Magnetic resonance of brain tumors: considerations of imaging contrast on the basis of relaxation measurements

Quantitative Synthetic Magnetic Resonance Imaging for Brain Metastases: A Feasibility Study

Simple Summary

This preliminary study aims to characterize brain metastases (BM) using T1 and T2 maps generated from newer, rapid, synthetic MRI (MAGnetic resonance image Compilation; MAGiC) in a clinical setting. In addition, synthetic MR could provide contrast images analogous to standard T1- and T2-weighted images. The reproducibility and repeatability of this method have been previously established for brain imaging. This study reports and analyzes the quantitative T1 and T2 values for 11 BM patients (17 BM lesions) with a total of 82 regions of interest (ROIs) delineated by an experienced neuroradiologist. The initial results, which need to be further validated in a larger patient cohort, demonstrated the ability of T1 and T2 metric values to characterize BMs and normal-appearing brain tissues. The T1 and T2 metrics could be potential surrogate biomarkers for BM free water content (cellularity) and tumor morphology, respectively.

Abstract

The present preliminary study aims to characterize brain metastases (BM) using T1 and T2 maps generated from newer, rapid, synthetic MRI (MAGnetic resonance image Compilation; MAGiC) in a clinical setting. We acquired synthetic MRI data from 11 BM patients on a 3T scanner. A multiple-dynamic multiple-echo (MDME) sequence was used for data acquisition and synthetic image reconstruction, including post-processing. MDME is a multi-contrast sequence that enables absolute quantification of physical tissue properties, including T1 and T2, independent of the scanner settings. In total, 82 regions of interest (ROIs) were analyzed, which were obtained from both normal-appearing brain tissue and BM lesions. The mean values obtained from the 48 normal-appearing brain tissue regions and 34 ROIs of BM lesions (T1 and T2) were analyzed using standard statistical methods. The mean T1 and T2 values were 1143 ms and 78 ms, respectively, for normal-appearing gray matter, 701 ms and 64 ms for white matter, and 4206 ms and 390 ms for cerebrospinal fluid. For untreated BMs, the mean T1 and T2 values were 1868 ms and 100 ms, respectively, and 2211 ms and 114 ms for the treated group. The quantitative T1 and T2 values generated from synthetic MRI can characterize BM and normal-appearing brain tissues.

Value of quantitative magnetic resonance imaging T1-relaxometry in predicting contrast-enhancement in glioblastoma patients

SUMMARIZING THE IMPORTANCE OF THE STUDY

The repetitive usage of gadolinium-based contrast agents (GBCA) is critical for magnetic resonance imaging (MRI) evaluation of tumor burden in glioblastoma patients. It is also a crucial tool for determination of radiographical response to treatment. GBCA injection, however, comes with a 2.4% rate of adverse events including life-threatening conditions such as nephrogenic systemic fibrosis (NSF). Moreover, GBCA have been shown to be deposited in brain tissue of patients even with an intact blood-brain barrier (BBB). The present study explores quantitative T1 relaxometry as an alternative non-invasive imaging technique detection of tumor burden and determination of radiographical response. This technique exploits specific properties of brain tissue with impaired BBB. With a sensitivity and specificity as high as 86% and 80%, respectively, quantitative T1-relaxometry allows for detecting contrast-enhancing areas without the use of GBCA. This method could make it unnecessary to subject patients to the risk of adverse events associated with the use of GBCA. Nonetheless, a large-scale analysis is needed to confirm our findings.

Background

Gadolinium-based contrast agents (GBCA) are crucial for magnetic resonance imaging (MRI)-based evaluation of tumor burden in glioblastoma (GBM). Serious adverse events of GBCA, even though uncommon, and gadolinium deposition in brain tissue could be avoided by novel imaging techniques not requiring GBCA. Altered tissue composition in areas with impaired blood-brain-barrier also alters the quantified T1 relaxation time (qT1), so that qT1 analysis could replace GBCA-based MRI for the analysis of tumor burden and response.

Methods

As a part of a prospective pilot MRI-relaxometry trial, patients with newly-diagnosed GBM who relapsed under standard radiochemotherapy were selected for this study. At recurrence, subtraction of qT1 maps pre and post-GBCA application (ΔqT1 maps) was used to determine areas of contrast-enhancement. With the contrast-enhancement on ΔqT1 maps as reference, ROC analysis served to detect an optimal qT1 cut-off on qT1 maps prior to GBCA to distinguish between contrast-enhancing tissue and its surroundings.

