissue, with a typical time from contrast injection to imaging of 1 to 3 minutes.
Changes in the Gd-DTPA uptake index maps were analyzed in regions of interest with respect to the distribution of the total accumulated planned dose and the gross tumor volume (GTV). GTV was defined on the postoperative, pre-RT, postcontrast T1-weighted MRI to include abnormal hyperintensity and the surgical cavity that would be treated by radiation (Figs 1 and 2), by a radiation oncologist in conjunction with a radiologist, both of whom were unaware of the image analysis plan. Within the GTV, voxels were classified into two groups as contrast enhanced and nonenhanced, by automatically thresholding signal intensities in the Gd-DTPA uptake index maps observed before RT. The voxels having indices one standard deviation (SD) above the mean (established from the brain region receiving < 10 Gy of the total accumulated planned dose) were defined as enhanced, and the remaining voxels in the GTV defined as nonenhanced. Furthermore, the resection cavity and necrosis were excluded from the nonenhanced GTV by rejecting voxels that had cerebral blood volumes (CBVs) two SDs below the mean of CBV defined in the contralateral healthy white matter region. The region adjacent to the GTV by 5 mm wide was targeted to received the full dose of radiation delivered to the GTV, and was classified as a periphery region of the GTV. The remainder of the brain was defined with respect to the distribution of the total accumulated planned dose as the regions receiving more than 68 Gy (excluding the GTV and the periphery of the GTV), between 60 and 68 Gy, and between 10-Gy intervals to a minimum of less than 10 Gy.
Statistical analysis was performed by using the analysis of variance. To adjust for multiple comparisons, only P values of < .01 were considered as significant. The correlation between the change in contrast uptake in the initially enhanced or nonenhanced GTV and overall survival was tested by linear regression.
RESULTS
We needed to validate several technical aspects of image analysis before we could conduct this study. First, it was important to generate the appropriate Gd-DTPA uptake index maps. We needed to focus on changes in contrast enhanced hyperintensity rather than on changes in T1 hyperintensity, whereas both types of hyperintensities were shown on post Gd-DTPA T1-weighted images. By using equation 2, we were able to eliminate T1 hyperintensity from the contrast uptake images (Fig 1). Second, we needed to define the regions of interest. Volumes of interest used in this analysis include the GTV that was drawn on the post-Gd-DTPA T1 weighted images, the contrast-enhanced GTV, the nonenhanced GTV, the periphery of the GTV, and the additional regions defined based on accumulated planned dose. Figure 2 shows the GTV (red inner contours pointed by red arrows), and isodose contours of 68 (yellow), 60 (dark pink), 50 (cyan), 40 (blue), 30 (green) and 20 Gy (outer red) overlaid on the postcontrast T1 weighted images (top row). In images on the bottom row of Figure 2, blue color marks the initially nonenhanced GTV but with CBV in the normal range compared to normal white matter while green marks the nonenhanced GTV with CBV 2 SDs lower than the mean of normal white matter. The regions enclosed by the cyan contours (cyan arrows) represent the periphery of the GTV. The unmarked regions pointed to by the red arrows represent the initially enhanced GTV. Before RT, the mean enhanced GTV occupied approximately 60% of the whole volume in the group. The median dose to the GTV was 68.4 Gy, with a range of 71.9 to 52.7 Gy (Table 1). The dose received by the GTV periphery was, as anticipated, similar to that of the GTV, and the difference between the two regions was no more than 0.5 Gy. Third, we needed to exclude the surgical cavity and necrosis in patients who underwent subtotal resection (Fig 2). As a result, volumes of the nonenhanced GTV ranged from 8.4 to 27.4 mL (median volume, 15.1 mL), which were adequate for analysis.
In addition to generating uptake maps and quantifying volumes, it was important to validate our image registration approach. Figure 3 shows two typical examples of postcontrast T1 weighted images acquired at week 3 of RT registered to those obtained before RT. Using mutual information and simplex optimization, we were able to produce excellent-quality image registration across different time points, which corrected for changes in head position and image slice orientation.
