Hypofractionated particle beam therapy for hepatocellular carcinoma-a brief review of clinical effectiveness.

Hepatocellular carcinoma (HCC) is the fifth most common malignancy and the second leading cause of cancer mortality worldwide. The cornerstone to improving the prognosis of HCC patients has been the control of loco-regional disease progression and the lesser toxicities of local treatment. Although radiotherapy has not been considered a preferred treatment modality for HCC, charged particle therapy (CPT), including proton beam therapy (PBT) and carbon ion radiotherapy (CIRT), possesses advantages (for example, it allows ablative radiation doses to be applied to tumors but simultaneously spares the normal liver parenchyma from radiation) and has emerged as an alternative treatment option for HCC. With the technological advancements in CPT, various radiation dosages of CPT have been used for HCC treatment via CPT. However, the efficacy and safety of the evolving dosages remain uncertain. To assess the association between locoregional control of HCC and the dose and regimen of CPT, we provide a brief overview of selected literature on dose regimens from conventional to hypofractionated short-course CPT in the treatment of HCC and the subsequent determinants of clinical outcomes. Overall, CPT provides a better local control rate compared with photon beam therapy, ranging from 80% to 96%, and a 3-year overall survival ranging from 50% to 75%, and it results in rare grade 3 toxicities of the late gastrointestinal tract (including radiation-induced liver disease). Regarding CPT for the treatment of locoregional HCC, conventional CPT is preferred to treat central tumors of HCC to avoid late toxicities of the biliary tract. In contrast, the hypo-fractionation regimen of CPT is suggested for treatment of larger-sized tumors of HCC to overcome potential radio-resistance.

Core tip: Charged particle therapy (CPT) for hepatocellular carcinoma (HCC), including proton beam therapy and carbon ion radiotherapy, offers physics-related advantages and results in better local control rates and lesser adverse effects. For peripherally large-sized HCC tumors, the hypo-fractionation regimen of CPT provides the benefit of increasing local control rates through overcoming radio-resistance, whereas conventional CPT is preferred for treating central tumors of HCC in terms of avoiding late toxicities of the biliary tract. Prospective data that will add to the accumulated evidence on the dosimetric constraints of hypofractionated CPT for the treatment of HCC are needed.

INTRODUCTION

Hepatocellular carcinoma (HCC), with a reported 5-year overall survival (OS) of 10% to 15%[1], is the fifth most common malignancy and the second leading cause of cancer mortality worldwide, with an estimated 782000 new cases and 745000 deaths in the year 2012[2,3]. The global increase in both the incidence and mortality of HCC is a major concern, and improving the management of HCC is a key challenge.

The cornerstone to improving the prognosis of HCC patients has relied on the control of loco-regional disease progression; loco-regional disease progression is the major cause of HCC-related death[4]. Surgical interventions, including liver resection and transplantation, are considered the first priority treatment modalities for patients with HCC[5]. However, only 10%-37% of patients are treated with surgery at the time of diagnosis because of their inability to tolerate the possible surgery-related complications because of underlying comorbidities[6-9].

Local ablation treatments, including percutaneous ethanol injection (PEI) and radiofrequency thermal ablation (RFA), have been recognized as alternative treatment options for patients with HCC, even though patients with large HCCs (> 5 cm in diameter) are not eligible to receive either PEI or RFA[10,11]. Transarterial chemoembolization (TACE) provides some benefits of locoregional control and a better prognosis for HCC patients in whom surgery or local ablation treatment is not feasible, although it is regarded as a non-curative treatment[12,13]. However, HCC patients who present with portal vein tumor thrombus (PVTT) are not advised to receive TACE treatment because of the increased risk of liver failure[14,15].

Radiotherapy (RT) has not been considered as a preferred treatment modality for HCC; instead, it is a complementary local treatment option for patients who are not candidates for surgery, local ablation, and TACE, mainly because the RT dose required for tumor ablation would be beyond the tolerance dose of the liver parenchyma and may induce liver injury, including classic and non-classic radiation-induced liver disease (RILD)[16]. In contrast to conventional RT, stereotactic body radiotherapy (SBRT), which combines image guidance technique and radiotherapy planning designation, not only provides highly conformal radiation delivery to allow ablative doses to be applied to tumors, but simultaneously spares the normal liver parenchyma from radiation[17-19]. SBRT, which is commonly performed using high radiation dose per fractions, has resulted in excellent local control (LC) for HCC in numerous retrospective and prospective studies[18-20].

