Water-equivalent pathlength reproducibility due to respiratory pattern variation in charged-particle pancreatic radiotherapy.

We evaluated the water-equivalent length (WEL) reproducibility due to variation in the external respiratory marker position when using a 4DCT scan in respiratory-gated charged-particle treatment. Two sets of pancreatic 4DCT data from two patients were acquired under free breathing conditions with 256-slice CT. The 4DCT data included two exhalation phases and the respiratory patterns in each patient differed, one being regular and the other irregular. The WEL calculation region is defined in the first respiratory cycle by two planes, one at the patient entrance surface and the other behind the target in the anterior-posterior (AP) and posterior-anterior (PA) directions. In the regular respiratory pattern, the WEL variation within the target region was less than 1.7 mm between the first and second exhalations in both AP and PA calculation directions. However, in the irregular breathing pattern, the respiratory amplitude at the second exhalation was 20% lower than that at the first exhalation; therefore, WEL variations from 8.1 to -9.1 mm and from 3.1 to -3.4 mm were observed within the target region in the AP and PA calculation directions, respectively. The WEL variation in the PA direction was smaller than that in the AP direction because the abdominal thickness is affected more in the AP direction. Respiratory pattern variation even affects WEL values in the respiratory-gated phase. This variation should be considered in treatment planning, and necessary improvements in respiratory reproducibility should be made.