Dataset Information

Deep learning-based predictive biomarker of pathological complete response to neoadjuvant chemotherapy from histological images in breast cancer.

ABSTRACT:

Background

Pathological complete response (pCR) is considered a surrogate endpoint for favorable survival in breast cancer patients treated with neoadjuvant chemotherapy (NAC). Predictive biomarkers of treatment response are crucial for guiding treatment decisions. With the hypothesis that histological information on tumor biopsy images could predict NAC response in breast cancer, we proposed a novel deep learning (DL)-based biomarker that predicts pCR from images of hematoxylin and eosin (H&E)-stained tissue and evaluated its predictive performance.

Methods

In total, 540 breast cancer patients receiving standard NAC were enrolled. Based on H&E-stained images, DL methods were employed to automatically identify tumor epithelium and predict pCR by scoring the identified tumor epithelium to produce a histopathological biomarker, the pCR-score. The predictive performance of the pCR-score was assessed and compared with that of conventional biomarkers including stromal tumor-infiltrating lymphocytes (sTILs) and subtype.

Results

The pCR-score derived from H&E staining achieved an area under the curve (AUC) of 0.847 in predicting pCR directly, and achieved accuracy, F1 score, and AUC of 0.853, 0.503, and 0.822 processed by the logistic regression method, respectively, higher than either sTILs or subtype; a prediction model of pCR constructed by integrating sTILs, subtype and pCR-score yielded a mean AUC of 0.890, outperforming the baseline sTIL-subtype model by 0.051 (0.839, P = 0.001).

Conclusion

The DL-based pCR-score from histological images is predictive of pCR better than sTILs and subtype, and holds the great potentials for a more accurate stratification of patients for NAC.

SUBMITTER: Li F

PROVIDER: S-EPMC8365907 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundInvasive breast cancer patients are increasingly being treated with neoadjuvant chemotherapy; however, only a fraction of the patients respond to it completely. To prevent overtreatment, there is an urgent need for biomarkers to predict treatment response before administering the therapy.MethodsIn this retrospective study, we developed hypothesis-driven interpretable biomarkers based on deep learning, to predict the pathological complete response (pCR, i.e., the absence of tumor cells in the surgical resection specimens) to neoadjuvant chemotherapy solely using digital pathology H&E images of pre-treatment breast biopsies. Our approach consists of two steps: First, we use deep learning to characterize aspects of the tumor micro-environment by detecting mitoses and segmenting tissue into several morphology compartments including tumor, lymphocytes and stroma. Second, we derive computational biomarkers from the segmentation and detection output to encode slide-level relationships of components of the tumor microenvironment, such as tumor and mitoses, stroma, and tumor infiltrating lymphocytes (TILs).ResultsWe developed and evaluated our method on slides from n = 721 patients from three European medical centers with triple-negative and Luminal B breast cancers and performed external independent validation on n = 126 patients from a public dataset. We report the predictive value of the investigated biomarkers for predicting pCR with areas under the receiver operating characteristic curve between 0.66 and 0.88 across the tested cohorts.ConclusionThe proposed computational biomarkers predict pCR, but will require more evaluation and finetuning for clinical application. Our results further corroborate the potential role of deep learning to automate TILs quantification, and their predictive value in breast cancer neoadjuvant treatment planning, along with automated mitoses quantification. We made our method publicly available to extract segmentation-based biomarkers for research purposes.

Project description:Convolutional neural networks (CNNs) are revolutionizing digital pathology by enabling machine learning-based classification of a variety of phenotypes from hematoxylin and eosin (H&E) whole slide images (WSIs), but the interpretation of CNNs remains difficult. Most studies have considered interpretability in a post hoc fashion, e.g. by presenting example regions with strongly predicted class labels. However, such an approach does not explain the biological features that contribute to correct predictions. To address this problem, here we investigate the interpretability of H&E-derived CNN features (the feature weights in the final layer of a transfer-learning-based architecture). While many studies have incorporated CNN features into predictive models, there has been little empirical study of their properties. We show such features can be construed as abstract morphological genes ("mones") with strong independent associations to biological phenotypes. Many mones are specific to individual cancer types, while others are found in multiple cancers especially from related tissue types. We also observe that mone-mone correlations are strong and robustly preserved across related cancers. Importantly, linear mone-based classifiers can very accurately separate 38 distinct classes (19 tumor types and their adjacent normals, AUC = [Formula: see text] for each class prediction), and linear classifiers are also highly effective for universal tumor detection (AUC = [Formula: see text]). This linearity provides evidence that individual mones or correlated mone clusters may be associated with interpretable histopathological features or other patient characteristics. In particular, the statistical similarity of mones to gene expression values allows integrative mone analysis via expression-based bioinformatics approaches. We observe strong correlations between individual mones and individual gene expression values, notably mones associated with collagen gene expression in ovarian cancer. Mone-expression comparisons also indicate that immunoglobulin expression can be identified using mones in colon adenocarcinoma and that immune activity can be identified across multiple cancer types, and we verify these findings by expert histopathological review. Our work demonstrates that mones provide a morphological H&E decomposition that can be effectively associated with diverse phenotypes, analogous to the interpretability of transcription via gene expression values. Our work also demonstrates mones can be interpreted without using a classifier as a proxy.

