A Noninvasive Multimodal Sono-Contrast NIR Spectroscopy System for Breast Cancer Diagnosis: Clinical Trial.

Kun Yan,Yu Yan,Lydia Liao
DOI: https://doi.org/10.1200/jco.2011.29.27_suppl.68
IF: 45.3
2011-01-01
Journal of Clinical Oncology
Abstract:68 Background: A noninvasive multimodal sono-constrast spectroscopy system was developed to detect breast cancer from benign mass. An IRB approved clinical study has been carried out to evaluate its diagnostic power. Fourteen patients with histologically-proven cancer and 26 with benign mass were enrolled with informed consent. Methods: This system combines three modules: diffuse reflectance spectroscopy, ultrasonography and low intensity focused ultrasound (LIFU). It reports the optical functional difference in cancer and benign mass after transient LIFU interaction with vascular flow. The optical intensity and LIFU signals were collected during each study. The ratio of optical intensities at wavelengths 685 nm and 830 nm was decomposed into low frequency and high frequency components using wavelet analysis. Results: Comparison of the high frequency component showed that LIFU stimulated transitory fluctuation in benign mass, but not in malignant tissue, as quantified by mean absolute deviation (MAD). The ratios of MAD during LIFU vs. baseline in cancer and benign mass are [mean (max/min/std)] 2.04 (3.27/1.41/0.43) and 3.00 (4.29/1.86/0.62). This suggests that high frequency fluctuation was amplified in benign mass during LIFU stimulation. There are a few outliers due to poor patient setup or patient movement during procedure. The MAD ratios are significantly different in cancer vs. benign mass (p<<0.01). An ROC curve was generated by varying threshold MAD values. The area under ROC curve (AUC) was calculated to be 0.91. With threshold 2.3, sensitivity of 92.8% and specificity of 80.8% can be achieved. Backward stepwise regression study reveals that system performance is not affected by patient age (hormone status), tumor depth and tumor size. Conclusions: Current results demonstrate the effectiveness of the multimodal system in characterizing breast cancer vs. benign mass. LIFU appears to exert more pronounced influence on blood flow in benign mass than in tumor. This lack of response to LIFU in breast cancer is possibly due to its abnormal blood vasculature. The analysis is ongoing to incorporate ultrasound imaging parameters with spectroscopy to further improve the diagnostic power.
What problem does this paper attempt to address?