Charles E Rutter1, Xiaopan Yao2, Brandon R Mancini3, Jenerius A Aminawung4, Anees B Chagpar5, Ozlen Saglam6, Erin W Hofstatter2, Maysa Abu-Khalaf7, Cary P Gross8, Suzanne B Evans9. 1. Department of Therapeutic Radiology, Yale School of Medicine, New Haven, CT. Electronic address: charles.rutter@yale.edu. 2. Department of Medical Oncology, Yale School of Medicine, New Haven, CT. 3. Department of Therapeutic Radiology, Yale School of Medicine, New Haven, CT. 4. Cancer Outcomes, Public Policy, and Effectiveness Research Center, Yale School of Medicine, New Haven, CT. 5. Cancer Outcomes, Public Policy, and Effectiveness Research Center, Yale School of Medicine, New Haven, CT; Department of Surgery, Yale School of Medicine, New Haven, CT. 6. Department of Pathology, Yale School of Medicine, New Haven, CT. 7. Department of Medical Oncology, Yale School of Medicine, New Haven, CT; Cancer Outcomes, Public Policy, and Effectiveness Research Center, Yale School of Medicine, New Haven, CT. 8. Cancer Outcomes, Public Policy, and Effectiveness Research Center, Yale School of Medicine, New Haven, CT; Department of Medicine, Yale School of Medicine, New Haven, CT. 9. Department of Therapeutic Radiology, Yale School of Medicine, New Haven, CT; Cancer Outcomes, Public Policy, and Effectiveness Research Center, Yale School of Medicine, New Haven, CT.
Abstract
BACKGROUND: We performed an analysis to determine the relative contribution of the Oncotype DX (ODX) recurrence score (RS) results in adjuvant therapy delivery compared with traditional pathologic factors. METHODS AND MATERIALS: We performed a retrospective review of women with stage I-IIIA breast cancer treated at the Yale Comprehensive Cancer Center from 2006 to 2012 with available ODX results. We constructed separate logistic models with the clinicopathologic factors alone and also integrating RS and compared these models using the likelihood ratio test and c-statistic to determine whether integration of the RS will result in better prediction of chemotherapy (CTx) delivery. RESULTS: We identified 431 women with a median age of 58 years. The RS was low (< 18), intermediate (18-30), and high (> 30) in 56%, 37%, and 7%, respectively. CTx was delivered to 30% of the patients. Age, differentiation, lymphovascular invasion, and progesterone receptor (PR) positivity < 50% were associated with CTx delivery in multivariable logistic regression of clinicopathologic factors alone (P < .05). In the model integrating the RS, an intermediate or a high RS was the most influential factor for CTx delivery (odds ratio, 7.87 vs. 265.35, respectively; P < .0001). The PR results and grade were no longer significant (P = .74 and P = .06, respectively). The integration of the RS resulted in improved model fit and precision, indicated by the likelihood ratio test (ΔG2, 100.782; df = 2; P < .0001) and an improved c-statistic (0.720 vs. 0.856). CONCLUSION: Gene expression profiling appears to account for a substantial amount of variability in CTx delivery in current practice. Further work is needed to ensure appropriate test usage and cost-effectiveness.
BACKGROUND: We performed an analysis to determine the relative contribution of the Oncotype DX (ODX) recurrence score (RS) results in adjuvant therapy delivery compared with traditional pathologic factors. METHODS AND MATERIALS: We performed a retrospective review of women with stage I-IIIA breast cancer treated at the Yale Comprehensive Cancer Center from 2006 to 2012 with available ODX results. We constructed separate logistic models with the clinicopathologic factors alone and also integrating RS and compared these models using the likelihood ratio test and c-statistic to determine whether integration of the RS will result in better prediction of chemotherapy (CTx) delivery. RESULTS: We identified 431 women with a median age of 58 years. The RS was low (< 18), intermediate (18-30), and high (> 30) in 56%, 37%, and 7%, respectively. CTx was delivered to 30% of the patients. Age, differentiation, lymphovascular invasion, and progesterone receptor (PR) positivity < 50% were associated with CTx delivery in multivariable logistic regression of clinicopathologic factors alone (P < .05). In the model integrating the RS, an intermediate or a high RS was the most influential factor for CTx delivery (odds ratio, 7.87 vs. 265.35, respectively; P < .0001). The PR results and grade were no longer significant (P = .74 and P = .06, respectively). The integration of the RS resulted in improved model fit and precision, indicated by the likelihood ratio test (ΔG2, 100.782; df = 2; P < .0001) and an improved c-statistic (0.720 vs. 0.856). CONCLUSION: Gene expression profiling appears to account for a substantial amount of variability in CTx delivery in current practice. Further work is needed to ensure appropriate test usage and cost-effectiveness.
Authors: Chalanda N Evans; Noel T Brewer; Susan T Vadaparampil; Marc Boisvert; Yvonne Ottaviano; M Catherine Lee; Claudine Isaacs; Marc D Schwartz; Suzanne C O'Neill Journal: Breast Cancer Res Treat Date: 2016-04-08 Impact factor: 4.872
Authors: Jinani Jayasekera; Susan T Vadaparampil; Susan Eggly; Richard L Street; Tanina Foster Moore; Claudine Isaacs; Hyo S Han; Bianca Augusto; Jennifer Garcia; Katherine Lopez; Suzanne C O'Neill Journal: JCO Oncol Pract Date: 2020-05-28
Authors: Andrew Dodson; David Okonji; Laura Assersohn; Anne Rigg; Amna Sheri; Nick Turner; Ian Smith; Marina Parton; Mitch Dowsett Journal: Breast Cancer Res Treat Date: 2017-11-11 Impact factor: 4.872