The three-dimensional (3D) structure of the cell and its nucleus is intricately associated with cellular function. Aberrations and heterogeneity in cellular and nuclear morphology serve as diagnostically important cues to identify malignant transformation. However, current cytopathology practice is predominantly based on a qualitative, 2D assessment of cellular morphology that may bias the diagnosis. We utilized the Cell-CT® platform to perform single-cell, optical tomography imaging and generate 3D cell images with isotropic, sub-micron spatial resolution. We developed custom image analysis tools to accurately quantify 3D cell and nuclear morphology. We used these measures to characterize the morphological heterogeneity within and among epithelial cell populations, and to identify morphological descriptors that are reliably predictive of cell state. We assessed the efficacy of our approach relative to conventional 2D assessment (at 40x magnification) to detect malignant transformation in immortalized and primary normal-cancer cell line pairs from human esophagus and breast epithelia. We observed the 3D method of assessment to provide a consistently superior performance (sensitivity and specificity) in predicting malignant transformation. Further, some widely used diagnostic parameters, such as nuclear size or nucleus-to-cytoplasm ratio, did not always exhibit the highest predictive power in 3D space. The availability of 3D morphological information also provided novel insights on intra-population morphological heterogeneity in both normal and cancer cell populations. Our results emphasize the need to incorporate 3D morphological information into standard pathological practice.
Citation Format: Vivek Nandakumar, Stephanie Helland, Jessica H. Han, Geriann Davis, Brian A. Ashcroft, Thai H. Tran, Ting Li, Karen S. Anderson, Rahul Pannala, Joseph R. Mikhael, David E. Fleischer, Roger H. Johnson, Kimberly J. Bussey, Deirdre R. Meldrum. Quantifying three-dimensional cellular morphology and its heterogeneity in epithelial cancers by single-cell optical tomography. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 4297. doi:10.1158/1538-7445.AM2014-4297