Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Mar 9;9(3):e14371.
doi: 10.1016/j.heliyon.2023.e14371. eCollection 2023 Mar.

A pathomic approach for tumor-infiltrating lymphocytes classification on breast cancer digital pathology images

Mario Verdicchio  1 , Valentina Brancato  1 , Carlo Cavaliere  1 , Francesco Isgrò  2 , Marco Salvatore  1 , Marco Aiello  1
Affiliations

A pathomic approach for tumor-infiltrating lymphocytes classification on breast cancer digital pathology images

Mario Verdicchio et al. Heliyon. .

Abstract

Background and objectives: The detection of tumor-infiltrating lymphocytes (TILs) could aid in the development of objective measures of the infiltration grade and can support decision-making in breast cancer (BC). However, manual quantification of TILs in BC histopathological whole slide images (WSI) is currently based on a visual assessment, thus resulting not standardized, not reproducible, and time-consuming for pathologists. In this work, a novel pathomic approach, aimed to apply high-throughput image feature extraction techniques to analyze the microscopic patterns in WSI, is proposed. In fact, pathomic features provide additional information concerning the underlying biological processes compared to the WSI visual interpretation, thus providing more easily interpretable and explainable results than the most frequently investigated Deep Learning based methods in the literature.

Methods: A dataset containing 1037 regions of interest with tissue compartments and TILs annotated on 195 TNBC and HER2+ BC hematoxylin and eosin (H&E)-stained WSI was used. After segmenting nuclei within tumor-associated stroma using a watershed-based approach, 71 pathomic features were extracted from each nucleus and reduced using a Spearman's correlation filter followed by a nonparametric Wilcoxon rank-sum test and least absolute shrinkage and selection operator. The relevant features were used to classify each candidate nucleus as either TILs or non-TILs using 5 multivariable machine learning classification models trained using 5-fold cross-validation (1) without resampling, (2) with the synthetic minority over-sampling technique and (3) with downsampling. The prediction performance of the models was assessed using ROC curves.

Results: 21 features were selected, with most of them related to the well-known TILs properties of having regular shape, clearer margins, high peak intensity, more homogeneous enhancement and different textural pattern than other cells. The best performance was obtained by Random-Forest with ROC AUC of 0.86, regardless of resampling technique.

Conclusions: The presented approach holds promise for the classification of TILs in BC H&E-stained WSI and could provide support to pathologists for a reliable, rapid and interpretable clinical assessment of TILs in BC.

Keywords: Breast cancer; Digital pathology; Machine learning; Pathomics; Tumor infiltrating lymphocytes.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Identifying stromal Tumor Infiltrating Lymphocyte (TIL) regions in a H&E-stained WSI of breast cancer. (A) H&E-stained WSI of breast cancer. (B) Example of a region of tissue. (C) Example of stromal TILs (green solid arrow) and cells that are no TILs (red dotted arrow). The image shows how it can be challenging to visually differentiate TILs from other cells. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)
Fig. 2
Fig. 2
Workflow of the approach developed in the study. Starting from H&E images from HER2+ and TN breast cancer, the approach starts with the segmentation of nuclei within tumor-associated associated stroma (A). Pathomic features were then extracted from cell segmentation within tumor-associated stroma (B). After the feature selection step, five classification models were trained (C) and their performance evaluated (D).
Fig. 3
Fig. 3
Selected pathomic features (n = 21) after the three steps of feature selection. Surviving features were reported in bold. Abbreviations: FSD = Fourier Shape Descriptor; MAD = Median Absolute Deviation; IDM = Inverse Difference Moment; IMC = Information Measure of Correlation.
Fig. 4
Fig. 4
Receiver Operating Characteristic (ROC) curves of prediction models without resampling (A), with SMOTE-resampling (B) and with downsampling (C). Abbreviations: LDA = Linear Discriminant Analysis; KNN = K-Nearest Neighbour; DT = Decision Tree; RF = Random Forest; MLP = Multi-layer Perceptron.
Fig. 5
Fig. 5
Comparison of AUCROCs between different classification models and different resampling techniques. Abbreviations: SMOTE = synthetic minority oversampling technique; LDA = Linear Discriminant Analysis; KNN = K-Nearest Neighbour; DT = Decision Tree; RF = Random Forest; MLP = Multi-layer Perceptron.
See this image and copyright information in PMC

References

    1. Salemme V., Centonze G., Cavallo F., Defilippi P., Conti L. The crosstalk between tumor cells and the immune microenvironment in breast cancer: implications for immunotherapy. Front. Oncol. 2021:11. https://www.frontiersin.org/articles/10.3389/fonc.2021.610303 accessed. - DOI - PMC - PubMed
    1. Li J.J., Tsang J.Y., Tse G.M. Tumor microenvironment in breast cancer—updates on therapeutic implications and pathologic assessment. Cancers. 2021;13:4233. doi: 10.3390/cancers13164233. - DOI - PMC - PubMed
    1. El Bairi K., Haynes H.R., Blackley E., Fineberg S., Shear J., Turner S., de Freitas J.R., Sur D., Amendola L.C., Gharib M., Kallala A., Arun I., Azmoudeh-Ardalan F., Fujimoto L., Sua L.F., Liu S.-W., Lien H.-C., Kirtani P., Balancin M., El Attar H., Guleria P., Yang W., Shash E., Chen I.-C., Bautista V., Do Prado Moura J.F., Rapoport B.L., Castaneda C., Spengler E., Acosta-Haab G., Frahm I., Sanchez J., Castillo M., Bouchmaa N., Md Zin R.R., Shui R., Onyuma T., Yang W., Husain Z., Willard-Gallo K., Coosemans A., Perez E.A., Provenzano E., Ericsson P.G., Richardet E., Mehrotra R., Sarancone S., Ehinger A., Rimm D.L., Bartlett J.M.S., Viale G., Denkert C., Hida A.I., Sotiriou C., Loibl S., Hewitt S.M., Badve S., Symmans W.F., Kim R.S., Pruneri G., Goel S., Francis P.A., Inurrigarro G., Yamaguchi R., Garcia-Rivello H., Horlings H., Afqir S., Salgado R., Adams S., Kok M., Dieci M.V., Michiels S., Demaria S., Loi S. The tale of TILs in breast cancer: a report from the international immuno-oncology biomarker working group. Npj Breast Cancer. 2021;7:1–17. doi: 10.1038/s41523-021-00346-1. - DOI - PMC - PubMed
    1. Kurozumi S., Inoue K., Matsumoto H., Fujii T., Horiguchi J., Oyama T., Kurosumi M., Shirabe K. Prognostic utility of tumor-infiltrating lymphocytes in residual tumor after neoadjuvant chemotherapy with trastuzumab for HER2-positive breast cancer. Sci. Rep. 2019;9:1583. doi: 10.1038/s41598-018-38272-1. - DOI - PMC - PubMed
    1. Salgado R., Denkert C., Demaria S., Sirtaine N., Klauschen F., Pruneri G., Wienert S., Van den Eynden G., Baehner F.L., Penault-Llorca F., Perez E.A., Thompson E.A., Symmans W.F., Richardson A.L., Brock J., Criscitiello C., Bailey H., Ignatiadis M., Floris G., Sparano J., Kos Z., Nielsen T., Rimm D.L., Allison K.H., Reis-Filho J.S., Loibl S., Sotiriou C., Viale G., Badve S., Adams S., Willard-Gallo K., Loi S., International TILs Working Group The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an International TILs Working Group 2014. Ann. Oncol. 2014;26(2015):259–271. doi: 10.1093/annonc/mdu450. - DOI - PMC - PubMed