Adaptive planning based on single beam optimization in passive scattering carbon ion radiotherapy for patients with pancreatic cancer

Yang Li,Tatsuya Ohno,Toshiaki Matsui,Shohei Okazaki,Masahiko Okamoto,Yoshiki Kubota,Mutsumi Tashiro,Takashi Nakano,Shintaro Shiba

doi:10.1186/s13014-021-01841-2

Abstract

BackgroundDaily anatomical deviations may distort the dose distribution in carbon ion radiotherapy (CIRT), which may cause treatment failure. Therefore, this study aimed to perform re-planning to maintain the dose coverage in patients with pancreatic cancer with passive scattering CIRT.MethodsEight patients with pancreatic cancer and 95 daily computed tomography (CT) sets were examined. Two types of adaptive plans based on new range compensators (RCs) (AP-1) and initial RCs (AP-2) were generated. In AP-2, each beam was optimized by manually adjusting the range shifter thickness and spread-out Bragg peak size to make dose reduction by < 3% of the original plan. Doses of the original plan with bone matching (BM) and tumor matching (TM) were examined for comparison. We calculated the accumulated dose using the contour and intensity-based deformable image registration algorithm. The dosimetric differences in respect to the original plan were compared between methods.ResultsUsing TM and BM, mean ± standard deviations of daily CTV V95 (%) difference from the original plan was − 5.1 ± 6.2 and − 8.8 ± 8.8, respectively, but 1.2 ± 3.4 in AP-1 and − 0.5 ± 2.1 in AP-2 (P < 0.001). AP-1 and AP-2 enabled to maintain a satisfactory accumulated dose in all patients. The dose difference was 1.2 ± 2.8, − 2,1 ± 1.7, − 7.1 ± 5.2, and − 16.5 ± 15.0 for AP-1, AP-2, TM, and BM, respectively. However, AP-2 caused a dose increase in the duodenum, especially in the left–right beam.ConclusionsThe possible dose deterioration should be considered when performing the BM, even TM. Re-planning based on single beam optimization in passive scattering CIRT seems an effective and safe method of ensuring the treatment robustness in pancreatic cancer. Further study is necessary to spare healthy tissues, especially the duodenum.

Highlights

Carbon ion radiotherapy (CIRT) has emerged as a viable treatment option for unresectable pancreatic cancer in the last decades [1,2,3,4]
Several studies have found that large position variation of pancreatic tumor and random gastrointestinal (GI) tract deformation exit between treatment fractions [7,8,9,10], unlike lung and liver tumors where good tumor reproducibility via tumor matching (TM) could account for interfractional deviations ensuring a robust treatment in carbon ion radiotherapy (CIRT) [11,12,13,14]
Houweling et al reported that using TM, the dose coverage was reduced by 10% on average and 17% for bone matching (BM) with CIRT [9]

Summary

Introduction

Carbon ion radiotherapy (CIRT) has emerged as a viable treatment option for unresectable pancreatic cancer in the last decades [1,2,3,4]. The carbon ion beam is highly sensitive to anatomical changes during treatment, which may cause a prominent dose distortion especially in pancreatic cancer. Some studies attempted to improve the dose distribution by modeling daily gastrointestinal organs variations [17] or optimizing the original plan (OP) (e.g., worst-case optimization) [18]. These efforts help maintain the robustness of the dose coverage; the single plan applied for the entire treatment course cannot fully account for the patient-specific interfractional organ deformation in pancreatic cancer [19]. Anatomical deviations may distort the dose distribution in carbon ion radiotherapy (CIRT), which may cause treatment failure.

Objectives

Methods

Results

Discussion

Conclusion