European Journal of Radiology 81S1 (2012) S104–S106 Contents lists available at SciVerse ScienceDirect European Journal of Radiology journa l h o me pa ge: www.elsevier.com/locate/ejrad MR spectroscopy in the breast clinic is improving Carolyn E. Mountford a,b, *, Christian Schuster c , Pascal A.T. Baltzer d , Peter Malycha e , Werner A. Kaiser d a Center for Clinical Spectroscopy, Brigham & Women’s Hospital, Boston 02115, USA b Centre for MR in Health, University of Newcastle, Callaghan, 2308 Australia c Siemens AG Healthcare Sector, 91052 Erlangen, Germany d Institute of Diagnostic and Interventional Radiology, Friedrich-Schiller-University Jena, D-07740, Germany e Department of Surgery, University of Adelaide, SA 5000, Australia article info Keywords: MR spectroscopy L-COSY Primary referral centre 1. Introduction Early studies suggested that in vivo proton MR spectroscopy could be added as an adjunct to dynamic contrast-enhanced MRI of the breast and increase the specificity [1]. Today there have been a wide range of results ranging from excellent to bad. It becomes obvious that these outcomes are dependent on the local clinical culture, hardware manufacturer and experience of the local MR team. Even when all aspects are covered the results are currently only acceptable for tumors 1 cm 3 and above. The discrepancy between the US and Australian sites was the assertion that the diagnosis of breast cancer could be undertaken based on one dimensional (1D) MRS in vivo if the total choline resonance is resolved into components at 3.23 and 3.28 ppm [2]. However, many sites could not achieve this resonance separation and report the “total” choline using the resonance intensity to deduce the pathology [3]; and monitor response to chemotherapy [4]. However it transpired that they were recruiting patients from tertiary referral centre who had undergone core biopsy. This was also the case with our Boston based Centre. An important difference was that the Australian surgeons had recruited patients where neither core biopsies nor clips were placed prior to the MR scan. Two technical issues also need attention. The first is the need for a highly trained operator due to the need to shim the magnet to acceptable levels of field homogeneity prior to data acquisition. The second is dependent on the first for success that is the capability to make a diagnosis on small tumors i.e. 0.3 to 1 cm 3 in size which often have a low cellularity. Resultant FID needs monitoring to ensure adequate signal to noise to allow a diagnosis to be made. The ability to continue collecting data whilst accumulating averages would be of great assistance. *Carolyn Mountford, DPhil, Center for Clinical Spectroscopy, Brigham & Women’s Hospital, Harvard Medical School, Boston 02115, USA. E-mail address: cmountford@partners.org (C. Mountford). We thus transferred the 2D L-COSY technology to the primary referral clinic of Professor Kaiser in Jena, Germany. The goal was to determine if the L-COSY technology could be made clinically applicable on tumors 1 cm 3 and above. In parallel Siemens Healthcare undertook improvements to the shimming capabilities and developed the capacity to monitor the FT spectrum during uninterrupted data acquisition. 2. Methods 2.1. Patients Patients were scanned between Dec-2009 and Mar-2011. All women with an MR determined malignancy had a subsequent biopsy or surgery whereupon the histopathology was undertaken. Those women with benign lesions on MRI did not undergo biopsy. The cohort of 22 consenting subjects included six infiltrating ductal carcinoma (IDC), five unclassified cancers and two infiltrating lobular carcinoma (ILC) in surgical follow up. Five women with benign lesions and four healthy controls were examined. 2.2. Equipment Data were accrued on a 3T MRI (Magnetom TIM Trio; Siemens AG, Erlangen, Germany) using a sixteen channel breast coil. 2.3. Diagnostic MR imaging Pre-contrast imaging consisted of an axial, fat suppressed, T2- weighted (T2w) turbo spin echo sequence (repetition time 4000- ms, echo time 96-ms, section thickness 3.0-mm) and an axial 3D T1-weighted gradient-echo sequence (repetition time 6.5-ms, echo time 2.6-ms, section thickness 1.0-mm). Followed by an axial 3D T1-weighted gradient-echo sequence, performed before and during the contrast (Omniscan). One pre-contrast phase was be initially acquired then followed by four phases acquired during, 0720-048X/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved.