Ensemble-based deep learning improves detection of invasive breast cancer in routine histopathology images.

Accurate detection of invasive breast cancer (IC) can provide decision support to pathologists as well as improve downstream computational analyses, where detection of IC is a first step. Tissue containing IC is characterized by the presence of specific morphological features, which can be learned by convolutional neural networks (CNN). Here, we compare the use of a single CNN model versus an ensemble of several base models with the same CNN architecture, and we evaluate prediction performance as well as variability across ensemble based model predictions. Two in-house datasets comprising 587 whole slide images (WSI) are used to train an ensemble of ten InceptionV3 models whose consensus is used to determine the presence of IC. A novel visualisation strategy was developed to communicate ensemble agreement spatially. Performance was evaluated in an internal test set with 118 WSIs, and in an additional external dataset (TCGA breast cancer) with 157 WSI. We observed that the ensemble-based strategy outperformed the single CNN-model alternative with respect to accuracy on tile level in 89 % of all WSIs in the test set. The overall accuracy was 0.92 (DICE coefficient, 0.90) for the ensemble model, and 0.85 (DICE coefficient, 0.83) for the single CNN alternative in the internal test set. For TCGA the ensemble outperformed the single CNN in 96.8 % of the WSI, with an accuracy of 0.87 (DICE coefficient 0.89), the single model provides an accuracy of 0.75 (DICE coefficient 0.78). The results suggest that an ensemble-based modeling strategy for breast cancer invasive cancer detection consistently outperforms the conventional single model alternative. Furthermore, visualisation of the ensemble agreement and confusion areas provide direct visual interpretation of the results. High performing cancer detection can provide decision support in the routine pathology setting as well as facilitate downstream computational analyses.

Unveiling the Power of Model-Agnostic Multiscale Analysis for Enhancing Artificial Intelligence Models in Breast Cancer Histopathology Images.

Developing AI models for digital pathology has traditionally relied on single-scale analysis of histopathology slides. However, a whole slide image is a rich digital representation of the tissue, captured at various magnification levels. Limiting our analysis to a single scale overlooks critical information, spanning from intricate high-resolution cellular details to broad low-resolution tissue structures. In this study, we propose a model-agnostic multiresolution feature aggregation framework tailored for the analysis of histopathology slides in the context of breast cancer, on a multicohort dataset of 2038 patient samples. We have adapted 9 state-of-the-art multiple instance learning models on our multi-scale methodology and evaluated their performance on grade prediction, TP53 mutation status prediction and survival prediction. The results prove the dominance of the multiresolution methodology, and specifically, concatenating or linearly transforming via a learnable layer the feature vectors of image patches from a high (20x) and low (10x) magnification factors achieve improved performance for all prediction tasks across domain-specific and imagenet-based features. On the contrary, the performance of uniresolution baseline models was not consistent across domain-specific and imagenet-based features. Moreover, we shed light on the inherent inconsistencies observed in models trained on whole-tissue-sections when validated against biopsy-based datasets. Despite these challenges, our findings underscore the superiority of multiresolution analysis over uniresolution methods. Finally, cross-scale analysis also benefits the explainability aspects of attention-based architectures, since one can extract attention maps at the tissue- and cell-levels, improving the interpretation of the model's decision. The code and results of this study can be found at github.com/tsikup/multiresolution_histopathology.

A population-based study on trajectories of HER2 status during neoadjuvant chemotherapy for early breast cancer and metastatic progression.

BACKGROUND: This study aimed to investigate the distribution and changes of HER2 status in untreated tumours, in residual disease and in metastasis, and their long-term prognostic implications. METHODS: This is a population-based cohort study of patients treated with neoadjuvant chemotherapy for breast cancer during 2007-2020 in the Stockholm-Gotland region which comprises 25% of the entire Swedish population. Information was extracted from the National Breast Cancer Registry and electronic patient charts to minimize data missingness and misclassification. RESULTS: In total, 2494 patients received neoadjuvant chemotherapy, of which 2309 had available pretreatment HER2 status. Discordance rates were 29.9% between primary and residual disease (kappa = 0.534), 31.2% between primary tumour and metastasis (kappa = 0.512) and 33.3% between residual disease to metastasis (kappa = 0.483). Adjusted survival curves differed between primary HER2 0 and HER2-low disease (p < 0.001), with the former exhibiting an early peak in risk for death which eventually declined below the risk of HER2-low. Across all disease settings, increasing the number of biopsies increased the likelihood of detecting HER2-low status. CONCLUSION: HER2 status changes during neoadjuvant chemotherapy and metastatic progression, and the long-term behaviours of HER2 0 and HER2-low disease differ, underscoring the need for obtaining tissue biopsies and for extended follow-up in breast cancer studies.

Real-world overall survival and characteristics of patients with ER-zero and ER-low HER2-negative breast cancer treated as triple-negative breast cancer: a Swedish population-based cohort study.

