Predict HER2 Status Research Articles

A radiomics-based interpretable machine learning model to predict the HER2 status in bladder cancer: a multicenter study

ObjectiveTo develop a computed tomography (CT) radiomics-based interpretable machine learning (ML) model to preoperatively predict human epidermal growth factor receptor 2 (HER2) status in bladder cancer (BCa) with multicenter validation.MethodsIn this retrospective study, 207 patients with pathologically confirmed BCa were enrolled and divided into the training set (n = 154) and test set (n = 53). Least absolute shrinkage and selection operator (LASSO) regression was used to identify the most discriminative features in the training set. Five radiomics-based ML models, namely logistic regression (LR), support vector machine (SVM), k-nearest neighbors (KNN), eXtreme Gradient Boosting (XGBoost) and random forest (RF), were developed. The predictive performance of established ML models was evaluated by the area under the receiver operating characteristic curve (AUC). The Shapley additive explanation (SHAP) was used to analyze the interpretability of ML models.ResultsA total of 1218 radiomics features were extracted from the nephrographic phase CT images, and 11 features were filtered for constructing ML models. In the test set, the AUCs of LR, SVM, KNN, XGBoost, and RF were 0.803, 0.709, 0.679, 0.794, and 0.815, with corresponding accuracies of 71.7%, 69.8%, 60.4%, 75.5%, and 75.5%, respectively. RF was identified as the optimal classifier. SHAP analysis showed that texture features (gray level size zone matrix and gray level co-occurrence matrix) were significant predictors of HER2 status.ConclusionsThe radiomics-based interpretable ML model provides a noninvasive tool to predict the HER2 status of BCa with satisfactory discriminatory performance.Critical relevance statementAn interpretable radiomics-based machine learning model can preoperatively predict HER2 status in bladder cancer, potentially aiding in the clinical decision-making process.Key PointsThe CT radiomics model could identify HER2 status in bladder cancer.The random forest model showed a more robust and accurate performance.The model demonstrated favorable interpretability through SHAP method.Graphical

Contrast-enhanced computed tomography-based radiomics nomogram for predicting HER2 status in urothelial bladder carcinoma.

To evaluate the performance of a clinical-radiomics model based on contrast-enhanced computed tomography (CE-CT) in assessing human epidermal growth factor receptor 2 (HER2) status in urothelial bladder carcinoma (UBC). From January 2022 to December 2023, 124 patients with UBC were classified into the training (n=100) and test (n=24) sets. CE-CT scans were performed on the patients. Univariate and multivariate analyses were conducted to identify independent predictors of HER2 status in patients with UBC. We employed eight machine learning algorithms to establish radiomic models. A clinical-radiomics model was developed by integrating radiomic signatures and clinical features. Receiver operating characteristic curves and decision curve analysis (DCA) were generated to evaluate and validate the predictive capabilities of the models. Among the eight classifiers, the random forest radiomics model based on CE-CT demonstrated the highest efficacy in predicting HER2 status, with area under the curve (AUC) values of 0.880 (95% CI: 0.813-0.946) and 0.814 (95% CI: 0.642-0.986) in the training and test sets, respectively. In the training set, the clinical-radiomics model achieved an AUC of 0.935, an accuracy of 0.870, a sensitivity of 0.881, and a specificity of 0.854. In the test set, the clinical-radiomics model achieved an AUC of 0.857, an accuracy of 0.760, a sensitivity of 0.643, and a specificity of 0.900. DCA analysis indicated that the clinical-radiomics model provided good clinical benefit. The radiomics nomogram demonstrates good diagnostic performance in predicting HER2 expression in patients with UBC.

A Comprehensive Model Outperformed the Single Radiomics Model in Noninvasively Predicting the HER2 Status in Patients with Breast Cancer

This study aimed to develop predictive models based on conventional magnetic resonance imaging (cMRI) and radiomics features for predicting human epidermal growth factor receptor 2 (HER2) status of breast cancer (BC) and compare their performance. A total of 287 patients with invasive BC in our hospital were retrospectively analyzed. All patients underwent preoperative breast MRI consisting of fat-suppressed T2-weighted imaging, axial dynamic contrast-enhanced MRI, and diffusion-weighted imaging sequences. From these sequences, radiomics features were derived. Three distinct models were established utilizing cMRI features, radiomics features, and a comprehensive model that amalgamated both. The predictive capabilities of these models were assessed using the receiver operating characteristic curve analysis. The comparative performance was then determined through the DeLong test and net reclassification improvement (NRI). In a randomized split, the 287 patients with BC were allotted to either training (234; 46 HER2-zero, 107 HER2-low, 81 HER2-positive) or test (53; 8 HER2-zero, 27 HER2-low, 18 HER2-positive) at an 8:2 ratio. The mean area under the curve (AUCs) for cMRI, radiomics, and comprehensive models predicting HER2 status were 0.705, 0.819, and 0.859 in training set and 0.639, 0.797, and 0.842 in test set, respectively. DeLong's test indicated that the combined model's AUC surpassed the radiomics model significantly (p<0.05). NRI analysis verified superiority of the combined model over the radiomics for BC HER2 prediction (NRI 25.0) in the test set. The comprehensive model based on the combination of cMRI and radiomics features outperformed the single radiomics model in noninvasively predicting the three-tiered HER2 status in patients with BC.

