A Novel Method to Set Appropriate Gating Window for Deep Inspiration Breath Hold Patients

Abstract

Patients receiving radiation therapy for left breast cancer (LBC) are at increased risk of cardiac toxicity. Deep inspiration breath-hold (DIBH) is used to reduce organ-at-risk (OAR) doses. The breath-hold window (BHW), as defined during patient simulation, is set qualitatively without accurate knowledge of potential dose benefits. Furthermore, a patient’s inability to reproducibly perform DIBH during treatment may compromise planned OAR doses. We aim to determine the ideal BHW and evaluate OAR dose consequences for patients unable to maintain DIBH. Twenty LBC patients treated with DIBH at our institution were identified in this IRB approved retrospective study. Using a respiration phase monitor, the window width, and level of breath-hold for each patient was set during CT simulation. Treatment plans with opposed tangents computed on free-breathing (FB) and DIBH CT scans were normalized to deliver >95% dose to >95% PTV. Minimum cardiac to chest wall distance (CCD) measured on the CT slice containing maximum heart in tangent field was recorded for both FB and DIBH. Using this data, we determined the change in OAR Dmean as a function of CCD. To correlate OAR doses with all ten phases of breathing cycle, we generated treatment plans on FB and 4DCT scans from 10 additional patients. These data were used to estimate dosimetric gains for different breath-hold levels: shallow, moderate, and deep inspiration and to set an optimal BHW. Plans were evaluated based on student t-test. Compared to FB plans, DIBH significantly reduced average heart Dmean: 1.8Gy vs 3.0Gy. The corresponding reductions in average Dmean for left ventricle (LV) and left anterior descending artery (LAD) were 2.5Gy vs 4.4Gy, and 8.4Gy vs 19.5Gy, respectively. These dose improvements correlated with an average increase of 16.1mm in CCD as measured on patient CT scans. The calculated rate of change (% dose/mm of CCD) for heart, LV, and LAD Dmean were 2.8%/mm, 2.6%/mm and 3.6%/mm, respectively. This linear correlation between OAR dose and CCD was validated with 4DCT patient data showing similar trend. We analyzed OAR dose change for three levels of breath-hold based on CCD range: shallow (0-5mm), moderate (5-10 mm) and deep (10-15mm). It was found that at these BH levels, 30%, 20%, and 10% of patients respectively had their mean heart dose exceed our institutional threshold (3Gy). In contrast, 50% of FB patients exceeded heart dose threshold. Hence, for our patients, a +/- 5 mm BH window at 15 mm above inspiration level was found to be appropriate. We have developed a novel method to set an optimal breath hold window for DIBH patients which should lead to more reproducible and consistent treatments with greater OAR dose benefit. The method also provides an estimate of dose consequences for patients unable to adequately breath-hold during treatment.