Role and Effectiveness of Hypofractionated Proton Beam Therapy and Combinations with Systemic Chemotherapy in Inoperable Extrahepatic Cholangiocarcinoma
Article information
Abstract
Purpose
This study aims to assess the clinical outcomes of hypofractionated proton beam therapy (PBT) for extrahepatic cholangiocarcinoma (EHCC) and to investigate the optimal sequencing for combining PBT with chemotherapy.
Materials and Methods
We retrospectively analyzed 59 consecutive patients with inoperable EHCC treated with PBT. The median prescribed dose of PBT was 50 GyE (range, 45 to 66 GyE) in 10 fractions. The combination sequences of PBT and chemotherapy were categorized as ‘Pre-PBT chemo’ (chemotherapy before PBT), ‘Post-PBT chemo’ (chemotherapy after PBT), and ‘No pre-/post-PBT chemo’ (no chemotherapy before or after PBT). Overall survival (OS), progression-free survival (PFS), and local PFS were estimated using the Kaplan-Meier method.
Results
All patients completed the planned treatments without any interruptions, and ≥ grade 3 acute adverse events were noted in 1.6% of the cases. The 1-year and 2-year freedom from local progression (FFLP) rates were 86.1% and 66.4%, respectively, with a median time of FFLP of 30.9 months. The 1- and 2-year OS rates were 74.5% and 25.3%, respectively, with a median survival time of 16.7 months. For prognostic factor analysis, pre- or post-PBT chemo was associated with a significantly reduced hazard ratio of 0.473 (95% confidence interval, 0.233 to 0.959; p=0.038) in the multivariate analysis. The median OS times for the groups receiving no pre-/post-PBT chemo, pre-PBT chemo, and post-PBT chemo were 14.6, 18.2, and 21.8 months, respectively (p < 0.05 for each).
Conclusion
Hypofractionated PBT for inoperable EHCC has demonstrated promising FFLP and OS rates with a safe toxicity profile. The combination of PBT with chemotherapy shows potential to improve clinical outcomes.
Introduction
In extrahepatic cholangiocarcinoma (EHCC), complete surgical resection offers the only chance for a cure, yet only a limited number of patients are considered suitable for this treatment [1]. The local progression of EHCC has a direct correlation with patient survival [2], and therefore, radiotherapy (RT) alone or combined with chemotherapy has been explored, showing improved median survival time from 2 to 4 months to 9 to 12 months [3,4]. To date, research on EHCC treatments mainly involves small, heterogeneous cohorts, often including other subtypes of bile duct cancers, due largely to the rarity of EHCC and the complexity of selecting appropriate treatment modalities and sequencing in the presence of diverse comorbid conditions in patients [5-15]. The efficacy and suitability of RT, particularly when combined with systemic therapy, are still subjects of ongoing research. Recently, proton beam therapy (PBT) has emerged as a promising treatment option due to its capacity to deliver higher doses of radiation to tumors while limiting exposure to surrounding healthy tissues, although data on the efficacy and safety of PBT remain scant [16]. Unfortunately, the infrequency and complexities of EHCC pose significant challenges in conducting randomized controlled trials in real-world scenarios to determine the efficacy and safety of PBT for inoperable EHCC patients. Hence, retrospective studies evaluating PBT for EHCC are considered invaluable.
The purpose of this study is to assess the clinical outcomes of hypofractionated PBT for EHCC regarding local tumor control, survival, and toxicity, and to also explore the optimal sequencing of PBT in combination with chemotherapy.
Materials and Methods
1. Patients
Patients who underwent PBT for EHCC between November 2012 and December 2021 were identified from a database and retrospectively reviewed. The inclusion criteria for the study were as follows: (1) histologically confirmed primary cholangiocarcinoma originating from the extrahepatic bile duct, extending from distal to hilar regions; (2) patients treated with PBT for extrahepatic biliary lesions; (3) absence of distant metastasis; and (4) no prior or current uncontrolled external malignancies. Exclusion criteria included patients who received PBT for non-extrahepatic lesions, had recurrent disease post-treatment, or displayed evidence of distant metastasis. All patients were assessed by a multidisciplinary team, including medical, surgical, and radiation oncologists, to thoroughly determine the optimal treatment approaches and the feasibility of surgical resection. Patients were staged according to the American Joint Committee on Cancer [17] guidelines, and, in addition to tumor factors such as stage and extent tumor, various clinical factors including performance status (PS), and underlying medical conditions were considered when deciding on the use, sequence, and regimens of systemic treatments and local treatments, including PBT.
