Abstract P2-11-19: Estimating the long-term risk of recurrence in patients receiving HER2-targeted agents in HER2+ early-stage breast cancer (ESBC)

Abstract Background: The APHINITY, KATHERINE and ExteNET pivotal trials demonstrated the benefits of HER2-targeted agents for disease-free survival in HER2+ ESBC. However, heterogeneity in patient populations and study designs make assessment of treatment benefits challenging. The objective of this study was to project absolute treatment benefits over a standardized 10-year time horizon. Methods: An epidemiologic model was developed to simulate numbers of distant recurrences for 1000 patients with newly diagnosed HER2+ ESBC in the ExteNET, APHINITY and KATHERINE patient populations and select subgroups. ExteNET evaluated neratinib vs. placebo from 2009 to 2012. APHINITY evaluated pertuzumab + trastuzumab vs. placebo + trastuzumab + standard adjuvant chemotherapy in high-risk patients without neoadjuvant therapy treated from 2011 to 2014. KATHERINE evaluated ado-trastuzumab emtansine (T-DM1) vs. trastuzumab in patients without pathologic complete response (pCR) following neoadjuvant therapy treated from 2013 to 2016. We used disease-free survival curves from each trial as model inputs to estimate the risk of distant recurrence and extrapolated these results to a standardized 10-year follow-up period. We assumed there were no further treatment benefits to prevention of recurrence beyond the follow-up period of the trial. Survival curve trends beyond each trial’s observation period were informed by the trastuzumab arm of the long-term HERA trial follow-up data. Model outputs were distant recurrences for the control arms for each study and the number of distant recurrences avoided from treatment. Results: Over a 10-year time horizon, treatment benefits ranged from 11 to 65 distant recurrences avoided. Treatment benefit differed according to baseline risk of the trial population and treatment effect, with higher-risk populations and greater treatment effect (e.g. KATHERINE trial) tending to yield greater absolute treatment benefit, with the exception of the ExteNET ITT population as changes in the treatment effect over time (non-proportional hazards) made modeling long-term results more complex. Absolute treatment benefits were similar for the intent-to-treat (ITT) group in the KATHERINE trial and the no pCR group in the ExteNET trial; and the ITT groups in the APHINITY and ExteNET trials. Refinement of model projections and inclusion of local recurrences in a more complex modeling structure are ongoing. Conclusions: Results of our simulation suggest that patients with HER2+ ESBC experience a range of reductions in distant recurrence from HER2-targeted agents, depending on baseline risk (pCR and hormone receptor status), and treatment effect. Epidemiologic modeling may be a useful method to estimate long-term outcomes and facilitate evaluation across disparate clinical trials. Table. Modeled benefits of HER2-targeted agents in a cohort of 1000 patients with HER2+ ESBC over a standardized 10-year timeframeSource data inputsModeled outputs (n=1000)Distant recurrences at 10 yearsClinical study characteristicsHazard ratio for distant recurrence from study1Baseline risk: Events in control armTreatment effect: Distant recurrences avoided from treatmentExteNET study (ITT)0.78 (0.60-1.01) at 5 years14318APHINITY study (ITT)0.76 (0.62-0.95) at 6 years10320ExteNET study (HR+; treated within 12 months of trastuzumab receipt)0.57 (0.39-0.83) at 5 years14727APHINITY study (HR+)0.78 (0.51-1.18) at 6 years26111KATHERINE study (no pCR, ITT)0.60 (0.45-0.79) at 3 years26464KATHERINE study (no pCR, HR+)0.57 (0.42-0.75) at 3 years223941ExteNET study (no pCR, HR+; treated within 12 months of trastuzumab receipt)0.61 (0.32-1.11) at 5 years265651Values in parentheses represent 95% confidence intervals. 2Hazard ratio not reported in clinical study publication. Estimated based on reported IDFS hazard ratio and proportion of distant to all invasive recurrences in ITT population. HR = hormone receptor; ITT = intent-to-treat; pCR = pathologic complete response. Citation Format: David L Veenstra, Nathaniel Hendrix, Chantal M Dolan, Kathryn A Fisher, Deepa Lalla, Nina Oestreicher, Adam Brufsky. Estimating the long-term risk of recurrence in patients receiving HER2-targeted agents in HER2+ early-stage breast cancer (ESBC) [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 P2-11-19.