Abstract PO5-13-06: Classifying HER2-low breast cancer using a combination of ERBB2 mRNA expression and altered genes

Abstract Background It has been recently demonstrated that some HER2-low, defined as immunohistochemistry (IHC) 1+ or 2+ with no gene amplification by FISH, breast cancer (BC) patients respond to trastuzumab deruxtecan (T-DXd). Results of the DAISY trial also suggested clinical activity with T-DXd in some of those classified as HER2-0 cancers by IHC, indicating an unmet need to better identify cancers that may truly respond to T-DXd. We compared ERBB2 (HER2) mRNA expression obtained from whole-transcriptome sequencing with IHC/FISH assignment of HER2 status. Furthermore, we constructed a classifier using ERBB2 expression and alterations in other genes as a proof of concept to illustrate classification of BC samples into HER2-0 versus HER2-low status. Methods We chose consecutive HR positive/HER2-negative BC samples that were submitted for OncoExTraTM testing between April 2020 and October 2022. The OncoExTra test uses tumor-normal whole exome and whole transcriptome sequencing to detect somatic single base substitutions, indels, copy number alterations, gene fusions and alternative transcripts. In addition, we investigated the expression profile of select genes for this analysis. We utilized OncoExTra data to examine the relationship between HER2 IHC status and gene expression, mutation and copy number profiles of selected genes, and built a classifier to predict HER2 status. Results A total of 279 BC samples were analyzed, including 106 HER2-IHC 0, 120 HER2-IHC 1+ and 53 HER2-IHC 2+. Of these, 273 were primary and 6 were metastatic samples. The distributions of patient age and tumor fraction were similar among these categories (p=0.54 and p=0.41, respectively; one-way ANOVA). ERBB2 expression differed among all three HER2-IHC categories and was lowest in HER2-0 and highest in HER2-2+ (HER2-0 vs HER2-1+, p&amp;lt; 0.001; HER2-0 vs HER2-2+, p&amp;lt; 0.0001; HER2-1+ vs HER2-2+, p=0.011; HER2-0 vs HER2-low, p&amp;lt; 0.0001; Mann-Whitney U test in all comparisons). Five genes were mutated in more than 10% of samples: PIK3CA (45.5%), GATA3 (15.4%), CDH1 (12.5%), MAP3K1 (11.1%) and TP53 (10.8%). Mutations in two genes, GATA3 and PTEN, appeared to differ in frequency between the HER2-0 and HER2-low categories (Table 1). We selected 4 genes with mutations, 2 genes with copy number changes, and expression from ERBB2, ESR1, CD274, HLA-A, HLA-B and HLA-C to train a logistic regression classifier to predict HER2 IHC status. The classifier was trained on 163 samples, hyperparameter tuning was done using 55 samples, and testing was performed on 61 samples. Preliminary results indicated that of the 28 HER2-IHC 0 samples in the test set, 17 (65%) were classified as HER2-low. Further development and evaluation of the classifier using a blinded testing set is currently in progress. Conclusion In this cohort of BC samples, ERBB2 mRNA expression was positively associated with HER2 protein IHC status. Additional studies are required to determine whether a classifier including ERBB2 mRNA expression could be used to identify patients for T-DXd therapy. The relative enrichment of PTEN alterations in HER2-IHC 0 samples suggests that mTOR/AKT inhibitors may sometimes be appropriate in patients who are not eligible for T-DXd therapy, and further investigation should be considered. Table 1. Number (frequency) of altered genes by HER2 IHC status. Citation Format: Gargi Basu, Niru Chennagiri, Turgut Dogruluk, David Hall, Jess Hoag, Cynthia Flannery, Snehal Thakkar, Melanie Palomares, Frederick Baehner, Lajos Pusztai. Classifying HER2-low breast cancer using a combination of ERBB2 mRNA expression and altered genes [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-13-06.

Abstract 4643: A novel epigenomic profiling assay reveals disease biology and therapeutic targets in gastro-esophageal cancer from 1 ml of plasma