Demographic, clinical, and treatment-related data were collected for each patient, including age, sex, Karnofsky PS, location, size, and stage of tumor, carbohydrate antigen 19-9 levels, prior biliary drainage history, therapies administered before, during, and after PBT, radiation doses prescribed for PBT, sites and timelines of disease progression, and survival status at the last follow-up. Systemic chemotherapy combinations with PBT were categorized as ‘Pre-PBT chemo’ (chemotherapy before PBT), ‘Post-PBT chemo’ (chemotherapy after PBT), and ‘No pre-/post-PBT chemo’ (no additional chemotherapy prior to and after PBT with/without concurrent chemotherapy).
2. PBT procedures
The procedures of PBT were meticulously described elsewhere [18]. Each patient underwent a contrast-enhanced four-dimensional computed tomography (CT) scan. The gross tumor volume (GTV) was marked on the average intensity projection CT images taken during the exhalation (gated) phases, accounting for 30% of the full respiratory cycle, combined with diagnostic imaging such as dynamic hepatobiliary CT and/or magnetic resonance imaging (MRI). The internal target volume (ITV), regarded as the clinical target volume, along with organs at risk (OARs), were delineated as the combination of GTV and each OAR on the CT images across all gated phases to account for their movements and extent changes. Further, the ITV with margins of 3-5 mm, cropped by 5 mm around gastrointestinal tract (GI) organs such as the stomach, duodenum, and both small and large bowels, was defined as the planning target volume 1 (PTV1). Similarly, the ITV with 5-7 mm margins was established as PTV2.
The PBT plan was executed using three non- or co-planar 230 MeV double-scattered proton beams (Proteus 235, Ion Beam Applications, S.A.) with the Eclipse software ver. 13.7 (Varian Medical System). The goal was to ensure that at least 90% of the PTVs were covered by 100% of the prescribed dose. The Gray equivalent (GyE) (GyE=physical dose×rela-tive biologic effectiveness of the proton beam [1.1]) was employed for expressing the radiation doses of PBT for each patient and the equivalent dose in a 2 Gy fraction (EQD2 ([GyE3 or GyE10]) was computed using the formula: EQD2 (GyE3 or GyE10)=total dose×[(fraction dose+α/β)/(2+α/β)], consideringα/β values of 10 (for tumors and acute responding tissues) or 3 (for late responding tissues). The median prescribed dose for PTV1 and PTV2 was 50 GyE (range, 45 to 66 GyE) in 10 fractions (EQD2; median, 62.5 GyE10; range, 54.4 to 91.3 GyE10) and 30 GyE in 10 fractions (EQD2, 32.5 GyE10), 5 times a week, respectively. Prescription doses for PTVs were determined based on the dose-volume constraints for OARs previously detailed [18]: constraints included radiation doses to the esophagus, stomach, and bowel (comprising duodenum, small and large intestine), which were capped at 39 GyE, 37 GyE, and 35 GyE, respectively, for a 2 cm3 volume; the maximum dose to the spinal cord was maintained below 39 GyE; and the relative volumes of the remaining residual liver (total liver–GTV) and the total liver receiving more than 27 GyE were kept below 50% and 60%, respectively. Patients were asked to fast for a minimum of 4 hours before undergoing PBT to reduce inter-fractional uncertainty. Irradiation was then administered during gated phases following localization with digital orthogonal and/or cone beam CT images. An example of PBT plan is presented in Fig. 1.
A 62-year-old woman diagnosed with unresectable perihilar cholangiocarcinoma received proton beam therapy (PBT) without systemic treatments due to her poor performance status. (A) Proton beam dose distribution. (B) Dose-volumetric histogram. PTV, planning target volume.