Charged particle therapies (CPT), including proton beam therapy (PBT) and carbon ion radiotherapy (CIRT), possess physics-related advantages, which allow for a better dose distribution than in photon beam therapy, especially for low- and medium-dose dosimetry in the normal liver parenchyma during the treatment of HCC[21-24]. The physics-related advantages of CPT resulted from the Bragg peak, a property of CPT, which refers to a sharp dose accumulation followed by rapid dose fall-off[25]. The numerous, stacked Bragg peaks of different energies form the spread-out Bragg peak (SOBP), which possesses dosimetry characteristic of the little exit doses of the clinical tumor target[21,24,26,27]. In the application of CPT in HCC treatment, the dosimetry benefit derived from SOBP of CPT has been confirmed in several studies[22,27,28] (Figure 1). In addition, the property of a higher relative biological effectiveness (RBE) for a charged particle beam, which is approximately 1.1 for a proton and 2-5 for a carbon ion[26,29], indicates higher radiobiological damage, with more DNA double strand breaks and more tumor ablation effects[26,30]. Moreover, the direct DNA damage effect produced via the CPT beam also had another radiobiological advantage in terms of the oxygen enhancement ratio (OER), which is defined as the ratio of radiation dose required to produce the same tumoricidal effect under hypoxic and normoxic conditions. The OER can be reduced to 1 by using the CPT beam, with a linear energy transfer more than 100 keV/μm for oxygen concentrations between 0% and 20%[31]. Consequently, increased application of CPT in HCC patients has been noticed in recent years, especially owing to the improved techniques of CPT and the increased numbers of CPT facilities[32,33].

Figure 1

The illustration of Bragg peak and spread-out Bragg peak.

With the technological advancements in CPT, it is reasonable that the protocol of the doses schedule shifts from conventional fractionation to hypofractionation, like the evolving process of the photon beam treatment SBRT for HCC. Several studies demonstrated that various radiation dose protocols, which ranged from 77 GyE (1 Gray equivalent protons is equivalent to delivering 1 Gy with photons) in 35 fractions to 66 GyE in 10 fractions for PBT[34-37] and 76 GyE in 20 fractions to 52.8 GyE in 4 fractions for CIRT[38,39], provide effective treatment results under different conditions in HCC patients. However, the optimal CPT dose and schedule for effective control of tumors in HCC patients with different comorbidities remain uncertain. The aim of the present systematic review is to evaluate the efficacy and safety of the different CPT dose regimens, leading to a conclusive summary of the adequate dose and fraction for clinical utilization in HCC treatment.

CLINICAL OUTCOMES FOR PARTICLE BEAM THERAPY WITH DOSE REGIMEN OF LESS THAN 5 GY PER FRACTION

The studies on cohorts treated with CPT for HCC are mainly from the United States, Japan, and Korea. We have reviewed 5 prospective and 2 retrospective studies, in which the dose protocols of CPT for treating HCC are less than 5 GyE per fraction. The clinical characteristics and outcomes of CPT for treating HCC using conventional fraction-size doses are summarized in Tables 1 and 2. In addition to the aforementioned characteristics, we summarized the target volume for HCC using CPT, including gross tumor volume (GTV), clinical target volume (CTV) extending from GTV, internal target volume (ITV), and planning target volume (PTV), as well as the toxicities that resulted from CPT (Table 2).