Abstract INTRODUCTION: Gastro-esophageal adenocarcinoma (GEA) is an aggressive and heterogeneous disease with a poor prognosis. Targeted therapies have failed largely due to challenges associated with biopsy collection, tumor antigen heterogeneity, and accurately distinguishing transcriptionally distinct esophageal, GE junction or gastric tumor types. We developed a novel, multimodal liquid biopsy assay that profiles genome-wide transcriptional activation. Here, we deploy this assay to define clinically relevant insights in a cohort of GEA patients. METHODS: 1 ml of plasma from whole blood samples was used to assess genome-wide enhancer, promoter and DNA methylation loci in 11 GEA patients (Stage III: 1, Stage IV: 10) and 7 healthy volunteers. 6 patients were confirmed as HER2+ based on standard tissue IHC/ISH methods. ctDNA estimation was performed using the ichorCNA algorithm on low pass whole genome sequencing data. A multimodal HER2 classifier developed previously using genome-wide features including HER2 was used to predict HER2 status of the GEA samples. RESULTS: Epigenomic profiling and comparison of GEA and healthy volunteer plasma samples revealed distinct genome-wide transcriptional activation patterns associated with gastroesophageal adenocarcinoma, including (i) the characteristic upregulation of HER2, FGFR2 and VEGF (ii) SOX2, KLF4, GATA6 transcription factor circuitry, and (iii) DNA methylation at the CHFR and MGMT promoters. We then explored if epigenomic profiling of liquid biopsies could infer the activation status of antibody-drug conjugate (ADC) targets currently under investigation in GEA cancers. Simultaneous examination of multiple target genes including HER2, HER3, TROP2, CLDN18, CEACAM5 and MET demonstrated patient-specific differences and dynamic range in transcriptional activation of a subset of these target genes. Finally, we investigated if HER2 status could be assigned accurately and quantitatively from a liquid biopsy using the multimodal HER2 classifier. HER2 positivity by IHC was predicted accurately in 8/11 samples corresponding to an area under the ROC curve (AUC) greater than 0.85. Discordance between plasma and tissue HER2 calls could be partially attributed to variability associated with tumor site/section used for tissue collection, IHC staining, and time between biopsy collection and blood draw. CONCLUSIONS: We demonstrate the feasibility of comprehensive epigenomic profiling from low input volume of 1ml of plasma to provide a snapshot of the tumor’s transcriptional state to derive clinically actionable insights. The HER2 classifier overcomes several of the challenges with IHC profiling by providing a uniform and quantitative measurement of HER2 activation. This non-invasive assay provides an accurate reflection of tumor transcriptional biology that can inform optimal patient selection strategies. Citation Format: Jonathan Beagan, Aparna Gorthi, Anthony D'Ippolito, Travis Clark, Michael Coyne, Tyrone Tamakloe, Amy Donahue, Baovy Tran, Hyun-Hwan Jeong, Kristian Cibulskis, Hathairat Sawaengsri, Corrie A. Painter, Juliann Chmielecki, J Carl Barrett, Mosammat Faria Afreen, Aditya Pandey, Ilexa Ashley Schechter, Martin S. Taylor, Matthew R. Strickland, Matthew Eaton, Samuel J. Klempner. A novel epigenomic profiling assay reveals disease biology and therapeutic targets in gastro-esophageal cancer from 1 ml of plasma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4643.

Intra- and peritumoral radiomics features based on multicenter automatic breast volume scanner for noninvasive and preoperative prediction of HER2 status in breast cancer: a model ensemble research

The aim to investigate the predictive efficacy of automatic breast volume scanner (ABVS), clinical and serological features alone or in combination at model level for predicting HER2 status. The model weighted combination method was developed to identify HER2 status compared with single data source model method and feature combination method. 271 patients with invasive breast cancer were included in the retrospective study, of which 174 patients in our center were randomized into the training and validation sets, and 97 patients in the external center were as the test set. Radiomics features extracted from the ABVS-based tumor, peritumoral 3 mm region, and peritumoral 5 mm region and clinical features were used to construct the four types of the optimal single data source models, Tumor, R3mm, R5mm, and Clinical model, respectively. Then, the model weighted combination and feature combination methods were performed to optimize the combination models. The proposed weighted combination models in predicting HER2 status achieved better performance both in validation set and test set. For the validation set, the single data source model, the feature combination model, and the weighted combination model achieved the highest area under the curve (AUC) of 0.803 (95% confidence interval [CI] 0.660–947), 0.739 (CI 0.556,0.921), and 0.826 (95% CI 0.689,0.962), respectively; with the sensitivity and specificity were 100%, 62.5%; 81.8%, 66.7%; 90.9%,75.0%; respectively. For the test set, the single data source model, the feature combination model, and the weighted combination model attained the best AUC of 0.695 (95% CI 0.583, 0.807), 0.668 (95% CI 0.555,0.782), and 0.700 (95% CI 0.590,0.811), respectively; with the sensitivity and specificity were 86.1%, 41.9%; 61.1%, 71.0%; 86.1%, 41.9%; respectively. The model weighted combination was a better method to construct a combination model. The optimized weighted combination models composed of ABVS-based intratumoral and peritumoral radiomics features and clinical features may be potential biomarkers for the noninvasive and preoperative prediction of HER2 status in breast cancer.

Machine learning-based model constructed from ultrasound radiomics and clinical features for predicting HER2 status in breast cancer patients with indeterminate (2+) immunohistochemical results.

