Recent advancements in radiotherapy for hepatocellular carcinoma

Abstract

Radiotherapy has traditionally been uncommonly used for hepatocellular carcinoma (HCC) due to concerns of radiation-induced liver damage. However, advancements including stereotactic body radiotherapy have improved target localization and reduced toxicity, leading to its inclusion as a locoregional therapy in recent international guidelines. Newer technologies, such as MR-Linac and particle radiotherapy, have further improved the therapeutic ratio of radiotherapy, showing early promising results. The increasing application of radiotherapy in HCC has led to broadening of its indication to bridging to transplant and palliation. Beyond being a standalone treatment, radiotherapy is increasingly being combined with systemic therapies to enhance treatment response and survival rates. This combination is particularly relevant for patients with portal vein tumour thrombosis, as other locoregional therapies are often unsuitable for them. Furthermore, with the evolving concepts of oligometastatic and oligoprogressive HCC, radiotherapy is being explored as a promising approach to manage these conditions, especially as HCC patients experience longer survival with the introduction of immunotherapy. In this review, we provide an overview of these recent advancements in radiotherapy for HCC.

Highlights

• •Stereotactic body radiotherapy is a safe and effective locoregional therapy for hepatocellular carcinoma.
• •MR-Linac and particle radiotherapy improve the therapeutic ratio, enhancing tumour control without increasing toxicity
• •Radiotherapy is also applicable in the setting of bridging to transplant and palliation.
• •Combining radiotherapy and immunotherapy has the potential of downstaging, treat portal vein tumour thrombosis and oligometastatic disease.

Introduction

Hepatocellular carcinoma (HCC) is the most common primary liver cancer. According to the GLOBACAN cancer statistics of 2022, HCC is the sixth most common cancer worldwide and the third leading cause of cancer mortality.¹ The leading cause of HCC is cirrhosis, which is predisposed by viral hepatitis, metabolic syndromes, or environmental factors.² Therefore, treatment of HCC needs to consider both the underlying liver function and tumour burden. The Barcelona Clinic Liver Cancer (BCLC) staging system is the most common treatment algorithm adopted globally that integrates tumour burden, performance status, and underlying liver function, providing guidance on the management of HCC at different stages.³ The complexity and challenges in HCC treatment underscores the need for a multidisciplinary approach to tailor management decisions for individual patients.^3,4

Historically, external beam radiotherapy was uncommonly used to treat HCC and was mainly reserved to use in the palliative setting. It was not as widely adopted as other locoregional therapies such as radiofrequency ablation (RFA) or transarterial chemoembolization (TACE) due to the fear of radiation-induced liver disease (RILD).^5,6 The risk of RILD is associated with the volume of liver irradiated and underlying liver function. In the past, due to poor resolution of the liver, treatment verification of tumour localization often utilized the diaphragm as a surrogate marker in the digitally reconstructed radiograph, which had been shown to correlate well with liver movement.⁷ However, craniocaudal movements of the diaphragm during the respiratory cycle can be up to 4 cm, therefore large margins are needed to ensure tumours are covered by the radiation field.⁸

However, concerns of RILD have largely been overcome by the introduction of stereotactic body radiotherapy (SBRT), which enables precise delivery of high focused radiation to a smaller volume. This is due to advancement in immobilization techniques, breathing control, and verification technique. In modern series, the risk of RILD is expected to be <5%.⁶ With more evidence supporting the role of SBRT in the management of HCC, it has now been incorporated in international guidelines as one of the treatment strategies for patients with localized HCC.^9,10

Recently, the introduction of magnetic resonance linear accelerator (MR-Linac) and particle radiotherapy has gained traction. These technologies further enhanced the therapeutic window of radiation, mainly by decreasing the radiation toxicity to neighbouring normal organs. In addition, the higher efficacy and longer survival brought by immunotherapy in the treatment of HCC has ignited exploration of combination treatment with systemic therapy and radiotherapy, in search for synergism. Moreover, as patients with HCC are surviving longer, it is increasingly observed that some patients experienced oligoprogression and may benefit from the addition of radiotherapy, similar to other cancers.^11,12 In this review, we will provide an overview of the recent advancement in radiotherapy surrounding these areas and highlight the crucial role of radiotherapy in the multidisciplinary management of HCC.