Results

Ten patients were included. A qT1 value >2051ms predicted contrast-enhancing tumor tissue with a sensitivity of 86% and specificity of 80% (AUC, 0.92; p<0.0001). Interestingly, qT1 prolongation >2051 ms that did not overlap with contrast-enhancing area transformed into contrast-enhancement later on (n=4).

Conclusion

T1-relaxometry may be a useful technique to assess tissue properties equivalent to contrast-enhancement without the need for GBCA application. It may also provide information on sites with future tumor progression. Nonetheless, large-scale studies are needed to confirm these findings.

Improved visibility of brain tumors in synthetic MP-RAGE anatomies with pure T1 weighting

Conventional MRI for brain tumor diagnosis employs T2 -weighted and contrast-enhanced T1 -weighted sequences. Non-enhanced T1 -weighted images provide improved anatomical details for precise tumor location, but reduced tumor-to-background contrast as elevated T1 and proton density (PD) values in tumor tissue affect the signal inversely. Radiofrequency (RF) coil inhomogeneities may further mask tumor and edema outlines. To overcome this problem, the aims of this work were to employ quantitative MRI techniques to create purely T1 -weighted synthetic anatomies which can be expected to yield improved tissue and tumor-to-background contrasts, to compare the quality of conventional and synthetic anatomies, and to investigate optical contrast and visibility of brain tumors and edema in synthetic anatomies. Conventional magnetization-prepared rapid acquisition of gradient echoes (MP-RAGE) anatomies and maps of T1 , PD and RF coil profiles were acquired in comparable and clinically feasible times. Three synthetic MP-RAGE anatomies (PD T1 weighting both with and without RF bias; pure T1 weighting) were calculated for healthy subjects and 32 patients with brain tumors. In healthy subjects, the PD T1 -weighted synthetic anatomies with RF bias precisely matched the conventional anatomies, yielding high signal-to-noise (SNR) and contrast-to-noise (CNR) ratios. Pure T1 weighting yielded lower SNR, but high CNR, because of increased optical contrasts. In patients with brain tumors, synthetic anatomies with pure T1 weighting yielded significant increases in optical contrast and improved visibility of tumor and edema in comparison with anatomies reflecting conventional T1 contrasts. In summary, the optimized purely T1 -weighted synthetic anatomy with an isotropic resolution of 1 mm, as proposed in this work, considerably enhances optical contrast and visibility of brain tumors and edema.

Maturational and aging effects on human brain apparent transverse relaxation

The goal of this study was to address the need for comprehensive reference data regarding maturational and aging effects on regional transverse relaxation rates (R₂) of the brain in normal humans. Regional R₂s were measured in twenty-five brain structures from a sample of seventy-seven normal volunteers 9 to 85 years of age. The relationships between regional R₂ and age were determined using generalized additive models, without the constraint of a specified a priori model. Data analysis demonstrated that the brain tissue R₂-age correlations followed various time courses with both linear and non-linear characteristics depending on the particular brain structure. Most anatomical structures studied exhibited non-linear characteristics, including the amygdala, hippocampus, thalamus, globus pallidus, putamen, caudate nucleus, red nucleus, substantia nigra, orbitofrontal white matter and temporal white matter. Linear trends were detected in occipital white matter and in the genu of corpus callosum. These results indicate the complexity of age-related R₂ changes in the brain while providing normative reference data that can be utilized in clinical examinations and studies utilizing quantitative transverse relaxation.

Fast radio-frequency enforced steady state (FRESS) spin echo MRI for quantitative T2 mapping: minimizing the apparent repetition time (TR) dependence for fast T2 measurement

Transverse relaxation time (T(2)) is a basic but very informative MRI parameter, widely used in imaging to examine a host of diseases, including multiple sclerosis, stroke, and tumor. However, short repetition time (TR) is often used to minimize scan time, which may introduce non-negligible errors in T(2) measurement. Specifically, due to the use of refocusing pulse, the steady state magnetization depends not only on TR but also on the TE. Hence, if the TE dependence is not properly accounted for, it may be mistaken as T(2)-induced signal attenuation, leading to non-negligible T(2) underestimation. Our study proposed a fast radio-frequency enforced steady state (FRESS) spin echo (SE) MRI sequence, which saturates the magnetization after the echo and ensures a TE-independent steady state. The proposed FRESS-SE MRI was evaluated with numerical simulation, implemented with echo planar imaging readout, and validated by both phantom and in vivo experiments. In summary, FRESS-SE T(2) MRI technique was developed for fast and accurate T(2) imaging, suitable for in vivo applications.