Project description:Apolipoprotein B mRNA editing enzyme catalytic polypeptide-like 3B (APOBEC3B) is a gene editing enzyme with cytidine deaminase activity and high expression of its mRNA in breast tumors have been shown to be associated with progressive cases and poor prognosis. In this study, we aimed to examine the relationship between the expression of APOBEC3B and the effect of neoadjuvant chemotherapy (NAC) using pretreatment biopsy tissue, and examined whether the expression of APOBEC3B influenced chemotherapy efficacy.We retrospectively selected a total of 274 patients with primary breast cancer who received NAC in more than 4 courses and underwent surgery at our institute. We assessed the expression of APOBEC3B mRNA using pretreatment biopsy specimens of NAC by quantitative real-time PCR (qRT-PCR) and examined the relationship between APOBEC3B mRNA expression and sensitivity to chemotherapy using pathological complete response (pCR) as an indicator. Further, we assessed the prognostic value of APOBEC3B in the patients receiving NAC.APOBEC3B mRNA expression levels were successfully assessed in 173 (63.1%) of the 274 specimens. The total pCR rate was 36.4% (n = 63). An association between APOBEC3B expression levels and pCR was observed (Wilcoxon test, P ? 0.0001). The patients were divided into two groups, low (n = 66) and high (n = 107), according to the APOBEC3B expression levels, using the cut-off value calculated by the receiver operating characteristics (ROC) curve for pCR. The rate of pCR was significantly higher among the patients in the high group than among those in the low group (47.7% vs 18.2%, P ? 0.0001). High APOBEC3B expression was significantly associated with high nuclear grade (P = 0.0078), high Ki-67 labeling index (P = 0.0087), estrogen receptor (ER) negativity (P ? 0.0001) and human epidermal growth factor receptor 2 (HER2) negativity (P = 0.032). Tumor size (P = 0.011), ER (P ? 0.0001), HER2 (P = 0.0013) and APOBEC3B expression (P = 0.037) were independent predictive factors for pCR in multivariate analysis. However, there was no association between APOBEC3B expression and prognosis.Our study showed that APOBEC3B mRNA expression correlated with sensitivity to NAC in breast cancer patients. In contrast to previous studies, APOBEC3B mRNA expression was not associated with breast cancer prognosis in patients receiving NAC.

Project description:BackgroundInflammation plays an important role in tumor proliferation, metastasis, and resistance to chemotherapy. The systemic inflammation response index (SIRI), has been reported to be closely related to prognosis in many tumors, such as breast and gastric cancers. However, the predictive value of pretreatment SIRI on pathological complete response (pCR) rates in patients with breast cancer treated with neoadjuvant chemotherapy (NAC) is unknown. This study examined the correlation between SIRI and pCR in patients with breast cancer receiving NAC and identified convenient and accurate predictive indicators for pCR.MethodsWe retrospectively analyzed the clinicopathological parameters and pretreatment peripheral blood characteristics of the 241 patients with breast cancer who received NAC between June 2015 and June 2020. Receiver operating characteristic (ROC) curves were used to determine the optimal cutoff of SIRI. ROC curves were also plotted to verify the accuracy of inflammatory markers for pCR prediction. The chi-squared test was used to explore the relationships of SIRI with pCR and other clinicopathological parameters. Multivariate analyses were performed using a logistic regression model.ResultsAmong the 241 patients, 48 (19.92%) achieved pCR. pCR was significantly related to SIRI, the neutrophil-lymphocyte ratio (NLR), the lymphocyte-monocyte ratio (LMR), molecular subtypes and other clinicopathological parameters, such as BMI, clinical T and N staging, and histological grade. Multivariate analyses indicated that the clinical T and N staging, SIRI, and NLR were independent prognostic factors for pCR in patients with breast cancer. The area under the ROC curve for SIRI was larger than that for NLR. Compared to patients with SIRI ≥0.72, patients with SIRI < 0.72 had a nearly 5-fold higher chance of obtaining pCR (odds ratio = 4.999, 95% confidence interval = 1.510-16.551, p = 0.000).ConclusionsPretreatment SIRI is predictive of pCR in patients with breast cancer receiving NAC, and the index can assist physicians in formulating personalized treatment strategies.