Concordance of local and central laboratory hormone and HER2 receptor status in ECOG 2197

Abstract

21022 Background: Central and local laboratory concordance for hormone and HER2 receptor measurement is of national interest. This study compares ER/PR/HER2 by local laboratories using immunohistochemistry (IHC) and central laboratories (IHC & quantitative RT-PCR). Methods: Of 2952 patients in E2197, a case-cohort sample of 776 patients who either did (N=179) or did not recur was studied. Central IHC for ER/PR/HER2 was performed using single 0.6 mm microarrays; Allred score (AS) was used for ER/PR (AS>2 = positive). Positive HER2 was 3+ staining in >10% cells for Central IHC and 2+ or 3+ for Local IHC. RT-PCR analysis by Oncotype DX™ for ER/PR/HER2 was performed using pre-defined cutoffs of 6.5, 5.5 and 11.5 units, respectively. Hormone receptor (HR) pos was defined as ER &/or PR pos. Results: Results from Local IHC (ER/PR in 776 & HER2 in 517 pts) were compared with Central IHC (760 pts) and RT-PCR results (776 pts). The discordance between HR positivity by Local IHC and RT-PCR was very low. However, 12% of HR neg pts by Local IHC (38/321) & Central IHC (39/326) were HR pos by RT-PCR. The relationship between ER and recurrence as a function of AS was examined. Patients with AS of 3–4 were found to be closer to the AS=2 group than to the AS>4 group Patients with AS of 3–4 were found to be closer to the AS ÿ 2 group than to the AS > 4 group (Est.HR for ER 0.97 for AS 3–4 vs. 0–2 and 0.46 for AS 5–8 vs. 0–2, and for PR were 0.84 for AS 3–4 vs. 0–2 and 0.41 for AS 5–8 vs. 0–2). Conclusions: There is a high degree of overall concordance among Local IHC, Central IHC, and Central RT-PCR for ER and PR. The degree of concordance is even greater for HR compared to ER or PR alone. Although the concordance with local labs for HER2 testing was poor, the concordance between Central IHC and RT-PCR was very high. The relatively high incidence (12%) of IHC HR neg pts who are HR pos by RT- PCR is notable. [Table: see text] [Table: see text]