Table 1

Clinical patient characteristics of the selected studies

Ref.	Study, Patient	Source/ Energy (MeV)	Liver function	Vascular invasion	Tumor size
Bush et al[40]	Phase II, 76	Proton	CPC score	4 patients	5.5 cm (mean)
			5-6 22
			7-9 36
			10-15 18
Hong et al[41]	Phase II, 44	Proton (230-250)	CPC	15 patients	5.0 cm (median,1.9-12.0)
			A 32
			B 9
			C 3
Chiba et al[42]	Retro, 162	Proton (250)	CPC	10 patients	3.8 cm (median, 1.5–14.5)
			A 82
			B 62
			C 10
Nakayama et al[37]	Retro, 47	Proton (155 to 250)	CPC	7 patients	N/A
			A 35
			B 9
			C 3
Kawashima et al[43]	Phase II, 30	Proton (235)	CPC	12 patients	45 cm (median,25-82)
			A 20
			B 10
Kim et al[44]	Phase I, 27	Proton (250)	CPC	N/A	2.3-3.2 cm (median, 1.3-7)
			A 24
			B 3
Kato et al[45]	Phase I/II, 24	Carbon‐ion (290-400)	CPC	3 patients	5.0 cm (median,2.1-8.5)
			A 16
			B 8
Mizumoto et al[34]	Retro, 266	Proton	CPC	N/A	< 3 cm 100
			A 203		3.0–4.9 cm 96
			B 60		50–99 cm 62
			C 3		> 100 cm 8
Komatsu et al[39]	Retro, 343	Proton, Carbon‐ion	CPC	92 patients	< 50 277 50-100
			A 262		80
			B 75		> 100 22
			C 6
Kim et al[46]	Retro, 71	Proton (230)	CPC	0	1.5 (median,1.0–8.5)
			A 68
			B 3
Shibuya et al[38]	Retro, 174	Carbon‐ion	CPC	0	3.0 (median,0.8‐10.3)
			A 153
			B 20

CPC: Child-pugh classification; Retro: Retrospective study; N/A: Non-analyses. CPC: Child-pugh classification; Retro: Retrospective study; N/A: Non-analyses.

Table 2

Main clinical results of the selected studies

Ref.	Dose fractionation (GyE/fractions)	Treatment planning	Local control	Survival outcome	Late severe adverse events (number or %)
Bush et al[40]	63 /15	PTV = GTV + 10-20 mm	80%	median PFS: 36 mo	G2 toxicities: 5/76
Hong et al[41]	58.05–67.5 /15	PTV = CTV +5-10 mm	94.8% (2 yr)	Median PFS: 13.9 mo	G3 toxicities: 4
				PFS: 39.9% (2 yr)
				OS: 63.2% (2 yr)
Chiba et al[42]	72 /16, 78 /20, 84 /28, 50 /10	CTV = GTV + 5–10 mm	86.9% (5 yr)	OS: 23.5% (5 yr)	Infection biloma: 1.1% Biliary duct stenosis: 0.5% GI bleeding:1.1%
Nakayama et al[37]	72.6/22, 77/ 35	PTV1 = CTV+ 5-10 mm PTV2 = PTV1 with alimentary tract avoiding	88% (3 yr)	OS: 50% (3 yr)
Kawashima et al[43]	76 /20	CTV= GTV+5 mm, PTV = CTV+3 mm	96% (2 yr)	OS: 62% (3 yr)	Hepatic insufficiencies : 8
Kim et al[44]	60/20 –72/24	PTV = ITV + 5-10 mm	71.4%–83.3% (3y)	OS: 42.3% (5 yr)	G2 toxicity: 0
Kato et al[45]	49.5–79.5/15	PTV= GTV+10 mm	81% (3 yr)	OS: 25% (5 yr)	No severe liver injury No > 2 points increase in CP score at any time
Mizmoto et al[34]	66/10, 72.6/22, 77/35	CTV= GTV+ 5-10 mm	81% (5 yr)	OS: 45 (5 yr)	G 2/3 GI toxicity: 6
Komatsu et al[39]	52.8–84.0 /4-38 (proton) 52.8–76.0 /4-20 (carbon ion)	CTV = GTV + 5 mm PTV = CTV + 5 mm	90.8% (5 yr)	OS: 38.2%	G 3: 12 RIHD: 4
Kim et al[36]	66/10	PTV = ITV + 0.5-0.7 cm	89.9% (3 yr)	PFS: 26.8% (3 yr) OS: 74.4% (3 yr)	no late GI toxicities or liver failure
Shibuya et al[38]	52.8/4, 60.0/4, 48/2,	CTV = GTV + 0.5 cm PTV = CTV+ 5‐15 mm	87.7% (3 yr)	73.3% (3 yr)	G 3-4: 5.7% (10) RIHD: 1.7% (3)

PFS: Progression-free survival; OS: Overall survival; PTV: Planned target volume; GTV: Gross tumor volume; CTV: Clinical target volume; G: Grade; RILD: Radiation-induced hepatic dysfunction; GI: Gastrointestinal.

