Longitudinal risk prediction for pediatric glioma with temporal deep learning.

Pediatric glioma recurrence can cause morbidity and mortality; however, recurrence pattern and severity are heterogeneous and challenging to predict with established clinical and genomic markers. Resultingly, almost all children undergo frequent, long-term, magnetic resonance (MR) brain surveillance regardless of individual recurrence risk. Deep learning analysis of longitudinal MR may be an effective approach for improving individualized recurrence prediction in gliomas and other cancers but has thus far been infeasible with current frameworks. Here, we propose a self-supervised, deep learning approach to longitudinal medical imaging analysis, temporal learning, that models the spatiotemporal information from a patient's current and prior brain MRs to predict future recurrence. We apply temporal learning to pediatric glioma surveillance imaging for 715 patients (3,994 scans) from four distinct clinical settings. We find that longitudinal imaging analysis with temporal learning improves recurrence prediction performance by up to 41% compared to traditional approaches, with improvements in performance in both low- and high-grade glioma. We find that recurrence prediction accuracy increases incrementally with the number of historical scans available per patient. Temporal deep learning may enable point-of-care decision-support for pediatric brain tumors and be adaptable more broadly to patients with other cancers and chronic diseases undergoing surveillance imaging.

Noninvasive Molecular Subtyping of Pediatric Low-Grade Glioma with Self-Supervised Transfer Learning.

Purpose To develop and externally test a scan-to-prediction deep learning pipeline for noninvasive, MRI-based BRAF mutational status classification for pediatric low-grade glioma. Materials and Methods This retrospective study included two pediatric low-grade glioma datasets with linked genomic and diagnostic T2-weighted MRI data of patients: Dana-Farber/Boston Children's Hospital (development dataset, n = 214 [113 (52.8%) male; 104 (48.6%) BRAF wild type, 60 (28.0%) BRAF fusion, and 50 (23.4%) BRAF V600E]) and the Children's Brain Tumor Network (external testing, n = 112 [55 (49.1%) male; 35 (31.2%) BRAF wild type, 60 (53.6%) BRAF fusion, and 17 (15.2%) BRAF V600E]). A deep learning pipeline was developed to classify BRAF mutational status (BRAF wild type vs BRAF fusion vs BRAF V600E) via a two-stage process: (a) three-dimensional tumor segmentation and extraction of axial tumor images and (b) section-wise, deep learning-based classification of mutational status. Knowledge-transfer and self-supervised approaches were investigated to prevent model overfitting, with a primary end point of the area under the receiver operating characteristic curve (AUC). To enhance model interpretability, a novel metric, center of mass distance, was developed to quantify the model attention around the tumor. Results A combination of transfer learning from a pretrained medical imaging-specific network and self-supervised label cross-training (TransferX) coupled with consensus logic yielded the highest classification performance with an AUC of 0.82 (95% CI: 0.72, 0.91), 0.87 (95% CI: 0.61, 0.97), and 0.85 (95% CI: 0.66, 0.95) for BRAF wild type, BRAF fusion, and BRAF V600E, respectively, on internal testing. On external testing, the pipeline yielded an AUC of 0.72 (95% CI: 0.64, 0.86), 0.78 (95% CI: 0.61, 0.89), and 0.72 (95% CI: 0.64, 0.88) for BRAF wild type, BRAF fusion, and BRAF V600E, respectively. Conclusion Transfer learning and self-supervised cross-training improved classification performance and generalizability for noninvasive pediatric low-grade glioma mutational status prediction in a limited data scenario. Keywords: Pediatrics, MRI, CNS, Brain/Brain Stem, Oncology, Feature Detection, Diagnosis, Supervised Learning, Transfer Learning, Convolutional Neural Network (CNN) Supplemental material is available for this article. © RSNA, 2024.

Automated temporalis muscle quantification and growth charts for children through adulthood.

Lean muscle mass (LMM) is an important aspect of human health. Temporalis muscle thickness is a promising LMM marker but has had limited utility due to its unknown normal growth trajectory and reference ranges and lack of standardized measurement. Here, we develop an automated deep learning pipeline to accurately measure temporalis muscle thickness (iTMT) from routine brain magnetic resonance imaging (MRI). We apply iTMT to 23,876 MRIs of healthy subjects, ages 4 through 35, and generate sex-specific iTMT normal growth charts with percentiles. We find that iTMT was associated with specific physiologic traits, including caloric intake, physical activity, sex hormone levels, and presence of malignancy. We validate iTMT across multiple demographic groups and in children with brain tumors and demonstrate feasibility for individualized longitudinal monitoring. The iTMT pipeline provides unprecedented insights into temporalis muscle growth during human development and enables the use of LMM tracking to inform clinical decision-making.

