KEY POINTS
- This exploratory retrospective study from Kobe University and Hyogo Ion Beam Medical Center, Japan, evaluated bioabsorbable spacer placement before high-dose proton beam therapy (PBT) for unresectable locally advanced pancreatic cancer involving the pancreatic body or tail. Fourteen consecutive patients underwent spacer surgery between August 2020 and January 2024. All had T4 disease, median age was 71 years, and median primary tumor diameter was 31 mm. One patient was found to have liver metastasis during surgery and subsequently received palliative treatment, leaving 13 patients in the definitive-PBT cohort.
- Bioabsorbable polyglycolic acid spacers were surgically inserted to separate the pancreatic tumor from adjacent gastrointestinal organs. Through laparotomy, trimmed PGA material was placed between the tumor and stomach and, when indicated, proximal jejunum. Tumor-to-GI separation ≥10 mm was achieved in all 14 patients. Median operative time was 163 minutes, median blood loss 37 mL, and median hospitalization ten days. The median interval between spacer implantation and proton treatment was 27 days, demonstrating that the additional surgical procedure imposed a substantial treatment-pathway commitment.
- Definitive proton therapy was delivered at 67.5 or 75 Gy(RBE) in 25 fractions. Four patients received 67.5 Gy(RBE) and nine received 75 Gy(RBE) using passive-scattering protons with end-expiratory respiratory gating. The principal gastrointestinal constraint was stomach/duodenum D0.5cm³ ≤48 Gy(RBE), with spinal cord maximum dose ≤45 Gy(RBE). Concurrent chemotherapy was administered to 10/13 patients, consisting of gemcitabine in six and S-1 in four. Elective para-aortic nodal irradiation was incorporated into the treatment volume.
- The composite primary safety endpoint occurred in four of 13 definitive-PBT patients (30.8%, 95% CI 9.1–61.4%). One patient developed grade 3 nausea requiring parenteral nutrition, and three experienced grade 3 leukopenia/neutropenia. No Clavien–Dindo grade ≥III surgical complications occurred among the 14 operated patients, although four (28.6%) experienced grade II complications, including delayed gastric emptying, paralytic ileus, or intra-abdominal abscess. Importantly, the composite safety endpoint was defined for the retrospective pooled analysis rather than uniformly prespecified across the contributing treatment protocols.
- No grade ≥3 structural gastrointestinal injury was observed, but the sample was too small to establish a low risk. The observed rate was 0/13, with an exact 95% CI of 0–24.7%. Three patients (23.1%) developed grade 1–2 duodenal mucosal injury, including one grade 2 ulcer and one symptomatic grade 2 duodenitis. The latter contributed to deferral of planned escalation to 82.5 Gy(RBE). Two patients also developed grade 2 vertebral compression fractures during follow-up, illustrating toxicity outside the gastrointestinal structures targeted by the spacer.
- Long-term local control remained limited despite high-dose irradiation. At median follow-up of 19.3 months, median overall survival from spacer placement was 24.8 months (95% CI 10.1 to not reached), and median progression-free survival was 11.8 months (95% CI 5.6–14.5). Estimated one- and two-year local control rates were 81.5% and 38.2%, respectively, while seven of 13 patients (53.8%) experienced local progression. The small cohort and limited numbers at risk make the two-year estimate particularly unstable.
- The study establishes feasibility, not the benefit of spacer insertion or dose escalation. No formal pre- versus post-spacer dosimetric comparison was performed across the cohort, and no non-spacer or photon-treated control group was available. Patients were selected for favorable performance status and specialized surgical and proton treatment, while pretreatment chemotherapy and radiation dose varied. The additional operative risks, short follow-up, and uncertainty surrounding uncommon late events prevent conclusions that bioabsorbable spacers reduce severe GI toxicity or improve survival.
CLINICAL TAKEAWAY
Bioabsorbable spacer implantation achieved sufficient anatomical separation to deliver proton therapy up to 75 Gy(RBE) in selected patients with unresectable pancreatic body or tail cancer. However, clinically significant adverse events occurred, and durable local control remained limited despite dose escalation. This small uncontrolled experience supports further prospective feasibility and safety evaluation, but does not establish a favorable benefit-risk balance for routine spacer surgery before high-dose proton therapy.