We aimed to predict human epidermal growth factor receptor 2 (HER2) 2+ status in patients with breast cancer by constructing and validating machine learning models utilizing ultrasound (US) radiomics and clinical features. We analyzed 203 breast cancer cases immunohistochemically determined as HER2 2+ and used fluorescence insitu hybridization (FISH) as the confirmation method. From each case, the study analyzed 840 extracted radiomics features and 11 clinicopathologic features. Cases were randomly split into training (n = 141) and validation sets (n = 62) at a 7:3 ratio. Univariate logistic regression analysis was first performed on the 11 clinicopathologic characteristics. The least absolute shrinkage and selection operator (LASSO) and decision tree (DT) techniques were employed for post-feature selection. Finally, 19 radiomics features were utilized in logistic regression (LR) and Naive Bayesian (NB) classifiers. Model performance was gauged using the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity. Our models exhibited notable diagnostic efficacy in differentiating HER2-positive from negative breast cancer cases. In the validation sets, the LR model outperformed the NB model with an AUC of 0.860 and accuracy of 83.8% compared to NB's AUC of 0.684 and accuracy of 79.0%. The LR model demonstrated higher sensitivity (92.3% vs. 46.2%) while the NB model had a better specificity (91.8% vs. 63.3%) in the validation set. Machine learning models grounded on radiomics efficiently predicted IHC HER2 2+ status in breast cancer patients, suggesting potential enhancements in clinical decision-making for treatment and management.

Predicting human epidermal growth factor receptor 2 status of patients with gastric cancer by computed tomography and clinical features.

There have been no studies on predicting human epidermal growth factor receptor 2 (HER2) status in patients with resectable gastric cancer (GC) in the neoadjuvant and perioperative settings. We aimed to investigate the use of preoperative contrast-enhanced computed tomography (CECT) imaging features combined with clinical characteristics for predicting HER2 expression in GC. We retrospectively enrolled 301 patients with GC who underwent curative resection and preoperative CECT. HER2 status was confirmed by postoperative immunohistochemical analysis with or without fluorescence in situ hybridization. A prediction model was developed using CECT imaging features and clinical characteristics that were independently associated with HER2 status using multivariate logistic regression analysis. Receiver operating characteristic curves were constructed and the performance of the prediction model was evaluated. The bootstrap method was used for internal validation. Three CECT imaging features and one serum tumor marker were independently associated with HER2 status in GC: enhancement ratio in the arterial phase (odds ratio [OR] = 4.535; 95% confidence interval [CI], 2.220-9.264), intratumoral necrosis (OR = 2.64; 95% CI, 1.180-5.258), tumor margin (OR = 3.773; 95% CI, 1.968-7.235), and cancer antigen 125 (CA125) level (OR = 5.551; 95% CI, 1.361-22.651). A prediction model derived from these variables showed an area under the receiver operating characteristic curve of 0.802 (95% CI, 0.740-0.864) for predicting HER2 status in GC. The established model was stable, and the parameters were accurately estimated. Enhancement ratio in the arterial phase, intratumoral necrosis, tumor margin, and CA125 levels were independently associated with HER2 status in GC. The prediction model derived from these factors may be used preoperatively to estimate HER2 status in GC and guide clinical treatment.

HER2/ERBB2 copy number analysis by targeted next-generation sequencing in breast cancer.

A combination of immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH) is the current standard of care for HER2 evaluation in breast cancer. Here, we investigate the potential clinical utility of next-generation sequencing (NGS)-derived HER2/ERBB2 copy number (CN) data for predicting HER2 status as defined by American Society of Clinical Oncology (ASCO)/College of American Pathologists (CAP) guidelines. In total, 294 locally recurrent and metastatic breast cancers previously tested by targeted hybrid capture-based NGS and by HER2 IHC/FISH were included. Analyses focused on the ERBB2 median log2 ratios and start-end genomic coordinates from NGS, average HER2 CN and HER2/CEP17 ratios from FISH, and the HER2 IHC scores. We also determined a more stringent log2 ratio cutoff to predict HER2-positive status with 100% specificity. Sixty-four (22%) cases were HER2 positive and 230 (78%) were HER2 negative by ASCO/CAP guidelines. The ERBB2 median log2 ratios from NGS strongly correlated with HER2 status by IHC/FISH (area under receiver operator characteristic curve = 0.951). ERBB2 log2 ratio more than 1.7 was 100% specific for HER2-positive results by IHC/FISH. Start and end genomic coordinates for regions of gain near ERBB2 by NGS also predicted HER2 status. Copy number data from our NGS panel strongly correlate with HER2 status. Using a stringent cutoff, ERBB2 log2 ratio accurately predicts HER2 positivity with high specificity. The NGS CN assessment may have utility in determining HER2 status in certain clinical settings.

Combining conventional ultrasound and ultrasound elastography to predict HER2 status in patients with breast cancer.

