Local Therapy Outcomes and Toxicity From the ATEMPT Trial (TBCRC 033), a Phase II Randomized Trial of Adjuvant T-DM1 vs. TH in Women With Stage I HER2 Positive Breast Cancer

Abstract

Anti-HER2 therapy improves local control in women with HER2+ breast cancer. This retrospective analysis evaluates the efficacy and safety of radiation therapy (RT) across a range of RT doses, targets, and schedules among patients receiving T-DM1 and TH in the ATEMPT trial.The ATEMPT trial is a prospective randomized phase II study of adjuvant ado-trastuzumab emtansine (T-DM1) vs. paclitaxel (T) and trastuzumab (H) following either mastectomy (M) or breast conserving surgery (BCS) in women with stage I HER2+ breast cancer. Eligible patients had T1N0 or T1N1mi HER2+ disease (3+ IHC or FISH ≥ 2) and were randomized 3:1 to either T-DM1 (3.6 mg/kg every three weeks x 17 treatments) or T (80 mg/m2 weekly for 3 weeks), H (4 mg/kg first week, then 2 mg/kg weekly x 12 weeks during T; then 6 mg/kg every 3 weeks x 39 weeks). Breast RT was required following BCS and permitted following M. Patients receiving T-DM1 began RT after 12 weeks of therapy and received RT concurrently with T-DM1; patients receiving TH began RT after T was complete and received RT concurrently with H. RT records were retrospectively reviewed to determine the fields, doses, and fraction size. Acute RT toxicity (CTC AE v4.0) was determined by two radiation oncologists; discrepancies were adjudicated by a third oncologist.Four hundred ninety-seven patients initiated protocol therapy (383 T-DM1, 114 TH). 75.1% had hormone receptor positive disease (HR+), and 3.8% micrometastatic nodal involvement. Among the 303 BCS patients, 41.7% in the T-DM1 arm and 42.6% of TH patients were treated with hypofractionation (HYPO; fraction size ≥ 2.5 Gy); nodal RT was given to 1.3% and 0%, respectively. 3.4% and 2.9% received partial breast irradiation (PBI). Of the 188 M patients, 4.3% underwent RT (all conventional fractionation (fx)). Grade 2 acute skin toxicity was seen in 33.2% of patients on the T-DM1 arm (20.3% HYPO), and 23.2% on the TH arm (25.0% HYPO). Only 2 patients experienced a grade 3 skin toxicity, both on the T-DM1 arm. One had conventional fx and the other HYPO. Five patients experienced pneumonitis following RT (T-DM1: n = 4, 1.0%; TH: n = 1, 0.9%). Two patients had grade 3 pneumonitis (both HYPO; 1 T-DM1, 1 TH); the remaining three patients had grade 2 pneumonitis (all T-DM1 and conventional fx). Two patients had an ipsilateral rib fracture following RT (one received TH, the other T-DM1). Three-year iDFS was 97.8% for the T-DM1 arm (95% CI 96.3-99.3), and 93.4% for TH (95% CI 88.7-98.2). Among the 18 iDFS events, 7 isolated local recurrences were seen (2 with T-DM1, 5 with TH; 5/7 HR+).RT was well-tolerated when given concurrently with either T-DM1 or H, regardless of fractionation. Local recurrences were unlikely, attesting to the efficacy of highly effective HER2-directed therapy with RT, and suggesting opportunities for future study of RT de-escalation.