Background: Estrogen receptor-low (ER-low) HER2-negative breast cancer has similar pathological and molecular characteristics as triple-negative breast cancer (TNBC), and it is questionable whether it should be considered a separate entity. When the international guidelines lowered the cutoff for ER positivity to ≥1% in 2010, the ≥10% threshold was kept in Sweden. ER-low breast cancer (ER 1-9%) is thus in Sweden treated as TNBC. We aimed to describe patient and tumor characteristics, treatment patterns and overall survival in a Swedish population-based cohort of patients with ER-zero and ER-low HER2-negative breast cancer treated as TNBC. Methods: All TNBC cases diagnosed in Sweden 2008-2020 were included in a population-based cohort study. Patient, tumor and treatment characteristics were analyzed by ER-status (ER 0% vs 1-9%), and associations between subgroups compared using χ2 test. Survival endpoint was overall survival (OS), and Kaplan-Meier curves were estimated. Cox proportional hazards models were used to estimate adjusted hazard ratios comparing ER-low to ER-zero. Findings: Of the 5655 tumors, 90.1% had an ER expression of 0%, while 9.9% were ER-low. ER-low tumors were grade III in 69.4% (80.8% in ER-zero tumors, p-value = 0.001), with a median Ki67 of 60% (63% in ER-zero tumors, p-value = 0.005). There were no significant differences in given chemotherapy (p = 0.546). A pathological complete response (pCR) was achieved in 28.1% of ER-low tumors (25.1% in ER-zero tumors). In the unadjusted analysis of OS, women with ER-low disease had a borderline but not significantly better OS than those with ER-zero disease (HR 0.84 (95% CI 0.71-1.00), p = 0.052). ER-status 1-9% vs 0% was not associated with OS in the multivariable analysis (HR 1.11 (0.90-1.36)). Distant disease-free survival did not differ by ER-status 0% vs 1-9% (HR 0.97 for ER-zero vs ER-low (0.62-1.53), p = 0.905). After preoperative treatment, the impact of pCR for OS did not significantly differ between ER-zero or ER-low disease. Interpretation: ER-low HER2-negative breast cancer has characteristics and prognosis similar to TNBC, when treated in the same way. Therefore, it seems reasonable to use a ≥10% threshold for ER positivity. This would provide patients with ER-low tumors the same treatment opportunities as patients with TNBC, within studies and within clinical routine. Funding: This work was financially supported by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA, in accordance with terms and conditions of a Master Collaboration Agreement between the company and Karolinska Institutet.

Clinical evaluation of deep learning-based risk profiling in breast cancer histopathology and comparison to an established multigene assay.

PURPOSE: To evaluate the Stratipath Breast tool for image-based risk profiling and compare it with an established prognostic multigene assay for risk profiling in a real-world case series of estrogen receptor (ER)-positive and human epidermal growth factor receptor 2 (HER2)-negative early breast cancer patients categorized as intermediate risk based on classic clinicopathological variables and eligible for chemotherapy. METHODS: In a case series comprising 234 invasive ER-positive/HER2-negative tumors, clinicopathological data including Prosigna results and corresponding HE-stained tissue slides were retrieved. The digitized HE slides were analysed by Stratipath Breast. RESULTS: Our findings showed that the Stratipath Breast analysis identified 49.6% of the clinically intermediate tumors as low risk and 50.4% as high risk. The Prosigna assay classified 32.5%, 47.0% and 20.5% tumors as low, intermediate and high risk, respectively. Among Prosigna intermediate-risk tumors, 47.3% were stratified as Stratipath low risk and 52.7% as high risk. In addition, 89.7% of Stratipath low-risk cases were classified as Prosigna low/intermediate risk. The overall agreement between the two tests for low-risk and high-risk groups (N = 124) was 71.0%, with a Cohen's kappa of 0.42. For both risk profiling tests, grade and Ki67 differed significantly between risk groups. CONCLUSION: The results from this clinical evaluation of image-based risk stratification shows a considerable agreement to an established gene expression assay in routine breast pathology.

Comparing deep learning and pathologist quantification of cell-level PD-L1 expression in non-small cell lung cancer whole-slide images.

