The quantitative study of lumbar vertebral bone marrow using T1 mapping and image analysis techniques: methodology and preliminary results

3-Tesla MR spectroscopy in patients subjected to bone marrow transplantation: clinical correlations

PURPOSE

This study evaluated the usefulness of 3-Tesla magnetic resonance (MR) spectroscopy in patients with non-Hodgkin's lymphoma (NHL) undergoing bone marrow transplantation (BMT).

MATERIALS AND METHODS

Twelve NHL patients who were candidates for BMT underwent three MR examinations of the lumbosacral spine: before ablative therapy for BMT, 15±4 days and 54±24 days after BMT. The MR study was supplemented by spectroscopic analysis. The lipid content was calculated and expressed as a percentage of lipid signal intensity relative to total signal intensity [fat fraction (FF)].

RESULTS

In the first MR study, the FF was 62.5±7%, in the second it was 70.75±5% and in the third it was 75±1%. We observed a statistically significant difference between FF values calculated at the various MR studies (p=0.02) and between red blood cell count (p=0.017), platelet count (p=0.003) and haematocrit (p<0.001) at the three MR studies. FF had a statistically significant correlation with the number of circulating platelets (p<0.01)

CONCLUSIONS

MR spectroscopy of the bone marrow of NHL patients undergoing BMT is noninvasive and highly sensitive for characterising and monitoring bone marrow after BMT.

MRI in the evaluation of late bone marrow changes following bone marrow transplantation

Measurements of MR spin-lattice (T1), and spin-spin (T2) relaxation times in lumbar vertebrae have been performed in a pilot study on six adult patients, treated for acute myeloid leukaemia (AML). All patients were treated with initial chemotherapy and then proceeded to bone marrow transplantation (BMT), conditioned with Melphalan and total body irradiation (TBI). MR measurements were made between 21 and 89 months after TBI. The relaxation times in the six patients were compared with those in six healthy age-matched volunteers to establish whether there were differences between the two groups. Average T1 values in the vertebrae of the treated patients are significantly shorter (p < 0.01) than in the healthy volunteers. This is consistent with the observation of a relatively hyperintense vertebral bone marrow in the T1 weighted images and is likely to be a consequence of treatment induced fatty replacement of marrow. Shorter T1 values tend to be distributed within the centre of the lumbar vertebrae compatible with observations, made by others, which suggest that the peripheral zone of the vertebral body has been repopulated with bone marrow cells whereas the central zone, around the basivertebral vein, is predominantly fat. Histogram displays of vertebral body relaxation time distributions (T1, T2) for both patients and healthy age-matched volunteers are similar in that both patients and volunteers give histograms that are only slightly skewed. This similarity is probably a reflection of the fact that the patients have been in remission for over a year and have generally healthy bone marrow.

Objective quantification of muscle and fat in human dystrophic muscle by magnetic resonance image analysis

Information about changes in muscle composition has to date been primarily restricted to histological examination of biopsy samples or qualitative assessment of images obtained using a variety of techniques (e.g., ultrasound, CT, and MRI). We describe the development of a quantitative method for the analysis of muscle composition using MR T2 relaxation time mapping and image analysis. This approach provides an objective means of studying muscle and, when used in conjunction with force production measurements, may provide an accurate measure of response to muscle therapy.

Magnetic resonance imaging in myelodysplastic syndromes

In this study we used magnetic resonance imaging (MRI) to examine the axial bone marrow of 11 patients with myelodysplastic syndromes (MDS). T1 weighted images showed a reduction in signal intensity in all patients. The degree of reduction in signal intensity correlated with bone marrow cellularity rather than percentage blasts or French-American-British Co-operative group (FAB) classification. T2 weighted images showed an increase in signal intensity in 10 out of 11 patients; however, reliable grading was not possible using this sequence. Both T1 and T2 weighted images indicated a homogeneous pattern of disease within the axial skeleton in MDS. We have also shown that iron deposition in multiply transfused patients is a significant factor influencing signal reduction. The results of this study suggest that MRI may have a diagnostic role in MDS but is unlikely to be of value in staging.

Characterization of bone marrow after transplantation by means of magnetic resonance

Magnetic resonance (MR) has become a new tool for noninvasive characterization of bone marrow in patients with hematological disorders in the past few years. Experiences gained from 1H MR imaging and spectroscopic investigations in 48 healthy volunteers and more than 130 patients with hematological disorders are reported and interpreted. Twenty-four of the patients underwent bone marrow transplantation (BMT) before the MR examinations. The findings in these studies provided noninvasive characterization and monitoring of vertebral marrow after BMT. Specifically, MR techniques were found to be suitable for studies of different aspects in physiological and pathological alterations of bone marrow: The water content within the marrow can be analyzed by chemical-shift selective-imaging techniques with good spatial resolution. Spectroscopic methods also allow more sensitive quantification of the signal fractions, as well as separate evaluation of the relaxation times of water and lipids. Relaxometry might be useful to characterize the cellular and extracellular portions of water molecules. Furthermore, the distribution of the magnetic field within small-volume elements of vertebral marrow can be measured. The field distribution is influenced by the trabecular density and the composition of the marrow. High amounts of hemosiderin in the marrow result in clearly broadened field distributions, demonstrated by increasing line widths in MR proton spectra. Magnetic resonance techniques can be used to assess not only the cellularity of the bone marrow, but also metabolic alterations in this compartment which result from cytotoxic treatment or immunological processes.

