MR Biomarkers: Ex Vivo Characterization of Human Tissue Samples

Room 714 A/B


Chairs: John R. Griffiths and Sabrina M. Ronen


Prog #

10:30   Invited Speaker: Characterization of Human Tissue Samples by Ex Vivo MR, When, Why, and How
Edward J. Delikatny, Ph.D.
University of Pennsylvania, Department of Radiology, Philadelpha, PA, USA
11:06 859. Metabolic Profiling of Changes in the Transition from Pre-Invasive to Invasive Cervical Cancer Using Magic Angle Spinning Magnetic Resonance Spectroscopy of Intact Tissues

Sonali S. De Silva1, Geoffery S. Payne1, Valerie A. Thomas2, Thomas E. Ind2, Nandita M. deSouza1

1Institute of Cancer Research and Royal Marsden NHS Foundation Trust, Sutton, UK; 2St Georges Hospital NHS Trust, Tooting Broadway, UK

The development of invasive cervical cancer is preceded by a well-defined pre-invasive stage, cervical intraepithelial neoplasia (CIN). Elevated levels of choline have previously been identified in a number of tumours including cervical cancer [1] but whether these metabolites are elevated in CIN tissue is still unclear. The study investigates the metabolic changes in the transition from normal-to-pre-invasive, and from pre-invasive to invasive cervical cancer using 1H and 31P HR-MAS MRS of intact tissues in which the tissue content was subsequently confirmed with histology. Targeted biopsies were obtained on colposcopy in cervical intraepithelial neoplasia, CIN, (low-grade n=5, high-grade n=40) and on visual inspection when invasive disease was clinically obvious (n=23). Cervical tissues were also obtained from women with normal smears undergoing hysterectomy for benign uterine disease (n=5). Mean concentrations of Aln, Cr, Cho, PC, Tau in 1H HR-MAS and PC, PE and Pi in 31P HR-MAS were calculated. HR-MAS spectral analysis shows significant differences between  invasive cervical cancer and high-grade CIN. Increased levels of choline-containing compounds are observed in cancer tissue. Our findings also suggest that normal tissue adjacent to tumour shows some metabolite depletion, while concentration of choline-containing metabolites is increased in CIN tissue adjacent to a malignancy.  Changes in choline metabolism may be a useful biomarker in identifying the transition of pre-invasive to invasive disease.

11:18  860. Quantitative Metabolic Profiling of Second and Third Trimester Human Amniotic Fluid Via High-Resolution MR Spectroscopy: Analysis for Biomarkers of Fetal Maturation

Brad R. Cohn1, Bonnie N. Joe1, Shoujun Zhao1, John Kornak1, Vickie Yi Zhang1, Rahwa Iman1, John Kurhanewicz1, Jingwei Yu1, Aaron B. Caughey1, Mark G. Swanson1

1UCSF, San Francisco, California , USA

High-resolution MR spectroscopy was used to quantify 21 metabolites in second and third trimester human amniotic fluid samples (N=50). Significant differences were found for 15 of 21 metabolites.  Glucose and several amino acid concentrations decreased, while creatinine and betaine increased with gestational age.  Stepwise linear regression models suggested that gestational age could be accurately predicted using combinations of alanine, glucose, and creatinine concentrations. This study provides normative data for assessing fetal development, evidence of pulmonary and renal maturation, and may aid in the application of in vivo MRS as a future non-invasive test for evaluating fetal development.

11:30   861. HR-MAS Spectroscopy of Human Testicular Biopsy Tissue Demonstrates Phosphocholine as a Biomarker of Male Fertility

Rahwa Iman1, Mark Swanson1, Paul J. Turek1, David Aaronson1, Helena Gurascier1, Dan Vigneron1, Sarah Nelson1, John Kurhanewicz1

1University of California San Francisco, San Francisco, California , USA

1H high resolution magic angle spinning (HR-MAS) spectroscopy was used to determine the metabolic profiles of normal, maturation arrested, and azoospermatic human testicular biopsy tissues.  Fourteen 1H HR-MAS spectra were acquired and quantified using the electronic standard ERETIC and the algorithm QUEST.  Phosphocholine (PC) concentrations were significantly higher in normal versus azoospermatic testes and there was no overlap between the two groups.  However, PC concentrations in maturation arrested testes overlapped both normal and azoospermatic testes.  This study suggests that PC could serve as a specific in vivo marker for fertility in 1H MRSI studies of human testes.

