Abstract PO1-01-03: Pathological response according to early metabolic remission in an interim FDG-PET scan and to tumor infiltrating lymphocytes - A secondary analysis of the phase II trial ABCSG 52/ATHENE investigating atezolizumab in early HER2+ BC

Abstract Background: ABCSG-52/ATHENE investigated a neoadjuvant chemotherapy de-escalation/immunotherapy escalation approach with trastuzumab, pertuzumab, atezolizumab and epirubicin for early-stage HER2-positive breast cancer. The pathological complete response rate (pCR; ypT0/Tis, ypN0) - the primary endpoint of the trial - was 60.3% (95% CI 47.5% - 71.9%), and was higher than the predefined threshold for positivity (≥ 40%). Interestingly, higher pCR rates were observed in PD-L1 immune cell (IC) negative tumors (pCR 69.2%) as compared to PD-L1 IC positive tumors (pCR 55.2%). No new safety concerns were observed. Primary outcome data were reported at the ESMO Breast Cancer Conference 2023. Here, we present the correlation of pathological outcome according to early metabolic response in an interim fluorodeoxyglucose F18 positron emission tomography (FDG-PET) scan, and according to stromal tumor infiltrating lymphocytes (TILs) at baseline. Methods: Patients (pts) with previously untreated, histologically confirmed HER2-positive early breast cancer (EBC; clinical prognostic stage cT1c-4a-d, N0–3, M0) were randomized 1:1 to two 3-weekly cycles of a chemotherapy-free induction phase (Part 1) with trastuzumab and pertuzumab (TP) plus 1200 mg atezolizumab (TP+A), or TP alone. Subsequently, all pts received 4 cycles of TP+A in combination with epirubicin 90 mg/m² (Part 2). Randomization was stratified according to baseline TILs ( 5% vs ≥ 5%), hormone receptor status (positive vs. negative), and prognostic stage (≤ IIA vs. ≥ IIB; AJCC v.8.0). An FDG-PET assessment was performed at baseline and after 2 cycles of study treatment. According to PERCIST, early metabolic responses were classified as metabolic complete response (mCR) and metabolic partial response (mPR) and association with outcome was tested with Fisher’s exact test. TIL assessment in pretreatment biopsy specimen were performed according to the recommendations from the International TILs Working Group. Lymphocytic predominant phenotype was defined as a TIL proportion of ≥ 50%. Correlation of TILs and pCR was modeled with logistic regression. Results: Overall, 58 pts were randomized to TP-A (n=29) or TP (n=29) at 9 sites. In 45 patients, a protocol predefined PERCIST assessment was available. An early mCR was achieved in 14 pts (n=14/45, 31.1%): 5 pts (n=5/21, 23.8%) in the TP group and 9 pts (n=9/24, 37.5%) in the TP-A group. Pts with and without an early mCR achieved a pCR in 92.9% (n=13/14) and 54.8% (n=17/31) of cases (p=0.016), respectively. An early mCR/mPR was achieved in 37 pts (n=37/45, 82.2%). Pts with and without an early mCR/mPR achieved a Residual Cancer Burden (RCB) class 0 or I in 89.2% (n=33/37) and 50.0% (n=4/8) of cases (p=0.024), respectively. The mean stromal TIL proportion was 23.9% in the overall population, with 22.8% in TP group and 25.0% in TP-A group. A lymphocytic predominant phenotype was seen in 10.3% of pts treated in the TP group, and 17.2% of pts treated in the TP-A group. No association between TIL proportion at baseline and pCR could be detected (OR for a 10%-point increase = 1.02, 95% CI 0.78-1.34). A moderate positive correlation between numeric values of TILs and PD-L1 ICs was observed (spearman r=0.57, P 0.0001). Exploratory correlation analyses of PD-L1 status, TIL proportion and metabolic responses in interim FDG-PET scans will be presented at the meeting. Conclusions: In ABCSG-52, chemotherapy de-escalation with intensified immunotherapy was effective and safe. Early metabolic response assessment was feasible and predicted pathologic response; therefore, incorporation of interim FDG-PET results into de-escalation trials merits further investigation. In contrast to regimens without immune-checkpoint inhibitors, TIL proportion at baseline did not correlate with pathological outcome. Funding: Provision of IMP and financial support by Roche Austria GmbH Citation Format: Gabriel Rinnerthaler, Daniel Egle, Rupert Bartsch, Clemens Schmitt, Andreas Petzer, Marija Balic, Edgar Petru, Ursula Denison, Christian F. Singer, Vesna Bjelic-Radisic, Simon Gampenrieder, Michael Knauer, Karl Sotlar, Christine Brunner, Florian Posch, Dominik Hlauschek, Lidija Sölkner, Zsuzsanna Bago-Horvath, Martin Filipits, Manuela Gili, Michael Gnant, Richard Greil. Pathological response according to early metabolic remission in an interim FDG-PET scan and to tumor infiltrating lymphocytes - A secondary analysis of the phase II trial ABCSG 52/ATHENE investigating atezolizumab in early HER2+ BC abstract. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO1-01-03.