3. Chemotherapy
The decision for chemotherapy timing and regimen before, during, and after PBT was tailored by physicians based on the patient’s PS and age. Prior to PBT, 20 patients (33.8%) underwent systemic chemotherapy, predominantly with regimes of gemcitabine/cisplatin (n=9) or gemcitabine/cisplatin/abraxane (n=9), usually for an average of 5 months. Concurrent chemotherapy during PBT, administered to 24 patients (40.7%), mostly involved fluoropyrimidine-based agents, especially 5-fluorouracil and capecitabine (n=21). Following PBT, continued treatment options including chemotherapy were evaluated for all patients depending on their PS and age. Post-PBT, 24 patients (40.7%) were administered chemotherapy, primarily with gemcitabine/cisplatin (n=17) and gemcitabine/cisplatin/abraxane (n=6), usually starting 3 months after PBT. Notably, 13 patients underwent both pre- and post-PBT chemotherapy.
4. Assessment and statistical analysis
Patients underwent weekly assessments during PBT. Upon completing PBT, they underwent clinical evaluations, laboratory tests, and imaging studies, such as dynamic hepatobiliary CT and/or MRI, at 1 month, then every 3 months for the first 2 years, and every 6 months thereafter. The evaluation of tumor response and adverse events (AEs) was conducted in accordance with the Response Evaluation Criteria in Solid Tumors (ver. 1.1) [19] and the Common Terminology Criteria for AEs (ver. 5.0) (https://ctep.cancer.gov/protocolDevelopment/electronic_applications/ctc.htm#ctc_50). Clinicians confirmed disease progression through pathological and/or radiological evidence, which showed an incremental growth in size over time. Definitions were specified for local, regional, and distant progression: local as progression at the primary tumor or tumor bed site, regional within lymph nodes, and distant as the emergence of distant metastases. Time for freedom from local progression (FFLP), progression-free survival (PFS), and overall survival (OS) were defined as the intervals beginning from the start of chemotherapy or PBT (whichever was earlier) until the detection of locoregional progression, any progression, or death, and the occurrence of death or last follow-up, respectively. Survival estimates were computed using the Kaplan-Meier method. The log-rank test facilitated univariate analyses to compare survival outcomes, while multivariate analysis incorporated a stepwise forward selection strategy, considering variables with univariate significance set at p < 0.1. Hazard ratios (HRs) were established using the Cox proportional hazards model. Due to the limited research on prognostic factors for radiotherapy in extrahepatic cholangiocarcinoma, most cut-off values were adopted as exploratory choices, guided by findings from prior studies on cholangiocarcinoma management [20-22]. Statistical significance was achieved at p < 0.05, with all statistical computations carried out on SPSS software ver. 27.0 (IBM Corp.).
Results
Between November 2011 and December 2021, a total of 89 consecutive patients with EHCC underwent PBT. Of these, 59 patients were eligible for analysis after excluding 17 individuals treated for metastatic hepatic lesions and 13 treated for locoregional recurrent diseases. Baseline patient and treatment characteristics are summarized in Table 1. The median age was 76 years, ranging from 48 to 88. A significant majority of the patients (86.4%) had a biliary drainage procedure performed prior to PBT, with (n=44, 74.6%) receiving internal drainage either through endoscopic nasobiliary drainage or endoscopic retrograde biliary drainage, and (n=7, 11.8%) receiving external drainage through percutaneous transhepatic biliary drainage. Dose-volumetric parameters in PBT planning were summarized in S1 Table.
Patient and treatment characteristics
The follow-up duration varied from 1 to 97 months, with a median follow-up time of 16.3 months. At the time of the analysis, 44 of the 59 patients (70.2%) had succumbed to causes such as disease progression (n=18), infectious conditions including acute cholangitis or liver abscess (n=16), and unknown causes (n=10). After PBT, the best objective response was complete response in no patient (0%), partial response in six (10.2%), stable disease in 44 (74.6%), progression of disease in four (6.8%), and not evaluable in five (8.5%). A total of 35 patients experienced some form of recurrence, with 12 (20.3%) encountering local, eight (13.6%) regional, and 24 (40.7%) distant progression (Fig. 2). Among the local recurrences, one case was identified as a marginal recurrence at the PBT field margin.
