AJR ARRS Membership
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sampson, J. H.
Right arrow Articles by Pedain, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sampson, J. H.
Right arrow Articles by Pedain, C.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?
DOI:10.2214/AJR.06.0428
AJR 2007; 188:703-709
© American Roentgen Ray Society

Original Research

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

John H. Sampson1, Raghu Raghavan2, James M. Provenzale3, David Croteau4, David A. Reardon1, R. Edward Coleman3, Inmaculada Rodríguez Ponce5, Ira Pastan6, Raj K. Puri7 and Christoph Pedain5

1 Department of Surgery, Duke University Medical Center, Durham, NC 27710.
2 Therataxis, Inc., Baltimore, MD 21218.
3 Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC 27710.
4 NeoPharm, Inc., Waukegan, IL 60085.
5 BrainLAB AG, Feldkirchen, Germany.
6 Laboratory of Molecular Biology, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892.
7 Division of Cellular and Gene Therapies, Center for Biologics Evaluation and Research, Food and Drug Administration, Rockville, MD 20852.

Received March 23, 2006; accepted after revision June 7, 2006.

 
Address correspondence to J. M. Provenzale (prove001{at}mc.duke.edu).

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.


Abstract
Top
 Abstract
Introduction
 Subjects and Methods
 Results
 Discussion
 References
 
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.

Keywords: brain • convection-enhanced delivery • drug delivery • glioma • MRI • oncology


Introduction
Top
 Abstract
 Introduction
Subjects and Methods
 Results
 Discussion
 References
 
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
Top
 Abstract
 Introduction
 Subjects and Methods
Results
 Discussion
 References
 
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.


Figure 1
View larger version (77K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 2
View larger version (64K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 3
View larger version (77K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 4
View larger version (68K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 5
View larger version (70K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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).

 

Figure 6
View larger version (61K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 7
View larger version (72K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 8
View larger version (49K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 9
View larger version (64K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 10
View larger version (64K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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).

 

Results
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
Discussion
 References
 
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).


Figure 11
View larger version (81K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 12
View larger version (91K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 13
View larger version (74K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 14
View larger version (69K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 15
View larger version (49K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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.

 

Figure 16
View larger version (38K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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).

 

View this table:
[in this window]
[in a new window]

  		TABLE 1: Prediction of Intraparenchymal Infusion with MRI

 

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.


Figure 17
View larger version (8K):
[in this window]
[in a new window]
[as a PowerPoint slide]
  		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).

 
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
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
References
 
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.

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.

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.

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.


References
Top
 Abstract
 Introduction
 Subjects and Methods
 Results
 Discussion
 References
 

  1. Groothuis DR. The blood-brain and blood-tumor barriers: a review of strategies for increasing drug delivery. Neuro-oncol2000; 2:45 -59[Abstract]
  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[Abstract/Free Full Text]
  3. Jain RK. Vascular and interstitial barriers to delivery of therapeutic agents in tumors. Cancer Metastasis Rev1990; 9:253 -266[CrossRef][Medline]
  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 Res2003; 9:5514 -5520[Abstract/Free Full Text]
  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[CrossRef][Medline]
  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 Physiol1999; 277:R1218 -R1229[Medline]
  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[Abstract/Free Full Text]
  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[Medline]
  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[Medline]
  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[Medline]
  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 Neurosurg1995; 82:1021 -1029[Medline]
  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 Res2000; 6:4148 -4153[Abstract/Free Full Text]
  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[CrossRef][Medline]
  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 Res2004; 64:6858 -6862[Abstract/Free Full Text]
  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 Res2004; 64:2572 -2579[Abstract/Free Full Text]
  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[Medline]
  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[Medline]
  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 Neurosurg2003; 99:893 -898[CrossRef][Medline]
  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[Medline]
  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. Neurosurgery2005; 56:1243 -1253[CrossRef][Medline]
  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[CrossRef][Medline]
  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 Neurooncol2004; 66:197 -201[CrossRef][Medline]
  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 Suppl2003; 88:105 -111[Medline]
  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 Suppl2003; 88:93 -103[Medline]
  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[CrossRef][Medline]
  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[Abstract/Free Full Text]
  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[Medline]
  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[Abstract]
  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[Abstract/Free Full Text]
  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[Abstract/Free Full Text]
  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 Med1997; 38:308 -314[Abstract/Free Full Text]
  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[Abstract/Free Full Text]
  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[CrossRef][Medline]
  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[CrossRef][Medline]
  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 Neurol2005; 57:298 -302[CrossRef][Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sampson, J. H.
Right arrow Articles by Pedain, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sampson, J. H.
Right arrow Articles by Pedain, C.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?