Before proceeding with a full analysis, we reviewed the clinical course of these patients. We noted that two patients (patients 5 and 6) had large tumor volume increases 3 months after completion of RT. For patient 5, the GTV increased by 50% at week 1 and by three-fold at 3 months after completion of RT compared to one observed before RT. For patient 6, the GTV increased by 50% at 1 month after completion of RT and by almost four-fold at 3 months after completion of RT. For the other 14 patients, changes in the GTV were much less (Table 1). Changes in the longest dimension of tumor at week 3 of RT versus pre-RT varied from 22% (patient 5) to –15% (patient 14). Given the large clinical difference observed, we elected to perform separate analyses of the 14 patients with relatively stable or responding tumors and the two patients whose tumor demonstrated rapid progression (patients 5 and 6). Also, for the patients who had decreases in tumor volumes during RT (eg, patients 7, 9, and 14), we manually readjusted the volume of the nonenhanced GTV region to ensure that the region did not overlap with the enhanced rim at a later time point due to a decrease in the tumor volume (Fig 4, top row). Similar adjustments were made to the enhanced volumes.
Having addressed these technical issues, we could then assess the effect of radiation on the permeability of different regions of the tumor and surrounding normal brain. Radiation affected the initially contrast-enhanced GTV differently from the other regions. Before RT, the mean Gd-DTPA uptake index in the enhanced GTV (28.6 ± 2.2) was significantly greater than in the region receiving less than 10 Gy (0.3 ± 0.1) and in the nonenhanced GTV (–0.8 ± 0.6) (P < .0005). The mean index in the initially enhanced GTV showed a nonsignificant decrease over the course of radiation treatment, and continued decreasing after completion of RT (Fig 5). The decrease became significant three months after completion of RT (P < .01).
An opposite temporal trend was observed in the regions of the GTV that were not enhanced initially (Fig 5). Before RT, the mean Gd-DTPA uptake index in the nonenhanced GTV region did not differ from the region receiving less than 10 Gy (P > .1). The mean index in the nonenhanced GTV increased significantly by week 1 of RT (P < .005), and reached a peak or plateau between week 3 during RT (P < .002) and 1 month after completion of RT (P < .01). Although the nonenhanced GTV region showed a significant uptake during the course of RT, the contrast uptake index remained significantly lower than in the initially contrast-enhanced GTV region during the course of treatment (P < .001). However, the difference between the two regions diminished over time, and was not significant by three months after completion of RT (P > .05). Both the increase in contrast uptake in the initially nonenhanced GTV and the decrease in contrast uptake in the initially enhanced GTV at week 3 of RT were visible on the Gd-DTPA uptake index maps (see Fig 4). The exclusion of the two patients in whom the corticosteroid dose decreased between pretreatment and week-3 MRI in the analysis did not alter the significant contrast uptake increase in the initially nonenhanced GTV (Fig 6). Furthermore, there were no significant differences in the contrast enhancement indices or changes in the indices between patients who had partial tumor resection and biopsy alone (P > .2).
In contrast to the effect of radiation on tumor enhancement, radiation had little effect on the periphery of the GTV. The Gd-DTPA uptake index in the periphery of the GTV did not differ from that of the region receiving less than 10 Gy (P > .1), and also no time-dependent change was detected (P > .4). Similarly, no time-dependent change in the contrast uptake was detected in the remaining regions of the brain, regardless of dose.
As it seemed possible that an increase in GD-DPTA uptake in the previously un-enhanced region could be a reflection of tumor progression rather than radiation-induced permeability changes, we performed two types of analyses to assess the relationship between uptake and progression. First, we analyzed separately the patients who showed clinical progression during or immediately after treatment and those who did not. Temporal changes in the contrast uptake of the 2 patients who showed rapid tumor progression during treatment differed from the group (Fig 7). For patient 6, an initial decrease in the pre-RT enhanced GTV was observed at week 1 during RT, similar to the results from the group. However, there was a large increase (two-fold greater than the group mean) 1 month after completion of RT. Furthermore, the increase in the Gd-DTPA uptake index in the initially nonenhanced GTV was greater than the group mean at week 1 during RT, and the extent of the increase was four-fold greater than the group mean 1 month after completion of RT. Patient 5 showed similar results: a marked increase of the contrast uptake in both initially enhanced and nonenhanced GTV. Therefore, although progression was associated with an increase in Gd-DPTA uptake in these two patients, this increase was easily distinguishable from the increase in permeability in the other 14 patients.