Summary of the 2022 ACR Intersociety Meeting.

The ACR Intersociety Committee meeting of 2022 (ISC-2022) was convened around the theme of "Recovering From The Great Resignation, Moral Injury and Other Stressors: Rebuilding Radiology for a Robust Future." Representatives from 29 radiology organizations, including all radiology subspecialties, radiation oncology, and medical physics, as well as academic and private practice radiologists, met for 3 days in early August in Park City, Utah, to search for solutions to the most pressing problems facing the specialty of radiology in 2022. Of these, the mismatch between the clinical workload and the available radiologist workforce was foremost-as many other identifiable problems flowed downstream from this, including high job turnover, lack of time for teaching and research, radiologist burnout, and moral injury.

Imaging of pediatric spine and spinal cord tumors: A COG Diagnostic Imaging Committee/SPR Oncology Committee/ASPNR White Paper.

Childhood spinal tumors are rare. Tumors can involve the spinal cord, the meninges, bony spine, and the paraspinal tissue. Optimized imaging should be utilized to evaluate tumors arising from specific spinal compartments. This paper provides consensus-based recommendations for optimized imaging of tumors arising from specific spinal compartments at diagnosis, follow-up during and after therapy, and response assessment.

Deep learning-based automatic tumor burden assessment of pediatric high-grade gliomas, medulloblastomas, and other leptomeningeal seeding tumors.

BACKGROUND: Longitudinal measurement of tumor burden with magnetic resonance imaging (MRI) is an essential component of response assessment in pediatric brain tumors. We developed a fully automated pipeline for the segmentation of tumors in pediatric high-grade gliomas, medulloblastomas, and leptomeningeal seeding tumors. We further developed an algorithm for automatic 2D and volumetric size measurement of tumors. METHODS: The preoperative and postoperative cohorts were randomly split into training and testing sets in a 4:1 ratio. A 3D U-Net neural network was trained to automatically segment the tumor on T1 contrast-enhanced and T2/FLAIR images. The product of the maximum bidimensional diameters according to the RAPNO (Response Assessment in Pediatric Neuro-Oncology) criteria (AutoRAPNO) was determined. Performance was compared to that of 2 expert human raters who performed assessments independently. Volumetric measurements of predicted and expert segmentations were computationally derived and compared. RESULTS: A total of 794 preoperative MRIs from 794 patients and 1003 postoperative MRIs from 122 patients were included. There was excellent agreement of volumes between preoperative and postoperative predicted and manual segmentations, with intraclass correlation coefficients (ICCs) of 0.912 and 0.960 for the 2 preoperative and 0.947 and 0.896 for the 2 postoperative models. There was high agreement between AutoRAPNO scores on predicted segmentations and manually calculated scores based on manual segmentations (Rater 2 ICC = 0.909; Rater 3 ICC = 0.851). Lastly, the performance of AutoRAPNO was superior in repeatability to that of human raters for MRIs with multiple lesions. CONCLUSIONS: Our automated deep learning pipeline demonstrates potential utility for response assessment in pediatric brain tumors. The tool should be further validated in prospective studies.

Response assessment in paediatric high-grade glioma: recommendations from the Response Assessment in Pediatric Neuro-Oncology (RAPNO) working group.

Response criteria for paediatric high-grade glioma vary historically and across different cooperative groups. The Response Assessment in Neuro-Oncology working group developed response criteria for adult high-grade glioma, but these were not created to meet the unique challenges in children with the disease. The Response Assessment in Pediatric Neuro-Oncology (RAPNO) working group, consisting of an international panel of paediatric and adult neuro-oncologists, clinicians, radiologists, radiation oncologists, and neurosurgeons, was established to address issues and unique challenges in assessing response in children with CNS tumours. We established a subcommittee to develop response assessment criteria for paediatric high-grade glioma. Current practice and literature were reviewed to identify major challenges in assessing the response of paediatric high-grade gliomas to various treatments. For areas in which scientific investigation was scarce, consensus was reached through an iterative process. RAPNO response assessment recommendations include the use of MRI of the brain and the spine, assessment of clinical status, and the use of corticosteroids or antiangiogenics. Imaging standards for brain and spine are defined. Compared with the recommendations for the management of adult high-grade glioma, for paediatrics there is inclusion of diffusion-weighted imaging and a higher reliance on T2-weighted fluid-attenuated inversion recovery. Consensus recommendations and response definitions have been established and, similar to other RAPNO recommendations, prospective validation in clinical trials is warranted.

