Abstract PD15-06: Estrogen receptor expression thresholds by IHC and mRNA for Ki67 response to aromatase inhibition: A POETIC study

Abstract Background. Estrogen receptor (ER) expression is the key determinant for endocrine treatment of breast cancer (BC). In clinical practise ER analysis is currently almost entirely by immunohistochemistry (IHC). ASCO/CAP recommends cut-offs of <1% (negative) and 1-10% (low) to minimise the risk of false negatives. There is uncertainty in the level of responsiveness of those with low ER. mRNA assessment could be acceptable or preferable if a similar or better threshold was shown to identify response to endocrine therapies. Change of Ki67 after 2 weeks’ presurgical treatment with aromatase inhibitors (AIs) relates to reduction in risk of recurrence. We aimed to define IHC and mRNA cut-points for predicting change in Ki67 in response to AI. Method. POETIC is a phase III trial in postmenopausal patients with ER+ (by local assessment) BC randomized to receive 2-weeks’ pre-surgical AI vs no presurgical treatment. Cases were selected from the AI treatment group. All HER2+ cases were analysed. To enrich the HER2- study set for lower ER values, we selected all of the 15% poorest Ki67 suppression (ie PR, <40% suppression) and 30% of the remainder categorised as intermediate (IR, 40-79%) and good-responders (GR, >79%). 770 baseline FFPE samples were analysed (582 HER2-, 188 HER2+). IHC used the NCL-L-ER6F11 (Novocastra, Leica) antibody with the DAKO Flex kit and were scored centrally. mRNA was assessed by rtPCR. Values were expressed according to an arbitrary scale (“A units”). Results. ER IHC was available from 517 HER2- and 188 HER2+ tumors and ER mRNA from 376 HER2- and 173 HER2+ tumors. The correlation between ER levels by IHC and rtPCR was moderately high (Rho=0.616, p<0.00001). In IHC <1%, 1-10% and >10% tumors median (range) mRNA levels were 1.290 (0.188-11.346), 2.368 (0.575-13.863) and 111.150 (0.826-1001.423), respectively. The tables show the %age of PR, IR and GR for each category of ER by IHC or mRNA. The number of cases was adjusted to cater for the analysis of only 30% of the HER2- IR and GR , by multiplying the HER2- IR and GR number of cases by 3.3, resulting in this not being an integer for some categories. IHC: 87% of cases had >60% of cells +ve and only 11-12% of these were PRs. For cases with <10% cells +ve (4.9% of the population) over 90% of cases were PRs and there were no GRs. Cases between 1 and 10% +ve were not less responsive than those <1% +ve. For cases with >10% <20% cells +ve there was a substantial number of GRs and IRs (14.6% and 19.0%, respectively). mRNA: 82% of cases had ≥ 40 A units and only 10-14% of these showed PR. For cases with 5 A units (4.2% of the population) there were 9.5% % IR and no GR. In the next category, ≥5 <10 A units, 34.4% of cases were IR or GR. The levels of ER that distinguished PRs from IR/GRs was no different between HER2+ and HER2- cases. Conclusions. Ki67 responsiveness increases with increasing levels of ER by both IHC and mRNA. There was little responsiveness at IHC<10% and no distinction in responsiveness between <1% and 1-10% cells positive. Above 10% positive cells substantial numbers of patients showed IR or GR. Similar separation of PRs from IR/GRs was achieved by IHC and mRNA. IHC (%)0 < 1≥1 < 10≥10 < 20≥20 < 40≥40 < 60≥60 < 80≥80No. of cases42.32922.625.671.7239.81010.6PR87.596.566.462.541.811.710.8IR12.53.419.020.741.640.933.1GR0.00.014.616.816.647.456.1 mRNA (A units)<5≥5 < 10≥10 < 20≥20 < 40≥40 < 80≥80 < 120≥120No. of cases45.318.336.589.2168.1177.6550.4PR90.565.638.426.913.710.710.7IR9.416.458.947.946.835.226.8GR0.018.02.725.239.554.162.5 Citation Format: Elena Lopez Knowles, Simone Detre, Margaret Hills, Gene F Schuster, Maggie CU Cheang, Holly Tovey, Lucy Kilburn, Judith Bliss, John Robertson, Ian Smith, Mitch Dowsett, POETIC investigators. Estrogen receptor expression thresholds by IHC and mRNA for Ki67 response to aromatase inhibition: A POETIC study [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD15-06.