T2 detection of tumor invasion within segmented components of glioblastoma multiforme

PURPOSE

To use T2-weighted images to detect tumor invasion when comparing normal individuals to groups of gliomablastoma multiforme (GBM) patients with varying levels of CXCR4, a chemokine receptor that promotes tumor migration.

MATERIALS AND METHODS

T2-weighted images were acquired preoperatively in 22 treatment-naïve GBM patients. Two groups were formed based on the expression levels of CXCR4. A third group of normal volunteers was used for comparison. Each image was segmented to obtain four different clusters for tissue types identified as white matter, basal ganglia, gray matter/edema and cerebrospinal fluid (CSF)/tumor. Signal intensity histograms were formed for each cluster and compared between groups.

RESULTS

In every cluster the GBM groups displayed significantly higher standard deviations of intensity distributions when compared to normal subjects. Significant differences in skewness were found between normal subjects and GBM patients in the white matter, basal ganglia, and CSF/tumor. Further, when the two groups of GBM patients were compared the CXCR4-high group was found to have a significant shift in the median intensity values in the cluster containing gray matter and peritumoral edema.

CONCLUSION

T2 signal intensity histograms in normal subjects differ significantly from those obtained from GBM groups, suggesting widespread dissemination of disease.

MRI texture analysis on texture test objects, normal brain and intracranial tumors

Texture analysis was performed in three different MRI units on T1 and T2-weighted MR images from 10 healthy volunteers and 63 patients with histologically confirmed intracranial tumors. The goal of this study was a multicenter evaluation of the usefulness of this quantitative approach for the characterization of healthy and pathologic human brain tissues (white matter, gray matter, cerebrospinal fluid, tumors and edema). Each selected brain region of interest was characterized with both its mean gray level values and several texture parameters. Multivariate statistical analyses were then applied in order to discriminate each brain tissue type represented by its own set of texture parameters. Texture analysis was previously performed on test objects to evaluate the method dependence on acquisition parameters and consequently the interest of a multicenter evaluation. Even obtained on different sites with their own acquisition routine protocol, MR brain images contain textural features that can reveal discriminant factors for tissue classification and image segmentation. It can also offer additional information in case of undetermined diagnosis or to develop a more accurate tumor grading.

Magnetic resonance imaging of human melanoma xenografts in vivo: proton spin-lattice and spin-spin relaxation times versus fractional tumour water content and fraction of necrotic tumour tissue

Proton nuclear magnetic resonance (1H-nmr) imaging is used routinely in clinical oncology to provide macroscopic anatomical information, whereas its potential to provide physiological information about tumours is not well explored. To evaluate the potential usefulness of 1H-nmr imaging in the prediction of tumour treatment resistance caused by unfavourable microenvironmental conditions, possible correlations between proton spin-lattice and spin-spin relaxation times (T1 and T2) and physiological parameters of the tumour microenvironment were investigated. Tumours from six human melanoma xenograft lines were included in the study. 1H-nmr imaging was performed at 1.5 T using spin-echo pulse sequences. T1- and T2-distributions were generated from the images. Fractional tumour water content and the fraction of necrotic tumour tissue were measured immediately after 1H-nmr imaging. Significant correlations across tumour lines were found for T1 and T2 versus fractional tumour water content (p < 0.001) as well as for T1 and T2 versus fraction of necrotic tumour tissue (p < 0.05). Tumours with high fractional water contents had high values of T1 and T2, probably caused by free water in the tumour interstitium. Fractional water content is correlated to interstitial fluid pressure in tumours, high interstitial fluid pressure being indicative of high vascular resistance. Tumours with high fractional water contents are thus expected to show regions with radiobiologically hypoxic cells as well as poor intravascular and interstitial transport of many therapeutic agents. T1 and T2 decreased with increasing fraction of necrotic tumour tissue, perhaps because complexed paramagnetic ions were released during development of necrosis. Viable tumour cells adjacent to necrotic regions are usually chronically hypoxic. Tumours with high fractions of necrotic tissue are thus expected to contain significant proportions of radiobiologically hypoxic cells. Consequently, quantitative 1H-nmr imaging has the potential to be developed as an efficient clinical tool in prediction of tumour treatment resistance caused by hypoxia and/or transport barriers for therapeutic agents. However, much work remains to be done before this potential can be adequately evaluated. One problem is that high fractional tumour water contents result in longer T1 and T2 whereas high fractions of necrotic tumour tissue result in shorter T1 and T2; i.e. the two parameters which are indicative of treatment resistance contribute in opposite directions. Another problem is that the correlations for T1 and T2 versus fraction of necrotic tumour tissue are not particularly strong.