Phase II study of peginterferon alpha-2b for patients with unresectable or recurrent craniopharyngiomas: a Pediatric Brain Tumor Consortium report.

BACKGROUND: Craniopharyngiomas account for approximately 1.2-4% of all CNS tumors. They are typically treated with a combination of surgical resection and focal radiotherapy. Unfortunately, treatment can lead to permanent deleterious effects on behavior, learning, and endocrine function. METHODS: The Pediatric Brain Tumor Consortium performed a multicenter phase 2 study in children and young adults with unresectable or recurrent craniopharyngioma (PBTC-039). Between December 2013 and November 2017, nineteen patients (median age at enrollment, 13.1 y; range, 2-25 y) were enrolled in one of 2 strata: patients previously treated with surgery alone (stratum 1) or who received radiation (stratum 2). RESULTS: Eighteen eligible patients (8 male, 10 female) were treated with weekly subcutaneous pegylated interferon alpha-2b for up to 18 courses (108 wk). Therapy was well tolerated with no grade 4 or 5 toxicities. 2 of the 7 eligible patients (28.6%) in stratum 1 had a partial response, but only one response was sustained for more than 3 months. None of the 11 stratum 2 patients had an objective radiographic response, although median progression-free survival was 19.5 months. CONCLUSIONS: Pegylated interferon alpha-2b treatment, in lieu of or following radiotherapy, was well tolerated in children and young adults with recurrent craniopharyngiomas. Although objective responses were limited, progression-free survival results are encouraging, warranting further studies.

Brainstem Injury in Pediatric Patients Receiving Posterior Fossa Photon Radiation.

PURPOSE: Brainstem necrosis is a rare, but dreaded complication of radiation therapy; however, data on the incidence of brainstem injury for tumors involving the posterior fossa in photon-treated patient cohorts are still needed. METHODS AND MATERIALS: Clinical characteristics and dosimetric parameters were recorded for 107 pediatric patients who received photon radiation for posterior fossa tumors without brainstem involvement from 2000 to 2016. Patients were excluded if they received a prescription dose <50.4 Gy, a brainstem maximum dose <50.4 Gy, or had fewer than 2 magnetic resonance imaging scans within 18 months after radiation. Post-radiation therapy magnetic resonance imaging findings were recorded, and brainstem toxicity was graded using National Cancer Institute Common Terminology Criteria for Adverse Events, version 5. RESULTS: The most common histologies were medulloblastoma (61.7%) and ependymoma (15.9%), and median age at diagnosis was 8.3 years (range, 0.8-20.7). Sixty-seven patients (62.6%) received craniospinal irradiation (median, 23.4 Gy; range, 18.0-39.6) as a component of their radiation therapy, and 39.3% and 40.2% of patients received an additional involved field or whole posterior fossa boost, respectively. Median prescribed dose was 55.8 Gy (range, 50.4-60.0). Median clinical and imaging follow-up were 4.7 years (range, 0.1-17.5) and 4.2 years (range, 0.1-17.3), respectively. No grade ≥2 toxicities were observed. The incidence of grade 1 brainstem necrosis was 1.9% (2 of 107). These patients were by definition asymptomatic and experienced resolution of imaging abnormality after 5.3 months and 2.1 years, respectively. CONCLUSIONS: Risk of brainstem necrosis was minimal in this multi-institutional study of pediatric patients treated with photon radiation therapy for tumors involving the posterior fossa with no cases of symptomatic brainstem injury, suggesting that brainstem injury risk is minimal in patients treated with photon therapy.

Response assessment in medulloblastoma and leptomeningeal seeding tumors: recommendations from the Response Assessment in Pediatric Neuro-Oncology committee.

