Supervised machine learning model to predict oncotype DX risk category in patients over age 50

Pawloski, Kate R; Gönen, Mithat; Wen, Hannah Y.; Tadros, Audree B; Thompson, Dola S.; Abbate, Kelly; Morrow, Monica; El‐Tamer, Mahmoud

doi:10.1007/s10549-021-06443-w

Cited by 7 publications

(7 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The predictive model developed by Pawloski et al showed performance power with specificity and negative predictive value for identifying low-risk patients at 96.3% and 92.9%. But the sensitivity and positive predictive value for predicting high-risk patients were lower (48.3% and 65.1%, respectively) [16]. Previous studies have divided data into low-and high-risk groups and developed models that predict each risk group.…”

Section: Discussionmentioning

confidence: 96%

“…The random forest model developed by Pawloski et al showed performance power with specificity and negative predictive value for identifying low-risk patients at 96.3% and 92.9%. But sensitivity and positive predictive value for predicting high-risk patients were lower (48.3% and 65.1%, respectively) [16]. Their predictive model with 500 trees was developed on the training cohort, using age, tumor size, histology, progesterone receptor (PR) expression, lympho-vascular invasion (LVI), and grade as predictors.…”

Section: Related Workmentioning

confidence: 99%

“…Recently, with the development of AI (artificial intelligence), several studies have reported the development of models to predict ODX risk based on deep learning [14][15][16]. However, there have not been many studies about AI predictive models.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Prediction of a Multi-Gene Assay (Oncotype DX and Mammaprint) Recurrence Risk Group Using Machine Learning in Estrogen Receptor-Positive, HER2-Negative Breast Cancer—The BRAIN Study

Ji,

Ahn,

Yoo

et al. 2024

Cancers

View full text Add to dashboard Cite

This study aimed to develop a machine learning-based prediction model for predicting multi-gene assay (MGA) risk categories. Patients with estrogen receptor-positive (ER+)/HER2− breast cancer who had undergone Oncotype DX (ODX) or MammaPrint (MMP) were used to develop the prediction model. The development cohort consisted of a total of 2565 patients including 2039 patients tested with ODX and 526 patients tested with MMP. The MMP risk prediction model utilized a single XGBoost model, and the ODX risk prediction model utilized combined LightGBM, CatBoost, and XGBoost models through soft voting. Additionally, the ensemble (MMP + ODX) model combining MMP and ODX utilized CatBoost and XGBoost through soft voting. Ten random samples, corresponding to 10% of the modeling dataset, were extracted, and cross-validation was performed to evaluate the accuracy on each validation set. The accuracy of our predictive models was 84.8% for MMP, 87.9% for ODX, and 86.8% for the ensemble model. In the ensemble cohort, the sensitivity, specificity, and precision for predicting the low-risk category were 0.91, 0.66, and 0.92, respectively. The prediction accuracy exceeded 90% in several subgroups, with the highest prediction accuracy of 95.7% in the subgroup that met Ki-67 <20 and HG 1~2 and premenopausal status. Our machine learning-based predictive model has the potential to complement existing MGAs in ER+/HER2− breast cancer.

show abstract

Section: Discussionmentioning

confidence: 96%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Prediction of a Multi-Gene Assay (Oncotype DX and Mammaprint) Recurrence Risk Group Using Machine Learning in Estrogen Receptor-Positive, HER2-Negative Breast Cancer—The BRAIN Study

Ji,

Ahn,

Yoo

et al. 2024

Cancers

View full text Add to dashboard Cite

show abstract

“…We have one of the larger datasets showing this association. Work by Pawloski K et al developed a supervised machine learning algorithm with a high degree of accuracy for predicting the RS, with the PR score being the strongest independent predictor [ 61 ]. The PR score is potentially a cheap and ready for use surrogate marker of the risk of recurrence, and more research to validate this finding is one of the outstanding questions.…”

Section: Discussionmentioning

confidence: 99%

Enhancing clinical decision support with genomic tools in breast cancer: A Scottish perspective

Peters,

Hall,

Jordan

et al. 2024

The Breast

View full text Add to dashboard Cite

“…Moreover, another recently published study, evaluating the cost-effectiveness of different multigene-expression assays in Germany, showed that all available assays (Oncotype DX, Mammaprint, Prosigna, Endopredict) reduce overall treatment costs [ 29 ]. Several statistical models using clinicopathologic risk factors to identify patients that will most likely benefit from RS testing are currently available, which might help to further reduce overall treatment costs [ 20 , 21 , 31 ].…”

Section: Discussionmentioning

confidence: 99%

Recurrence Score® Result Impacts Treatment Decisions in Hormone Receptor-Positive, HER2-Negative Patients with Early Breast Cancer in a Real-World Setting—Results of the IRMA Trial

Dannehl

Engler

Volmer

et al. 2022

Cancers

View full text Add to dashboard Cite

Background: Patients with hormone receptor-positive (HR+), HER2-negative (HER2-) early breast cancer (eBC) with a high risk of relapse often undergo adjuvant chemotherapy. However, only a few patients will gain benefit from chemotherapy. Since classical tumor characteristics (grade, tumor size, lymph node involvement, and Ki67) are of limited value to predict chemotherapy efficacy, multigene expression assays such as the Oncotype DX® test were developed to reduce over- and undertreatment. The IRMA trial analyzed the impact of Recurrence Score® (RS) assessment on adjuvant treatment recommendations. Materials and methods: The RS result was assessed in patients with HR+/HER2− unilateral eBC with 0–3 pathologic lymph nodes who underwent primary surgical treatment at the Department for Women’s Health of Tuebingen University, Germany. Therapy recommendations without knowledge of the RS result were compared to therapy recommendations with awareness of the RS result. Results: In total, 245 patients underwent RS assessment. Without knowledge of the RS result, 92/245 patients (37.6%) would have been advised to receive chemotherapy. After RS assessment, 56/245 patients (22.9%) were advised to undergo chemotherapy. Chemotherapy was waived in 47/92 patients (51.1%) that were initially recommended to receive it. Chemotherapy was added in 11/153 patients (7.2%) that were recommended to not receive it initially. Summary: Using the RS result to guide adjuvant treatment decisions in HR+/HER2− breast cancer led to a substantial reduction of chemotherapy. In view of the results achieved in prospective studies, the RS result is among other risk-factors suitable for the individualization of adjuvant systemic therapy.

show abstract

Supervised machine learning model to predict oncotype DX risk category in patients over age 50

Cited by 7 publications

References 29 publications

Prediction of a Multi-Gene Assay (Oncotype DX and Mammaprint) Recurrence Risk Group Using Machine Learning in Estrogen Receptor-Positive, HER2-Negative Breast Cancer—The BRAIN Study

Prediction of a Multi-Gene Assay (Oncotype DX and Mammaprint) Recurrence Risk Group Using Machine Learning in Estrogen Receptor-Positive, HER2-Negative Breast Cancer—The BRAIN Study

Enhancing clinical decision support with genomic tools in breast cancer: A Scottish perspective

Recurrence Score® Result Impacts Treatment Decisions in Hormone Receptor-Positive, HER2-Negative Patients with Early Breast Cancer in a Real-World Setting—Results of the IRMA Trial

Contact Info

Product

Resources

About