Variation of the magnetic relaxation rate 1/T1 of water protons with magnetic field strength (NMRD profile) of untreated, non-calcified, human astrocytomas: correlation with histology and solids content

The magnetic relaxation rate 1/T1 of tissue water protons was measured over a wide range of magnetic field strengths (NMRD profile) for 92 fresh surgical specimens of astrocytomas to search for correlations of 1/T1 with tumor histology, as determined by light microscopy, and to assess the diagnostic potential of NMRD profiles for grading astrocytomas. A third goal was to elucidate the molecular determinants of 1/T1. Each specimen was histologically graded and inspected for evidence of mineral deposits (Ca, Fe); its dry weight was determined and expressed in % of original wet weight. To minimize variability not directly related to tumor grade, this initial report is limited to NMRD profiles of 47 non-calcified, non-hemorrhagic, untreated astrocytomas. For these, the mean value of 1/T1 at very low magnetic field strengths was found to increase with increasing grade of malignancy; no clear correlation could be demonstrated at high fields where most imaging is done. The spread of 1/T1 for different grades of malignancy is large, however, and the overlap significant, even at the lowest field, so that astrocytomas can not be graded by NMRD profiles alone. Average 1/T1 and average dry weight increase with grade of malignancy; but the variability of 1/T1 among specimens of the same dry weight is large, indicating that at least one other cellular parameter, not variable in normal tissue, influences 1/T1 strongly. We hypothesize that this parameter reflects changes at the molecular level in size distribution, mobility, or intermolecular interaction of cytoplasmic proteins. Which specific changes are induced by malignant transformation in astrocytomas remains to be investigated.

The dependence of proton longitudinal and transverse relaxation times on cell-cycle phase: mouse MCA-transformed 10T1/2 TCL-15 cells

Attempts to determine proton NMR longitudinal relaxation times (T1) as a function of cell-cycle stage using cells synchronized by chemical methods have yielded conflicting results (P. T. Beall, C. F. Hazlewood, and P. N. Rao, Science 192, 904 (1976); R. N. Muller et al., FEBS Lett. 114, 231 (1980); D. N. Wheatley, et al., J. Cell Sci. 88, 13 (1987]. This has raised the question whether a true dependence of T1 on cell-cycle phase exists. In the present study, the centrifugal elutriation technique was used to obtain relatively pure, synchronized cell populations of TCL-15 cells (a methylcholanthrene-transformed line of mouse 10T1/2 cells) for measurement of proton NMR relaxation rates. This technique provides a means to procure synchronized cell populations without the use of chemical agents as in the above-cited investigations and therefore avoid possible effects caused by the chemical agents of the NMR relaxation processes. Both T1 and the transverse relaxation time, T2, of water protons in synchronized-cell pellets obtained in this study, exhibited a dependence on cell-cycle phase at least for the first half of the cell cycle (G1 to S). Cells in G1 phase exhibited quantitatively higher T1 and T2 relaxation times compared to those measured for cells in mid S phase. Such changes were found to correlate with changes in water content. The distribution of cell-cycle phases of each cell population was determined by the DNA histogram using flow cytometric methods. Possible relaxation mechanisms which may contribute to the cell-cycle-specific phenomena of the intracellular T1 and T2 times are discussed.