Development of stereotactic body radiotherapy in hepatocellular carcinoma

SBRT is a technique adapted from stereotactic radiosurgery. It delivers highly conformal, ultra-high doses per fraction (i.e. ultra-hypofractionated radiotherapy) to extracranial targets. Apart from the use of much larger doses per fraction, SBRT distinguishes from conventional three-dimensional radiotherapy in two main ways: it is generally applied to well-circumscribed tumours, and it allows for smaller margins to account for localization uncertainties. Because of the larger dose per fraction, SBRT requires stringent and reproducible patient immobilization and respiratory motion management for every treatment session.

When radiotherapy was first developed, early attempts were made to deliver high doses per fraction in order to shorten overall treatment time and reduce machine usage and bed occupancy per patient. Unfortunately, these attempts resulted in severe late toxicity. In his retirement editorial, Prof. Gilbert Fletcher, former Chairman of the Department of Radiology at MD Anderson Cancer Center, strongly cautioned against the use of large-dose-per-fraction treatments.¹³ Together with the emergence of more effective chemotherapy since the 1950s, the development of hypofractionated radiotherapy was essentially paused for almost 50 years. However, in the early 2000s, advances in motion management revived interest in ultra-hypofractionated radiotherapy. This concept was successfully reintroduced for early-stage lung cancer yielding remarkable results. In a phase I study that explored the use of 24-60 Gy, divided in three fractions, in 33 patients with medically inoperable stage I non-small-cell lung cancer, there was no maximum tolerated dose reached, with 87% showing response (complete response in 27%).¹⁴

Since then, there has been a plethora of trials exploring the use of SBRT for tumours at various sites, including the liver. Although initial attempts resulted in high rates of RILD, investigators from the Princess Margeret Hospital in Toronto eventually identified the liver’s volume tolerance limits and established safe dose constraints, significantly reducing the incidence of RILD.¹⁵ Because of the fragile nature of cirrhotic livers, the most critical constraints for liver SBRT relate to the total dose received by the normal liver, and the volume of normal liver that can be preserved. However, because the liver moves substantially with respiration, it was impossible to deliver precise radiation until the introduction of SBRT. Its advanced motion management technologies have made precise delivery of radiation to the liver feasible, by reducing the irradiated volume, and thereby enabling safer delivery of high doses to liver tumours.

There are several techniques for motion management that can be used to minimize tumour motion during treatment, including abdominal compression, breath-hold, respiratory gating, and the use of four-dimensional computed tomography planning (4D-CT) (Figure 1). Abdominal compression involves attaching a device to the treatment couch at a relocatable position, and applying light pressure to the patient’s abdomen just below the xiphoid process. This minimizes respiratory movement and thus allows for a smaller treatment margin given for respiratory tumour motion. The technique is particularly useful for upper abdominal targets such as liver tumours, and has the advantage of being relatively simple to use, while allowing the beam to operate continuously during the treatment.

Figure 1.

Techniques to control breathing motion in radiotherapy. (A) Abdominal compressor is applied lightly on the upper abdomen to reduce respiratory motion. (B) An infrared camera is set up to track patients’ breathing motion (bottom left). The breathing pattern can be tracked and visualized on the screen. (C) Radiation is being delivered at near maximal exhalation (i.e. respiratory gating) of each breathing cycle (yellow bars).

For patients who are physically fit with good pulmonary function, breath-hold techniques can be employed. These methods restrict radiation delivery to specific phases of the respiratory cycles, minimizing the margins needed to cover the entire respiratory cycle. This results in more precise treatment and less irradiation of normal tissue. One common approach includes instructing the patient to hold their breath at or near maximum inspiration, with radiation interrupted if the breath-hold threshold cannot be maintained by the patient. Respiratory gating is another widely used technique. The patient is coached to breathe in a reproducible pattern, while their breathing cycle is being monitored electronically by spirometry or externally by infrared camera tracking. Radiation is delivered only during a specific ‘gate’ or window in the breathing phase, typically end-exhalation when motion is minimal. The radiation beam is automatically turned off outside this window (Figure 1).

Finally, the adoption of 4D-CT helps achieve precise tumour delineation. The ‘fourth dimension’ refers to time. When patients undergo 4D-CT scanning, a series of CT images is acquired corresponding to different phases of the respiratory cycle, allowing visualization of the full extent of tumour motion. In the past, without 4D-CT, a hypothetical large margin would be applied around the tumour to avoid missing the target due to breathing motion. With 4D-CT, tumour movement during respiratory cycle can be accurately visualized, enabling a substantial reduction in the treated volume while still ensuring adequate coverage.

