Can MR Spectroscopy Predict Lymph Node Status in Breast cancer

Abstract

Experimental studies have shown increased concentrations of glycerophosphocholine‐phosphocholine (GPC‐PC), choline, lactate, alanine and uridine diphosphoglucose and higher GPC‐PC/threonine (Thr) in lymph nodes with metastases. As biochemical changes in tumor infiltrated lymph nodes occur earlier than morphological changes and Proton (1H) Magnetic Resonance Spectroscopy MRS) can detect these chemical changes, it can be a potentially useful investigation in evaluation of sentinel lymph node (SLN) in breast cancer for early detection of metastasis. Aim: To evaluate the role of MR Spectroscopy (MRS) in identifying biochemical changes in sentinel lymph node in breast cancer and correlate its findings with histopathology to predict presence of metastasis. Materials &MethodsSLN were obtained from 59 breast cancer patients who underwent SLN biopsy. Each LN was bisected into two equal halves. One half was snap frozen in liquid nitrogen (‐196°C) so that all the enzymatic reactions are arrested and no more metabolic changes occur thereby, keeping the physiological state of the tissues intact. The tissues were then stored at −80°C until perchloric acid extraction was carried out. The other half was subjected to histopathological evaluation (HPE). Conventional HPE of LN was performed using standard Eosin & Hematoxylin staining. MRS: Water‐soluble metabolites from the tissue samples were extracted using PCA extraction procedure as reported in the literature. Proton (1H) MRS of all specimen was performed by an investigator from Department of NMR who was blinded to results of histopathology. Experiments were carried out on a narrow bore spectrometer operating at 700 MHz (Agilent Technologies, Santa Clara, CA, USA) equipped with 5 mm dedicated multinuclear broadband inverse probe at 25°C. Spectrum with water suppression was acquired with a single 60° pulse (Ernst angle). The concentration of of GPC, PC, Thr & Lac were estimated. Statistical Analyses: Levels of concentrations of various metabolites were compared between involved and non‐involved LNs using Wilcoxon rank sum (Mann Whitney) test. Diagnostic test indices were calculated taking histopathology as ‘gold standard’.ResultsMean concentrations of GPC, PC, CHO, Threonine and lactate were significantly increased in involved as compared to non‐involved nodes (Table 1). A cut‐off value of 0.80 for the GPC‐PC/Thr ratio was chosen to obtain a maximum accuracy of 89% for the classification of the lymph nodes based on our previous study. MRS accurately predicted metastasis in 21 of the 24 patients who had lymph node metastasis on histopathology. Out of 35 patients with no lymph node metastasis, MRS correlated accurately with histopathology in 30 of them. Sensitivity, specificity, PPV, NPV and overall accuracy for MRS in detecting LN metastasis were 87.5%; 88%; 80.7%; 90.9% and 86.4% respectively. Likelihood ratio for positive test and negative test for MRS to detect LN metastasis were 6.1 and 0.14 respectively.ConclusionMRS can accurately predict presence of metastasis in SLN in breast cancer. Further studies with a larger sample size are essential before incorporating MR Spectroscopy in routine clinical practice to detect lymph node metastasis in breast cancer.