Original Research
Neuroradiology
March 2007

Induction of Hyperintense Signal on T2-Weighted MR Images Correlates with Infusion Distribution from Intracerebral Convection-Enhanced Delivery of a Tumor-Targeted Cytotoxin

Abstract

OBJECTIVE. Convection-enhanced delivery is a promising approach to intracerebral drug delivery in which a fluid pressure gradient is used to infuse therapeutic macromolecules through an indwelling catheter into the interstitial spaces of the brain. Our purpose was to test the hypothesis that hyperintense signal changes on T2-weighted images produced by such infusions can be used to track drug distribution.
SUBJECTS AND METHODS. Seven adults with recurrent malignant glioma underwent concurrent intracerebral infusions of the tumor-targeted cytotoxin cintredekin besudotox and 123I-labeled human serum albumin. The agents were administered through a total of 18 catheters among the seven patients. Adequacy of distribution of drug was determined by evidence of distribution of 123I-labeled human serum albumin on SPECT images coregistered with MR images. Qualitative analysis was performed by three blinded observers. Quantitative analysis also was performed.
RESULTS. Infusions into 12 catheters produced intraparenchymal distribution as seen on SPECT images, but infusions into six catheters did not. At qualitative assessment of signal changes on MR images, reviewers correctly predicted which catheters would produce extraparenchymal distribution and which catheters would produce parenchymal distribution. Of the 12 infusions that produced intraparenchymal distribution, four catheters had been placed in regions of relatively normal signal intensity and produced regions of newly increased signal intensity, the volume of which highly correlated with the volume and geometry of distribution on SPECT (r2 = 0.9502). Eight infusions that produced intraparenchymal distribution were performed in regions of preexisting hyperintense signal. In these brains, additional signal changes were always produced, but quantitative correlations between areas of newly increased signal intensity and the volume and geometry of distribution on SPECT could not be established.
CONCLUSION. Convection-enhanced infusions frequently do not provide intraparenchymal drug distribution, and these failures can be identified with MRI soon after infusion. When infusions are performed into regions of normal signal intensity, development of hyperintense signal change strongly correlates with the volume and geometry of distribution of infusate.

Introduction

Drug delivery remains a major limitation to effective therapy for malignant brain tumors. Systemic delivery of drugs to tumors within the intracerebral compartment is hampered by the restrictive blood-brain barrier and high intratumoral pressure [1-4]. Conventional regional intracerebral drug delivery bypasses these barriers but relies on diffusion of high concentrations of drug away from the site of injection, which is severely limited by molecular weight and tissue clearance.
Convection-enhanced delivery is a novel intracerebral drug delivery technique in which a fluid pressure gradient is used to infuse therapeutic macromolecules through an indwelling catheter directly into the interstitial spaces of brain parenchyma. In preclinical animal models, convection-enhanced delivery has provided broad and homogeneous intracerebral distribution of macromolecular therapeutic agents [4-19]. Although early clinical studies [20-28] have shown the efficacy of convection-enhanced delivery, it is apparent that delivery is a complex process and that with current approaches to convection-enhanced delivery, optimum drug delivery may occur in as few as 20% of patients. Our work imaging the distribution of 123I-labeled human serum albumin (123I-HSA) with SPECT after intracerebral convection-enhanced delivery, along with one other report of a similar technique [20], represent the only available data on the distribution of drugs with this promising technique in humans. SPECT of radiotracers for this application is limited by the availability of appropriate radiotracers and by the paucity of anatomic information. A better method of tracking drug distribution by convection-enhanced delivery is needed.
We and other investigators [27, 29] have postulated that signal changes on MRI studies may be useful for estimating distribution of molecules delivered with convection enhancement. In this study, we provide evidence, first, that lack of production of hyperintense signal on T2-weighted images can be used to clearly identify infusions that fail to provide intraparenchymal drug distribution and, second, that development of hyperintense signal abnormality in regions with previously normal signal intensity strongly correlates with the volume and geometry of drug distribution.