Improved localization and dose delivery with MR-Linac and particle radiotherapy

A successful radiotherapy delivers a maximum dose to the tumour while minimizing the dose to other organs. Much of the work in the advancement in radiotherapy is to optimize and balance these two factors. While SBRT provides an opportunity to improve precision to deliver a higher dose to the tumour and minimize dosing to surrounding organs, it is limited by uncertainties of tumour location due to intra-abdominal organ movements and respiration during and between each fraction. Utilization of daily cone-beam CT (CBCT) has improved target localization and is particularly helpful if the target lesion does not move significantly with respiration or if it can be clearly visible in CT. However, as the daily CBCT is a plain CT, HCCs are typically not visible on CBCT (Figure 2). Therefore, the precise delivery of radiotherapy to HCC using modern linacs relies heavily on surrogate markers to gauge the lesion location, such as with the insertion of fiducial markers, identification of any prior lipoidal stains from previous TACE, or surrounding anatomy (e.g. diaphragm for liver dome lesions) (Figure 2). The uncertainty of the tumour during radiotherapy necessitates a wider margin to ensure dose to the tumour is adequate. However, the liver is surrounded by many radiosensitive organs including the stomach, duodenum, and small bowel, which only have moderate tolerance to radiation. Encompassing these organs in the ablative treatment field may result in serious acute and late toxicities including gastrointestinal perforation or fistulation.¹⁶ This is one of the reasons why it can be difficult to deliver ablative doses to some HCC lesions close to the luminal gastrointestinal organs.

Figure 2.

Image quality of diagnostic computed tomography (CT) and daily cone-beam CT. (A) Sagittal images of liver tumour from diagnostic CT and (B) daily cone-beam CT (right). The imaging quality is better in diagnostic CT than in cone-beam CT due to the use of contrast, faster scanning times, and a narrow fan-shaped beam, resulting in higher resolution and less artifacts. For patients treated with transarterial chemoembolization, the resulting lipiodol stain can serve as a helpful marker to guide daily image verification. However, it often produces artifacts and degrades image quality. The orange line indicates the internal gross tumour volume (iGTV), which is the integral gross tumour volumes across the entire respiratory cycle. The red line indicates the planning target volume, which is created by applying a 5 mm margin around the iGTV to account for setup uncertainties.

MR-Linac is a novel radiotherapy delivery system that utilizes the excellent soft tissue resolution of MRI and combines with real-time online adaptive planning workflow that enables the precise delivery of an ablative dose to the tumour without damage to neighbouring tissues^17,18 (Figure 3). The ability of MR-Linac to adapt to the shape of the day allows for the reduction of the extra margins that are added to the treating volume to account for organ motion, respiratory movements, and setup uncertainties in CT-Linac treatments, offering advantages of improving local tumour control and reducing adverse events (Table 1).19, 20, 21, 22, 23, 24 For instance, in a retrospective report that compared the margins added to the tumour volumes between CT-Linac and MR-Linac radiotherapy plans, it has been shown that smaller margins can be added in tumours planned using MR-Linac, despite larger tumour size.²⁵ In one of the largest retrospective studies evaluating the role of MR-guided SBRT in 99 patients with localized liver tumours (52 HCCs and 47 cholangiocarcinoma/combined hepatocellular-cholangiocarcinoma) of median size of 3 cm, the adoption of daily online adaptive MR-guided SBRT enabled lesions with a planning target volume (PTV) within 1 cm of luminal organs to be treated to 50 Gy over five fractions, which is typically not possible using conventional CT-Linac.²⁶ It is worthwhile to note that in the study described, 2-year local control was >90% even in these anatomically challenging areas.²⁶ Although the optimal dose-response relationship has not been clearly established in HCC, a radiation dose in the range of 40-50 Gy in five fractions has been suggested to improve local control.^27,28 Nonetheless, the selection of radiotherapy dose needs to be individualized rather than being a standard, taking account of the heterogeneity of liver function, to balance the benefit of local control and the risks of severe gastrointestinal toxicities and liver decompensation, especially in the context of a cirrhotic liver.

Figure 3.

