Accuracy of breast cancer axillary lymph node treatment plans based on 2-dimensional imaging: what we should know before interpreting 2-dimensional treatment-planning era studies.

Objective: To investigate axillary 2-dimensional treatment planning accuracy.

Methods: Computed tomography (CT) simulation data for 16 breast cancer cases taken after level I-II axillary dissection were analyzed. An additional 6 patients underwent CT simulation using the historical 90-degree position (HP), and the standard-bore CT position (CT-P). Two physicians identified the lateral and medial borders of the coracoid process (CCP) on digitally reconstructed radiography (DRR). The DRR-identified x coordinates were compared with the CT-measured x coordinates. x coordinates differences between the most medial surgical clip and the borders of the CCP as identified on CT were analyzed. Fields were designed to cover various amounts of the axilla, and treatment plans were generated to compare doses to the most medial surgical clip.

Results: In 11 and 6 cases for each physician, respectively (lateral border), and in all cases for both physicians (medial border), the DRR identification of the CCP was medial to that on CT. In 9 and 8 cases, the most medial surgical clip was lateral to the medial and lateral borders of the CCP, respectively. In all data sets, the average difference was larger in the HP compared with CT position. The number of patients who received more than 90% of the prescribed dose when using the plans with the mid humeral head border, lateral border of the CCP, and medial border of the CCP were as follows: 6, 1, and 0, respectively.

Conclusions: When using 2-dimensional treatment planning, the dose to the undissected axilla can vary depending on the anatomic landmark used to define the lateral border of the axillary field. This may account for outcome differences found in older radiotherapy studies.