Introduction: Identifying the HER2 status of breast cancer patients is important for treatment options. Previous studies have shown that ultrasound features are closely related to the subtype of breast cancer. Methods: In this study, we used features of conventional ultrasound and ultrasound elastography to predict HER2 status. Results and Discussion: The performance of model (AUROC) with features of conventional ultrasound and ultrasound elastography is higher than that of the model with features of conventional ultrasound (0.82 vs. 0.53). The SHAP method was used to explore the interpretability of the models. Compared with HER2- tumors, HER2+ tumors usually have greater elastic modulus parameters and microcalcifications. Therefore, we concluded that the features of conventional ultrasound combined with ultrasound elastography could improve the accuracy for predicting HER2 status.

Characteristics and predictive modeling of HER2 mutation in patients with non-small cell lung cancer.

e13559 Background: Recent advances in targeted therapies have increased the survival and quality of life for non-small cell lung cancer (NSCLC) patients. Understanding clinical characteristics and predicting HER2 status may facilitate efficient use of next-generation sequencing (NGS) testing and promote precision medicine/targeted therapies. Methods: NSCLC patients ≤65 years old, diagnosed between 01/2011 and 06/2021 who have NGS test results were identified in from two National Cancer Institute designated cancer centers in the Northeastern region of the United States. Demographics, behavioral, and clinical characteristics were extracted from electronic health record (EHR) data and EHR-linked cancer registry. Prevalence prediction modelling of HER2/ERBB2 mutation was performed using machine learning (ML) methods including logistic regression (LR), random forest (RF), naïve Bayes (NB), extreme gradient boosting (XGBoost) and multi-layer perceptron (MLP). 10- fold cross validation and k-nearest neighbor method were used to impute missing values. Under-sampling, oversampling and feature selection methods were tried. Results: Among 761 NSCLC patients with NGS reports available, 353 (46%) were ≤65 years old. Among 353 (mean age 57, 58% female, 88% non-Hispanic, 69% white, 57% had metastasis, 58% prescribed chemotherapy) 25 (7%) had a HER2 mutation. Based on univariate significance tests with LR on the complete data: cerebrovascular disease, BMI between 27.5-29.9, KRAS, EGFR and MET were significant with odds ratios (ORs) ranging from 0.2-3.9. In our prediction models, BMI, cerebrovascular disease, weight loss, hepatic metastases, diabetes, congestive heart failure, and KRAS, EGFR, MET and RET alteration status were significant predictors in multivariate analyses, suggesting the potential usefulness of these features in predicting HER2 mutation among patients with NSCLC. Area under the ROC curve (AUC) varied across different models from 70.9% (RF), 70% (XGB and MLP), 69.5% (LR) to 68.2% (Naïve Bayes). The best performing model with RF yielding an AUC ~71% can provide relative high specificity and acceptable sensitivity depending on the cut off. Conclusions: Our prediction models yielded AUC around 70% with the best performing model’s AUC being 71% in a subgroup aged ≤65. After further validation, our prediction model may be used to identify patients appropriate for repeat NGS testing who did not have HER2 testing in first molecular profiling. Future work is needed to predict status in repeat testing with previously test negative patients who relapse or have progression of disease. Random forest (RF) Model performance across different cutoff for the predicted probability. [Table: see text]

Abstract P5-02-33: Deep learning-based assessment of HER2-low expression on breast cancer H&amp;E digital whole slide images

Abstract Background: Antibody drug conjugates (ADCs) against HER2 have shown meaningful clinical activity in HER2 low breast cancers, defined as 1+ or 2+ staining on immunohistochemistry (IHC) without gene amplification by in situ hybridization (ISH) techniques. Given that these methods were originally developed for an accurate detection of HER2 3+, their sensitivity and robustness for the detection of low and ultra-low levels of HER2 are questionable. We have recently described a deep learning algorithm that can detect signatures of HER2 expression based on training utilizing scanned H&amp;E whole slide images (WSI) of breast cancers for which IHC and mRNA expression levels of HER2 were available. Here, we report the application of our algorithm to two independent breast cancer cohorts. Methods: A model was developed based on recognition of invasive breast cancer in whole slide images of H&amp;E staining, and then trained via computational neural network with multiple instance learning for binary classification of cases as HER2 “negative” and HER2 “expressed” (low). For training, true negatives were defined as having HER2 IHC-0 and mRNA level &amp;lt; 7.6. HER2-low cases were defined as IHC-1+/2+ and mRNA &amp;gt;9. IHC-0 cases with mRNA &amp;gt;7.6 were excluded from the training cohorts. The resulting model (HER2Complete) was able to distinguish HER2-negatives from HER2-low cases with an AUC of 0.91 (+/- 0.08). Here we use Her2Complete to assess HER2 in two additional cohorts that include 901 ER+/HER2 IHC-0 and 52 HER2 IHC 0+ breast cancers from MSK and TCGA cohorts, respectively. For the TCGA cohort, concomitant transcriptomics data (RNASeq) as a reference for HER2 mRNA expression were retrieved and “HER2 expressed” defined as RNASeq expression of HER2 greater than the 90th percentile of the geometric mean of expression of three reference genes not expressed in breast tissues (TTN, MUC13, OR10A6). Values less than this reference cut-off in the TCGA cohort were considered “HER2 not expressed.” Results: Among the 901 IHC-0 test cases from the MSK cohort, the model identified 82 as ‘negative’, whereas 819 were found to have features of HER2 expression (HER2-Low). Of the 82 negative cases in the MSK cohort, all except 13 cases expressed mRNA levels &amp;lt; 9, and 786/819 of the HER2-low cases expressed mRNA levels &amp;gt;8. Of the 52 IHC 0+ cases in the TCGA cohort, 33 also had “HER2 not expressed” by our reference based RNASeq expression cut-off. Our model identified 15 of these 33 as ‘negative’, while 15 of the 19 TCGA cases with IHC 0+ and HER2 ‘expressed’ by our cut-off were identified as ‘HER2-Low’ by our model. Conclusions: AI tools based on the analysis of WSIs of routinely prepared H&amp;E sections may predict HER2 status in breast cancer. This work requires further investigation using treatment response data to demonstrate that cases with morphologic features of low level HER2 expression will respond to ADCs. Citation Format: Gerard Oakley III, Jorge Reis-Filho, David Klimstra, Marc Goldfinger, Yikan Wang, Antonio Marra. Deep learning-based assessment of HER2-low expression on breast cancer H&amp;E digital whole slide images [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P5-02-33.

