The REMAR (Rhein-Main-Registry) real-world study: prospective evaluation of the 21-gene breast recurrence score® assay in addition to Ki-67 for adjuvant treatment decisions in early-stage breast cancer

Abstract

Purpose: Ki-67 is recommended by international/national guidelines for risk stratification in early breast cancer (EBC), particularly for defining "intermediate risk," despite inter-laboratory/inter-observer variability and cutoff uncertainty. We investigated Ki-67 (> 10%- < 40%, determined locally) as a prognostic marker for intermediate/high risk in EBC, pN0-1 patients.

Methods: This prospective, non-interventional, real-world study included females ≥ 18 years, with pN0/pN1mi/pN1, HR+ , HER2-negative EBC, and locally determined Ki-67 ranging 10%-40%. The primary outcome was changes in treatment recommendations after disclosing the Oncotype DX Breast Recurrence Score^®(RS) assay result.

Results: The analysis included 567 patients (median age, 57 [range, 29-83] years; 70%/1%/29%/ with pN0/pN1mi/pN1 disease; 81% and 19% with RS results 0-25 and 26-100, respectively). The correlations between local and central Ki-67, local Ki-67, and the RS, and central Ki-67 and the RS results were weak (r = 0.35, r = 0.3, and r = 0.46, respectively), and discrepancies were noted in both directions (e.g., local Ki-67 was lower or higher than central Ki-67). After disclosing the RS, treatment recommendations changed for 190 patients (34%). Changes were observed in pN0 and pN1mi/pN1 patients and in patients with centrally determined Ki-67 ≤ 10% and > 10%. Treatment changes were aligned with RS results (adding chemotherapy for patients with higher RS results, omitting it for lower RS results), and their net result was 8% reduction in adjuvant chemotherapy use (from 32% pre-RS results to 24% post-RS results).

Conclusion: The Oncotype DX^® assay is a tool for individualizing treatments that adds to classic treatment decision factors. The RS result and Ki-67 are not interchangeable, and Ki-67, as well as nodal status, should not be used as gatekeepers for testing eligibility, to avoid under and overtreatment.

Keywords: 21-gene assay; Adjuvant treatment; Breast cancer; Ki-67; Recurrence Score,; Registry.

Fig. 1

Sankey diagram displaying the correlation between Ki-67 determined centrally and locally (n = 562 patients with both local and central testing). The number of patients in each Ki-67 group is shown on both sides of the diagram. Local Ki-67 evaluations ranged from 10 to 40% and used routine grouping by 5% increments

Fig. 2

Scatter plots displaying the correlation between the RS results and Ki-67 determined centrally (a) or centrally (b). Local Ki-67 evaluations ranged from 10 to 40% per the inclusion threshold for REMAR and used routine grouping by 5% increments. The blue line represents Ki-67 levels of 10%

Fig. 3

Sankey diagram displaying the absolute changes in 190 patients in treatment recommendations following the disclosure of the RS result for the entire cohort. CET chemoendocrine therapy, ET endocrine therapy, RS recurrence score

Fig. 4

Sankey diagrams displaying the overall changes in treatment recommendations following the disclosure of the RS result by nodal status (a) and central Ki-67 levels (b). CET chemoendocrine therapy, ET endocrine therapy, RS recurrence score