11:42 862.  Metabolomics-Based Viability Assessment of Cystic Echinococcosis Using High-Field 1H-MRS of Biopsies

William E. Hull1, Waldemar Hosch2, Günter W. Kauffmann2, Thomas Junghanss2

1German Cancer Research Center (DKFZ), Heidelberg, Germany; 2University Clinic Heidelberg, Heidelberg, Germany

Cystic echinococcosis (CE) or hydatid disease is caused by larval stages of the cestode parasite Echinococcus granulosus (canine tapeworm) and is a zoonosis of worldwide distribution. Clinical staging of cyst development by ultrasonography (US) requires accurate assessment of cyst viability, assuming a correlation between cyst morphology and parasite viability (light microscopy). Using high-field 1H-MRS of 50 cyst biopsies, 48 metabolite parameters were determined, and multivariate analysis provided excellent discrimination between viable and nonviable cysts, regardless of US class, as well as a rationale for discriminating between individual US classes.

11:54  863. Identification of Amide Protons of GSH in MR Spectra of Tumor Cells

Sveva Grande1, Anna Maria Luciani1, Antonella Rosi1, Laura Guidoni1, Vincenza Viti1

1Istituto Superiore di Sanita', Roma, Italy

MRS can be used to monitor intracellular pH in in vitro and in vivo samples. Amide signals were chosen to this purpose, because their intensity is related to pH value. The origin of these signals is not yet clear. In the present work, through the comparison of spectra from tumor cells and from GSH solution, we assign two main peaks of the 1H MR cell spectra to NH protons of gly and cys of GSH. Assignment is confirmed by treatment of cells with buthionine sulfoximine (BSO), that inhibits the activity of the enzyme γ-glutamyl cysteine synthetase.

12:06 864. 1H MRS of Sputum for the Non-Invasive Diagnosis of Lung Cancer

Tedros Bezabeh1, Celia Marginean2, Garth Nicholas2, Omkar B. Ijare1

1National Research Council Institute for Biodiagnostics, Winnipeg, Canada; 2The Ottawa Hospital, Ottawa, Canada

Application of 1H MR spectroscopy for the detection of lung cancer has been investigated. Sputum samples from lung cancer patients and normal controls were analyzed by 1H MRS. Glucose was found to be absent in sputum samples obtained from lung cancer patients. The absence of glucose could be due to an increased rate of glycolysis in the lung cancer cells in these patients. Cancer cells rely heavily on glycolysis for their metabolic demands since they cannot produce energy from pyruvate, due to a defective Krebs cycle. This observation may have a diagnostic potential for the rapid and non-invasive diagnosis of lung cancer.

12:18 865. Temperature Effect on the HRMAS Spectra of Human Brain Tumour Biopsies, Their Pattern Recognition Analysis and Their Post-HRMAS Histopathology

Daniel Valverde-Saubí1, 2, Ana Paula Candiota, 12, Maria Antonia Molins3, Miguel Feliz3, Oscar Godino4, Juan Martino4, Juan José Acebes4, Carles Arús1, 2

1Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Spain; 2Centro de Investigación Biomédica en Red de Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN), Cerdanyola del Vallès, Spain; 3Universitat de Barcelona, Barcelona, Spain; 4Hospital Universitari de Bellvitge, L'Hospitalet del Llobregat, Spain

We investigated the recording temperature effects on the HRMAS spectral pattern and on supervised classifier development. HRMAS spectra were acquired at 0ºC and 37ºC and analysed using  pattern recognition techniques. Post-HRMAS histopathology was carried out. Major reversible temperature dependent changes were detected on the spectral pattern of the samples analysed. A better classification was achieved using the spectra acquired at 0ºC. Sample morphology was essentially not affected by 37ºC HRMAS analysis. The temperature dependent changes explain the better classification at 0ºC. Higher dispersion of data at 37ºC could be due to the existence of molecular subtypes of glioblastoma.