1
|
Cordova C, Muñoz R, Olivares R, Minonzio JG, Lozano C, Gonzalez P, Marchant I, González-Arriagada W, Olivero P. HER2 classification in breast cancer cells: A new explainable machine learning application for immunohistochemistry. Oncol Lett 2022; 25:44. [PMID: 36644146 PMCID: PMC9811637 DOI: 10.3892/ol.2022.13630] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/15/2022] [Accepted: 10/31/2022] [Indexed: 12/15/2022] Open
Abstract
The immunohistochemical (IHC) evaluation of epidermal growth factor 2 (HER2) for the diagnosis of breast cancer is still qualitative with a high degree of inter-observer variability, and thus requires the incorporation of complementary techniques such as fluorescent in situ hybridization (FISH) to resolve the diagnosis. Implementing automatic algorithms to classify IHC biomarkers is crucial for typifying the tumor and deciding on therapy for each patient with better performance. The present study aims to demonstrate that, using an explainable Machine Learning (ML) model for the classification of HER2 photomicrographs, it is possible to determine criteria to improve the value of IHC analysis. We trained a logistic regression-based supervised ML model with 393 IHC microscopy images from 131 patients, to discriminate between upregulated and normal expression of the HER2 protein. Pathologists' diagnoses (IHC only) vs. the final diagnosis complemented with FISH (IHC + FISH) were used as training outputs. Basic performance metrics and receiver operating characteristic curve analysis were used together with an explainability algorithm based on Shapley Additive exPlanations (SHAP) values to understand training differences. The model could discriminate amplified IHC from normal expression with better performance when the training output was the IHC + FISH final diagnosis (IHC vs. IHC + FISH: area under the curve, 0.94 vs. 0.81). This may be explained by the increased analytical impact of the membrane distribution criteria over the global intensity of the signal, according to SHAP value interpretation. The classification model improved its performance when the training input was the final diagnosis, downplaying the weighting of the intensity of the IHC signal, suggesting that to improve pathological diagnosis before FISH consultation, it is necessary to emphasize subcellular patterns of staining.
Collapse
Affiliation(s)
- Claudio Cordova
- Cell Function and Structure Laboratory (EFC Lab.), Faculty of Medicine, Universidad de Valparaíso, Valparaíso 2341386, Chile,PhD Program in Health Sciences and Engineering, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile
| | - Roberto Muñoz
- PhD Program in Health Sciences and Engineering, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile,School of Informatics Engineering, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile
| | - Rodrigo Olivares
- School of Informatics Engineering, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile,Center for Research and Development in Health Engineering, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile
| | - Jean-Gabriel Minonzio
- PhD Program in Health Sciences and Engineering, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile,School of Informatics Engineering, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile,Center for Research and Development in Health Engineering, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile,Millennium Institute for Intelligent Healthcare: iHEALTH, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile
| | - Carlo Lozano
- Pathological Anatomy Service, Carlos Van Buren Hospital, Valparaíso 2340105, Chile
| | - Paulina Gonzalez
- Pathological Anatomy Service, Carlos Van Buren Hospital, Valparaíso 2340105, Chile,School of Medical Technology, Andrés Bello National University (UNAB), Viña del Mar, 2520000, Chile
| | - Ivanny Marchant
- Medical Modeling Laboratory, Faculty of Medicine, Universidad de Valparaíso, Valparaíso 2362735, Chile
| | - Wilfredo González-Arriagada
- Faculty of Dentistry, Universidad de los Andes, Santiago 7620086, Chile,Biomedical Research and Innovation Center (CIIB), Universidad de los Andes, Santiago 7620086, Chile
| | - Pablo Olivero