Subjects and Methods

Patient Selection

Patients ≥ 18 years old with recurrent or progressive and resectable supratentorial malignant glioma (World Health Organization grade III or IV) and a Karnofsky performance status score > 70 were eligible for this study. A solid contrast-enhancing tumor nodule ≥ 1.0 cm and ≤ 5.0 cm was required for the study. Patients enrolled also had to have completed external beam radiation therapy ≥ 8 weeks before study entry and recovered from toxicity of previous local therapies. Patients were excluded if they had signs of impending cerebral herniation, multifocal disease, tumor crossing the midline, or subependymal or leptomeningeal spread or if they had another significant uncontrolled medical illness. Follow-up visits were completed for all patients. The institutional review board (4774-03-4R0) and the U.S. Food and Drug Administration (BB-IND-8959) approved the protocol. Informed consent was obtained from all patients who participated after the nature of all procedures was explained.
Eight patients were enrolled, and seven received infusions. The seventh patient enrolled (in stage 2) experienced extensive vasogenic edema with aphasia and right-sided hemiparesis after craniotomy and before drug infusion and was removed from the study. Four patients were treated in stage 1 and three patients in stage 2 of the study. The median age of the patients was 46.5 years (range, 19-62 years), and the mean Karnofsky performance score was 90 (range, 80-100). Four patients had recurrent glioblastoma multiforme, three had recurrent anaplastic astrocytoma, and one initially had low-grade astrocytoma that underwent malignant transformation to glioblastoma multiforme. Seven of eight patients underwent gross total resection during this study. One patient, who received drug infusion in stage 1, did not undergo subsequent tumor resection because of rapid tumor progression. All patients had undergone surgical resection, external beam radiation therapy, and cytotoxic chemotherapy. Cytostatic agents were also used in two patients, intracranial implantation of a carmustine wafer (Gliadel wafer) in two patients, stereotactic radiosurgery in one patient, and immunotherapy in one patient.

Cintredekin Besudotox and 123I-HSA

Cintredekin besudotox is a recombinant chimeric protein consisting of a genetically engineered, mutated, and truncated form of the cytotoxic Pseudomonas aeruginosa exotoxin fused to interleukin-13. The full sequence encoding cintredekin besudotox has been described [30]. Before delivery, cintredekin besudotox was diluted with 0.2% HSA (Plasbumin-25, Bayer) in 0.9% saline solution. For the first 48 hours of each infusion, SPECT with 123I-HSA was used as a surrogate marker for imaging the infusion distribution because the maximally tolerated dose of cytotoxin is too small to radiolabel and image directly. The HSA was purified to homogeneity with ion-exchange high-pressure liquid chromatography and radiolabeled with 123I (MDS Nordion International) by a modified iodogen method with a targetspecific activity of 80 mCi/10 mg. Iodine-123-HSA was chosen because its size, shape, and molecular weight are similar to those of the cytotoxin, and HSA forms an otherwise essential component of cytotoxin drug formulations.

Treatment

The study had a two-stage design. Patients enrolled in stage 1 received a combined preresection and postresection continuous infusion of 123I-HSA coinfused with cintredekin besudotox. Patients in stage 2 received only postresection continuous infusion. Patients enrolled in stage 1 underwent stereotactic biopsy to confirm the existence of viable malignant glioma before stereotactic placement of two infusion catheters. At least one catheter was always placed into the contrast-enhancing component of the tumor, and infusions standardly began 1 day after catheter placement. Before resection, cintredekin besudotox was infused at a concentration of 0.5 μg/mL for 96 hours in a fixed total infusion volume of 51.8 mL at a total infusion rate of 0.540 mL/h divided by the number of catheters placed. A craniotomy for tumor resection was performed 15 ± 7 days after the end of the preresection infusion with stereotactically guided, postresection, intraoperative placement of one to three infusion catheters into parenchyma surrounding the resection cavity.
In the postresection setting, cintredekin besudotox was infused at a concentration of 0.5 μg/mL over 96 hours in a fixed total infusion volume of 72.0 mL at a fixed total infusion rate of 0.750 mL/h divided by the number of catheters placed. Patients enrolled in stage 2 did not receive a preresection infusion but underwent craniotomy followed by a postresection infusion identical to the postresection infusion used in stage 1 except that postoperative catheter placement was performed 3-7 days after resection through a small burr hole.
Patients in both stages underwent the following sequence of imaging studies. First, a postresection MR image or CT scan was obtained within 24 hours of catheter placement but before infusion to document catheter positions. Next, SPECT was performed 6, 24, and 48 hours after the start of infusion, MRI being performed within 90 minutes of the 24- and 48-hour SPECT scans. SPECT scans of the head were obtained with a three-head scanner (Trionix Research Laboratory) fitted with two fan beam collimators (Triad LESR, Trionix Research Laboratory) and a precise pinhole collimator. MRI was performed on a 3-T unit with an eight-channel dedicated head coil (Trio, Siemens Medical Solutions). Finally, MRI was performed 72 hours after the start of the infusion. No SPECT beyond 48 hours was possible because of the half-life (13.2 h) of 123I and the limited radiation dose that could be delivered to the patient. Several fiduciary markers were placed on the patient's head to coregister MR images with SPECT images.
Open-ended, barium-impregnated silicon infusion catheters (CD-435, Vygon Neuro) with a 1.0-mm inner diameter and a 2.0-mm outer diameter were used. The positions of all catheters were documented within 24 hours with CT or MRI. At drug administration, a compression hub was used to connect the proximal end of the infusion catheter to a three-way Luer-lock stopcock connector with an air filter and an antisiphon valve. One pump was used for each catheter-tubing unit along with a 30-mL syringe to infuse cintredekin besudotox. Patients were maintained on high-dose corticosteroids during and after the infusion.

Imaging Assessments

The volume of distribution (Vd) for SPECT was determined with a threshold pixel method that has proved accurate at our institution for calculating the volume of small spheres ranging in diameter from 1.3 to 5.3 cm in a brain phantom model. Three-dimensional discrete Fourier transform convolution was used to calculate isodose contours for the infusion through each catheter [31, 32]. The Vd was taken as the volume contained within the diameters at 50% of the maximum SPECT signal and always was found at the tip of the infusion catheter. Measurement of Vd for SPECT was performed by a single observer, who was blinded to degree of change in signal intensity on T2-weighted images. The observer used iPlan cranial analysis software (Brain-LAB) on a 1.5-GHz computer (Pentium Mobile).
The following MRI protocol was used. Acquisition of unenhanced T1-weighted, T2-weighted, and FLAIR images in the axial plane was followed by contrast-enhanced 1-mm 3D spoiled gradient-recalled acquisition in the steady state. The Vd for signal abnormality produced on T2-weighted images was performed by a single observer blinded to Vd for SPECT results. This observer used the same analysis software used for SPECT image analysis. On these MR images, increased signal intensity on pretherapy T2-weighted images was subtracted from posttherapy images after coregistration of the two data sets so that only the increase in signal intensity on T2-weighted images could be measured and compared with SPECT distribution.

Analysis and Statistics

We performed both qualitative analysis of SPECT and MR images and quantitative analysis of SPECT and MR images. In the qualitative analysis, SPECT images were coregistered with T1-weighted MR images showing the catheter trajectories and were graded by three reviewers in consensus about whether the images showed intraparenchymal infusion for each catheter. A successful infusion is shown in Figure 1A, 1B, 1C, 1D, 1E, 1F, 1G, 1H, 1I, 1J. T2-weighted images also showing the catheter trajectories were evaluated separately by three blinded reviewers in consensus. The images were graded on whether baseline hyperintense signal was present near the catheter tip before infusion and whether images after the start of infusion showed substantial change in signal intensity at each catheter site. The results of these analyses were compared in a two-by-two format with Fisher's exact test, and 95% CIs were calculated for sensitivity, specificity, and positive and negative predictive values. In the quantitative analysis of SPECT and MR images, volumetric measurement of the comparisons between T2-weighted MR images and SPECT images was analyzed with iPlan software. A Pearson's correlation coefficient was calculated for the least-squares fit of the relation between Vd for SPECT and Vd for signal abnormality produced on T2-weighted images 24 and 48 hours after the start of infusion.

Results

A total of 21 catheters were placed in the seven patients. Three catheters were excluded from evaluation because of a delay in infusion that prohibited SPECT owing to radionuclide decay (one catheter) or obvious extraparenchymal infusion identified with SPECT before MRI was performed (two catheters). The remaining 18 catheters were used in our analysis.

Qualitative Assessment

Reviewers identified six catheters in four patients on SPECT images for which no intraparenchymal distribution of agent was seen either 24 or 48 hours after the start of infusion because the infusions leaked essentially completely into the subarachnoid or intraventricular CSF spaces. An unsuccessful infusion is depicted in Figure 2A, 2B, 2C, 2D, 2E, 2F. For these six catheters, reviewers found that the baseline MR images had normal signal intensity at the catheter tip in five cases and abnormal signal intensity at the catheter tip in one case. In none of these six catheters did the reviewers find an increase in signal intensity on T2-weighted images around the catheter during the infusion at either 24 or 48 hours. Thus lack of signal intensity increase on T2-weighted images led to correct identification of all six infusions that failed to provide notable intraparenchymal drug distribution (p < 0.001) (Table 1).
TABLE 1: Prediction of Intraparenchymal Infusion with MRI
New Hyperintense Signal on T2-Weighted MR ImagesNo. of Catheters with Solely Intraparenchymal Drug Distribution on SPECTNo. of Catheters with Extraparenchymal Drug Distribution on SPECT
Present120
Absent
0
6
Note—p < 0.001; sensitivity, 100% (95% CI, 74-100%); specificity, 100% (95% CI, 54-100%); positive predictive value, 100%; negative predictive value, 100%.
For the remaining 12 catheters, reviewers identified evidence of intraparenchymal distribution of infused 123I-HSA on SPECT images. Reviewers found that the baseline MR images had abnormal signal intensity at the catheter tip in eight cases and normal signal intensity at the catheter tip in four cases. For all 12 of these catheters, reviewers found an increase in signal intensity on T2-weighted images around the catheter during the infusion. Thus, in our small series, visual assessment of signal intensity change on T2-weighted images had 100% sensitivity (95% CI, 74-100%), specificity (95% CI, 54-100%), and positive and negative predictive values for identification of catheters through which drug was being infused into brain parenchyma (p < 0.001) (Table 1).