Our second method of assessing the relationship between Gd-DPTA uptake and progression was to attempt to correlate uptake changes with survival. If increases in Gd-DPTA were simply a reflection of progression, then patients showing the largest changes would be anticipated to have worse survival. In fact, the changes in Gd-DTPA uptake of the GTV, observed at week 3 of RT or 1 month after completion of RT, were not correlated to survival, especially in the initially nonenhanced GTV. The cross correlation coefficients between survival and the changes of Gd-DTPA uptake in the nonenhanced GTV at week 3 of RT or 1 month after completion of RT versus pre RT was –0.01 (P > .9), or –0.43 (P > .1). Similarly, no significant correlation was found between changes in contrast uptake in the enhanced GTV at week 3 of RT or 1 month after completion of RT versus pre-RT (cross correlation coefficients –0.14 and P > .5, and –0.53 and P > .05, respectively). Taken together, these findings suggest that although progression is associated with increased Gd-DPTA uptake, the specific changes that we describe in this study in the initially nonenhanced region reflect increased permeability.
DISCUSSION
To our knowledge, this is the first study is to use quantitative and high-resolution imaging to assess the effect of radiation on the permeability of tumor and normal brain to a molecule in the size range of chemotherapeutic agents in high-grade gliomas. We found that permeability in the regions of the GTV that initially showed little enhancement increases significantly during a course of RT. The onset of the increase in Gd-DTPA uptake occurs after approximately 1 week (10 Gy), reaches its maximum or plateau after 3 weeks (30 Gy), and subsides 1 month after completion of RT. In contrast, although Gd-DTPA uptake in the initial enhanced GTV remains high compared to the nonenhanced regions during treatment, uptake decreases over the course of RT and after completion of RT. In fact, 3 months after the completion of RT, Gd-DTPA uptake does not differ significantly between the initially enhanced and nonenhanced GTV regions.
There are several limiting factors in our study. First, only a relatively small number of patients have been studied. Secondly, given the nature of clinical high-grade gliomas, tumor volumes can change during and immediately following the course of RT, and in some cases demonstrate rapid tumor growth. Rapid tumor growth is often associated with angiogenesis, which could confound our results. We have tried to address this weakness by excluding the two patients in whom the tumor volumes showed a rapid increase during our evaluation period. Also, we have tested the correlation between the contrast-uptake change in the initially nonenhanced tumor region and survival, which shows no significant correlation. Third, the MR images used in the current study were acquired via a standard clinical protocol. The contrast uptake index is calculated from the clinical pre- and postcontrast T1 weighted images. This index might not have the sensitivity to detect the subtle BBB opening in the normal brain tissue regions that received high radiation dose.
Our findings may have important implications for how to combine chemotherapy and radiation therapy. For chemotherapy to be effective, both alone and as synergistic with radiation, it must gain access to tumor. One reason for the overall modest efficacy of chemotherapy may be that the blood-tumor barrier prevents delivery of these agents to all tumor cells. In most high-grade gliomas, the blood-brain/tumor barrier is partially disrupted, as evidenced by the post Gd-DTPA T1-weighted images and histopathology.12-14 In the postcontrast images, there is an enhanced tumor region, usually (but not always) located in the periphery of tumor, indicating the existence of highly permeable vessels (Figs 1 and 2). However, the contrast enhancement in high-grade gliomas is not uniform, and in our study, the nonenhanced region occupied approximately 40% of the GTV. This suggests that even in highly malignant gliomas there are regions that are not accessible to drug. We would hypothesize that chemotherapeutic agents may not be able to pass through the blood-brain/tumor barrier to reach these tumor cells.
Our analysis also reveals that the effect of radiation on the blood-brain barrier and blood-tumor barrier is selective. In the periphery of the GTV, we did not observe a significant change in the contrast agent uptake, even though this region received the same dose as the GTV. The median volume of the regions that received more than 68 Gy but excluded the GTV was 89 mL, range from 26 to 330 mL, which was twice as great as the median volume of the GTV. Likewise, the healthy brain more distant from the high-dose region also showed no effect. Although the underlying mechanism of our finding remains undetermined, our observation implies that the radiation-induced permeability change is more selective for tumor than normal brain vasculature. This finding is encouraging, in that it offers hope that the radiation-induced increase in blood-brain barrier disruption may improve tumor selectivity of chemotherapy.
