BACKGROUND: Accurate determination of hormonal receptors (HR) status is crucial in therapeutic planning for women with breast cancer. Since the introduction of Oncotype DX, discordance with immunohistochemical (IHC) analysis has emerged to pose a clinical dilemma for the treating oncologist. The purpose of the study was to determine the accuracy of Oncotype DX RT-qPCR assay compared with IHC. MATERIAL AND METHODS: Consecutive breast carcinoma cases (n=114) that had Oncotype DX scoring for HR from 2009 to 2011 were included in the study. All cases were stained with Dako antibodies and scored using image analysis. Thereafter, cases with low RT-qPCR scores were stained with antibody clones from Ventana and Leica. RESULTS: Estrogen receptor (ER) and progesterone receptor (PR) tested by RT-qPCR were positive in 109 of 114 (97.3%) and in 85 (74.6%) samples, respectively. The Spearman correlation was 65% for ER and 91% for PR. For ER, the recommended H-score, percentage score, and the Allred score cutoffs had high concordance rate with RT-qPCR. These cutoffs were inaccurate to determine PR status, with false results seen in 14 (12.3%), 12 (10.5%), and 17 (14.9%) cases, respectively. Although the established cutoff for ER detected by RT-qPCR of 6.5 had high accuracy, PR cutoff of 5.5 was inaccurate. The best cutoffs to correlate with different scoring systems were 3.8, 4.9, and 3.5, respectively. In the low HR expression group, marked variability was detected using different antibody clones. CONCLUSIONS: We conclude that although there was minimal difference in ER status between these 2 assays, PR had considerable discordance.
BACKGROUND: Accurate determination of hormonal receptors (HR) status is crucial in therapeutic planning for women with breast cancer. Since the introduction of Oncotype DX, discordance with immunohistochemical (IHC) analysis has emerged to pose a clinical dilemma for the treating oncologist. The purpose of the study was to determine the accuracy of Oncotype DX RT-qPCR assay compared with IHC. MATERIAL AND METHODS: Consecutive breast carcinoma cases (n=114) that had Oncotype DX scoring for HR from 2009 to 2011 were included in the study. All cases were stained with Dako antibodies and scored using image analysis. Thereafter, cases with low RT-qPCR scores were stained with antibody clones from Ventana and Leica. RESULTS:Estrogen receptor (ER) and progesterone receptor (PR) tested by RT-qPCR were positive in 109 of 114 (97.3%) and in 85 (74.6%) samples, respectively. The Spearman correlation was 65% for ER and 91% for PR. For ER, the recommended H-score, percentage score, and the Allred score cutoffs had high concordance rate with RT-qPCR. These cutoffs were inaccurate to determine PR status, with false results seen in 14 (12.3%), 12 (10.5%), and 17 (14.9%) cases, respectively. Although the established cutoff for ER detected by RT-qPCR of 6.5 had high accuracy, PR cutoff of 5.5 was inaccurate. The best cutoffs to correlate with different scoring systems were 3.8, 4.9, and 3.5, respectively. In the low HR expression group, marked variability was detected using different antibody clones. CONCLUSIONS: We conclude that although there was minimal difference in ER status between these 2 assays, PR had considerable discordance.
Authors: Satbir Singh Thakur; Haocheng Li; Angela M Y Chan; Roxana Tudor; Gilbert Bigras; Don Morris; Emeka K Enwere; Hua Yang Journal: PLoS One Date: 2018-01-05 Impact factor: 3.240