Lack of standard response criteria in clinical trials for medulloblastoma and other seeding tumors complicates assessment of therapeutic efficacy and comparisons across studies. An international working group was established to develop consensus recommendations for response assessment. The aim is that these recommendations be prospectively evaluated in clinical trials, with the goal of achieving more reliable risk stratification and uniformity across clinical trials. Current practices and literature review were performed to identify major confounding issues and justify subsequently developed recommendations; in areas lacking scientific investigations, recommendations were based on experience of committee members and consensus was reached after discussion. Recommendations apply to both adult and pediatric patients with medulloblastoma and other seeding tumors. Response should be assessed using MR imaging (brain and spine), CSF cytology, and neurologic examination. Clinical imaging standards with minimum mandatory sequence acquisition that optimizes detection of leptomeningeal metastases are defined. We recommend central review prior to inclusion in treatment cohorts to ensure appropriate risk stratification and cohort inclusion. Consensus recommendations and response definitions for patients with medulloblastomas and other seeding tumors have been established; as with other Response Assessment in Neuro-Oncology recommendations, these need to now be prospectively validated in clinical trials.

Pediatric brain tumors.

Among all causes of death in children from solid tumors, pediatric brain tumors are the most common. This article includes an overview of a subset of infratentorial and supratentorial tumors with a focus on tumor imaging features and molecular advances and treatments of these tumors. Key to understanding the imaging features of brain tumors is a firm grasp of other disease processes that can mimic tumor on imaging. We also review imaging features of a common subset of tumor mimics.

18F-FDG PET and MR imaging associations across a spectrum of pediatric brain tumors: a report from the pediatric brain tumor consortium.

UNLABELLED: The purpose of this study was to describe (18)F-FDG uptake across a spectrum of pediatric brain tumors and correlate (18)F-FDG PET with MR imaging variables, progression-free survival (PFS), and overall survival (OS). METHODS: A retrospective analysis was conducted of children enrolled in phase I/II clinical trials through the Pediatric Brain Tumor Consortium from August 2000 to June 2010. PET variables were summarized within diagnostic categories using descriptive statistics. Associations of PET with MR imaging variables and PFS and OS by tumor types were evaluated. RESULTS: Baseline (18)F-FDG PET was available in 203 children; 66 had newly diagnosed brain tumors, and 137 had recurrent/refractory brain tumors before enrolling in a Pediatric Brain Tumor Consortium trial. MR imaging was performed within 2 wk of PET and before therapy in all cases. The (18)F-FDG uptake pattern and MR imaging contrast enhancement (CE) varied by tumor type. On average, glioblastoma multiforme and medulloblastoma had uniform, intense uptake throughout the tumor, whereas brain stem gliomas (BSGs) had low uptake in less than 50% of the tumor and ependymoma had low uptake throughout the tumor. For newly diagnosed BSG, correlation of (18)F-FDG uptake with CE portended reduced OS (P = 0.032); in refractory/recurrent BSG, lack of correlation between (18)F-FDG uptake and CE suggested decreased PFS (P = 0.023). In newly diagnosed BSG for which more than 50% of the tumor had (18)F-FDG uptake, there was a suggestion of lower apparent diffusion coefficient (P = 0.061) and decreased PFS (P = 0.065). CONCLUSION: (18)F-FDG PET and MR imaging showed a spectrum of patterns depending on tumor type. In newly diagnosed BSG, the correlation of (18)F-FDG uptake and CE suggested decreased OS, likely related to more aggressive disease. When more than 50% of the tumor had (18)F-FDG uptake, the apparent diffusion coefficient was lower, consistent with increased cellularity. In refractory/recurrent BSG, poor correlation between (18)F-FDG uptake and CE was associated with decreased PFS, which may reflect concurrent tissue breakdown at sites of treated disease and development of new sites of (18)F-FDG-avid malignancy.

Infantile postoperative encephalopathy: perioperative factors as a cause for concern.

We report on 6 infants who underwent elective surgery and developed postoperative encephalopathy, which had features most consistent with intraoperative cerebral hypoperfusion. All infants were <48 weeks' postmenstrual age and underwent procedures lasting 120 to 185 minutes. Intraoperative records revealed that most of the measured systolic blood pressure (SBP) values were <60 mm Hg (the threshold for hypotension in awake infants according to the Pediatric Advanced Life Support guidelines) but that only 11% of the measured SBP values were <1 SD of the mean definition of hypotension (<45 mm Hg) as reported in a survey of members of the Society for Pediatric Anesthesia in 2009. Four infants also exhibited prolonged periods of mild hypocapnia (<35 mm Hg). One infant did not receive intraoperative dextrose. All infants developed new-onset seizures within 25 hours of administration of the anesthetic, with a predominant cerebral pathology of supratentorial watershed infarction in the border zone between the anterior, middle, and posterior cerebral arteries. Follow-up of these infants found that 1 died, 1 had profound developmental delays, 1 had minor motor delays, 2 were normal, and 1 was lost to follow-up. Although the precise cause of encephalopathy cannot be determined, it is important to consider the role that SBP hypotension (as well as hypoglycemia, hyperthermia, hyperoxia, and hypocapnia) plays during general anesthesia in young infants in the development of infantile postoperative encephalopathy. Our observations highlight the lack of evidence-based recommendations for the lower limits of adequate SBP and end-tidal carbon dioxide in anesthetized infants.