Our findings are overall consistent with an earlier report using a radioisotope.10 In that study, patients received whole-brain radiation of 2 Gy per fraction via parallel opposed and paired portals. Before radiation, 99mTc-GH counts in the tumor regions were 30% greater than in the normal tissue region. After patients received 30 to 40 Gy, 99mTc-GH counts increased 25% and 22% in the healthy tissue and tumor regions, respectively, compared to before radiation. The use of only a single time point during treatment and the low spatial resolution of this technique did not permit the types of quantitative measurements used in the current study. Experiments in animal models also support our findings. In a study of radiation-induced blood-brain barrier permeability changes in a rat glioma model, MRI-derived Gd-DTPA transport coefficient increased approximately 2.5-fold after receiving 25 Gy whole brain irradiation compared with before radiation.15 Another study assessed the dynamic change of the blood-brain barriar permeability produced by radiation using fluorescein isothiocyanate (FITC)-dextran molecules with molecule sizes 4.4, 10, 38.2, 70, and 150 kDa in the healthy rat.16 A single dose of 20 Gy significantly increased the permeability of the blood-brain barrier to FITC-dextran in the pial vessels, an effect that peaked 24 hours after radiation and depended on the molecule size. Note, however, that, in contrast to these studies, we did not find an increase in permeability of the normal brain. It is possible that the single large fraction used in the animal studies does not produce the same effect on the blood-brain barrier as fractionated low doses of radiation used in the clinic.
Our study reveals that the permeability of the blood-vessel barrier in the high-grade gliomas increases in response to radiation. This finding potentially supports the survival advantage of concomitant and adjuvant temozolomide and RT versus RT alone in glioblastoma multiform patients as reported from a randomized phase III trial by the European Organisation for Research and Treatment of Cancer (EORTC) Brain & RT Groups and National Cancer Institute of Canada (NCIC) Clinical Trials Group.17 However, many questions still remain. It would be interesting to determine if these results will apply to low-grade gliomas, especially grade 2 tumors, in which the use of chemotherapy is now being extensively explored, as well as to assess whether radiation affects the permeability of the blood-tumor barrier in brain metastases. In addition, new studies using dynamic contrast enhanced imaging could complement the current study by providing a more quantitative assessment of permeability and offer the possibility of pharmacokinetic modeling (at the cost of lower spatial resolution and signal-to-noise ratio).18-23 Furthermore, the use of contrast agents with different molecule sizes may allow us to estimate the maximum size of therapeutic agents that is able to cross the opening barrier, and the time that it takes for the agent to across the barrier. We feel these types of studies have high potential to help us design more rationale and, we hope, effective schedules combining chemotherapy and radiation therapy for the treatment of high-grade brain tumors.
Authors' Disclosures of Potential Conflicts of Interest
The following authors or their immediate family members have indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. Honoraria: Theodore S. Lawrence, Astro, Center for Biomedical Communications, Elsevier, National Institutes of Health, NMCR, S.G. Madison, University of Pennsylvania, William Vlasek. Research Funding: Theodore S. Lawrence, AstraZeneca, Eli Lilly, MedImmune. For a detailed description of these categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and Disclosures of Potential Conflicts of Interest found in Information for Contributors in the front of each issue.
Acknowledgment
We thank Diana Gomez-Hassan, MD, PhD, for advice on radiological evaluation.
NOTES
Supported in part by NIH grants 2 PO1 CA59827 and PO1 CA85878, and the Walther Foundation.
