Researchers at the Paul Scherrer Institute (PSI) in Switzerland and Rigshospitalet in Denmark have conducted a study to determine whether the breath-hold technique is suitable for use with pencil-beam scanned (PBS) intensity-modulated proton therapy (IMPT). They determined that this technique would be clinically feasible for patients with tumours smaller than 200 cm3 (Int. J. Radiat. Oncol. Biol. Phys. doi: 10.1016/j.ijrobp.2017.08.023).

Their study was based on prior research conducted at the PSI. These studies showed that changes in water-equivalent path length (WEPL) and baseline shifts of a tumour were good indicators of breath-hold treatment plan robustness, for PBS proton therapy of patients with early-stage lung cancer (Acta Oncol. 56 853). Prior studies had also shown that the robustness of breath-hold treatment plans was better for patients with large tumours, principal investigator Jenny Gorgisyan told medicalphysicsweb.

Proton simulation

For the proton therapy simulation, Gorgisyan and colleagues utilized breath-hold CT data from 15 NSCLC patients who had received 66 Gy of image-guided photon radiotherapy in 33 fractions. They evaluated data from 10 breath-hold CT scans for each patient: one acquired at baseline, and three consecutive breath-hold images acquired at days 2, 16 and 31. They constructed three-field IMPT treatment plans based on the planning breath-hold CT scan, using multi-field optimization.

The researchers selected a robust beam arrangement so that each beam would traverse regions presenting small anatomical difference between the planning CT and the repeated breath-hold CT scan at day 2. These treatments were initially normalized to the mean planning target volume (PTV) dose, then increased 1–3% to maintain PTV V95% (the volume receiving 95% of the prescribed dose) around 95%. Dose constraints were met for all cases, with the exception of the oesophagus for four patients.

The researchers used their 4D dose calculation software based on the deformed dose grid approach. They also performed deformable image registration, which was a key aspect of the simulation in order to cumulative sum the dose distributions calculated on repeated CT scans. In total, they computed 1485 4D dose distributions (three fields x 33 fractions x 15 patients). During the evaluation, the researchers assessed the baseline shift of the tumour, changes in lung volume and changes in WEPL. They computed WEPL as the path-length to the distal edge voxels of the target, multiplied by the ratio of the relative stopping power of protons in the traversed material, to that in water.

Clinically feasible

IMPT treatment could be provided for 12 of the 15 cases, based on the clinical feasibility requirement of three breath-holds or less per treatment. All 12 patients had tumour sizes of less than 200 cm3. The three patients who required between four and nine breath-holds had clinical target volumes (CTV) of 371 cm3, 509 cm3 and 958 cm3. For nine of the 15 cases, IMPT in breath-hold was dosimetrically robust and feasible to deliver regarding the treatment time.

The authors reported that the mean CTV V95% was 99.8% for the planned and 99.7% for the simulated dose distributions. Dose degradation to the target and dose increase to organs-at-risk due to inter- and intra-fractional residual breath-hold motion was minor for most patients. The good plan robustness was verified by the fact that 12 cases had a CTV V95% degradation of less than 5%.

They note that the differences in WEPL appeared to be an important predictor of plan robustness. Because WEPL calculation is faster to perform than a full dose calculation, the authors believe that "the combination of evaluating WEPL online and the visual inspection of the daily verification imaging can potentially lead to more clinical confidence in the dose delivery for mobile lung cancers."

Gorgisyan advised that calculating WEPL on repeated CT scans would determine breath-hold reproducibility before treatment for patient selection. "A long treatment time results in more repeated breath-holds, which may both cause discomfort for the patient and larger uncertainty in the tumour position," she explained. "Using WEPL, we can identify patients for whom the breath-hold technique would be clinically inappropriate."

"Because prior studies showed that the breath-hold technique was superior to treatment in free breathing in terms of the surrounding organs-at-risk when delivering photon radiotherapy, we expect to see this tendency in proton therapy as well," Gorgisyan said. "However, this needs to be confirmed by multiple simulation studies for proton therapy, and our study represents initial investigation. We expect that a combination of breath-hold with other motion management techniques such as rescanning will further increase the robustness of the treatment plan."