Recommendations for imaging tumor response in neurofibromatosis clinical trials.

OBJECTIVE: Neurofibromatosis (NF)-related benign tumors such as plexiform neurofibromas (PN) and vestibular schwannomas (VS) can cause substantial morbidity. Clinical trials directed at these tumors have become available. Due to differences in disease manifestations and the natural history of NF-related tumors, response criteria used for solid cancers (1-dimensional/RECIST [Response Evaluation Criteria in Solid Tumors] and bidimensional/World Health Organization) have limited applicability. No standardized response criteria for benign NF tumors exist. The goal of the Tumor Measurement Working Group of the REiNS (Response Evaluation in Neurofibromatosis and Schwannomatosis) committee is to propose consensus guidelines for the evaluation of imaging response in clinical trials for NF tumors. METHODS: Currently used imaging endpoints, designs of NF clinical trials, and knowledge of the natural history of NF-related tumors, in particular PN and VS, were reviewed. Consensus recommendations for response evaluation for future studies were developed based on this review and the expertise of group members. RESULTS: MRI with volumetric analysis is recommended to sensitively and reproducibly evaluate changes in tumor size in clinical trials. Volumetric analysis requires adherence to specific imaging recommendations. A 20% volume change was chosen to indicate a decrease or increase in tumor size. Use of these criteria in future trials will enable meaningful comparison of results across studies. CONCLUSIONS: The proposed imaging response evaluation guidelines, along with validated clinical outcome measures, will maximize the ability to identify potentially active agents for patients with NF and benign tumors.

Exploratory evaluation of MR permeability with 18F-FDG PET mapping in pediatric brain tumors: a report from the Pediatric Brain Tumor Consortium.

UNLABELLED: The purpose of this study was to develop a method of registering (18)F-FDG PET with MR permeability images for investigating the correlation of (18)F-FDG uptake, permeability, and cerebral blood volume (CBV) in children with pediatric brain tumors and their relationship with outcome. METHODS: Twenty-four children with brain tumors in a phase II study of bevacizumab and irinotecan underwent brain MR and (18)F-FDG PET within 2 wk. Tumor types included supratentorial high-grade astrocytoma (n = 7), low-grade glioma (n = 9), brain stem glioma (n = 4), medulloblastoma (n = 2), and ependymoma (n = 2). There were 33 cases (pretreatment only [n = 12], posttreatment only [n = 3], and both pretreatment [n = 9] and posttreatment [n = 9]). (18)F-FDG PET images were registered to MR images from the last time point of the T1 perfusion time series using mutual information. Three-dimensional regions of interest (ROIs) drawn on permeability images were automatically transferred to registered PET images. The quality of ROI registration was graded (1, excellent; 2, very good; 3, good; 4, fair; and 5, poor) by 3 independent experts. Spearman rank correlations were used to assess correlation of maximum tumor permeability (Kps(max)), maximum CBV (CBV(max)), and maximum (18)F-FDG uptake normalized to white matter (T/W(max)). Cox proportional hazards models were used to investigate associations of these parameters with progression-free survival (PFS). RESULTS: The quality of ROI registration between PET and MR was good to excellent in 31 of 33 cases. There was no correlation of baseline Kps(max) with CBV(max) (Spearman rank correlation = 0.018 [P = 0.94]) or T/W(max) (Spearman rank correlation = 0.07 [P = 0.76]). Baseline CBV(max) was correlated with T/W(max) (Spearman rank correlation = 0.47 [P = 0.036]). Baseline Kps(max), CBV(max), and T/W(max) were not significantly associated with PFS (P = 0.42, hazard ratio [HR] = 0.97, 95% confidence interval [CI] = 0.90-1.045, and number of events [n(events)] = 15 for Kps(max); P = 0.41, HR = 0.989, 95% CI = 0.963-1.015, and n(events) = 14 for CBV(max); and P = 0.17, HR = 1.49, 95% CI = 0.856-2.378, and n(events) = 15 for T/W(max)). CONCLUSION: (18)F-FDG PET and MR permeability images were successfully registered and compared across a spectrum of pediatric brain tumors. The lack of correlation between metabolism and permeability may be expected because these parameters characterize different molecular processes. The correlation of CBV and tumor metabolism may be related to an association with tumor grade. More patients are needed for a covariate analysis of these parameters and PFS by tumor histology.

