Carbon Ion Radiotherapy Versus Perioperative Adjuvant Chemotherapy and Curative Surgery for Resectable Pancreatic Cancer

Patients and Methods

Ethical approval. The study was approved by the Institutional Review Board (IRB) of Kanagawa Cancer Center prior to initiation (Approval No.: 2022 Epidemiological Study-21). Written informed consent for the use of non-personally identifiable clinical data was obtained from each patient prior to surgery.

Patients. Pancreatic cancer patients who underwent NAC + curative surgery + adjuvant chemotherapy or CIRT monotherapy at Kanagawa Cancer Center from September 2018 to September 2021 were retrospectively investigated. The inclusion criteria for chemotherapy cases were: 1) patients with histopathologically proven pancreatic cancer; 2) patients with clinical cT1-2, N0-1, M0, or cStage I-II pancreatic cancer diagnosed by the International Union Against Cancer 8th edition TNM classification (8); 3) patients who received NAC with gemcitabine and S-1 + curative resection + postoperative adjuvant chemotherapy with S-1; and 4) patients with Eastern Cooperative Oncology Group (ECOG) performance status 0–1. The exclusion criteria were: 1) patients who did not receive perioperative chemotherapy for any reason or whose chemotherapy was stopped during the procedure, 2) patients who underwent residual (R1 or R2) resection instead of radical (R0) resection, 3) patients who had multiple cancers or a history of cancer, and 4) patients who withdrew consent.

The inclusion criteria for CIRT-treated cases were: 1) patients with histopathologically proven pancreatic cancer; 2) patients with cT1-2, N0-1, M0, or cStage I-II pancreatic cancer diagnosed by the International Union Against Cancer 8^th edition TNM classification; 3) patients who underwent and completed CIRT treatment; and 4) patients with an Eastern Cooperative Oncology Group (ECOG) performance status 0–1. The exclusion criteria were: 1) patients in whom CIRT was discontinued for some reason, 2) patients with multiple cancers or a history of cancer, and 3) patients who withdrew consent. As a result, 110 NAC + curative surgery + adjuvant chemotherapy cases and 35 CIRT-treated cases were included in the study (Table I). These cases were matched and compared using the PSM method described below.

CIRT. The eligibility criteria for CIRT monotherapy for pancreatic cancer were a pathological or clinical diagnosis of pancreatic cancer, no distant metastases, and no proximity to the gastrointestinal tract. The treatment plan for CIRT included 4D computed tomography (CT) with a 2-mm slice thickness to account for respiratory migration. Contrast-enhanced CT was also performed to distinguish between tumour and normal tissue. In addition, fluorodeoxyglucose-positron emission tomography (FDG-PET) was performed prior to radiotherapy for target identification. The clinical target volume (CTV) included GTV as well as tumour invasion that could not be identified grossly. The internal target volume was defined as the range of movement of the CTV by respiration identified from the 4D CT. In addition, a setup margin of at least 5 mm in 3D was used as the PTV; the PTV was corrected for anatomical boundaries such as the gastrointestinal tract and non-infiltrated bone.

The prescribed dose for CIRT was set at 55.2 Gy relative biological effectiveness (RBE) with a margin of at least 95% of the PTV. The maximum dose for the gastrointestinal tract was 44 Gy RBE, and the irradiated area was adjusted for each case. CIRT was administered in 12 fractions over a 3-week period.

NAC. NAC was administered intravenously with gemcitabine at a dose of 1,000 mg/m² on days 1 and 8, and S-1 was administered at the following doses according to body surface area (BSA): 40 mg for BSA <1.25 m², 50 mg for BSA 1.25–1.5 m², 60 mg for BSA >1.50 m² for 21 days. Twice daily doses were administered on days 1–14 of the cycle. Patients with creatinine clearance of 50–60 ml/min received a reduced dose of 20 mg/day of S-1. Neoadjuvant treatment was repeated for two cycles unless unacceptable toxicity developed. If Grade 3 hematologic toxicity or Grade 2 nonhematologic toxicity was observed, both gemcitabine and S-1 were withheld until recovery; patients who exhibited Grade 4 hematologic toxicity or Grade 3 nonhematologic toxicity were administered gemcitabine (−200 mg/m²/day) and S-1 (−20 mg/day) were both administered at reduced doses.

Surgery. Surgery was performed 2–4 weeks after the last administration of S-1, usually within a maximum of 6 weeks, and curative resection with regional lymph node dissection was performed depending on the location of the individual pancreatic cancer and its progression, if the patient was diagnosed as cT1-2, N0-1, or M0 on imaging evaluation after preoperative chemotherapy, and if curative resection was feasible. Intraoperative laparoscopic cytology was mandatory. Unexpected intraoperative findings of unresectability, such as distant metastasis or tumour extension into the hepatic, celiac, or superior mesenteric arteries, were not considered curative resection, and appropriate bypass surgery was performed as necessary. Patients who did not undergo curative resection were excluded from the study.

Postoperative adjuvant chemotherapy. Patients who had undergone radical resection were treated with S-1 for 6 months within 10 weeks after curative surgery as adjuvant therapy, and postoperative adjuvant therapy with S-1 was administered to patients who had histologically confirmed pancreatic cancer by histological examination of the resected specimen, had a histologically R0 curative gross resection, had a negative peritoneal lavage cytology for cancer cells, had no distant metastases such as extra-regional metastases, and had no early recurrence confirmed by CT scan performed before the scheduled postoperative adjuvant chemotherapy. Patients were defined as those with no distant metastasis such as extra-regional nodal metastasis and no early recurrence confirmed by CT scan performed prior to adjuvant chemotherapy. S-1 was administered orally at 80 mg/m²/day twice daily for 4 weeks, followed by a 2-week rest period and four cycles. Treatment was continued unless unacceptable toxicity developed. Patients who developed grade 3 hematologic toxicity or grade 2 nonhematologic toxicity had S-1 discontinued until recovery. For patients who developed grade 4 hematologic toxicity or grade 3 nonhematologic toxicity, the dose of S-1 was reduced by 20 mg/day when treatment was resumed.

PSM. To evaluate the outcome of surgery/CIRT monotherapy in this study cohort, 1:1 matching between the surgery and CIRT monotherapy groups was performed. Patients were stratified by laboratory parameters such as age, sex, tumour site, tumour diameter, UICC8 T stage, pre-treatment CA19-9 level, pre-treatment CEA level, and presence of hypertension, diabetes, or COPD to reduce patient-related and confounding group bias. Matching was performed using the EZR program (Saitama Medical Center, Jichi Medical University, Tochigi, Japan), a graphical user interface (GUI) of R (The R Foundation for Statistical Computing, Vienna, Austria), using a calliper with a width equal to 0.2 of a standard deviation.

Statistical analysis. The χ² test or Fisher’s exact test was used to compare differences between categorical variables, and the Mann–Whitney U-test was used for continuous or nominal values; RFS and OS were calculated from the date of first surgery for surgery cases and from the last day of treatment for CIRT cases, and curves were generated using the Kaplan–Meier method. Differences between curves were analysed by the log-rank test. Variables identified as significant (p<0.05) in the univariate analysis for survival were candidates for the multivariate regression analysis, and results were presented as hazard ratios (HR) and 95% confidence intervals (CI); values of p<0.05 were considered statistically significant. All statistical analyses were performed using EZR (Saitama Medical Center), a GUI of R (The R Foundation for Statistical Computing).