First, two prospective phase II trials were conducted in the US to evaluate the efficacy and safety of PBT for treating HCC. Bush et al[40], in Loma Linda, published their results using the regimen of 63 GyE in 15 fractions of PBT in the treatment of HCC. They recruited a total of 76 HCC patients, of which 58 patients had underlying liver functions characterized by Child-Pugh A or B and mean tumor sizes of 5.5 cm[40]. The local control (LC) rate was 80%, and the median progression-free survival (PFS) was 36 months; only 5 patients had grade 2 gastrointestinal (GI) adverse effects after PBT treatment[40]. The 3-year overall survival (OS) was 70% for patients (n = 18) who underwent subsequent liver transplantation after PBT, of which 33% (n = 6) and 39% of the patients had complete remission (CR) (n = 7) and microscopic residues only, respectively[40]. Hong et al[41] enrolled 44 patients with HCC from the Massachusetts General Hospital, MD Anderson Cancer Center, and University of Pennsylvania who were administered PBT using the regimen of 67.5 GyE in 15 fractions (58.05 GyE in 15 fractions for location of tumors within 2 cm of the porta hepatis), of which 41 patients had liver function of Child-Pugh A or B and median tumor size of 5.0 cm[41]. The 2-year LC rate and the median PFS for all the patients were 94.8% and 13.9 months, respectively[41]. In their study, 4 patients experienced grade 3 radiation-related toxicities, including thrombocytopenia, liver failure and ascites, gastric ulcer, and elevated bilirubin. A higher occurrence (29.5%) of vascular thrombosis was reported in patients in their study compared to the 5% occurrence that was reported in Bush et al[40]’s cohort.

Chiba et al. reported the clinical experience of PBT for 162 patients with median HCC tumor size of 3.8 cm at the University of Tsukuba using a PBT dose regimen of 50 to 88 GyE in 10-24 fractions with a median fraction dose of 4.5 GyE[42]. Of these, 88.9% of patients had liver function of Child-Pugh A or B, and 6.1% patients had vascular thrombosis[42]. The 5-year LC and OS rates were 86.9% and 23.5%, respectively. The 5-year OS rate for patients with a solitary tumor and Child–Pugh class A was 53.5%[42]. The late toxicities included infected biloma (2 patients), common bile duct stenosis (1 patient), and GI tract bleeding (2 patients)[42]. Nakayama et al[36] updated the clinical outcomes of the University of Tsukuba, and reported a study of 47 patients, whose HCC tumor locations were within 2 cm of the GI tract. They used the PBT regimen of 72.6 GyE in 22 fractions and 77 GyE in 35 fractions in order to avoid GI tract toxicity. The 3-year LC and OS rates were 88% and 50%, respectively, and 4 patients experienced grade 2 or 3 GI bleeding during follow-up[36].

Kawashima et al[43] conducted one phase II study, which enrolled 30 HCC patients with Child-Pugh A or B, to evaluate the safety and efficacy of PBT using a dose regimen of 76 GyE in 20 fractions. The median tumor size of the study was 4.5 cm and 40% patients had macroscopic vascular invasion[43]. The 2-year LC and OS rates were 96% and 66%, respectively; and 8 patients developed hepatic insufficiencies after PBT, of which 4 cases died of hepatic insufficiency-related complications 6 to 9 mo later.

Kim et al[44] reported one phase I dose escalation study of 27 HCC patients, using a PBT dose regimen of 60 GyE in 20 fractions (dose level 1, n = 8), 66 GyE in 22 fractions (dose level 2, n = 7), and 72 GyE in 24 fractions (dose level 3, n = 12). The median tumor size of patients entering into dose level 3 was 2.5 cm. The CR rates of primary tumors after PBT for patients receiving dose levels 1, 2, and 3 were 62.5%, 57.1%, and 100%, respectively (P = 0.039)[44]. The 3-year LC and OS rates for all patients were 79.9% and 56.4%, respectively[44]. Regarding liver toxicity, 4 cases had a 1-point of decrease in the Child-Pugh score, and 1 case had a 1-point increase in the Child-Pugh score, whereas the other 22 cases showed no change in the Child-Pugh score[44].