Challenges with defining response to antitumor agents in pediatric neuro-oncology: a report from the response assessment in pediatric neuro-oncology (RAPNO) working group.

Criteria for new drug approval include demonstration of efficacy. In neuro-oncology, this is determined radiographically utilizing tumor measurements on MRI scans. Limitations of this method have been identified where drug activity is not reflected in decreased tumor size. The RANO (Response Assessment in Neuro-Oncology) working group was established to address limitations in defining endpoints for clinical trials in adult neuro-oncology and to develop standardized response criteria. RAPNO was subsequently established to address unique issues in pediatric neuro-oncology. The aim of this paper is to delineate response criteria issues in pediatric clinical trials as a basis for subsequent recommendations.

Phase I trial of single-dose temozolomide and continuous administration of o6-benzylguanine in children with brain tumors: a pediatric brain tumor consortium report.

PURPOSE: To estimate the maximum tolerated dose (MTD) and dose-limiting toxicity (DLT) of escalating doses of temozolomide combined with O(6)-benzylguanine in patients < or =21 years with recurrent brain tumors. EXPERIMENTAL DESIGN: Treatment strata consisted of patients who had previously received no or local radiotherapy (Str1) and patients who had undergone craniospinal radiotherapy or myeloablative chemotherapy (Str2). One-hour i.v. administration of O(6)-benzylguanine at 120 mg/m(2) was followed by 48-h continuous infusion at 30 mg/m(2)/day. Single-dose temozolomide at five dosage levels (267, 355, 472, 628, and 835 mg/m(2)) was given at least 6 h after completion of O(6)-benzylguanine bolus. Treatment was repeated after recovery from toxicities at least 4 weeks apart for a maximum of 12 courses. Dose escalation followed the modified continual reassessment method. Pharmacokinetic analyses of temozolomide and 5-triazeno imidazole carboxamide (MTIC) were done in 28 patients. RESULTS: A total of 44 and 26 eligible patients were enrolled on Str1 and Str2, respectively. Median age at study entry in each stratum was 8.6 and 11.3 years, respectively. Predominant diagnoses were high-grade/brainstem glioma in Str1 and medulloblastoma in Str2. Whereas the estimated MTDs of temozolomide for Str1 and Str2 were 562 and 407 mg/m(2), respectively, the doses recommended for phase II investigations are 472 and 355 mg/m(2), respectively. DLTs were predominantly neutropenia and thrombocytopenia. Three patients with gliomas experienced centrally confirmed partial responses to therapy. Four patients completed all planned therapy. Temozolomide and MTIC exposures were statistically associated with temozolomide dosage. CONCLUSIONS: The current schedule of temozolomide and O(6)-benzylguanine is safe and showed modest activity against recurrent brain tumors in children.

Perfusion MRI of U87 brain tumors in a mouse model.

Continuous arterial spin labeling (CASL) was used to obtain an index of cerebral blood flow (ICBF) in the normal mouse brain and in an orthotopic mouse model of human U87 high-grade glioma at 8.5 T. Under the assumption of a constant tissue:blood partition coefficient for water in different tissues, the mean ICBF (n = 14) was found to be 50 +/- 9 mL/100g/min for tumor core and 209 +/- 11 mL/100g/min for normal tissue. The apparent T(1) (T(1app)) was 2.01 +/- 0.06 sec for tumor core and 1.66 +/- 0.03 sec for normal tissue. The ICBF and the T(1app) values were significantly different (P < 0.001) between these two regions. The detailed changes of ICBF and T(1app) in the transition from the tumor core through the tumor periphery to surrounding tissue were studied. Immunohistochemistry indicated that tumor vascularity was not uniform, with microvessel density highest in normal brain and the tissue surrounding the tumor and lowest in the tumor core. The large difference in ICBF between the tumor core and normal tissue suggests that this index might be useful for the assessment of the efficacy of antiangiogenic therapy.