Potential of magnetic resonance spectroscopy to detect metastasis in axillary lymph nodes in breast cancer

Focused pathological evaluation of axillary lymph nodes in breast cancer is gaining importance. Nuclear magnetic resonance (NMR) spectroscopy that assesses the whole of the specimen has the potential in evaluating micrometastases. The biochemical changes associated with breast cancer metastases in axillary nodes by in vitro NMR and its use in the detection of axillary metastases in a clinical setting in comparison with conventional histopathology is presented in this study. Eighty-eight lymph nodes obtained from 30 patients with breast cancer were investigated. Histopathology revealed metastases in 20 nodes from 11 patients, while in vitro NMR spectroscopy revealed metastases in 22 nodes. Out of these 22 nodes, 16 were the same, which showed metastases on bistopathology, while 6 nodes have shown metastases only on in vitro magnetic resonance spectroscopy (MRS). These 6 nodes with suspicion of metastases on MRS were subjected to reevaluation with serial sectioning and immunohistochemistry, but no additional metastases were revealed. Forty metabolites could be identified from the MR spectrum of lymph nodes. The levels of the glycerophosphocboline-phosphocholine (GPC-PC), choline, lactate, alanine and uridine diphosphoglucose were elevated significantly in nodes with metastases. In addition, the intensity ratio of GPC-PC/threonine (Thr) was higher in nodes with metastases, and using this as marker, MRS detected the axillary metastases with a sensitivity, specificity and accuracy of 80%, 91% and 88%, respectively. Neoadjuvant chemotherapy (NACT) lowered the concentrations of GPC-PC and GPC-PC/Thr ratio. The accuracy of MRS in detecting metastases was 75% in patients who received NACT (n=9) as compared to 96% in those who did not (n=21). Our results demonstrate the potential of in vitro MRS in characterizing the metabolite profile of the axillary nodes with breast cancer metastases. It detected axillary metastases with reasonable accuracy and can be complementary to histopathological evaluation and immunohistochemistry. (c) 2005 Elsevier Inc. All rights reserved.