Session Item

Regarding salvage radiation therapy (RT) for breast cancer, lymphedema is one of the most concerning complications. However, optimal dose constraints to minimize subsequent arm lymphedema (SAL) are not well studied. This study aimed to identify risk factors for SAL including dosimetric parameters in patients treated with salvage repeated RT for locoregional breast cancer.

We retrospectively reviewed 65 patients who received salvage repeated RT for locoregional recurrent breast cancer between 2003 and 2017. Regional nodal areas were divided into 7 distinct subregions and retrospectively contoured per atlas suggested by Gross et al (Int J Radiat Oncol Biol Phys, 2019). Initial and repeated RT plans for each patient were deformed and merged to calculate dose-volume histogram at nodal area. Dosimetric analysis was performed using EQD2 dose of Dmax, Dmean, Dmin, V30Gy, V40Gy, V45Gy, and V50Gy, V60Gy of each subregion. Maximally selected rank statistics were used to identify optimal cut points of dosimetric variables. Multivariate Cox proportional hazard regression and Akaike information criterion model were used to identify optimal model to predict SAL.

Salvage repeated RT was combined with salvage surgery in 34 patients (52.3%) and salvage taxane-based chemotherapy was administered in 29 patients (44.6%). Total number of RT course were 2 in 59 patients (90.6%) and 3 in 6 (9.4%). Median EQD2 of accumulated prescription dose was 113.0 Gy (interquartile range [IQR] 105.0-121.6). Salvage RT modalities were 3DCRT/electron in 28 patients (43.1%), IMRT in 26 patients (40.0%), and proton beam therapy in 11 patients (16.9%). With a median follow-up of 30.4 months (IQR 19.3-52.5), 15 patients (23.1%) experienced SAL with a median interval of 6.43 months (IQR 13.3-16.5) after secondary RT. Patients with SAL frequently received salvage taxane-based chemotherapy and higher dose to axillary-lateral thoracic vessel junction (ALTJ), lateral to ALTJ, supraclavicular/superior-medial, and superior-lateral region compared with patients without SAL. In multivariate analysis for SAL without dosimetric parameters, salvage taxane-based chemotherapy remained significant (Hazard ratio [HR], 4.19; P= 0.041). After incorporating dosimetric parameters, an final optimal model selected lateral to ALTJ V30Gy≥55.8% (HR 5.35; P = 0.001) and salvage taxane-based chemotherapy (HR, 4.73; P = 0.021).

Salvage repeated RT for locoregional recurrent breast cancer induced SAL in 23.1% of the patients. Salvage taxane-based chemotherapy and lateral to ALTJ V30Gy≥55.8% significantly increased the risk of SAL. Incorporating lateral to ALTJ V30Gy into dose constraints and early consult to the department of rehabilitation could be helpful to reduce SAL when planning salvage repeated RT for locoregional breast cancer.