Presented at the 46th Annual Meeting of the American Society for Therapeutic Radiation and Oncology (ASTRO), Atlanta, GA, October 3-7, 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Curran WJ Jr, Scott CB, Horton J, et al: Recursive partitioning analysis of prognostic factors in three Radiation Therapy Oncology Group malignant glioma trials. J Natl Cancer Inst 85:704-710, 1993
Walker MD, Green SB, Byar DP, et al: Randomized comparisons of radiotherapy and nitrosoureas for the treatment of malignant glioma after surgery. N Engl J Med 303:1323-1329, 1980
Wallner KE, Galicich JH, Krol G, et al: Patterns of failure following treatment for glioblastoma multiforme and anaplastic astrocytoma. Int J Radiat Oncol Biol Phys 16:1405-1409, 1989
Loeffler JS, Alexander E III, Shea WM, et al: Radiosurgery as part of the initial management of patients with malignant gliomas. J Clin Oncol 10:1379-1385, 1992
Werner-Wasik M, Scott CB, Nelson DF, et al: Final report of a phase I/II trial of hyperfractionated and accelerated hyperfractionated radiation therapy with carmustine for adults with supratentorial malignant gliomas: Radiation Therapy Oncology Group Study 83-02. Cancer 77:1535-1543, 1996
Randomized trial of procarbazine, lomustine, and vincristine in the adjuvant treatment of high-grade astrocytoma: A Medical Research Council Trial. J Clin Oncol 19:509-518, 2001
van der Kogel AJ: Radiation-induced damage in the central nervous system: An interpretation of target cell responses. Br J Cancer Suppl 7:207-217, 1986
Rubin P, Gash DM, Hansen JT, et al: Disruption of the blood-brain barrier as the primary effect of CNS irradiation. Radiother Oncol 31:51-60, 1994
Schultheiss TE, Kun LE, Ang KK, et al: Radiation response of the central nervous system. Int J Radiat Oncol Biol Phys 31:1093-1112, 1995
Qin DX, Zheng R, Tang J, et al: Influence of radiation on the blood-brain barrier and optimum time of chemotherapy. Int J Radiat Oncol Biol Phys 19:1507-1510, 1990
Salazar AM, Levy HB, Ondra S, et al: Long-term treatment of malignant gliomas with intramuscularly administered polyinosinic-polycytidylic acid stabilized with polylysine and carboxymethylcellulose: An open pilot study. Neurosurgery 38:1096-1103, 1996
Weindel K, Moringlane JR, Marme D, et al: Detection and quantification of vascular endothelial growth factor/vascular permeability factor in brain tumor tissue and cyst fluid: The key to angiogenesis Neurosurgery 35:439-448, 1994
Takano S, Yoshii Y, Kondo S, et al: Concentration of vascular endothelial growth factor in the serum and tumor tissue of brain tumor patients. Cancer Res 56:2185-2190, 1996
Chaudhry IH, O'Donovan DG, Brenchley PE, et al: Vascular endothelial growth factor expression correlates with tumour grade and vascularity in gliomas. Histopathology 39:409-415, 2001
Krueck WG, Schmiedl UP, Maravilla KR, et al: MR assessment of radiation-induced blood-brain barrier permeability changes in rat glioma model. AJNR Am J Neuroradiol 15:625-632, 1994
Yuan H, Gaber MW, McColgan T, et al: Radiation-induced permeability and leukocyte adhesion in the rat blood-brain barrier: Modulation with anti-ICAM-1 antibodies. Brain Res 969:59-69, 2003
Stupp R, Mason WP, Van den Bent MJ, et al: Concomitant and adjuvant temozolomide (TMZ) and radiotherapy (RT) for newly diagnosed glioblastoma multiform (GBM): Conclusive results of a randomized phase III trial by the EORTC brain & RT groups and NCIC clinical trials group. American Society of Clincial Oncology, New Orleans, LA
p. 1, 2004
Tofts PS, Brix G, Buckley DL, et al: Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: Standardized quantities and symbols. J Magn Reson Imaging 10:223-232, 1999
Tofts PS: Modeling tracer kinetics in dynamic Gd-DTPA MR imaging. J Magn Reson Imaging 7:91-101, 1997
Larsson HB, Tofts PS: Measurement of blood-brain barrier permeability using dynamic Gd-DTPA scanning—A comparison of methods. Magn Reson Med 24:174-176, 1992
Larsson HB, Stubgaard M, Frederiksen JL, et al: Quantitation of blood-brain barrier defect by magnetic resonance imaging and gadolinium-DTPA in patients with multiple sclerosis and brain tumors. Magn Reson Med 16:117-131, 1990
Brix G, Semmler W, Port R, et al: Pharmacokinetic parameters in CNS Gd-DTPA enhanced MR imaging. J Comput Assist Tomogr 15:621-628, 1991
Ewing JR, Knight RA, Nagaraja TN, et al: Patlak plots of Gd-DTPA MRI data yield blood-brain transfer constants concordant with those of 14C-sucrose in areas of blood-brain opening. Magn Reson in Med 50:283-292, 2003(Yue Cao, Christina I. Tsi)