- Cell Function and Structure Laboratory (EFC Lab.), Faculty of Medicine, Universidad de Valparaíso, Valparaíso 2341386, Chile,PhD Program in Health Sciences and Engineering, Faculty of Engineering, Universidad de Valparaíso, Valparaíso 2362735, Chile,Correspondence to: Dr Pablo Olivero, Cell Function and Structure Laboratory (EFC Lab.), Faculty of Engineering, Universidad de Valparaíso, 2664 Hontaneda, Valparaíso 2341386, Chile, E-mail:
| |
Collapse
|
2
|
Wang L, Jiang Q, He MY, Shen P. HER2 changes to positive after neoadjuvant chemotherapy in breast cancer: A case report and literature review. World J Clin Cases 2022; 10:260-267. [PMID: 35071526 PMCID: PMC8727267 DOI: 10.12998/wjcc.v10.i1.260] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 02/28/2021] [Revised: 06/10/2021] [Accepted: 07/12/2021] [Indexed: 02/06/2023] Open
Abstract
BACKGROUND As the most common cancer in women, breast cancer is the leading cause of death. Most patients are initially diagnosed as stage I-III. Among those without distant metastases, 64% are local tumors and 27% are regional tumors. Patients in stage IIA-IIIC and those who meet the breast-conserving criterion with the exception of tumor size can consider neoadjuvant chemotherapy (NACT). It is worth noting that the status of tumor cell biomarkers is not consistently static. Endocrine-related estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2) encoded by erythroblastic leukemia viral oncogene homolog 2 gene can all alter from positive to negative or vice versa, especially in luminal B subtype after NACT. In addition, determination of HER2 status currently mainly relies on immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH), but FISH is commonly used when the result of IHC is uncertain. HER2 is regarded as negative when the IHC result is 0/1+ without the addition of FISH. To the best of our knowledge, this is the first report of a case harboring HER2 status transformation and IHC1+ with positive amplification by FISH after NACT.
CASE SUMMARY A 49-year-old woman discovered a mass in her right breast and underwent diagnostic workup. Biopsies of the right breast lesion and axillary lymph nodes were obtained. The results pointed to invasive ductal carcinoma with the IHC result for ER (80%), PR (60%), Ki-67 (20%) and ambiguous expression of HER2 (IHC 2+) with negative amplification by FISH (HER2/CEP17 ratio of 1.13). She underwent surgery after NACT. The pathological findings of the surgically resected sample supported invasive ductal carcinoma with the tumor measuring 1.1 cm × 0.8 cm × 0.5 cm and had spread to one of fifteen dissected lymph nodes. Retesting of the specimen showed that the tumor was positive for ER (2+, 85%) and PR (2+, 10%) but negative for HER2 by IHC (1+). Also Ki-67 had dropped to 2%. The patient was regularly monitored every 3 mo without evidence of recurrence.
CONCLUSION Biomarker status should be reassessed after NACT especially in luminal subtypes.
Collapse
Affiliation(s)
- Luo Wang
- Department of Medical Oncology, The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou 310003, Zhejiang Province, China
| | - Qi Jiang
- Department of Medical Oncology, The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou 310003, Zhejiang Province, China
| | - Meng-Ye He
- Department of Medical Oncology, The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou 310003, Zhejiang Province, China
| | - Peng Shen
- Department of Medical Oncology, The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou 310003, Zhejiang Province, China
| |
Collapse
|
3
|
Hariri N, Zare S, Murphy J, Fadare O. Cost-effectiveness of a Dual (Immunohistochemistry and Fluorescence In Situ Hybridization) HER2/neu Testing Strategy on Invasive Breast Cancers. Appl Immunohistochem Mol Morphol 2021; 29:42-48. [PMID: 32205742 DOI: 10.1097/pai.0000000000000849] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/21/2022]
Abstract