Patterns of disease progression at the time of analysis: (A) initial patterns of disease progression and (B) cumulative patterns of disease progression. NED, no evidence of disease progression.
The FFLP, PFS, and OS curves for all patients were depicted in Fig. 3. The FFLP rate stood at 86.1% at 1 year and 66.4% at 2 years, and the median time for FFLP was 30.9 months (Fig. 3A). The 1- and 2-year PFS rates were 57.7% and 26.8%, respectively, with a median PFS duration of 14.1 months (Fig. 3B). The 1- and 2-year OS rates were 74.5% and 25.3%, respectively, with a median survival time of 16.7 months (Fig. 3C). The results of the univariate analysis for FFLP, PFS, and OS were summarized in Table 2. No significant prognostic factors were identified for FFLP. For PFS, both PS and pre- or post-PBT chemotherapy showed statistical significance (p < 0.05 each). For OS, both PS and pre- or post-PBT chemotherapy with PBT also were statistically significant (p < 0.05 each). The results of the multivariate analysis for PFS and OS were summarized in Table 3. Neither PS nor the chemotherapy combination was a significant prognostic factor for PFS. However, pre- or post-PBT chemotherapy with PBT significantly reduced the HR to 0.473 (95% confidence interval, 0.233 to 0.959; p=0.038) for OS. The OS curves, categorized by the combination sequence of chemotherapy with PBT, are depicted in Fig. 4. The median OS times for no pre-/post-PBT chemotherapy, pre-PBT chemotherapy, and post-PBT chemotherapy were 14.6, 18.2, and 21.8 months, respectively (p < 0.05 each). Patient distributions between no pre-/post-PBT chemotherapy, pre-PBT chemotherapy, and post-PBT chemotherapy are summarized in S2 Table. The median follow-up durations for the no pre-/post-PBT chemotherapy, pre-PBT chemotherapy, and post-PBT chemotherapy groups were 12.0, 16.8, and 21.6 months, respectively. The results of the multivariate analysis for OS were summarized in S3 Table. Post-PBT chemotherapy significantly reduced the HR to 0.299 (95% confidence interval, 0.113 to 0.792; p=0.015) for OS.
The duration of freedom from local progression (FFLP) (A), progression-free survival (PFS) (B), and overall survival (OS) (C) curves for all patients analyzed.
Univariate analysis of the pre-treatment characteristics for the FFLP, PFS, and OS
Multivariate analysis of PFS and OS
Comparison of survival curves based on the combination of chemotherapy (CTx) with proton beam therapy (PBT). Progression-free survival (PFS) (A) and overall survival (OS) (B) curves of PBT alone vs. pre/post-PBT CTx. PFS (C) and OS (D) curves of PBT alone vs. pre-PBT CTx vs. post-PBT CTx.
The AEs associated with PBT are summarized in Table 4. Eleven patients (18.7%) experienced acute grade 1-2 toxicities related to PBT, either combined with or without concurrent chemotherapy, with hematologic toxicity being the most prevalent. One patient developed grade 3 neutropenia one month after PBT during concurrent 5-fluorouracil treatment. Additionally, two patients experienced a grade 1 duodenal ulcer and a grade 2 gastric ulcer, respectively, which resolved with supportive care.
AEs related with proton beam therapy
Discussion
Surgical resection remains the only curative treatment option for EHCC, yet less than one-third of patients are suitable for curative resection at diagnosis [23]. Patients with inoperable EHCC generally experience dismal prognosis, with survival primarily influenced by local progression of the disease [12,24]. EHCC, a malignancy localized to a specific subsite of the biliary tract, follows a natural course and has an etiology distinct from other subsites [2,25]. However, previous studies focusing on EHCC treatment are often limited due to their small scale and frequently include other bile duct cancer subtypes (i.e., gallbladder, intrahepatic, and periampullary cancer) owing to the rare incidence of this disease [5-15]. Given the uncommon nature of EHCC and similar biliary tract cancers, it is challenging to conduct randomized clinical trials to evaluate the effectiveness of local treatment options, resulting in non-established standard therapies. Furthermore, the proximity of extrahepatic bile duct cancer to gastrointestinal organs and the presence of biliary comorbidities (i.e., biliary obstruction and cholangitis) can restrict the available treatment options. Consequently, fewer studies have investigated RT as a local treatment for extrahepatic bile duct cancer relative to other subtypes, and these studies are primarily retrospective and involve relatively few participants (Table 5) [5,7,8,13,16,26-28]. Nonetheless, the various limitations and complexities associated with EHCC render conducting randomized clinical trials for radiotherapy (RT) in real-world settings challenging. Hence, retrospective analyses of real-world clinical data, such as this study, are crucial and clinically valuable for assessing the efficacy of different RT techniques, including PBT.
