The patient needs to be immobilized during C-ion RT for pancreatic cancer as is the case for other tumor sites. C-ion RT at facilities with the most powerful heavy-ion beams is performed with fixed beam directions that are vertical, horizontal, or at a 45-degree oblique angle to the target. However, recurrent rectal tumors can be located anywhere in the pelvis, including the presacral region and iliac and mesenteric lymph nodes, and therefore, the direction of the carbon ion beam needs to be varied. Furthermore, it is usually necessary to use more than two ports in order to obtain a dose distribution that avoids the bowels and the skin. Consequently, complex immobilization is frequently required.

At the National Institute of Radiological Sciences (NIRS), this was achieved using relatively thick, low-temperature thermoplastic (Shellfitter; Keraray Co., Ltd., Osaka, Japan), a sufficiently large customized cradle (Moldcare; Alcare, Tokyo, Japan), and a rotatable couch on the treatment bed.

Management of respiratory motion is also essential in most cases of recurrent rectal cancer (see Chap.​ 9, Motion Management). There are many different methods for motion management. At the NIRS, synchronization with respiration is adopted during both computed tomography (CT) image acquisition and during every C-ion RT session, irrespective of the direction of the beam. The effect of respiratory motion when using a vertical beam is less than that when using a horizontal beam for a pelvic tumor. However, this is mainly due to differences in the length of the air gap, and it is desirable to irradiate the tumor using similar conditions to those established by dose calculations using CT images taken during the expiratory phase.

A set of 2.5-mm-thick CT images along the whole abdomen were taken for treatment planning, with the patient placed in immobilizing devices and with respiratory gating.

Field arrangements were generally designed using a 3- or 4-field plan. The gross tumor volume (GTV) was established using CT images and margins for possible subclinical tumor spread, and magnetic resonance imaging (MRI) and fludeoxyglucose (FDG)-positron emission tomography were used to diagnose tumor spread. The clinical target volume (CTV) was defined as the gross volume plus 0.5 or 0 cm (in contact with the duodenum or stomach). The CTV included the GTV, neuroplexus, and regional lymph nodes. The latter consisted of the celiac, superior mesenteric, common hepatic, and peri-pancreatic lymph nodes (Fig. 26.1). At least 50 % of the functioning renal parenchyma received a dose of 15 GyE or less, and the spinal cord dose was limited to 30 GyE or less. Planning target volume (PTV) is defined as the CTV with a margin of at least 5 mm in all directions, although these margins are modified if critical organs, such as the skin or bowels, are close to the tumor.

The current recommended dose is 55.2 GyE delivered in 12 fractions for locally advanced pancreatic cancer and 35.2 GyE delivered in 8 factions for resectable pancreatic cancer. The whole of the prescribed dose is given at the maximum dose point of each portal, and treatment planning is performed so that the PTV is covered with at least 90 % of the prescribed dose. For this purpose, additional margins are necessary to determine the shape of the collimator and compensation bolus. Figure 26.1 shows a representative dose distribution for C-ion RT as it is currently used at the NIRS.

Major organs at risk during this treatment are the stomach, duodenum, intestine, kidney, and spinal cord. It is particularly important to spare the bowels from the high-dose area of the carbon ion beam. At the NIRS, dose constraints for the digestive tract are established on the basis of the treatment results for prostate, uterine, cervical, and recurrent rectal cancers, where the maximum dose to the colon should not exceed 83 % of the prescribed dose.

At the NIRS, C-ion RT is performed once a day and in 4 fractions per week (from Tuesday until Friday). The field is verified at every treatment session using a computer-aided online positioning system to prevent the positioning error exceeding 2 mm. A fluoroscopic image is taken during the expiration phase of respiration and compared with the reference images, including the image taken during the previous simulation or the digitally constructed radiograph. C-ion RT is performed with respiration gating as described above, where the carbon ion beam is delivered only during the expiration phase.