Image quality of planning computed tomography (CT) and daily on-board MR images. (A-C) Axial, coronal, and sagittal planning CT and (D-F) daily magnetic resonance (MR) images using T1 3DVane SPAIR sequence demonstrating liver tumour. Note the liver tumour is barely visible on CT images but clearly visible on MR images.

Table 1.

Selected studies on magnetic resonance-guided radiotherapy for the treatment of HCC

Author (year)	Study nature	N	Median size	Median dose	Local control	PFS	OS	TRAE
Su et al. (2025)19	Retrospective	103 HCC	41 HCC: <3 cm 34 HCC: 3-5 cm 28 HCC: ≥5 cm	50 Gy, number of fractions not provided	ORR: 54.4%	21.7 months	40.7 months	No grade ≥3 toxicity
Bordeau et al. (2023)20	Prospective	56 liver tumours, 48 HCC	2.3 cm	50 Gy in 5 Fr	1-year: 98.1%	1 year: 32.4%	1-year: 85.1% 2-year: 74.2%	No RILD No grade ≥3 toxicity
Boldrini et al. (2021)21	Retrospective	10 HCC	1.9 cm	50 Gy in 5 Fr	90%	Not reported	Not reported	One patient with CP-A to CP-B
Yoon et al. (2021)22	Retrospective	106 patients with intra-abdominal tumours 11 HCC, 16 CC	3.0 cm	40 Gy in 5 Fr	2-year: 74%	1-year: 36% 2-year: 8%	2-year: 57%	Acute grade ≥3 toxicity: 0.9% and 0% Acute grade ≥3 toxicity: 5.2% and 2.1%
de Mol van Otterloo et al. (2021)23	Retrospective	702 patients 33 (5%) liver tumour	Not reported	Not reported	Not reported	Not reported	Not reported	Grade ≥3 toxicity for liver tumours: one umbilical hernia and one peripheral ischemia
Feldman et al. (2019)24	Retrospective	26 HCC	Not reported	45-50 Gy in 5 Fr	1-year: 96.5%	Not reported	1-year: 85%	No grade ≥3 toxicity

CC, cholangiocarcinoma; CP, Child-Pugh grade; FR, fraction; Gy, Gray; HCC, hepatocellular carcinoma; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; RILD, radiation-induced liver disease; TRAE, treatment-related adverse event.

Another technological advancement in radiotherapy is the introduction of proton and particle radiotherapy. In contrast to photon therapy, charged particle radiotherapy has the unique characteristic of the Bragg’s peak, whereby the maximum energy deposition occurs at the end of their path. This allows for a highly focused radiation beam to be delivered to the tumour, at the same time sparing nontumorous tissue both before and after the tumour, which has been proposed to reduce toxicity and improve tumour control²⁹ (Figure 4). In a recent systematic review and meta-analysis including a total of 1858 HCC patients receiving proton-based radiotherapy (PBT) from 22 studies between 2004 and 2023, the pooled rates of 3- and 5-year local progression-free survival (PFS) after PBT were 88% and 86% respectively, demonstrating excellent tumour control with PBT. The pooled 3- and 5-year overall survival (OS) rates were 60% and 46%. Grade ≥3 hepatic toxicity, classic RILD and nonclassic RILD occurred in only 1%, 2%, and 1% of the study subjects, respectively, highlighting its safety.³⁰ When compared with other locoregional treatment modalities, PBT also demonstrated noninferiority. For example, PBT was compared with RFA in a randomized study, enrolling 144 patients with recurrent HCCs of size < 3 cm, limited to a maximum of two lesions. The 2-year local PFS rate for PBT was 92.8%, compared with RFA, which was 83.9% in the per protocol population. Of note, none of the patients treated with PBT had any grade ≥3 treatment-related adverse events (TRAE) compared with 16.1% in patients treated with RFA.³¹ In another study comparing PBT (n = 36) with TACE (n = 39) in HCC lesions that are not suitable for resection or ablation, PBT demonstrated improved PFS (not reached versus 12 months, P = 0.002) and local control [hazard ratio (HR) 5.64, P = 0.003], but similar 2-year OS at ∼68%.³² Nonetheless, high-grade TRAEs were notably considerably fewer in the PBT arm, compared with TACE. Patients treated with TACE were hospitalized more than patients with PBT, suggesting PBT is a less toxic treatment than TACE.

Figure 4.

