Breathing Effects in Proton Scanning for Hodgkin's Lymphoma

Abstract: Purpose/Background Proton beam scanning (PBS) for mediastinal Hodgkin's Lymphoma (HL) is limited when treating in free-breathing (FB) due to breathing motion in the thoracic region. Tumour motion and the time structure of the beam delivery will cause interplay and dose blurring effects, resulting in discrepancies of dose in the planned dose distribution both inside and at the edges of tumour volume (TV). Current national guidelines only allow tumour motion less than $5$ mm in any direction to ensure target dose coverage. This study assesses the effects of simulated breathing motions in PBS delivery for various mediastinal HL dose distributions using different tumour motion amplitudes and beam delivery parameters. Specifically, this study aims to quantify the dosimetric effects of large breathing amplitudes on PBS delivery in the cases included in the study and identify parameters that may mitigate breathing effects. Material and methods The impact of breathing was estimated by computing relative dose differences between dose distributions during motion and corresponding static conditions. Dose distributions were measured using a 2D ionisation chamber array detector on top of a modified motion platform at the Swedish proton facility in Uppsala. Planar dose distributions from three HL patients with clinical robust treatment plans were measured in solid water for different target depths by allocating various amounts of solid water plates on top of the detector. The respiratory tumour motion was simulated using an asymmetrical sine-curve with amplitudes of 5, 10 and 15 mm in the superior-inferior (SI) direction. Tumour motion parameters were assessed by changing period time and motion regularity. We also studied a range of beam delivery parameters by modifying the air gap, spot spacing and accumulating fractions. Breathing effects were quantified by computing a 98th percentile of the absolute value of . Results Breathing effects increased with larger amplitudes but could be mitigated by accumulating fractions, increasing the air gap and reducing the spot spacing. The interplay effect dominated the overall breathing effect and varied drastically between patients at 5 mm motion amplitude, with a single-fraction 98th percentile value ranging from 10.3% in the worst-case scenario, representing a heterogeneous dose distribution, to 2.2% in the best-case scenario, corresponding to a homogeneous dose distribution. Conclusion For the patients considered, interplay effects are highly patient-specific, highlighting that the impact of respiratory motion has a high dependency on the beam delivery and plan characteristics and not only on the tumour motion pattern. PBS may be allowed for tumour motion above 5 mm in the SI direction. However, to reduce the interplay effect, care should be given to avoid a heterogeneous dose distribution in the tumour volume, keeping the spot intensities low if achievable.