Association of lesion contour and lesion composition on MR with HER2 status in breast cancer brain metastases

Background and purposeWith the development of HER2-directed therapies, identifying non-invasive imaging biomarkers of HER2 status in breast cancer brain metastases has become increasingly important, particularly given the risks of tissue sampling within the brain and the possibility of a change in receptor expression from the primary tumor to the brain metastasis. The purpose of this study was to evaluate whether lesion contour and composition on MR could help identify the HER2 status of breast cancer brain metastases. Materials and methodsWe derived a cohort of 34 women with a mean age of 55 years (range: 31–81 years) with a total of 47 distinct histologically proven breast cancer brain metastases with preoperative contrast-enhanced brain MR and HER2 immunohistochemistry and/or fluorescent in-situ hybridization (FISH) of the resected/biopsied brain specimens from 2018 to 2021. Two fellowship-trained neuroradiologists evaluated the lesion contour and lesion composition of each lesion. Logistic regression analyses were performed. ResultsIn a logistic regression model, an irregular contour had an odds ratio of 170 (p = 0.007) in differentiating HER2-positive from HER2-negative lesions. In a logistic regression model, when compared to a predominantly cystic lesion composition, a solid lesion composition had an odds ratio of 17 (p = 0.016) in differentiating HER2-positive from HER2-negative lesions. ConclusionLesion contour and lesion composition on MR were significantly associated with the HER2 status of breast cancer brain metastases. Current assessment of HER2 status requires tissue sampling and immunochemical and/or FISH analyses. A non-invasive imaging biomarker that may help predict HER2 status may be of great clinical value.

Development and validation of multivariate models integrating preoperative clinicopathological and radiographic findings to predict HER2 status in gastric cancer

The combination of trastuzumab and chemotherapy is recommended as first-line therapy for patients with human epidermal growth factor receptor 2 (HER2) positive advanced gastric cancers (GCs). Successful trastuzumab-induced targeted therapy should be based on the assessment of HER2 overexpression. This study aimed to evaluate the feasibility of multivariate models based on hematological parameters, endoscopic biopsy, and computed tomography (CT) findings for assessing HER2 overexpression in GC. This retrospective study included 183 patients with GC, and they were divided into primary (n = 137) and validation (n = 46) cohorts at a ratio of 3:1. Hematological parameters, endoscopic biopsy, CT morphological characteristics, and CT value-related and texture parameters of all patients were collected and analyzed. The mean corpuscular hemoglobin concentration value, morphological type, 3 CT value-related parameters, and 22 texture parameters in three contrast-enhanced phases differed significantly between the two groups (all p < 0.05). Multivariate models based on the regression analysis and support vector machine algorithm achieved areas under the curve of 0.818 and 0.879 in the primary cohort, respectively. The combination of hematological parameters, CT morphological characteristics, CT value-related and texture parameters could predict HER2 overexpression in GCs with satisfactory diagnostic efficiency. The decision curve analysis confirmed the clinical utility.