Dosimetric comparison. Between (A) PBT and (B) IMRT. Note that IMRT results in larger area irradiated with low dose. IMRT, intensity-modulated radiotherapy (photon); PBT, proton beam radiotherapy.

Although PBT offers dosimetric advantages over photon-based radiotherapy, differences in on-board image verification processes may influence the applicability of PBT. Historically, PBT has relied primarily on two-dimensional orthogonal X-ray imaging for treatment verification, which depends on bony landmarks or implanted fiducial markers.³³ This approach inherently introduces greater uncertainty in daily tumour localization compared with volumetric CT-based methods, such as CBCT, which have become standard in modern photon-based treatment platforms. Consequently, larger margins have traditionally been required in PBT to compensate for these positional uncertainties. This requirement can be particularly challenging when radiosensitive gastrointestinal luminal organs are located immediately distal to the tumour along the beam pathway, as PBT often necessitates additional distal margins to account for beam range uncertainties.^33,34 However, contemporary proton machines have increasingly incorporated advanced CT-based volumetric image guidance (including on-board CBCT capabilities), thereby reducing uncertainties in tumour localization to levels comparable to those achieved in photon-based therapy.

A separate consideration regarding the widespread adoption of PBT in HCC is the absence of high-quality evidence to support it demonstrating superior survival compared with photon-based approaches. Although multiple retrospective studies have reported improved overall survival and reduced incidence of liver toxicity with PBT over photon-based therapy, high-level evidence from prospective randomized studies remains limited.^35,36 The ongoing phase III randomized trial (NRG-GI003, NCT03186898) is currently recruiting to explore the efficacy and toxicity profiles of proton versus photon radiotherapy in patients with liver cancer. On a different note, a recently published multicentre phase III randomized trial has provided support to the role of proton therapy in locally advanced oropharyngeal cancer. This study demonstrated that PBT is noninferior to photon therapy, with respect to PFS [5-year rates: 81.3% for PBT versus 76.2% for photon therapy; HR 0.88, 95% confidence interval (CI) 0.57-1.35, P = 0.005]. Moreover, PBT is associated with a significant improvement in OS (5-year rates: 90.9% versus 81.0%; HR 0.58, 95% CI 0.34-0.99, P = 0.045).³⁷ High-grade toxicities, including severe lymphopenia, dysphagia, xerostomia, and gastrostomy tube dependence, were significantly reduced with PBT.³⁷ Together, these findings highlight the evolving technological landscape and the potential of adoption of PBT in the management of HCC.

Carbon ion radiotherapy (CIRT) is an emerging charged particle radiotherapy that has gained substantial interest in the treatment of intra-abdominal tumours such as HCC. In addition to its characteristic Bragg’s peak that spares normal tissue from radiation toxicity, its stronger biological effectiveness and higher linear-energy transfer (LET) properties, compared with PBT, enable better tumour control with CIRT in radioresistant and hypoxic conditions.³⁸ To date, a few prospective studies have evaluated the treatment outcomes of HCC using CIRT in Japan,^39,40 China,⁴¹ and Europe.⁴² In general, CIBT offered excellent local control, with nearly 100% local control for lesions <5 cm, but slightly high rates of local relapse for larger tumours. Nonetheless, high-grade TRAEs including RILD were rarely observed. Furthermore, retrospective propensity score analyses comparing CIBT with RFA and TACE consistently demonstrated superior local control with CIBT.^43,44

Broadening indications of radiotherapy: Bridging to transplant and palliation

The renewed interest in applying radiotherapy for HCC has also led to exploration of its role as a bridging therapy before liver transplant.²⁷ The rationale for local therapy in this setting is to stabilize the disease, prevent progression, and thereby reduce the risk of patients dropping off the list while waiting for organ availability. The decision to select radiotherapy is generally made when other locoregional therapies are unsuitable (e.g. tumour located in an unfavourable location such as at the liver dome or close to large vessels, making ablation inappropriate). One of the earliest reports on the application of radiotherapy in this setting is from the University of Toronto. In this case series of 10 patients with tumour diameters between 2.5 cm and 10.8 cm, a median dose of 33 Gy (range 8.5-54 Gy in 1-6 fractions) was delivered. Five out of 10 patients successfully underwent transplantation without complications.⁴⁵ Multiple retrospective studies have reported pathological response rates of 40%-60%, with acceptable toxicities including 60%-70% maintained stable liver function after SBRT.46, 47, 48 When compared with other locoregional therapies such as RFA and TACE, there is no difference in dropout rates and actuarial survival from the time of listing in patients received SBRT as bridging.⁴⁹

