Practice-changing use of 21-gene oncotype DX breast recurrence score assay in a public hospital in Brazil: Results of 155 cases.

Abstract

e12518 Background: The 21-gene assay estimates risk of recurrence expressed as a Recurrence Score result between 0 and 100. The assay is clinically validated to predict chemotherapy (CT) benefit in node-negative (N0) and node-positive (up to three axillary nodes, N1) hormone receptor-positive (HR+) early-stage breast cancer (ESBC). The TAILORx study, which randomized 6711 women with N0 ESBC and Recurrence Score result 11-25 to hormonal therapy (HT) ± CT, demonstrated that patients older than 50 years derived no benefit from CT+HT, although younger patients with Recurrence Score result 16-25 may derive some benefit. The 21-gene assay has been incorporated in practice guidelines worldwide, based on evidence of clinical utility in N0 and N1 ESBC. We evaluated the impact of the 21-gene test on treatment decisions for patients with N0 and N1 ESBC at Sistema Único de Saúde in Brazil. Methods: Eligible patients were post-surgery with T1/T2 tumors, had HR+, HER2−, N0 or N1 ESBC, and were candidates for adjuvant systemic therapy. Treatment recommendations, CT+HT or HT alone, were captured before and after 21-gene test results. All patients were seen at Pérola Byington Hospital, a public hospital in São Paulo, Brazil. TAILORx results were used to guide decisions for or against CT for each patient. Results: From 08/2018 to 04/2019, 155 women were enrolled. Patient mean age was 57.6 years (29-78), 116 (75%) were postmenopausal, and 53 (34%) had N1 breast cancer. Based on clinical data alone, 151 patients had pre-assay recommendations of CT. Post-assay, 106 of 151 patients (70%) had changes in CT recommendation: 104 (69%) initially recommended CT received HT alone, and 2 (1%) initially recommended HT alone received CT+HT (Table). Using the modified Adjuvant!Online criteria for clinical risk classification, 109 of 155 patients (70%) had high risk, 48 (44%) of whom received CT. Of 46 patients with low clinical risk, 10 (22%) received CT. CT use trended with histologic grade: 11% with grade 1, 69% with grade 2, and 61% with grade 3. Most of our patients had tumor bigger than 2 cm (61%) with 8% bigger than 4 cm. Conclusions: The change in clinical practice at this public hospital was greater than expected: 69% of initial treatment recommendations were changed with the Recurrence Score result to omit CT. Clinicopathologic criteria did not correlate well with Recurrence Score results and did not identify those most likely to benefit from CT. A cost-effectiveness study is underway at our institution. [Table: see text]