Faculty Bibliography

OBJECTIVES: The ability to differentiate between brain tumor progression and radiation therapy induced necrosis is critical for appropriate patient management. In order to improve the differential diagnosis, we combined fluorine-18 2-fluoro-deoxyglucose positron emission tomography ((18)F-FDG PET), proton magnetic resonance spectroscopy ((1)H MRS) and histological data to develop a multi-parametric machine-learning model. METHODS: We enrolled twelve post-therapy patients with grade 2 and 3 gliomas that were suspicious of tumor progression. All patients underwent (18)F-FDG PET and (1)H MRS. Maximal standardized uptake value (SUVmax) of the tumors and reference regions were obtained. Multiple 2D maps of choline (Cho), creatine (Cr), and N-acetylaspartate (NAA) of the tumors were generated. A support vector machine (SVM) learning model was established to take imaging biomarkers and histological data as input vectors. A combination of clinical follow-up and multiple sequential MRI studies served as the basis for assessing the clinical outcome. All vector combinations were evaluated for diagnostic accuracy and cross validation. The optimal cutoff value of individual parameters was calculated using Receiver operating characteristic (ROC) plots. RESULTS: The SVM and ROC analyses both demonstrated that SUVmax of the lesion was the most significant single diagnostic parameter (75% accuracy) followed by Cho concentration (67% accuracy). SVM analysis of all paired parameters showed SUVmax and Cho concentration in combination could achieve 83% accuracy. SUVmax of the lesion paired with SUVmax of the white matter as well as the tumor Cho paired with the tumor Cr both showed 83% accuracy. These were the most significant paired diagnostic parameters of either modality. Combining all four parameters did not improve the results. However, addition of two more parameters, Cho and Cr of brain parenchyma contralateral to the tumor, increased the accuracy to 92%. CONCLUSION: This study suggests that SVM models may improve detection of glioma progression more accurately than single parametric imaging methods. RESEARCH SUPPORT: National Cancer Institute, Cancer Center Support Grant Supplement Award, Imaging Response Assessment Teams.

Background: Simultaneous PET/MR is a new technology that may be used in the evaluation of dementia patients. There are few data in the literature regarding quantitative differences between PET data obtained at PET/CT vs PET/ MR and how this may impact image interpretation. This study compared the PET interpretation of PET/CT vs PET/ MR by two independent experienced nuclear medicine physicians. Methods: Forty-five minutes following injection of 10 mCi of FDG, 19 patients with clinically-suspected dementia underwent a 15-min clinical brain PET/CT. Simultaneous PET/MR scanning was subsequently performed (60 min listmode) at approximately 90 min post-injection. Two experienced nuclear medicine physicians blindly interpreted the PET portion of all PET/CT scans, attributing a specific diagnosis (normal, AD, FTD, LBD, other dementia, mixed phenotype or unspecified disease) and severity scale (mild, moderate or severe abnormality). The readers then blindly interpreted the PET data obtained from PET/MR. Concordance between PET/CT (reference standard) and PET/ MR with respect to diagnosis and disease severity was assessed for each reader. Results: Reader A classified 12 PET/CT scans as AD, 5 as unspecified dementia, 1 as LBD and 1 as normal with a mean severity score of 2.0. Reader B classified 10 PET/CT scans as AD, 3 as unspecified, 1 as LBD and 5 as normal with mean severity score of 2.1. PET/MR interpretations with comparison to PET/CT yielded an 84% (16/19) intrareader concordance of diagnosis, with 95% (18/19) of severity scores varying by one point or less. Reader B exhibited 84% intra-reader concordance of dementia pattern diagnosis, with 89% (17/19) of all scores varying by one point or less. Conclusions: Our preliminary analysis in clinically-suspected dementia patients showed a relatively high concordance of intra-reader assignment of diagnosis and severity of findings between PET/CT and PET/MR when evaluated by two blinded experienced nuclear medicine physicians. These results suggest PET/MR!

Tendon stem cells (TSCs) may be used to effectively repair or regenerate injured tendons. However, the fates of TSCs once implanted in vivo remain unclear. This study was aimed to determine the feasibility of labeling TSCs with super-paramagnetic iron oxide (SPIO) nano-particles to track TSCs in vivo using MRI. Rabbit TSCs were labeled by incubation with 50 mug/mL SPIO. Labeling efficiency, cell viability, and proliferation were then measured, and the stemness of TSCs was tested by quantitative real time RT-PCR (qRT-PCR) and immunocytochemistry. We found that the labeling efficiency of TSCs reached as high as 98%, and that labeling at 50 mug/mL SPIO concentrations did not alter cell viability and cell proliferation compared to non-labeled control cells. Moreover, the expression levels of stem cell markers (Nucleostemin, Nanog, and Oct-4) did not change in SPIO-labeled TSCs compared to non-labeled cells. Both labeled and non-labeled cells also exhibited similar differentiation potential. Finally, labeled TSCs could be detected by MRI both in vitro and in vivo. Taken together, the findings of this study show that labeling TSCs with SPIO particles is a feasible approach to track TSCs in vivo by MRI, which offers a non-invasive method to monitor repair of injured tendons.

This study addresses the spatial relation between local Na(+) and K(+) imbalances in the ischemic core in a rat model of focal ischemic stroke. Quantitative [Na(+)] and [K(+)] brain maps were obtained by (23)Na MRI and histochemical K(+) staining, respectively, and calibrated by emission flame photometry of the micropunch brain samples. Stroke location was verified by diffusion MRI, by changes in tissue surface reflectivity and by immunohistochemistry with microtubule-associated protein 2 antibody. Na(+) and K(+) distribution within the ischemic core was inhomogeneous, with the maximum [Na(+)] increase and [K(+)] decrease typically observed in peripheral regions of the ischemic core. The pattern of the [K(+)] decrease matched the maximum rate of [Na(+)] increase ('slope'). Some residual mismatch between the sites of maximum Na(+) and K(+) imbalances was attributed to the different channels and pathways involved in transport of the two ions. A linear regression of the [Na(+)]br vs. [K(+)]br in the samples of ischemic brain indicates that for each K(+) equivalent leaving ischemic tissue, 0.8+/-0.1 Eq, on average, of Na(+) enter the tissue. Better understanding of the mechanistic link between the Na(+) influx and K(+) egress would validate the (23)Na MRI slope as a candidate biomarker and a complementary tool for assessing ischemic damage and treatment planning.

Repeatability of in vivo measurement of multicomponent T2* relaxation in articular cartialges in human knee is important to clinical use. This study evaluated the repeatability of two-component T2* relaxation on seven healthy human subjects. The left knee was scanned once a day in three consecutive days, on a clinical 3T MRI scanner with eight-channel knee coil and ultrashort echo time pulse sequence at 11 echo times=0.6-40 ms. The intrasubject and intersubject repeatability was evaluated via coefficient of variation (CV=standard deviation/mean) in four typical cartilage regions: patellar, anterior articular, femoral, and tibial regions. It was found that the intrasubject repeatability was good, with CV<10% for the short- and long-T2* relaxation time in the layered regions in the four cartilages (with one exception) and CV<13% for the component intensity fraction (with two exceptions). The intersubject repeatability was also good, with CV approximately 8% (range 1-15%) for the short- and long-T2* relaxation time and CV approximately 10% (range 2-20%) for the component intensity fraction. The long-T2* component showed significantly better repeatability (CV approximately 8%) than the short-T2* component (CV approximately 12%) (P<0.005). These CV values suggest that in vivo measurement of two-component T2* relaxation in the knee cartilages is repeatable on clinical scanner at 3 T, with a signal-to-noise ratio of 90.