Quantitative Assessment

Twelve of the 18 catheters showed intraparenchymal distribution of infused 123I-HSA on visual assessment of the SPECT images. Four of these 12 catheters had been placed in areas of normal signal intensity on baseline T2-weighted images. Hyperintense signal on T2-weighted images surrounding these four catheters always completely encompassed the Vd and closely matched the geometric configuration of the coregistered SPECT signal for 123I-HSA at the 50% isodose level (Fig. 2A, 2B, 2C, 2D, 2E, 2F). Quantitative analysis of the changes in signal intensity caused by these catheters showed a mean Vd on SPECT of 3.84 ± 1.30 cm3 (SD) at 24 hours and 11.34 ± 4.91 cm3 at 48 hours. An increase in signal intensity on T2-weighted images was seen in all of these cases. The mean Vd of the subsequently increased signal intensity was 5.45 ± 1.52 cm3 at 24 hours and 11.44 ± 2.90 cm3 at 48 hours. There was also a strong correlation between the 50% isodose volume of the distribution imaged with SPECT and the volume of region of abnormal signal intensity depicted on T2-weighted MR images (r2 = 0.9502 for all data points, r2 = 0.9094 for the 24-hour data points, and r2 = 0.9412 for the 48-hour data points) (Fig. 3). Therefore, in patients with minimal or no preinfusion signal abnormality on T2-weighted images, the presence and geometry of change in signal intensity on T2-weighted images not only ensured that the infusate was not completely leaking into the subarachnoid or intraventricular CSF spaces but also accurately depicted the geometry and volume of macromolecule distribution.
Eight catheters were placed into areas of preexisting hyperintense signal on T2-weighted images. A change in the volume and geometry of the T2 signal hyperintensity during infusion was seen in all these cases. However, accurate depiction of the Vd and geometric distribution of the infusate could not be determined from the T2-weighted MR images.
Our data indicate that a change in signal intensity on T2-weighted images should be expected in catheters through which drug is being infused in an intraparenchymal location after delivery of ≥ 6 mL of infusate and that lack of visualization of such changes likely indicates failed intraparenchymal drug delivery with leak of the infusate into CSF spaces.

