Abstract P2-03-10: Non-invasively measured Warburg effect: Optically measured tissue oxygenation and its correlation with Ki67 proliferation

Abstract

Abstract Clinically, the level of Ki67 expression is used as a biomarker for cancer proliferation. In this clinical study, we focus on malignant tumor properties and investigate the correlations between macroscopically measured Diffuse Optical Tomography (DOT) physiological parameters with the Ki67 proliferation marker. The DOT tumor-to-normal tissue parameters were previously shown to have excellent tumor sensitivity and specificity by Choe and co-workers (JBO, 14(2):024020). In order to quantify cell proliferation, the percent of Ki67-expressing nuclei (i.e., from all the nuclei in an ROI) was used to specify Ki67 expression in cancer and normal tissues. Then, the tumor-to-normal ratio of Ki67-expressing nuclei was calculated to derive “relative” Ki67 values (i.e., rKi67). For this analysis, only normal glandular tissues with Ki67 expression were used. For these determinations of rKi67, the range of the Ki67-nuclei present in cancer tissues was 0.36-23.45% (N = 8), and in normal tissues it was 0.19-7.41% (N = 8). Ki67-expression in cancer-only tissues was also compared to DOT parameters. The range of the Ki67 present in cancer used for the analysis of the cancer-only tissues was 0.36-27.77% (N = 15). Among the DOT parameters, rStO2 (relative tissue oxygenation) and rHbO2 (relative oxy-hemoglobin concentration) were highly correlated with rKi67 as shown in table 1 (Pearson correlation: 0.92, p-value: 0.001 for rStO2 and correlation: 0.93, p-value: 0.002 for rHbO2). Also, for cancer-only Ki67, rHb (relative deoxy-hemoglobin concentration) showed a weak inverse correlation with cancer Ki67%: correlation: -0.58, p-value: 0.026. We additionally tested if the Diffuse Optical Tomography parameters are significantly different in more proliferative cancer compared to the less proliferative cancer (as determined by the 15% cutoff point). For this purpose, only rHb differentiated Ki67-positive from Ki67-negative cancer, with lower values of rHb occurring for the Ki67-positive cancer (p-value: 0.018, Wilcoxon-ranked-sum test). In this correlation study, we observed that rKi67 was highly correlated with rStO2 and rHbO2. Further, the cancer-only Ki67 showed lower correlations with the relative DOT parameters. However, we found that rHb was inversely correlated with cancer-only Ki67 percent values. Additionally, rHb was lower in the Ki67-positive cancer compared to the Ki67-negative cancer (using the 15% cutoff point). Taken together, these results suggest that in more proliferative cancers, although more oxy-hemoglobin might be supplied to the cancer, the level of oxygenated-hemoglobin remains high and less oxygen is utilized for cancer metabolism (i.e., consistent with lower rHb). This finding appears to be consistent with the Warburg effect, which accounts for the fact that proliferative cells will go through glycolysis to increase biomass without using oxygen, despite sufficient presence of oxygen in the tissue environment. Overall, the results of this study corroborate expectations that macroscopic measurement of breast cancer physiology using DOT can reveal information microscopic pathological properties of breast cancer. Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P2-03-10.