Simplified intensity-modulated radiotherapy using pre-defined segments to reduce cardiac complications in left-sided breast cancer

Background and purpose: Left-sided breast cancer patients pose a difficult clinical challenge when significant heart and contralateral breast irradiation are present, particularly with tangential uniform beams. The aims of the study are: (1) to design and evaluate a simplified intensity-modulated radiotherapy (IMRT) (SI) solution using pre-defined segments, (2) to compare the SI technique with a conformal (CN) and a full fluence IMRT (FI) approach using two sets of beam orientations, clinical (-C) and optimal (-O), and (3) to quantify the benefits of treatment technique and beam orientation. Patients and methods: Nine left-sided breast cancer patients with a maximum heart distance of at least 2.0 cm were planned using three different techniques and two different beam orientations. All three techniques were planned using clinical orientations (i.e. CN-C, FI-C and SI-C). Two techniques were planned using more optimal orientations (i.e. FI-O and SI-O). Dose-volume histograms and radiobiologic modelling are used for plan evaluation. Results: The average mean planning target volume (PTV) doses are 91.6 +/- 4.5, 98.4 +/- 6.3, 102.0 +/- 8.7, 100.0 +/- 5.9 and 103.9 +/- 8.3% for the CN-C. FI-C. SI-C, FI-O and SI-O plans, respectively. The average normal tissue complication probabilities for late excess cardiac mortality are 2.1 +/- 0.6 +/- 0.2 +/- 0.1, 0.2 +/- 0.1, 0.1 +/- 0.0 and 0.1 +/- 0.0%, respectively. For a given beam orientation, FI plans are the best and CN plans are the worst. The dose distributions for the SI-C and FI-C plans are almost identical with significant heart sparing but at a cost of some target underdosage. The dose distributions are better conformed around the PTV with more optimal beam orientations, resulting in better sparing of adjacent organs at risk. FI-C plans are inferior to SI-O plans. Conclusions: For clinical uniform two-beam orientations, significant heart sparing is possible with the addition of intensity modulation but at the expense of worsening target coverage. Simplified IMRT can, for all intents, be substituted for full IMRT with clinical beam orientations. Applying more optimal non-Uniform beam orientations improves PTV coverage while maintaining significant heart sparing but increases the PTV dose heterogeneity. (C) 2003 Elsevier lreland Ltd. All rights reserved.