EVALUATING THE TRANSPORTABILITY OF NON-INFERIORITY EVIDENCE FOR CHEMOTHERAPHY DE-ESCALATION IN N+ HR+HER2- BREAST CANCER: A HTA SIMULATION STUDY
Author(s)
Tobias Vogelmann, PhD1, York Francis Zoellner, Prof. Dr.2, Martin Bostrup, MSc, PhD3.
1LinkCare GmbH, Ludwigsburg, Germany, 2Hamburg University of Applied Sciences, Hamburg, Germany, 3VERACYTE, Charlottenlund, Denmark.
1LinkCare GmbH, Ludwigsburg, Germany, 2Hamburg University of Applied Sciences, Hamburg, Germany, 3VERACYTE, Charlottenlund, Denmark.
OBJECTIVES: In early hormone receptor-positive, HER2-negative (HR+/HER2−) breast cancer, gene expression tests guide whether postmenopausal patients with one to three positive lymph nodes (N1) can safely omit adjuvant chemotherapy. Non-inferiority (NI) verdicts for omitting chemotherapy are often established in pooled N1 populations. However, routine axillary-staged care may include a higher proportion of patients with 2-3 positive nodes than clinical trial populations. If treatment effects differ by nodal burden, pooled NI evidence may therefore not be directly transportable to real-world practice. We assessed how robust a de-escalation NI verdict is to real-world node distributions and whether node-resolved trial evidence supports transportability across the nodal range.
METHODS: We compared published trial node distributions with a real-world axillary-staged population and examined node-resolved NI evidence from the OPTIMA trial (ISRCTN42400492), which evaluated gene expression tests-guided chemotherapy de-escalation across pN1, 1-3 nodes, and pN2, 4-9 nodes. In a simulation study, stratum-specific treatment effects were re-weighted across plausible real-world node distributions. A probabilistic sensitivity analysis with 20,000 iterations propagated uncertainty in hazard ratios and baseline node distribution.
RESULTS: The real-world node distribution carried more 2-3-node disease than a previously published trial population: 43% versus 34%. OPTIMA provided node-resolved NI evidence across the pN1-pN2 range, with concordant treatment effects in pN1 and pN2: HR 1.11 and HR 1.19, respectively. The five-year absolute difference was approximately 1.6 percentage points in both strata, with no apparent clinically meaningful heterogeneity by nodal status. Re-weighting toward higher-node disease left the NI verdict materially unchanged. In probabilistic sensitivity analysis, NI was retained across the prespecified range of plausible real-world node distributions.
CONCLUSIONS: Node-resolved,full-range NI evidence supports the transportability of chemotherapy de-escalation verdicts across real-world nodal distributions, including the higher-node disease seen in routine axillary-staged care. Node-resolved evidence therefore reduces reliance on untested assumptions of homogeneous treatment effect across nodal strata.
METHODS: We compared published trial node distributions with a real-world axillary-staged population and examined node-resolved NI evidence from the OPTIMA trial (ISRCTN42400492), which evaluated gene expression tests-guided chemotherapy de-escalation across pN1, 1-3 nodes, and pN2, 4-9 nodes. In a simulation study, stratum-specific treatment effects were re-weighted across plausible real-world node distributions. A probabilistic sensitivity analysis with 20,000 iterations propagated uncertainty in hazard ratios and baseline node distribution.
RESULTS: The real-world node distribution carried more 2-3-node disease than a previously published trial population: 43% versus 34%. OPTIMA provided node-resolved NI evidence across the pN1-pN2 range, with concordant treatment effects in pN1 and pN2: HR 1.11 and HR 1.19, respectively. The five-year absolute difference was approximately 1.6 percentage points in both strata, with no apparent clinically meaningful heterogeneity by nodal status. Re-weighting toward higher-node disease left the NI verdict materially unchanged. In probabilistic sensitivity analysis, NI was retained across the prespecified range of plausible real-world node distributions.
CONCLUSIONS: Node-resolved,full-range NI evidence supports the transportability of chemotherapy de-escalation verdicts across real-world nodal distributions, including the higher-node disease seen in routine axillary-staged care. Node-resolved evidence therefore reduces reliance on untested assumptions of homogeneous treatment effect across nodal strata.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
EE98
Topic
Economic Evaluation, Health Technology Assessment, Methodological & Statistical Research
Disease
Oncology