Programmed death-ligand 1 (PD-L1) expression is currently used in the clinic to assess eligibility for immune-checkpoint inhibitors via the tumor proportion score (TPS), but its efficacy is limited by high interobserver variability. Multiple papers have presented systems for the automatic quantification of TPS, but none report on the task of determining cell-level PD-L1 expression and often reserve their evaluation to a single PD-L1 monoclonal antibody or clinical center. In this paper, we report on a deep learning algorithm for detecting PD-L1 negative and positive tumor cells at a cellular level and evaluate it on a cell-level reference standard established by six readers on a multi-centric, multi PD-L1 assay dataset. This reference standard also provides for the first time a benchmark for computer vision algorithms. In addition, in line with other papers, we also evaluate our algorithm at slide-level by measuring the agreement between the algorithm and six pathologists on TPS quantification. We find a moderately low interobserver agreement at cell-level level (mean reader-reader F1 score = 0.68) which our algorithm sits slightly under (mean reader-AI F1 score = 0.55), especially for cases from the clinical center not included in the training set. Despite this, we find good AI-pathologist agreement on quantifying TPS compared to the interobserver agreement (mean reader-reader Cohen's kappa = 0.54, 95% CI 0.26-0.81, mean reader-AI kappa = 0.49, 95% CI 0.27-0.72). In conclusion, our deep learning algorithm demonstrates promise in detecting PD-L1 expression at a cellular level and exhibits favorable agreement with pathologists in quantifying the tumor proportion score (TPS). We publicly release our models for use via the Grand-Challenge platform.

Training pathologists to assess stromal tumour-infiltrating lymphocytes in breast cancer synergises efforts in clinical care and scientific research.

A growing body of research supports stromal tumour-infiltrating lymphocyte (TIL) density in breast cancer to be a robust prognostic and predicive biomarker. The gold standard for stromal TIL density quantitation in breast cancer is pathologist visual assessment using haematoxylin and eosin-stained slides. Artificial intelligence/machine-learning algorithms are in development to automate the stromal TIL scoring process, and must be validated against a reference standard such as pathologist visual assessment. Visual TIL assessment may suffer from significant interobserver variability. To improve interobserver agreement, regulatory science experts at the US Food and Drug Administration partnered with academic pathologists internationally to create a freely available online continuing medical education (CME) course to train pathologists in assessing breast cancer stromal TILs using an interactive format with expert commentary. Here we describe and provide a user guide to this CME course, whose content was designed to improve pathologist accuracy in scoring breast cancer TILs. We also suggest subsequent steps to translate knowledge into clinical practice with proficiency testing.

Immune cell infiltrate in ductal carcinoma in situ and the risk of dying from breast cancer: case-control study.

BACKGROUND: Studies identifying risk factors for death from breast cancer after ductal carcinoma in situ (DCIS) are rare. In this retrospective nested case-control study, clinicopathological factors in women treated for DCIS and who died from breast cancer were compared with those of patients with DCIS who were free from metastatic disease. METHODS: The study included patients registered with DCIS without invasive carcinoma in Sweden between 1992 and 2012. This cohort was linked to the National Cause of Death Registry. Of 6964 women with DCIS, 96 were registered with breast cancer as cause of death (cases). For each case, up to four controls (318; women with DCIS, alive and without metastatic breast cancer at the time of death of the corresponding case) were selected randomly by incidence density sampling. Whole slides of tumour tissue were evaluated for DCIS grade, comedo necrosis, and intensity of periductal lymphocytic infiltrate. Composition of the immune cell infiltrate, expression of oestrogen receptor, progesterone receptor, human epidermal growth factor receptor 2, and proliferation marker Ki-67 were scored on tissue microarrays. Clinical information was obtained from medical records. Information on date, site, and histological characteristics of local and distant recurrences was obtained from medical records for both cases and controls. RESULTS: Tumour tissue was analysed from 65 cases and 195 controls. Intense periductal lymphocytic infiltrate around DCIS was associated with an increased risk of later dying from breast cancer (OR 2.21. 95% c.i. 1.01 to 4.84). Tumours with more intense lymphocytic infiltrate had a lower T cell/B cell ratio. None of the other biomarkers correlated with increased risk of breast cancer death. CONCLUSION: The immune response to DCIS may influence the risk of dying from breast cancer.

Image-based multiplex immune profiling of cancer tissues: translational implications. A report of the International Immuno-oncology Biomarker Working Group on Breast Cancer.

Recent advances in the field of immuno-oncology have brought transformative changes in the management of cancer patients. The immune profile of tumours has been found to have key value in predicting disease prognosis and treatment response in various cancers. Multiplex immunohistochemistry and immunofluorescence have emerged as potent tools for the simultaneous detection of multiple protein biomarkers in a single tissue section, thereby expanding opportunities for molecular and immune profiling while preserving tissue samples. By establishing the phenotype of individual tumour cells when distributed within a mixed cell population, the identification of clinically relevant biomarkers with high-throughput multiplex immunophenotyping of tumour samples has great potential to guide appropriate treatment choices. Moreover, the emergence of novel multi-marker imaging approaches can now provide unprecedented insights into the tumour microenvironment, including the potential interplay between various cell types. However, there are significant challenges to widespread integration of these technologies in daily research and clinical practice. This review addresses the challenges and potential solutions within a structured framework of action from a regulatory and clinical trial perspective. New developments within the field of immunophenotyping using multiplexed tissue imaging platforms and associated digital pathology are also described, with a specific focus on translational implications across different subtypes of cancer. © 2024 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.