Discussion

Although convection-enhanced delivery has become a routine method of drug delivery to the brain in human trials, including a few phase 3 trials of novel therapeutic agents, drug distribution with this method is rarely monitored and has not been optimized. Unfortunately, this lack of information on drug distribution has the potential to cause failure of these clinical studies independent of the efficacy of the therapeutic agent. The purpose of this study was to determine whether routine T2-weighted MRI signal changes would provide clinically useful information regarding the distributions of these infusions. Previous findings [27, 29] have suggested a correlation between MR signal change and drug distribution, but no quantitative comparisons have been performed. It has also been apparent in the evaluation of our data and the work of others that change in signal intensity on T2-weighted images does not always correlate with the geometry and Vd of a drug. For these reasons, in this study we sought to identify a subset of patients in whom such a correlation might be established.
Fig. 1A —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Preinfusion axial T2-weighted image shows catheter trajectory (green line) and catheter tip in occipital lobe.
Fig. 1B —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Axial T2-weighted image 24 hours after beginning of infusion shows development of new hyperintense signal abnormality at catheter tip.
Fig. 1C —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Axial T2-weighted image 48 hours after beginning of infusion shows increase in hyperintense signal at catheter tip.
Fig. 1D —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Axial T2-weighted image 96 hours after beginning of infusion shows prominent region of hyperintense signal at catheter tip.
Fig. 1E —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Superimposition on D of intraparenchymal distribution from 48-hour 123I-human serum albumin SPECT at 50% isodose level (orange).
Fig. 1F —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Preinfusion axial T2-weighted image shows catheter trajectory (green line) and catheter tip in occipital lobe.
Fig. 1G —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Axial T2-weighted image 24 hours after beginning of infusion shows development of new hyperintense signal abnormality at catheter tip.
Fig. 1H —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Axial T2-weighted image 48 hours after beginning of infusion shows increase in hyperintense signal at catheter tip.
Fig. 1I —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Axial T2-weighted image 96 hours after beginning of infusion shows prominent region of hyperintense signal at catheter tip.
Fig. 1J —19-year-old man (A-E) and 56-year-old man (F-J) with glioblastoma multiforme. Serial increases in MR signal intensity after successful convection-enhanced infusion in two patients with little or no preexisting signal abnormality directly at catheter tip. In both patients morphology of region of hyperintense signal abnormality on MRI closely matches distribution of infusate seen on SPECT. Human figure indicates imaging plane. Blue line indicates plane perpendicular to tip of catheter. Green line indicates catheter trajectory. Superimposition on I of intraparenchymal distribution from 48-hour 123I-human serum albumin SPECT at 50% isodose level (green).
Our findings indicate that in most patients, examination of T2-weighted images within 48 hours after the start of infusion can be useful in determining whether infused fluid is leaking completely into the subarachnoid space or is becoming distributed within the cerebral parenchyma. This finding is important given that even in this small and carefully monitored trial, 40% (8/20, including patients with extraparenchymal infusion identified with SPECT alone) of infusions were completely ineffective. Therefore, use of MRI to determine that infusions are intraparenchymal in nature is an important step in implementing this therapy.
Fig. 2A —19-year-old man with glioblastoma multiforme. Two types of infusion failure were seen in this patient. Forty-eight hours after initiation of convection-enhanced infusion, leakage of infusate into subarachnoid CSF space occurred, and hyperintense signal did not develop on T2-weighted MR images. A-C, Images show failure due to catheter crossing sulcus within backflow region. D-F, Images show failure due to placement of catheter tip within subarachnoid space. Human figure indicates imaging plane. Blue line and indicates plane perpendicular to catheter shown in C and F. Thick red and thicker and lighter blue indicate catheter trajectories. Oblique sagittal T2-weighted image shows trajectory of catheter (red line) and superimposed coregistered SPECT signal (yellow outline) around tip of catheter at 50% isodose level. Catheter tip is inappropriately adjacent to sulcus.
Fig. 2B —19-year-old man with glioblastoma multiforme. Two types of infusion failure were seen in this patient. Forty-eight hours after initiation of convection-enhanced infusion, leakage of infusate into subarachnoid CSF space occurred, and hyperintense signal did not develop on T2-weighted MR images. A-C, Images show failure due to catheter crossing sulcus within backflow region. D-F, Images show failure due to placement of catheter tip within subarachnoid space. Human figure indicates imaging plane. Blue line and indicates plane perpendicular to catheter shown in C and F. Thick red and thicker and lighter blue indicate catheter trajectories. Oblique coronal T2-weighted image oriented 90° to A shows catheter trajectory (red line), superimposed coregistered SPECT signal at 50% isodose level, and portions of resection cavity medial to catheter trajectory. A sulcus, through which infusate has leaked into subarachnoid CSF spaces, is seen extending approximately 1 cm along distal end of catheter tip. Long axis of region of SPECT signal extends toward subarachnoid space (rather than circumferentially surrounding catheter tip) indicating leakage of infusate into subarachnoid space (yellow outline). Leakage accounts for absence of development of hyperintense signal adjacent to catheter tip.
Fig. 2C —19-year-old man with glioblastoma multiforme. Two types of infusion failure were seen in this patient. Forty-eight hours after initiation of convection-enhanced infusion, leakage of infusate into subarachnoid CSF space occurred, and hyperintense signal did not develop on T2-weighted MR images. A-C, Images show failure due to catheter crossing sulcus within backflow region. D-F, Images show failure due to placement of catheter tip within subarachnoid space. Human figure indicates imaging plane. Blue line and indicates plane perpendicular to catheter shown in C and F. Thick red and thicker and lighter blue indicate catheter trajectories. Oblique axial T2-weighted image oriented 90° to A and B shows tip of catheter (red dot) en face in sulcus. Because of catheter placement in sulcus, trajectory of infusate (yellow outline) is not intraparenchymal but extends across multiple sulci in subarachnoid space.
Fig. 2D —19-year-old man with glioblastoma multiforme. Two types of infusion failure were seen in this patient. Forty-eight hours after initiation of convection-enhanced infusion, leakage of infusate into subarachnoid CSF space occurred, and hyperintense signal did not develop on T2-weighted MR images. A-C, Images show failure due to catheter crossing sulcus within backflow region. D-F, Images show failure due to placement of catheter tip within subarachnoid space. Human figure indicates imaging plane. Blue line and indicates plane perpendicular to catheter shown in C and F. Thick red and thicker and lighter blue indicate catheter trajectories. Axial T2-weighted image depicting trajectory of second catheter (turquoise line) shows end of catheter projecting into subarachnoid space and wholly extraparenchymal infusate volume (outline). Hyperintense signal is not present around catheter tip.
Fig. 2E —19-year-old man with glioblastoma multiforme. Two types of infusion failure were seen in this patient. Forty-eight hours after initiation of convection-enhanced infusion, leakage of infusate into subarachnoid CSF space occurred, and hyperintense signal did not develop on T2-weighted MR images. A-C, Images show failure due to catheter crossing sulcus within backflow region. D-F, Images show failure due to placement of catheter tip within subarachnoid space. Human figure indicates imaging plane. Blue line and indicates plane perpendicular to catheter shown in C and F. Thick red and thicker and lighter blue indicate catheter trajectories. Reconstructed image in coronal plane shows catheter (turquoise line) piercing pial surface and infusate accumulation (outline) within basilar cisterns.
Fig. 2F —19-year-old man with glioblastoma multiforme. Two types of infusion failure were seen in this patient. Forty-eight hours after initiation of convection-enhanced infusion, leakage of infusate into subarachnoid CSF space occurred, and hyperintense signal did not develop on T2-weighted MR images. A-C, Images show failure due to catheter crossing sulcus within backflow region. D-F, Images show failure due to placement of catheter tip within subarachnoid space. Human figure indicates imaging plane. Blue line and indicates plane perpendicular to catheter shown in C and F. Thick red and thicker and lighter blue indicate catheter trajectories. Reconstructed image in sagittal plane shows catheter tip (dot) within center of extraparenchymal infusate volume (outline).
Some limitations of our study are worth noting. First, our approach provides limited information on the distribution of infusate in patients with substantial preexisting hyperintense signal abnormality. This limitation may be important in the care of patients with malignant glioma because even after resection of intraparenchymal tumors, marked hyperintense signal often persists in the peritumoral tissue. Second, because our study was focused on tumors primarily within white matter, we were not able to determine whether development of hyperintense signal changes gives information about drug delivery within gray matter. Convection-enhanced delivery techniques have been used in gray matter structures in the management of neurodegenerative diseases [5, 33-35]. Information about infusate distribution will be important in those settings. Third, our choice of the inert radiotracer HSA as a surrogate for imaging drug distribution precluded evaluation of the potential influence of affinity-based or isoelectric binding kinetics, which may influence therapeutic drug distribution. However, such information about binding kinetics is not relevant if the therapeutic agents are used at concentrations that greatly exceed that needed to fully saturate all potential binding sites. Such is the case for the drug used in our trial and for most other therapeutic macromolecules being considered for administration by convection-enhanced delivery. If drugs that may be susceptible to binding-site barriers are delivered by convection-enhanced delivery, these forces must be further considered and evaluated. Even in that setting, however, our findings would have clinical relevance. Even if the lack of signal change on T2-weighted images indicated only catheters with infusions that were leaking completely into the CSF space, this information would still be useful because it might prompt repositioning of the catheter or cessation of the infusion to reduce the potential toxicity of misdirected infusions.
Fig. 3 —Line plot shows volume of distribution of 123I-human serum albumin measured at 50% isodose level with SPECT on x-axis and volume of region of hyperintense signal on T2-weighted MR images 24 and 48 hours after start of infusion on y-axis for catheters placed into regions without substantial hyperintensity before infusion. Plot shows correlation between these two measures (r2 = 0.9502 for all data points, r2 = 0.9094 for 24-hour data points, and r2 = 0.9412 for 48-hour data points).
In conclusion, our findings indicate that hyperintense signal changes on T2-weighted MR images give useful information about the outcome of infusions by convection-enhanced delivery in the human brain. The widespread availability of the routine imaging sequences we used allows MRI to play a fundamental role in assessment of intraparenchymal drug delivery with this technique.