Regarding CIRT, Kato et al. conducted the first phase I-II trial with 24 HCC patients with Child-Pugh A or B liver function, a median tumor size of 5.0 cm, and vascular invasion of 12.5%[45]. Escalated CIRT doses of 49.5 to 79.5 GyE in 15 fractions were used in their study[45]. The overall tumor response, 3-year LC, and 3-year OS rates were 71%, 81%, and 50%, respectively[45]. Patients treated with doses ≥ 72.0 Gy (RBE) did not develop recurrence [45]. No severe liver injury occurred, except in 1 case of grade 2 late lung reaction, 1 case of grade 2 late GI complication, and 2 cases of grade 2 late skin reactions after the completion of CIRT[45].

Altogether, these findings indicate that conventional fraction-size CPT with varying target volume, including CTV, ITV, and PTV designation, could provide excellent local control for patients with relatively small, isolated tumors and concomitant Child-Pugh class A/B/C liver disease. For central tumors and tumors adjacent to the bowel, conventional fractionation of CPT is a safe approach that not only provides good local tumor control but also lessens the adverse effect (Figure 2).

Figure 2

Illustrations of doses and regimens of charged particle therapy in the treatment of different locations of hepatocellular carcinoma. For central tumors and tumors adjacent to the bowel, conventional fractionation of charged particle therapy (CPT) is a safe approach that not only provides good local tumor control but also lessens adverse effects. In contrast to that for central tumors, short-course hypofractionation of CPT might provide better outcomes for larger-sized tumors that are located at peripheral areas of the liver. PBT: Proton beam therapy; CIRT: Carbon ion radiotherapy.

CLINICAL OUTCOMES FOR PARTICLE BEAM THERAPY WITH DOSE REGIMENS OF MORE THAN 5 GYE PER FRACTION

Regarding dose regimens using fractionation size larger than 5 GyE, we reviewed 4 retrospective studies and summarized the underlying clinicopathological features: patients’ number, liver function, and size of the tumor, as well as the treatment characteristics (PBT or CIRT) (Table 1). Table 2 summaries the dose, fraction size, treatment plan (including GTV, CTV, and PTV), late toxicities, LC, PFS, and OS.

Mizumoto et al[34] reported the LC and OS of a cohort of 266 HCC patients at the Proton Medical Research Center in Tsukuba who were treated with PBT using three different treatment protocols according to the tumor location[35]. The dosage regimen protocols of PBT included 66 GyE in 10 fractions for peripheral tumors (tumor located 2 cm away from hilum), 72.6 GyE in 22 fractions for central tumors (tumor located within 2 cm of the hilum), and 77 GyE in 35 fractions for central tumors which were adjacent to the GI tract[34,35]. The median tumor size was 3.4 cm, and 99% of patients were characterized by a cirrhosis status of Child-Pugh A or B. The 3-year LC, PFS, and OS rates for all patients were 87%, 21% and 61%, respectively[34]. Among these three different dosage regimens, there were no significant differences in the LC and PFS of patients. In all, 12 patients experienced symptomatic late toxicity, which included rib fracture (3 patients), dermatitis (grade 1: 2; grade 3:1 patients), and perforation, bleeding, or inflammation of the GI tract (grade 2: 3 patients grade 3: 3 patients)[34]. For patients whose tumors were located adjacent to the porta hepatis, the PBT (72.6 GyE in 22 fractions or 77 GyE in 35 fractions) resulted in a 3-year LC and OS rates of 86% and 50%, respectively, and no subsequent bile duct stenosis was observed in them[34].

