Yuka Kuramoto1, Natsumi Wada1, Yoshikazu Uchiyama2. 1. Graduate School of Health Sciences, Kumamoto University, 4-24-1 Kuhonji, Chuo-ku, Kumamoto, Kumamoto, 862-0976, Japan. 2. Department of Medical Image Sciences, Faculty of Life Sciences, Kumamoto University, 4-24-1 Kuhonji, Chuo-ku, Kumamoto, Kumamoto, 862-0976, Japan. y_uchi@kumamoto-u.ac.jp.
Abstract
PURPOSE: Neoadjuvant pharmacotherapy is essential for patients with breast cancer who wish to preserve the breast by shrinking the malignant tumor, allowing breast-conserving surgery. It may eliminate cancer cells completely, which is known as pathologic complete response (pCR). Patients with pCR have a lower risk of recurrence. The purpose of this study was to develop a method for predicting patients who achieve pCR by neoadjuvant pharmacotherapy using radiomic features in MR images. METHODS: Fat-suppressed T2-weighted MR images of 64 cases were identified from the ISPY1 dataset. There were 26 cases of pCR and 38 cases of non-pCR. The image slice with the largest tumor diameter was selected from MR images, and the tumor region was manually segmented. A total of 371 radiomic features were calculated from the tumor region. We selected nine radiomic features using Lasso in this study. A support vector machine (SVM) with nine radiomic features was used for predicting patients with pCR. RESULTS: The result of the ROC analysis showed that the area under the curve of SVM was 0.92 for distinguishing between pCR and non-pCR. Although the input data contain data that were misclassified by SVM, the survival curve classified into the pCR group was at a higher position than the non-pCR group. However, the log-rank test was [Formula: see text]. CONCLUSIONS: We developed a method to predict patients with pCR by neoadjuvant pharmacotherapy using noninvasive MR images. The survival curve of patients classified as having pCR by the proposed method was higher than those classified as non-pCR. Since the proposed method predicts patients who achieve pCR by neoadjuvant pharmacotherapy, it enhances the value of preoperative image information.
PURPOSE: Neoadjuvant pharmacotherapy is essential for patients with breast cancer who wish to preserve the breast by shrinking the malignant tumor, allowing breast-conserving surgery. It may eliminate cancer cells completely, which is known as pathologic complete response (pCR). Patients with pCR have a lower risk of recurrence. The purpose of this study was to develop a method for predicting patients who achieve pCR by neoadjuvant pharmacotherapy using radiomic features in MR images. METHODS: Fat-suppressed T2-weighted MR images of 64 cases were identified from the ISPY1 dataset. There were 26 cases of pCR and 38 cases of non-pCR. The image slice with the largest tumor diameter was selected from MR images, and the tumor region was manually segmented. A total of 371 radiomic features were calculated from the tumor region. We selected nine radiomic features using Lasso in this study. A support vector machine (SVM) with nine radiomic features was used for predicting patients with pCR. RESULTS: The result of the ROC analysis showed that the area under the curve of SVM was 0.92 for distinguishing between pCR and non-pCR. Although the input data contain data that were misclassified by SVM, the survival curve classified into the pCR group was at a higher position than the non-pCR group. However, the log-rank test was [Formula: see text]. CONCLUSIONS: We developed a method to predict patients with pCR by neoadjuvant pharmacotherapy using noninvasive MR images. The survival curve of patients classified as having pCR by the proposed method was higher than those classified as non-pCR. Since the proposed method predicts patients who achieve pCR by neoadjuvant pharmacotherapy, it enhances the value of preoperative image information.
Authors: Laura Heacock; Alana Lewin; Abimbola Ayoola; Melanie Moccaldi; James S Babb; Sungheon G Kim; Linda Moy Journal: Acad Radiol Date: 2019-08-20 Impact factor: 3.173
Authors: Elizabeth Hope Cain; Ashirbani Saha; Michael R Harowicz; Jeffrey R Marks; P Kelly Marcom; Maciej A Mazurowski Journal: Breast Cancer Res Treat Date: 2018-10-16 Impact factor: 4.872
Authors: Hui Li; Yitan Zhu; Elizabeth S Burnside; Erich Huang; Karen Drukker; Katherine A Hoadley; Cheng Fan; Suzanne D Conzen; Margarita Zuley; Jose M Net; Elizabeth Sutton; Gary J Whitman; Elizabeth Morris; Charles M Perou; Yuan Ji; Maryellen L Giger Journal: NPJ Breast Cancer Date: 2016-05-11
Authors: Nathaniel M Braman; Maryam Etesami; Prateek Prasanna; Christina Dubchuk; Hannah Gilmore; Pallavi Tiwari; Donna Plecha; Anant Madabhushi Journal: Breast Cancer Res Date: 2017-05-18 Impact factor: 6.466
Authors: Nathaniel Braman; Prateek Prasanna; Jon Whitney; Salendra Singh; Niha Beig; Maryam Etesami; David D B Bates; Katherine Gallagher; B Nicolas Bloch; Manasa Vulchi; Paulette Turk; Kaustav Bera; Jame Abraham; William M Sikov; George Somlo; Lyndsay N Harris; Hannah Gilmore; Donna Plecha; Vinay Varadan; Anant Madabhushi Journal: JAMA Netw Open Date: 2019-04-05