Acknowledgments

We acknowledge and thank Laura O. Thomas for her critical review of the manuscript. We also acknowledge the contributions of Denise Lally-Goss, Lisa Tansey, Sharon McGehee, Neil A. Petry, Kim Greer, B. H. Joshi, Martin L. Brady, Amy Grahn, Jeffrey Sherman, Terence Wong, Allan H. Friedman, Henry S. Friedman, Darell D. Bigner, and James E. Herndon II.

Footnotes

Address correspondence to J. M. Provenzale ([email protected]).
The employment status of R. Raghavan at Therataxis, D. Croteau at NeoPharm, and C. Pedain and I. Rodríguez-Ponce at BrainLAB did not influence the data in this study.
This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute, Center for Cancer Research.
Research supported by NIH/NCRR K23 RR16065 (Sampson), NIH/NCI R01 CA97611 (Sampson), 2P50-NS20023 (Bigner/Sampson, 5P50-CA108786 (Bigner/Sampson), Accelerate Brain Cancer Cure (ABC2) (Sampson), BrainLAB AG, and NeoPharm, Inc. Experimental data were acquired using shared instrumentation funded by the National Center for Research Resources of the National Institutes of Health (S10 RR15697). The views presented in this article do not necessarily reflect those of the Food and Drug Aministration. These studies were conducted as part of collaboration between the Food and Drug Administration and NeoPharm, Inc., under a Cooperative Research and Development Agreement (CRADA). This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute, and Center for Cancer Research.