The response of the KHT sarcoma to radiotherapy as measured by water proton NMR relaxation times: relationships with tumor volume and water content

The potential application of magnetic resonance imaging (MRI) to predict tumor response to radiotherapy is investigated. The water proton spin-lattice and spin-spin relaxation times (T2 and T2, respectively) of murine sarcomas (designated KHT) were measured shortly after excision. This study has demonstrated significantly different responses in T1 and T2 between the control and the irradiated tumors at various times following single doses of X rays. Quite generally, the changes in relaxation times correlated with the changes in tumor water content, indicating that the MR relaxation-time probes are fairly sensitive to radiation-induced edema and dehydration. The possible relationships between the T1 and T2 responses and radiobiological effects such as those on tumor blood flow, vascular permeability, physiological state of cells, and cell death are discussed. It is conceivable that the findings obtained from this investigation could be extended to in situ studies for potential applications in clinical radiotherapy.

Characterization of normal brain tissue using seven calculated MRI parameters and a statistical analysis system

A statistical analysis system for classifying normal brain tissue has been applied to the analysis of MRI scans on 45 volunteers. The Bayes Maximum Likelihood method was used to achieve a discrimination accuracy of 84% for 13 tissue types among three age group sets, with classification accuracies for individual regions ranging from 50 to 100%. In order to attain this level of discrimination a set of seven derived relaxation-type parameters was used to categorize the tissue types. Values for these experimentally estimated parameters were derived from the MRI intensities of eight images in the following pulse sequences: (1) a Carr-Purcell-Meiboom-Gill (CPMG) four-echo train, (2) a single-echo inversion recovery, and (3) three single-echo sequences with varying repetition times, TR, and echo delays, TE. The T2 values derived from ratios of single-echo intensities showed better discrimination power than those from the four-echo CPMG train. The general precision of the seven estimated parameters was excellent, with percentage standard deviations ranging from 4 to 18% for the various regions studied. The tissue discrimination achieved by use of just three relaxation parameters, T1, T2, and proton density, calculated from intensities of images from a four-echo sequence, an inversion recovery sequence, and a short TR single-echo sequence, was not as good, being only 55%.

Tumours of the central nervous system. Proton magnetic resonance relaxation times T1 and T2 and histopathologic correlates

Proton MR relaxation times T1 and T2 were determined in vitro in 136 small specimens of astrocytomas grades I-IV, of oligodendrogliomas, metastases of adenocarcinomas, meningiomas and acoustic neuromas. In addition, 7 samples of peritumoural white matter were analysed. The analysed specimens were studied microscopically in their entirety regarding tumour type and occurrence of necrosis and non-tumour tissue admixture, such as fibrosis and haemorrhage. Most of the gliomas had longer relaxation times than normal white matter and T2 was significantly longer than in the other three tumour groups. The metastases had longer T1 than normal white matter, while T2 varied. The astrocytomas tended to show shorter relaxation times with increasing degree of malignancy, and shortening of T1 and T2 correlating with the proportion of tissue necrosis. Similarly, the metastases with tissue necrosis had shorter T1 and T2 than non-necrotic samples. The meningiomas had T1 values comparable with normal cortex, while the T2 values varied. Tumours containing a large proportion of fibrous tissue had shorter relaxation times than the others. Acoustic neuromas had only slightly longer T1 than normal white matter, while T2 was not prolonged. Both T1 and T2 were significantly shorter than in all other tumours studied. Peritumoural white matter had prolonged relaxation times compared with normal white matter, correlating to increased water content. These in vitro differences regarding relaxation times in various types of tumours of the central nervous system, dependent on various types of tissue alterations, should be of interest for the interpretation of in vivo images.

Effect of induced field inhomogeneity on transverse proton NMR relaxation in tissue water and model systems

The effect of induced field inhomogeneity (IFI) on transverse NMR relaxation of water protons in tissue has been investigated by examining the field dependence of the effective transverse relaxation rates (1/T2 eff) for in vitro canine brain tissue samples. At fields of 0.47, 2.35, 7.05 T (corresponding to 20, 100, and 300 MHz, respectively) the transverse relaxation rates for both white and gray matter samples follow a field dependence of the form 1/T2 eff = C0 + C1 B0, where B0 is the applied field. The linearly dependent term, C1 B0, which reflects the IFI contribution, does not contribute much (i.e., less than 20%) at fields less than 2.0 T. However, at greater field strengths the contribution is appreciable, e.g., greater than 60% at 7.0 T. Results from model systems of glass beads are also reported to illustrate IFI effects. For both the model systems and canine brain tissue samples, the effects of restricted diffusion are qualitatively evident in Hahn spin-echo experiments.