SlideGraph+: Whole slide image level graphs to predict HER2 status in breast cancer

Human epidermal growth factor receptor 2 (HER2) is an important prognostic and predictive factor which is overexpressed in 15-20% of breast cancer (BCa). The determination of its status is a key clinical decision making step for selection of treatment regimen and prognostication. HER2 status is evaluated using transcriptomics or immunohistochemistry (IHC) through in-situ hybridisation (ISH) which incurs additional costs and tissue burden and is prone to analytical variabilities in terms of manual observational biases in scoring. In this study, we propose a novel graph neural network (GNN) based model (SlideGraph+) to predict HER2 status directly from whole-slide images of routine Haematoxylin and Eosin (H&E) stained slides. The network was trained and tested on slides from The Cancer Genome Atlas (TCGA) in addition to two independent test datasets. We demonstrate that the proposed model outperforms the state-of-the-art methods with area under the ROC curve (AUC) values > 0.75 on TCGA and 0.80 on independent test sets. Our experiments show that the proposed approach can be utilised for case triaging as well as pre-ordering diagnostic tests in a diagnostic setting. It can also be used for other weakly supervised prediction problems in computational pathology. The SlideGraph+ code repository is available at https://github.com/wenqi006/SlideGraph along with an IPython notebook showing an end-to-end use case at https://github.com/TissueImageAnalytics/tiatoolbox/blob/develop/examples/full-pipelines/slide-graph.ipynb.

HEROHE Challenge: Predicting HER2 Status in Breast Cancer from Hematoxylin-Eosin Whole-Slide Imaging.

Breast cancer is the most common malignancy in women worldwide, and is responsible for more than half a million deaths each year. The appropriate therapy depends on the evaluation of the expression of various biomarkers, such as the human epidermal growth factor receptor 2 (HER2) transmembrane protein, through specialized techniques, such as immunohistochemistry or in situ hybridization. In this work, we present the HER2 on hematoxylin and eosin (HEROHE) challenge, a parallel event of the 16th European Congress on Digital Pathology, which aimed to predict the HER2 status in breast cancer based only on hematoxylin–eosin-stained tissue samples, thus avoiding specialized techniques. The challenge consisted of a large, annotated, whole-slide images dataset (509), specifically collected for the challenge. Models for predicting HER2 status were presented by 21 teams worldwide. The best-performing models are presented by detailing the network architectures and key parameters. Methods are compared and approaches, core methodologies, and software choices contrasted. Different evaluation metrics are discussed, as well as the performance of the presented models for each of these metrics. Potential differences in ranking that would result from different choices of evaluation metrics highlight the need for careful consideration at the time of their selection, as the results show that some metrics may misrepresent the true potential of a model to solve the problem for which it was developed. The HEROHE dataset remains publicly available to promote advances in the field of computational pathology.

Predicting HER2 Status in Breast Cancer on Ultrasound Images Using Deep Learning Method.

PurposeThe expression of human epidermal growth factor receptor 2 (HER2) in breast cancer is critical in the treatment with targeted therapy. A 3-block-DenseNet-based deep learning model was developed to predict the expression of HER2 in breast cancer by ultrasound images.MethodsThe data from 144 breast cancer patients with preoperative ultrasound images and clinical information were retrospectively collected from the Shandong Province Tumor Hospital. An end-to-end 3-block-DenseNet deep learning classifier was built to predict the expression of human epidermal growth factor receptor 2 by ultrasound images. The patients were randomly divided into a training (n = 108) and a validation set (n = 36).ResultsOur proposed deep learning model achieved an encouraging predictive performance in the training set (accuracy = 85.79%, AUC = 0.87) and the validation set (accuracy = 80.56%, AUC = 0.84). The effectiveness of our model significantly exceeded the clinical model and the radiomics model. The score of the proposed model showed significant differences between HER2-positive and -negative expression (p < 0.001).ConclusionsThese results demonstrate that ultrasound images are predictive of HER2 expression through a deep learning classifier. Our method provides a non-invasive, simple, and feasible method for the prediction of HER2 expression without the manual delineation of the regions of interest (ROI). The performance of our deep learning model significantly exceeded the traditional texture analysis based on the radiomics model.

Quantitative Parameters of Diffusion Spectrum Imaging: HER2 Status Prediction in Patients With Breast Cancer.