Magnetic resonance relaxation time mapping in multiple sclerosis: normal appearing white matter and the "invisible" lesion load

Prolonged T1 and/or T2 relaxation times (RT) in the normal appearing white matter (NAWM) of patients with multiple sclerosis (MS) have been attributed either to a diffuse abnormality, or to "small lesions" undetected by visual inspection of conventional MR images. In a comparison of brain slices from five MS patients and five healthy control subjects, we have confirmed that the average T1 and T2 RTs obtained from NAWM in patients with MS are significantly prolonged (p < .04). Quantitative pixel-by-pixel mapping shows that this overall prolongation is due to the averaging of RTs from two subfractions of NAWM. In all patients a proportion (average 54% for T1 and 63% for T2) of the total white matter pixel sample from each MR brain slice had RT values indistinguishable from those found in the white matter of matched healthy control subjects (i.e., "normal normal appearing white matter," NNAWM). Scattered throughout the NAWM were multiple small areas, often of only one or two pixels, with abnormal RT values. These lesions, which were revealed only by pixel-by-pixel mapping of RT, made up a significant proportion (average 47% for T1 or 57% for T2 estimates) of the total (visible plus "invisible") lesion load per slice, and of the NAWM (average 36% for T1, 27% for T2), with wide interpatient variability. Further studies of these minute lesions are required to determine their total volume in the brain, their precise nature, evolution and relevance to the functional deficit in MS.

Leukemic red bone marrow changes assessed by magnetic resonance imaging and localized 1H spectroscopy

Red bone marrow of healthy persons has considerable contents of water and lipids. The cellularity and the corresponding fat-water ratio within the marrow show clear changes in hematological diseases. Magnetic resonance (MR) methods use the signals of the protons of water and lipids. This paper gives a comparison between different standard MR techniques and recently developed fat- and water-selective imaging methods, addressing their sensitivity to bone marrow changes in leukemia. Additionally, 1H results of spectroscopic methods are presented. The results and conclusions are based on the examination of 26 healthy volunteers and 106 patients with general or focal bone marrow alterations. Standard T1-weighted images did not distinguish bone marrow of young healthy volunteers with relatively high cellularity from acute leukemia. Using fat- and water-selective methods, patients with untreated leukemia showed only water proton signals and no lipid signals from red bone marrow of vertebral bodies and the pelvis. This phenomenon was never observed in healthy volunteers. Following chemotherapy, lipid and water contents normalize in successfully treated patients. Nonresponders did not show significant changes of the fat-water ratio after up to 3 weeks of therapy. Phase contrast imaging provides information about the difference between fat and water fractions within the bone marrow, but quantitative determination of the absolute fat and water fractions requires acquisition of several images and suffers from the susceptibility effects in trabecular bone marrow. The fat-water ratio and additional qualities of water and lipid protons (relaxation times) can be evaluated by volume-selective MR spectroscopy. Typical results of spectra from small-volume elements in hypercellular vertebral bone marrow of leukemic patients before cytotoxic treatment and of normocellular or hypocellular marrow after therapy are demonstrated.

Detection of bone marrow abnormalities in patients with Hodgkin's disease by T1 mapping of MR images of lumbar vertebral bone marrow

Pixel by pixel T1 mapping of MR images has been used in conjunction with image analysis techniques to study the lumbar vertebral bone marrow of 20 patients with refractory or relapsed Hodgkin's disease prior to salvage chemotherapy, or high dose chemotherapy with autologous bone marrow rescue. Compared to 18 age matched controls, seven patients had significantly abnormal lumbar vertebral marrow T1 histograms with median, 5% and 95% centile T1 values that lay outside a three dimensional 95% probability region obtained for the normals. Six of these patients had increased areas of high T1 consistent with bone marrow involvement with lymphoma; although, only in two of them had Hodgkin's disease been detected by bilateral iliac crest bone marrow biopsy. Post-treatment studies were performed in four patients who had abnormal areas of high T1 in the lumbar vertebral bone marrow. All showed normalisation of the T1 histogram, with a reduction of T1 values to within the normal range as defined by the studies of age matched controls. T1 mapping of bone marrow offers potential for detecting bone marrow infiltrates in patients with lymphoma, and also for the assessment of treatment response.