The American Society of Clinical Oncology and the College of American Pathologists recommend that human epidermal growth factor receptor 2 (HER2)/neu status be determined for all invasive breast cancers. Although the most commonly used modalities to determine HER2/neu status, immunohistochemistry (IHC) and fluorescence in situ hybridization (FISH), generally give concordant results, a small but consistent discordance rate between them has been demonstrated in prior studies. Most institutions in the United States use a "reflex testing strategy" for determining HER2/neu status. In a reflex testing strategy, cancers are screened with IHC, and FISH is reflexively performed only if the IHC results are classified as equivocal. Other institutions perform both tests on all cancers to maximize diagnostic accuracy (dual testing strategy). The cost-effectiveness of dual testing and reflex testing strategies are comparatively assessed herein. A decision analysis model was established comparing these 2 testing strategies. Model probabilities were obtained from an institutional invasive breast cancer cohort. Quality-adjusted life-years (QALYs) and cost estimates were extracted from published literature. All costs were converted to 2018 US$ values using the consumer price index. One-way sensitivity analysis was performed, as well as probabilistic sensitivity analyses with acceptability curves. A total of 1247 consecutive invasive breast cancer specimens were tested with a dual strategy. By IHC, 65%, 28%, and 10% were negative, equivocal, and positive, respectively. By FISH, 87% and 13% were HER2/neu-negative and HER2/neu-positive, respectively. Six patients whose cancers were IHC-positive (3+) were found to be FISH-negative. These 6 represented 5% of the 119 cases with HER2/neu 3+ scores and 0.55% of the 1082 cases with HER2/neu-negative results by FISH. Sixteen (2%) of 809 cases with a negative IHC result were ultimately classified as HER2/neu-amplified by FISH. These 16 cases constituted 10% of all cases that were ultimately classified as HER2/neu-amplified by FISH. Overall, a reflex testing strategy was found to be less costly than a dual testing strategy ($44,470.99 vs. $45,908.86, respectively), but was also less effective (10.28 vs. 10.30 QALYs). The incremental cost-effectiveness ratio was $70,051.55/QALY. In conclusion, in this single institutional cohort of breast cancers, a dual testing strategy to determine HER2/neu status was found to be more cost-effective than a reflex testing strategy.
Collapse
Affiliation(s)
- Nosaibah Hariri
- Department of Pathology, Anatomic Pathology Division, University of California San Diego Health
- Department of Pathology, University of Tabuk School of Medicine, Tabuk, Saudi Arabia
| | - Somaye Zare
- Department of Pathology, Anatomic Pathology Division, University of California San Diego Health
| | - James Murphy
- Department of Radiation Medicine and Applied Sciences, University of California San Diego, La Jolla, CA
| | - Oluwole Fadare
- Department of Pathology, Anatomic Pathology Division, University of California San Diego Health
| |
Collapse
|
4
|
Shuai Y, Ma L. Prognostic value of pathologic complete response and the alteration of breast cancer immunohistochemical biomarkers after neoadjuvant chemotherapy. Pathol Res Pract 2018; 215:29-33. [PMID: 30501932 DOI: 10.1016/j.prp.2018.11.003] [Citation(s) in RCA: 10] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 11/01/2018] [Accepted: 11/06/2018] [Indexed: 01/08/2023]
Abstract
Neoadjuvant chemotherapy(NCT) has become the standard treatment for breast cancer. The information about the tumor's sensitivity to chemotherapy and prognostic significance based on response to therapy can be provided after individualized neoadjuvant treatment. The biomarkers are key factors in the decision-making process regarding treatment as well as important prognostic indicators. Studies have shown that patients who achieve pathological complete response(pCR) after NCT have a better prognosis. For patients who do not achieve pCR, the pathological characteristics of the residual tumor can make an effect on the survival. Furthermore, the immunohistochemical (IHC) markers of the residual diseases after primary systemic therapy might be different from the primary tumor. Estrogen receptor (ER), progesterone receptor (PR), and Ki67 can usually change after NCT, while human epidermal growth factor receptor 2(HER2) seems to be more stable. The relationship between changes in breast cancer molecular biomarkers and the prognosis after neoadjuvant therapy is not yet clear. The article will make a review about it.
Collapse
Affiliation(s)
- Yanjie Shuai
- Hebei Medical University, Fourth Affiliated Hospital, Hebei Province Tumor Hospital, China
| | - Li Ma
- Hebei Medical University, Fourth Affiliated Hospital, Hebei Province Tumor Hospital, China.
| |
Collapse
|