References

1.
Groothuis DR. The blood-brain and blood-tumor barriers: a review of strategies for increasing drug delivery. Neuro-oncol 2000; 2:45-59
2.
Zalutsky MR, Moseley RP, Coakham HB, Coleman RE, Bigner DD. Pharmacokinetics and tumor localization of 131I-labeled anti-tenascin monoclonal antibody 81C6 in patients with gliomas and other intracranial malignancies. Cancer Res 1989; 49:2807-2813
3.
Jain RK. Vascular and interstitial barriers to delivery of therapeutic agents in tumors. Cancer Metastasis Rev 1990; 9:253-266
4.
Grossi PM, Ochiai H, Archer GE, et al. Efficacy of intracerebral microinfusion of trastuzumab in an athymic rat model of intracerebral metastatic breast cancer. Clin Cancer Res 2003; 9:5514-5520
5.
Hamilton JF, Morrison PF, Chen MY, et al. Heparin coinfusion during convection-enhanced delivery (CED) increases the distribution of the glial-derived neurotrophic factor (GDNF) ligand family in rat striatum and enhances the pharmacological activity of neurturin. Exp Neurol 2001; 168:155-161
6.
Morrison PF, Chen MY, Chadwick RS, Lonser RR, Oldfield EH. Focal delivery during direct infusion to brain: role of flow rate, catheter diameter, and tissue mechanics. Am J Physiol 1999; 277:R1218-R1229
7.
Bobo RH, Laske DW, Akbasak A, Morrison PF, Dedrick RL, Oldfield EH. Convection-enhanced delivery of macromolecules in the brain. Proc Natl Acad Sci U S A 1994; 91:2076-2080
8.
Laske DW, Morrison PF, Lieberman DM, et al. Chronic interstitial infusion of protein to primate brain: determination of drug distribution and clearance with single-photon emission computerized tomography imaging. J Neurosurg 1997; 87:586-594
9.
Laske DW, Ilercil O, Akbasak A, Youle RJ, Oldfield EH. Efficacy of direct intratumoral therapy with targeted protein toxins for solid human gliomas in nude mice. J Neurosurg 1994; 80:520-526
10.
Morrison PF, Laske DW, Bobo H, Oldfield EH, Dedrick RL. High-flow microinfusion: tissue penetration and pharmacodynamics. Am J Physiol 1994; 266:R292-R305
11.
Lieberman DM, Laske DW, Morrison PF, Bankiewicz KS, Oldfield EH. Convection-enhanced distribution of large molecules in gray matter during interstitial drug infusion. J Neurosurg 1995; 82:1021-1029
12.
Heimberger AB, Archer GE, McLendon RE, et al. Temozolomide delivered by intracerebral microinfusion is safe and efficacious against malignant gliomas in rats. Clin Cancer Res 2000; 6:4148-4153
13.
Sanftner LM, Sommer JM, Suzuki BM, et al. AAV2-mediated gene delivery to monkey putamen: evaluation of an infusion device and delivery parameters. Exp Neurol 2005; 194:476-483
14.
Saito R, Bringas JR, Panner A, et al. Convection-enhanced delivery of tumor necrosis factor-related apoptosis-inducing ligand with systemic administration of temozolomide prolongs survival in an intracranial glioblastoma xenograft model. Cancer Res 2004; 64:6858-6862
15.
Saito R, Bringas JR, McKnight TR, et al. Distribution of liposomes into brain and rat brain tumor models by convection-enhanced delivery monitored with magnetic resonance imaging. Cancer Res 2004; 64:2572-2579
16.
Mamot C, Nguyen JB, Pourdehnad M, et al. Extensive distribution of liposomes in rodent brains and brain tumors following convection-enhanced delivery. J Neurooncol 2004; 68:1-9
17.
Cunningham J, Oiwa Y, Nagy D, Podsakoff G, Colosi P, Bankiewicz KS. Distribution of AAV-TK following intracranial convection-enhanced delivery into rats. Cell Transplant 2000; 9:585-594
18.
Degen JW, Walbridge S, Vortmeyer AO, Oldfield EH, Lonser RR. Safety and efficacy of convectionenhanced delivery of gemcitabine or carboplatin in a malignant glioma model in rats. J Neurosurg 2003; 99:893-898
19.
Groothuis DR, Ward S, Itskovich AC, et al. Comparison of 14C-sucrose delivery to the brain by intravenous, intraventricular, and convection-enhanced intracerebral infusion. J Neurosurg 1999; 90:321-331
20.
Patel SJ, Shapiro WR, Laske DW, et al. Safety and feasibility of convection-enhanced delivery of Cotara for the treatment of malignant glioma: initial experience in 51 patients. Neurosurgery 2005; 56:1243-1253
21.
Laske DW, Youle RJ, Oldfield EH. Tumor regression with regional distribution of the targeted toxin TF-CRM107 in patients with malignant brain tumors. Nat Med 1997; 3:1362-1368
22.
Rainov NG, Heidecke V. Long term survival in a patient with recurrent malignant glioma treated with intratumoral infusion of an IL4-targeted toxin (NBI-3001). J Neurooncol 2004; 66:197-201
23.
Kunwar S. Convection enhanced delivery of IL13-PE38QQR for treatment of recurrent malignant glioma: presentation of interim findings from ongoing phase 1 studies. Acta Neurochir Suppl 2003; 88:105-111
24.
Weber FW, Floeth F, Asher A, et al. Local convection enhanced delivery of IL4-Pseudomonas exotoxin (NBI-3001) for treatment of patients with recurrent malignant glioma. Acta Neurochir Suppl 2003; 88:93-103
25.
Weber F, Asher A, Bucholz R, et al. Safety, tolerability, and tumor response of IL4-Pseudomonas exotoxin (NBI-3001) in patients with recurrent malignant glioma. J Neurooncol 2003; 64:125-137
26.
Rand RW, Kreitman RJ, Patronas N, Varricchio F, Pastan I, Puri RK. Intratumoral administration of recombinant circularly permuted interleukin-4-Pseudomonas exotoxin in patients with high-grade glioma. Clin Cancer Res 2000; 6:2157-2165
27.
Lidar Z, Mardor Y, Jonas T, et al. Convection-enhanced delivery of paclitaxel for the treatment of recurrent malignant glioma: a phase I/II clinical study. J Neurosurg 2004; 100:472-479
28.
Sampson JH, Reardon DA, Friedman AH, et al. Sustained radiographic and clinical response in patient with bifrontal recurrent glioblastoma multiforme with intracerebral infusion of the recombinant targeted toxin TP-38: case study. Neuro-oncol 2005; 7:90-96
29.
Mardor Y, Roth Y, Lidar Z, et al. Monitoring response to convection-enhanced Taxol delivery in brain tumor patients using diffusion-weighted magnetic resonance imaging. Cancer Res 2001; 61:4971-4973
30.
Joshi BH, Plautz GE, Puri RK. Interleukin-13 receptor alpha chain: a novel tumor-associated transmembrane protein in primary explants of human malignant gliomas. Cancer Res 2000; 60:1168-1172
31.
Akabani G, Hawkins WG, Eckblade MB, Leichner PK. Patient-specific dosimetry using quantitative SPECT imaging and three-dimensional discrete Fourier transform convolution. J Nucl Med 1997; 38:308-314
32.
Akabani G, Reist CJ, Cokgor I, et al. Dosimetry of 131I-labeled 81C6 monoclonal antibody administered into surgically created resection cavities in patients with malignant brain tumors. J Nucl Med 1999; 40:631-638
33.
Gill SS, Patel NK, Hotton GR, et al. Direct brain infusion of glial cell line-derived neurotrophic factor in Parkinson disease. Nat Med 2003; 9:589-595
34.
Love S, Plaha P, Patel NK, Hotton GR, Brooks DJ, Gill SS. Glial cell line-derived neurotrophic factor induces neuronal sprouting in human brain. Nat Med 2005; 11:703-704
35.
Patel NK, Bunnage M, Plaha P, Svendsen CN, Heywood P, Gill SS. Intraputamenal infusion of glial cell line-derived neurotrophic factor in PD: a two-year outcome study. Ann Neurol 2005; 57:298-302

