Size and Shape Filtering of Malignant Cell Clusters within Breast Tumors Identifies Scattered Individual Epithelial Cells as the Most Valuable Histomorphological Clue in the Prognosis of Distant Metastasis Risk

Abstract

Survival and life quality of breast cancer patients could be improved by more aggressive chemotherapy for those at high metastasis risk and less intense treatments for low-risk patients. Such personalized treatment cannot be currently achieved due to the insufficient reliability of metastasis risk prognosis. The purpose of this study was therefore, to identify novel histopathological prognostic markers of metastasis risk through exhaustive computational image analysis of 80 size and shape subsets of epithelial clusters in breast tumors. The group of 102 patients had a follow-up median of 12.3 years, without lymph node spread and systemic treatments. Epithelial cells were stained by the AE1/AE3 pan-cytokeratin antibody cocktail. The size and shape subsets of the stained epithelial cell clusters were defined in each image by use of the circularity and size filters and analyzed for prognostic performance. Epithelial areas with the optimal prognostic performance were uniformly small and round and could be recognized as individual epithelial cells scattered in tumor stroma. Their count achieved an area under the receiver operating characteristic curve (AUC) of 0.82, total area (AUC = 0.77), average size (AUC = 0.63), and circularity (AUC = 0.62). In conclusion, by use of computational image analysis as a hypothesis-free discovery tool, this study reveals the histomorphological marker with a high prognostic value that is simple and therefore easy to quantify by visual microscopy.

Analysis of Spatial Distribution and Prognostic Value of Different Pan Cytokeratin Immunostaining Intensities in Breast Tumor Tissue Sections

“…With both TNM and histologic grade being ineffective in this early breast cancer patient group without lymph node involvement or metastasis (N0M0), only tumor size showed a prognostically significant association by an AUC of 0.65 (95% CI = 0.51–0.78; p = 0.04). For the distribution of clinicopathological parameters in this group of patients, including age, tumor size, histologic grade, estrogen receptor (ER), progesterone receptor (PR), HER2 status and metastasis location, please refer to our previous report [ 17 ].…”

“…Multivariate binary logistic regression analysis ( Table 3 ) included the demographic (age), clinicopathological (ER and tumor size), and the 38 image analysis features calculated in the original images and narrow ranges satisfying the selection entry criterion of p ≤ 0.05 obtained by the ROC analysis ( Table 1 and [ 17 ]). We used continuous values for this test to avoid the bias introduced by categorization.…”

“…Immunostaining of keratin reveals the shapes, sizes, and distribution of malignant cell patches in cancers caused by the neoplastic transformation of epithelial cells and thus presents a paramount prognostic potential. While the cytokeratin staining of malignant nests [ 16 ] and their prognostic value in breast tumors [ 17 , 18 , 19 , 20 , 21 ] have been established, this is the first systemic analysis of the pan cytokeratin immunostaining based on its intensity. We revealed the spatial distribution and prognostic value for each of the produced seven narrow ranges of pan cytokeratin immunostaining intensities.…”

“…Still lighter shades of grey in the range of 200–220 mostly stained the borders of tumor nests, while the lightest 220–255 staining was found only in the stroma. As cytokeratins are expressed in epithelial cells which populate tumor nests and only sporadically found in the stroma [ 17 , 31 ], we assumed that the intensity level 220 was the precise cutoff between the specific 0–220 pan cytokeratin staining that was limited to tumor nests and the non-specific 220–255 staining that was almost entirely restricted to the stroma. Interestingly, while image analysis features exerted good prognostic performance within narrow grey level ranges, their pooling into a single wide 0–220 range could not improve the prognostic performance.…”

“…Both normal and malignant epithelial cells can be stained with AE1 and AE3 monoclonal antibodies that are well-defined and have high specificity for keratin fibers [ 16 ]. Malignant epithelial cells are usually organized in tumor nests of variable sizes and shapes scattered throughout stroma which contains mostly fibroblasts, some immune cells, and few epithelial buds [ 17 ]. We have previously found that low pan cytokeratin staining intensity is associated with a poor breast cancer outcome [ 18 ].…”