ObjectiveTo explore the value of quantitative parameters derived from diffusion spectrum imaging (DSI) in preoperatively predicting human epidermal growth factor receptor 2 (HER2) status in patients with breast cancer.MethodsIn this prospective study, 114 and 56 female patients with invasive ductal carcinoma were consecutively included in a derivation cohort and an independent validation cohort, respectively. Each patient was categorized into HER2-positive or HER2-negative groups based on the pathologic result. All patients underwent DSI and conventional MRI including dynamic contrast-enhanced MRI (DCE-MRI) and diffusion-weighted imaging (DWI). The tumor size, type of the time-signal intensity curve (TIC) from DCE-MRI, apparent diffusion coefficient (ADC) from DWI, and quantitative parameters derived from DSI, including diffusion tensor imaging (DTI), diffusion kurtosis imaging (DKI), mean apparent propagator (MAP), and neurite orientation dispersion and density imaging (NODDI) of primary tumors, were measured and compared between the HER2-positive and HER2-negative groups in the derivation cohort. Univariable and multivariable logistic regression analyses were used to determine the potential independent predictors of HER2 status. The discriminative ability of quantitative parameters was assessed by receiver operating characteristic (ROC) curve analyses and validated in the independent cohort.ResultsIn the derivation cohort, the tumor size, TIC type, and ADC values did not differ between the HER2-positive and HER2-negative groups (p = 0.126–0.961). DSI quantitative parameters including axial kurtosis of DKI (DKI_AK), non-Gaussianity (MAP_NG), axial non-Gaussianity (MAP_NGAx), radial non-Gaussianity (MAP_NGRad), return-to-origin probability (MAP_RTOP), return-to-axis probability of MAP (MAP_RTAP), and intracellular volume fraction of NODDI (NODDI_ICVF) were lower in the HER2-positive group than in the HER2-negative group (p ≤ 0.001–0.035). DSI quantitative parameters including radial diffusivity (DTI_RD), mean diffusivity of DTI (DTI_MD), mean squared diffusion (MAP_MSD), and q-space inverse variance of MAP (MAP_QIV) were higher in the HER2-positive group than in the HER2-negative group (p = 0.016–0.049). The ROC analysis showed that the area under the curve (AUC) of ADC was 0.632 and 0.568, respectively, in the derivation and validation cohorts. The AUC values of DSI quantitative parameters ranged from 0.628 to 0.700 and from 0.673 to 0.721, respectively, in the derivation and validation cohorts. Logistic regression analysis showed that only NODDI_ICVF was an independent predictor of HER2 status (p = 0.001), with an AUC of 0.700 and 0.721, respectively, in the derivation and validation cohorts.ConclusionsDSI could be helpful for preoperative prediction of HER2, but DSI alone may not be sufficient in predicting HER2 status preoperatively in patients with breast cancer.

Deep learning trained on hematoxylin and eosin tumor region of Interest predicts HER2 status and trastuzumab treatment response in HER2+ breast cancer

The current standard of care for many patients with HER2-positive breast cancer is neoadjuvant chemotherapy in combination with anti-HER2 agents, based on HER2 amplification as detected by in situ hybridization (ISH) or protein immunohistochemistry (IHC). However, hematoxylin & eosin (H&E) tumor stains are more commonly available, and accurate prediction of HER2 status and anti-HER2 treatment response from H&E would reduce costs and increase the speed of treatment selection. Computational algorithms for H&E have been effective in predicting a variety of cancer features and clinical outcomes, including moderate success in predicting HER2 status. In this work, we present a novel convolutional neural network (CNN) approach able to predict HER2 status with increased accuracy over prior methods. We trained a CNN classifier on 188 H&E whole slide images (WSIs) manually annotated for tumor Regions of interest (ROIs) by our pathology team. Our classifier achieved an area under the curve (AUC) of 0.90 in cross-validation of slide-level HER2 status and 0.81 on an independent TCGA test set. Within slides, we observed strong agreement between pathologist annotated ROIs and blinded computational predictions of tumor regions / HER2 status. Moreover, we trained our classifier on pre-treatment samples from 187 HER2+ patients that subsequently received trastuzumab therapy. Our classifier achieved an AUC of 0.80 in a five-fold cross validation. Our work provides an H&E-based algorithm that can predict HER2 status and trastuzumab response in breast cancer at an accuracy that may benefit clinical evaluations.

The Value of Predicting Human Epidermal Growth Factor Receptor 2 Status in Adenocarcinoma of the Esophagogastric Junction on CT-Based Radiomics Nomogram.

PurposeWe developed and validated a CT-based radiomics nomogram to predict HER2 status in patients with adenocarcinoma of esophagogastric junction (AEG).MethodA total of 101 patients with HER2-positive (n=46) and HER2-negative (n=55) esophagogastric junction adenocarcinoma (AEG) were retrospectively analyzed. They were then randomly divided into a training cohort (n=70) and a verification cohort (n=31). The radiomics features were obtained from the portal phase of the CT enhanced scan. We used the least absolute shrinkage and selection operator (LASSO) logistic regression method to select the best radiomics features in the training cohort, combined them linearly, and used the radiomics signature formula to calculate the radiomics score (Rad-score) of each AEG patient. A multivariable logistic regression method was applied to develop a prediction model that incorporated the radiomics signature and independent risk predictors. The prediction performance of the nomogram was evaluated using the training and validation cohorts.ResultIn the training (P<0.001) and verification groups (P<0.001), the radiomics signature combined with seven radiomics features was significantly correlated with HER2 status. The nomogram composed of CT-reported T stage and radiomics signature showed very good predictive performance for HER2 status. The area under the curve (AUC) of the training cohort was 0.946 (95% CI: 0.919–0.973), and that of the validation group was 0.903 (95% CI: 0.847–0.959). The calibration curve of the radiomics nomogram showed a good degree of calibration. Decision-curve analysis revealed that the radiomics nomogram was useful.ConclusionThe nomogram CT-based radiomics signature combined with CT-reported T stage can better predict the HER2 status of AEG before surgery. It can be used as a non-invasive prediction tool for HER2 status and is expected to guide clinical treatment decisions in clinical practice, and it can assist in the formulation of individualized treatment plans.