Optimisation and Individualisation of Heart-Sparing Breast Radiotherapy Techniques (The HeartSpare Study)
Session type: Poster / e-Poster / Silent Theatre session
The current priority in breast radiotherapy is to reduce mortality and morbidity without compromising local control. Of primary concern is reducing the dose received by cardiac tissues. Breath-hold techniques, during which the diaphragm pulls the heart downwards and away from the radiotherapy field, have been shown to reduce the dose to cardiac tissues. Two techniques in current use are active breathing control deep inspiratory breath-hold (ABC_DIBH) and voluntary deep inspiratory breath-hold (v_DIBH). v_DIBH is likely to be cheaper and easier to implement than ABC_DIBH as specialised equipment is not required. However, data on its reproducibility are lacking. This single centre, randomised, non-blinded crossover study (The HeartSpare Study) tests the hypothesis that v_DIBH will be equivalent to ABC_DIBH in terms of set-up reproducibility.
Total recruitment number is 23 patients (providing a power of 95% to rule out a mean displacement difference of 2mm). CT scans are acquired in each technique after appropriate patient training. Patients receive 40Gy/15# over 3 weeks; they are randomised to the technique with which they start their treatment and switch to the other technique for their final 8 fractions. Daily electronic portal imaging (EPI) is performed for each tangential beam. The population errors, including standard deviation (SD) of the random error (σ), the SD of the systematic error (Σ) and the group systematic error (M), are calculated for the tangential beams in AP and sup/inf directions.
Interfraction reproducibility of chest wall position has been analysed from the first 6 patients. The median errors (in mm) observed for either beam in either direction for v_DIBH and ABC_DIBH respectively were: σ (0.96, 0.90), Σ (1.42, 2.12) and M (0.23, 0.33).
Early results from this study suggest that v_DIBH and ABC_DIBH are equivalent in terms of interfraction reproducibility. Full results will be presented in November 2012.