Lymphedema Risk Following Regional Nodal Irradiation in Breast Cancer: Implications for Field Arrangement and Treatment Volume

Abstract

Upper extremity lymphedema (LE) is an important complication with long-term quality-of-life impact for breast cancer survivors. Rates of LE following regional nodal irradiation (RNI) range from 9% to 65% based on known risk factors such as BMI and extent of nodal dissection. The effect of other factors such as beam arrangement and axillary volume encompassed in the radiotherapy (RT) field has not been well characterized. This study sought to identify factors associated with the development of LE, including detailed information on RT beam arrangement and treatment volume. Between 1999 and 2015, 492 women received RNI for breast cancer treatment at our institution following surgery. Arm circumferences were prospectively measured prior to the initiation of RT and during follow-up. RT fields were defined as follows: Field 1 = exclusion of the dissected axilla, defined as anterior beam encompassing < 1/3 of the humeral head at the lateral field edge, +/- a supplementary posterior axillary (PA) field. Field 2 = anterior beam encompassing ≥ 1/3 of the humeral head, +/-a supplementary PA field. Field 3 = parallel-opposed beams encompassing the full axilla. LE was defined as arm circumference on the treated side ≥ 2.5cm or ≥ 2cm on 2 visits either 10cm above or 15cm below the olecranon process compared to the opposite arm, or LE documented after evaluation by a LE specialist. Cox proportional hazards models were used to analyze risk factors for LE development. Median post-operative follow-up was 66 months (range, 7-204 months). The overall 2-year and 5-year cumulative incidence (CI) of LE was 18.5% and 24.8% respectively. The proportion of patients receiving fields 1, 2 and 3 were 20.5% (n = 101), 41.1% (n = 202) and 38.4% (n = 189) respectively. On univariate analysis, Field 1 was associated with lower 5-year CI of LE (6.93%) compared to field 2 (28.2%; P<0.0001) and field 3 (30.7%; P<0.0001). There were no significant differences in year of treatment, pre-treatment BMI, handedness, age, type of breast surgery, type of axillary surgery, number of LN removed or use of chemotherapy between groups. On multivariate analysis, pre-treatment BMI (hazard ratio [HR]: 1.02, P = 0.0002), number of LN dissected (HR: 1.03, P = 0.002), field 2 (HR: 4.73, P<0.001) and field 3 (HR: 3.37, P = 0.002) were associated with LE risk. Overall rate of regional recurrences (RR) was 1.62%. Despite equal proportion of locally advanced disease, field 1 was not associated with higher rate of RR (P = 0.43). In addition to other risk factors, the volume of irradiated axilla was the most important risk factor for LE following RNI. Restricting the volume of axilla being treated to undissected axilla following axillary dissection conveys a lower risk of lymphedema without increased risk for RR. By contrast, the beam arrangement was not associated with LE risk. Adherence to expert guidelines regarding omission of the dissected axilla appears to be integral to minimizing the rate of LE after breast radiotherapy.