Information & Authors

Information

Published In

American Journal of Roentgenology
Pages: 703 - 709
PubMed: 17312057

History

Submitted: March 23, 2006
Accepted: June 7, 2006

Keywords

  1. brain
  2. convection-enhanced delivery
  3. drug delivery
  4. glioma
  5. MRI
  6. oncology

Authors

Affiliations

John H. Sampson
Department of Surgery, Duke University Medical Center, Durham, NC 27710.
Raghu Raghavan
Therataxis, Inc., Baltimore, MD 21218.
James M. Provenzale
Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC 27710.
David Croteau
NeoPharm, Inc., Waukegan, IL 60085.
David A. Reardon
Department of Surgery, Duke University Medical Center, Durham, NC 27710.
R. Edward Coleman
Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC 27710.
Inmaculada Rodríguez Ponce
BrainLAB AG, Feldkirchen, Germany.
Ira Pastan
Laboratory of Molecular Biology, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892.
Raj K. Puri
Division of Cellular and Gene Therapies, Center for Biologics Evaluation and Research, Food and Drug Administration, Rockville, MD 20852.
Christoph Pedain
BrainLAB AG, Feldkirchen, Germany.

Metrics & Citations

Metrics

Citations

Export Citations

To download the citation to this article, select your reference manager software.

Articles citing this article

View Options

View options

PDF

View PDF

PDF Download

Download PDF

Media

Figures

Other

Tables

Share

Share

Copy the content Link

Share on social media