Rofo 2005; 177 - A11
DOI: 10.1055/s-2005-931825

In vivo spectroscopy pf the ovary at 3 Tesla

C Mountford 1, P Stanwell 1, J Carter 1, A Ferrier 1, S Doran 1, C Lean 1, P Russell 1
  • 1Institute for Magnetic Resonance Research and Department of Magnetic Resonance in Medicine, University of Sydney, Australia

Purpose: The goal is to improve screening for ovarian cancer preoperatively by non-invasively confirming or excluding malignancy of a suspicious pelvic mass and, if malignant, determining extent of disease by using chemical shift magnetic resonance imaging at 3T. The spectral alterations associated with ovarian pathology on biopsy are already established (1).

Methods: Patients were recruited from those presenting with primary ovarian neoplasms to the Gynaecological Oncology Unit at Royal North Shore and Royal Prince Alfred Hospitals in Sydney. In vivo MR data was collected on a commercially available whole-body 3T system (Magnetom Trio, Siemens AG, Erlangen, Germany) combined with an 8-element phased-array coil (USA Instruments, Aurora, USA). Ovaries were identified on T2-weighted (TE/TR 98/4000ms, FOV 22-cm, 0.43×0.83×3.0-mm resolution)) axial and coronal images. Coronal 2D-CSI (TE 135ms, TR 1340ms, 12.5×12.5×10.0-mm resolution) data was collected with outer volume suppression (OVS) pulses, offset to the frequency of lipid, used to limit lipid bleed into the region of interest (2,3) Coronal spectroscopic imaging data was reconstructed in Figure 1, using only the choline resonance and overlaid on the coronal T2-weighted reference image. During subsequent surgery the tumours are removed but the orientation maintained to allow registration with the features obtained in the MR images. Tissue was fixed using standard protocols and subtypes identified according to the WHO Histological Classification (4).

Results: At 3T MRI identifies an ovarian mass clearly. It can also distinguish a solid mass from a cystic component and identifies any discrete solid nodules projecting into the cyst. The volume of interest, when examined by 2D-CSI identifies the pathology due to the presence of diagnostic metabolites such as choline. The independent histopathological diagnosis confirms the spectral diagnoses.

Conclusion: MRI identifies the complexity of ovarian masses, including cysts and the presence of solid nodules projecting into a cyst. The preoperative pathological diagnosis is made by spectroscopic imaging using the presence of diagnostic metabolites. Access to such MR technology should allow more appropriate triaging of patients with malignant disease to a specialist gynaecological oncology unit for primary treatment and more confident triaging of patients with benign disease to general gynaecological units or to conservative management.

References:

1. Mountford C, Doran S, Lean C, Russell P. Chemical Reviews 2004; 104:3677–704.

2. Maudsley A, Hilal SK, Simon HE, Wittekoek S. Radiology 1984; 153:745–50.

3. Vigneron D, Kurhanewicz J, Nelson S. In: Higgins C, Hricak H, Helms C, eds. Magnetic resonance imaging of the body. New York: Raven Press, 1997; 205–20.

4. Russell P, Robboy S, Anderson M. Ovarian tumours: classification and clinical perspective. London: Churchill Livingstone, 2002.