Komatsu et al[39] reported the clinical outcome of a large cohort of HCC patients who were treated at the Hyogo Ion Beam Medical Center (HIBMC), including 242 and 101 patients (108 tumors) who underwent PBT (278 tumors) and CIRT (108 tumors), respectively, for HCC. The dosage regimens of CPT included 8 and 4 different protocols of PBT (52.8-84.0 GyE in 4-38 fractions) and CIRT (52.8-76.0 GyE in 4-20 fractions), respectively[39]. The percentage of tumor sizes < 5 cm, within 5-10 cm, and > 10 cm were 37.8%, 37.4% and 41.1%, respectively[39]. The 5-year LC rate for all patients receiving PBT and CIRT were 90.2% and 93%, respectively[39]. The 5-year LC rate for patients with tumor < 5 cm, within 5 to 10 cm, and > 10 cm were 95.3%, 84.4%, and 42.2%, respectively[39]. The PBT and CIRT resulted in equivalent 5-year LC rates of 95.5% and 94.5%, respectively, in the treatment of patients with tumors < 5 cm. For patients whose tumors were within 5 to 10 cm, the PBT and CIRT resulted in equivalent 5-year LC rates of 84.1% and 90.9%, respectively[39]. In those whose tumors were > 10 cm, CIRT resulted in a better 5-year LC rate of 80% compared to 43.4% for PBT[39]. Four patients developed RILD after CPT, and no patients died of CPT treatment-related toxicities.

Kim et al[46] designed a study to assess the optimal time of tumor response after PBT for 71 patients with HCC, which comprised 68 patients with Child-Pugh A, and 3 patients with Child-Pugh B; their median tumor size was 1.5 cm. Use of a PBT regimen of 66 GyE in 10 fractions resulted in the CR rate of 93%, and most patients (93.9%) achieved this within one year after PBT. Overall, the PBT resulted in 3-year local progression-free survival (LPFS), relapse-free survival (RFS), and OS rates of 89.9%, 26.8%, and 74.4%, respectively. Within 3 months after treatment, 3 patients had a 1-point increase in Child–Pugh score, 3 patients experienced grade 1 elevated liver function, and no patients experienced > grade 3 toxicities.

In a multicenter retrospective study conducted by the Japan Carbon Ion Radiation Oncology Study Group (J‐CROS), Shibuya et al[38] reported the effectiveness and safety of short-course CIRT for 174 patients with HCC (median tumor size; 3.0 cm). Of these, 153 patients had Child-Pugh A, and 20 patients with Child-Pugh B. The prescription radiation doses of CIRT included 48.0 GyE in 2 fractions (n = 46), 52.8 GyE (n = 108) in 4 fractions, and 60.0 GyE (n = 20) in 4 fractions[38]. After a median follow-up period of 20.3 (range, 2.9-103.5) months, the 3-year LC and OS rates for all patients were 81.0% and 73.3%, respectively[38]. Multivariate analysis also disclosed that Eastern Cooperative Oncology Group performance status 1-2, Child-Pugh class B, maximum tumor diameter ≥ 3 cm, multiple tumors, and serum alpha fetoprotein level > 50 ng/mL were significant prognostic factors for a worse OS. Regarding CIPT-related toxicities, 10 patients (5.7%) experienced grade 3 or 4 treatment-related toxicities, and 3 patients (1.7%) experienced RIHD.

Altogether, a larger fraction size of CPT radiation dose possesses similar treatment outcomes to those of CPT with a conventional fraction size, without compromising normal organ toxicities. For larger size tumors, short-course CPT might provide better outcomes (Figure 2). For central tumors, hypofractionation CPT is not preferred according to the protocol proposed by Mizumoto et al[34] (Figure 2). Further dosimetric constraints for avoiding late toxicities of biliary stenosis are warranted to expand the utilization of hypofractionation CPT.

CONCLUSION

CPT, including PBT and CIRT, could be used to deliver ablative doses to HCC tumors with normal liver sparing. Overall, conventional CPT or hypofractionated CPT (including short-course CPT) could not only provide LC rates > 90% but also result in < 5% grade 3 toxicities. For large-sized HCC tumors, the hypo-fractionation regimen of CPT may provide the benefit of increasing LC through overcoming radio-resistance, whereas conventional CPT is preferred for treating central tumors of HCC by avoiding late toxicities of the biliary tract. Prospective data are still warranted to accumulate evidence on the dosimetric constraints of hypofractionated or short-course CPT in the treatment of HCC.