Abstract GS2-03: Insights of applied radiotherapy among patients undergoing breast-conserving surgery with or without axillary sentinel lymph node biopsy: secondary results from the INSEMA trial

Abstract Background: The de-escalation of axillary surgery during breast-conserving surgery (BCS) must be considered alongside radiotherapy to accurately interpret oncological outcomes. The INSEMA trial demonstrated that omitting axillary sentinel lymph node biopsy (SLNB) in clinically node-negative early breast cancer (BC) patients (pts) undergoing BCS is oncologically safe regarding 5-year invasive disease-free survival (iDFS). Analysis from other de-escalation trials (ACOSOG Z0011, SENOMAC) revealed that a significant percentage of pts received regional nodal irradiation (RNI). INSEMA documented dose distribution in ipsilateral axillary levels I-III and captured RNI. This study investigates how pts parameters, surgical axillary extent, and radiation techniques affect ipsilateral axillary dose distribution. Material and methods: INSEMA (NCT02466737), a surgical trial on SLNB omission in early invasive BC and BCS, randomized 5502 pts between September 2015 and April 2019 in Germany and Austria. The protocol mandated whole-breast irradiation (WBI) while RNI was only permitted in patients with 4 or more involved lymph nodes. This pre-planned secondary analysis includes 5154 pts from 108 radiotherapy (RT) facilities. Contouring of the ipsilateral axilla (level I-III) followed the Radiation Therapy Oncology Group consensus definitions. Dose parameters are presented as relative doses (in % of the prescribed breast dose) to avoid different absolute doses between conventionally and hypofractionated cases. Results: Of 5154 pts, 4890 (95.7%) received postoperative WBI per INSEMA protocol. The majority of pts (N=2800, 58.0%) were treated with 3D-conformal RT-technique using standard tangential fields, other pts received various modern intensity-modulated radiotherapy techniques. Deep inspiration breath-hold was used in 102 pts (2.1%). Conventional fractionation was more common (N=3163, 66.0%) than moderate hypofractionation (N=1630, 34.0%). A tumor bed boost was applied in 3807 pts (74.9%), delivered either simultaneously (N=1973, 52.6%) or sequentially (N=1781, 47.4%); 257 pts (6.8%) received an intraoperative boost irradiation. No differences were observed between randomized groups regarding the technique used, fractionation schedule, and boost application. Median and average dose for each axillary level differed significantly between arms with higher median doses in the SLNB arm compared to the no SLNB arm for level I (91.4% vs. 86.3%; p