Summary of previous studies for radiotherapy in inoperable extrahepatic cholangiocarcinoma
Several retrospective studies have demonstrated that conventional fractionated RT improves OS compared to no treatment (median, 9 vs. 4 months; p < 0.05) (Table 5) [4,7]. Combining conventional fractionated RT with concurrent chemotherapy has shown potential in enhancing both locoregional control and OS, proving feasible and well-tolerated in patients with inoperable and non-metastatic EHCC (Table 5) [4,7,26,27]. Additionally, data from the Surveillance, Epidemiology, and End Results database and the National Cancer Database have indicated that chemoradiotherapy is associated with improved survival compared to RT alone [6] and chemotherapy alone [29], respectively. In a phase I/II trial examining conventional fractionated RT of 49.6 Gy combined with concurrent gemcitabine followed by gemcitabine and capecitabine, the median OS time was 7.9 months, with 36.4% (4 out of 11 patients) experiencing grade 3 or 4 cholangitis [30]. These findings indicate the need of application and research of more sophisticated and modern RT techniques to safely and effectively treat EHCC patients.
Recently, stereotactic body radiotherapy (SBRT), which delivers an EQD2 ranging from 50 to 93.8Gy10 in 3-5 fractions, has been explored to improve local tumor control in EHCC patients (Table 4) [5,8,13,28]. Numerous studies on SBRT for inoperable EHCC patients, involving relatively small study populations, have shown encouraging outcomes in terms of local control and OS. However, they have also reported a relatively higher occurrence of grade 3 or higher AEs, including 2-year FFLP rates of 47%-73%, a median OS of 5-15 months, and grade 3 or higher AE rates of 11%-16%, mainly comprising GI and biliary AEs [5,8,9]. A specific analysis of 27 patients with inoperable cholangiocarcinoma, including 26 with perihilar tumors, who were treated with SBRT delivering 45 Gy in 3 fractions, was reported by Kopek et al. [8]. This study also demonstrated significant local control, with a 1-year FFLP of 84%, but noted a high incidence of grade 3 or higher AEs of 30% (Table 5) [8]. SBRT appears to enhance local tumor control by administering higher biological radiation doses in brief fractions, yet GI and biliary complications remain major concerns when using SBRT for EHCC patients. PBT, similar to SBRT, represents a contemporary radiation therapy technique that improves local tumor control and lowers the likelihood of AEs in EHCC patients. Yamazaki et al. [16] discussed the clinical outcomes of 93 patients with inoperable and recurrent EHCC treated with either conventional or moderately hypofractionated PBT, delivering doses of 67.5 GyE (range, 50 to 72.6 Gy) in 25 fractions (range, 22 to 30). They noted favorable survival outcomes with safe toxicity profiles: a median survival time of 20.1 months and and ≥ grade 3 acute and late AEs of 5.4% and 4.3%, respectively. In the current study, hypofractionated PBT, administering a dose of 50 GyE (range, 45 to 66 GyE) in 10 fractions, without pre- or post-PBT chemotherapy, exhibited promising survival outcomes with median OS of 14.6, 18.2, and 21.8 months, respectively. It is important to note that the earlier study by Yamazaki et al. [16] likely included patients with more favorable prognoses, such as those clinically diagnosed without biopsy and those with locoregional recurrent disease after surgical resection; hence, the survival outcomes observed in the present study are particularly promising. Moreover, the findings of this study suggest PBT can achieve high local tumor control with an acceptable toxicity profile, including 1-year and 2-year FFLP rates of 86.1% and 66.4%, respectively, and a grade 3 AE rate of 1.6%.
Chemotherapy has demonstrated an increasing survival benefit in advanced biliary tract cancer, including EHCC [14,15]. In this study, PBT achieved notably high local control, with 1-year and 2-year FFLP rates of 86.1% and 66.4%, respectively. However, the primary pattern of failure was distant metastasis (Fig. 2), and the application of systemic treatment both before and after PBT was significantly associated with enhanced OS (Tables 2 and 3). These findings indicate that, if a patient’s PS permits, integrating a treatment strategy that combines chemotherapy with local treatment is a reasonable approach. To effectively and safely combine radiotherapy and chemotherapy, utilizing the hypofractionated regimen of modern RT techniques to minimize interruptions of chemotherapy, reduce the risk of additional AEs, and increase patient compliance is advantageous. Given these considerations, the 2-week hypofractionated proton therapy utilized in this study emerges as a suitable therapeutic option for EHCC patients. The study has inherent limitations due to its retrospective design and relatively small study population, potentially leading to the overestimation of treatment effectiveness and underestimation of AEs. In the subgroup analysis examining chemotherapy sequence in combination with PBT, caution is warranted in interpreting results, as there were differences in patient distribution by age, PS, and stage between groups (S2 Table). Nevertheless, multivariate analysis was used to adjust for these confounding factors, and results indicated that the post-PBT chemotherapy group was associated with a reduced HR. The heterogeneity in chemotherapy regimens, primarily doublet or triplet combinations, could confound the results; however, regimens were tailored based on patient PS and age, so we expect minimal impact on clinical outcomes following PBT. Meanwhile, conducting large-scale clinical trials for therapeutic options in inoperable EHCC patients is challenging due to high prevalence among the elderly and complex conditions influencing treatment decisions, such as age, PS, and the presence or absence of biliary obstruction and cholangitis. Hence, despite their limitations, small retrospective studies in the real world possess meaningful clinical significance as they provide supporting data for clinical practice and further research. In this study, patients who received PBT combined with pre- and/or post-PBT chemotherapy exhibited better OS than those who underwent PBT alone. Moreover, patients who underwent PBT followed by chemotherapy demonstrated improved survival rates compared to those who received chemotherapy after PBT. These findings suggest that initiating PBT before chemotherapy might represent a rational treatment sequence for combinations of PBT and chemotherapy. Although the combination of PBT and chemotherapy appears to be safe and effective, further research is warranted to determine the optimal sequence of treatments.
In conclusion, hypofractionated PBT for inoperable EHCC has demonstrated encouraging results in terms of local tumor control and survival rates, with favorable toxicity profiles, even when integrated with chemotherapy before, during, and after PBT sessions. This study indicates that hypofractionated PBT may serve as an effective and safe local treatment option in conjunction with systemic therapy, thereby minimizing disruptions to systemic treatment while managing acceptable AEs and promoting patient adherence. Furthermore, initiating PBT prior to systemic therapy could be a viable treatment sequence. Nonetheless, further extensive prospective studies are essential to confirm the efficacy and safety of hypofractionated PBT, and to establish the most effective treatment sequencing for PBT and systemic therapy.
Electronic Supplementary Material
Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).
Notes
Ethical Statement
Patient data were anonymized following case number assignment, and all study methods complied with the Good Clinical Practice guidelines and the Declaration of Helsinki. The institutional review board of the National Cancer Center (NCC) (NCC2024-0067) approved this study, and the requirement for written informed consent was waived due to its retrospective design.
Author Contributions
Conceived and designed the analysis: Kim TH.
Collected the data: Lee SU, Kim TH.
Contributed data or analysis tools: Lee SU, Kim TH, Woo SM, Chun JW, Shin H, Cho YR, Kim BH, Koh YH, Kim SS, Suh YG, Moon SH, Lee WJ.
Performed the analysis: Lee SU.
Wrote the paper: Lee SU, Kim TH, Woo SM, Chun JW, Shin H, Cho YR, Kim BH, Koh YH, Kim SS, Suh YG, Moon SH, Lee WJ.
Conflict of Interest
Conflict of interest relevant to this article was not reported.
Funding
This study was supported by National Cancer Center grants (NCC 2410941). The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.