At the other end of the spectrum, patients with terminal HCC often experience abdominal distension and pain, but there have been limited treatment options to alleviate these symptoms due to poor liver function. Radiotherapy has a long history of use to palliate symptoms caused by liver tumours.^50,51 More recently, several studies have explored the use of single-fraction radiotherapy at 8 Gy in this setting with favourable outcomes.52, 53, 54 The use of single fraction is particularly attractive as it is more comfortable and convenient for patients already burdened by symptoms of terminal HCC. A recent multicentre, randomized, controlled phase III study led by Dawson et al. reported significant improvement in hepatic pain with single-fraction radiotherapy at 1 month after radiotherapy, compared with best supportive care alone in patients with end-stage liver cancer, providing strong evidence to support the role of radiotherapy in symptom palliation in this population.⁵⁴

Combination of radiation with systemic therapy to improve response and survival

In the past two decades, little progress has been made in the development of effective systemic therapy for HCC compared with other malignancies. Many trials have failed to demonstrate efficacy of novel drugs or combinations that were shown to be effective in other cancers.55, 56, 57, 58, 59 In this background, combination of systemic therapy and locoregional therapy has become an active area of clinical research to boost response and survival. The addition of radiation to systemic therapy has a strong scientific rationale. For instance, in an HCC cell line model it was shown that radiation followed by sorafenib enhanced radiosensitivity, downregulated vascular endothelial growth factor receptor 2 (VEGFR2) and suppressed the ERK pathway which could promote cell proliferation.⁶⁰ In another HCC cell line model that explored the synergism of radiotherapy and lenvatinib, it was found that lenvatinib enhanced radiotherapy-induced toxicity, apoptotic activity, and DNA damage via suppression of Src/STAT3/NF-κB signaling.⁶¹ The effect of immunotherapy can also be augmented with radiotherapy following radiation-induced cell death and exposure of tumour antigens, enabling effective immune cell priming.⁶²

One of the largest trials to date studying the combination of radiotherapy with systemic therapy is the RTOG-1112 trial.⁶³ The study is a multicentre phase III randomized clinical trial comparing SBRT followed by sorafenib, to sorafenib alone, in patients with advanced, unresectable HCC. Eligible patients were deemed unsuitable or refractory to standard locoregional therapies and were candidates for first-line systemic therapy. The study was prematurely closed due to slow accrual and a shift in the standard of care from sorafenib to atezolizumab plus bevacizumab. Nonetheless, a total of 193 patients were recruited and randomly assigned. Patient population was balanced in performance status, aetiology, BCLC stage, baseline liver function, and macrovascular invasion. In the final report, patients treated with SBRT plus sorafenib had an improved OS at 15.8 months compared with 12.3 months, in spite of a marginal statistical insignificance (HR 0.77, 90% CI 0.59-1.01, one-sided P value = 0.06). PFS was also significantly improved with the addition of SBRT, resulting in a median PFS of 9.2 months, compared with 5.5 months with sorafenib alone. Of note, high-grade TRAEs were comparable between the two arms, including grade 3 gastrointestinal bleeding (5% in each arm).⁶³

In addition to targeted therapy, combining locoregional therapy such as radiotherapy with immunotherapy, which is the current standard of care for unresectable HCC, has also been actively explored.⁹ Notably, three recent studies (EMERALD-1, LEAP-012, TALENTACE) have already demonstrated locoregional therapy with TACE, and combined with immunotherapy improved survival and response in patients with unresectable HCC compared with TACE alone.64, 65, 66 While these trials laid down the foundation of the beneficial effect of combining immunotherapy with TACE, the presumed synergistic effect was not apparent from the trial results. Rather, the combination of TACE and immunotherapy appear to exert subadditive effects. For instance, the response rates in the combination arm in these trials hovered at ∼45% by the RECIST 1.1 criteria, while TACE alone offered response rates of 30%. The 15% increase in response with the addition of immunotherapy was much lower than the typical response rates for immunotherapy in the unresectable HCC setting.67, 68, 69

Indeed, TACE exerts its effect mainly by ischemic necrosis, and thus the immunogenic effect of TACE was very short-lived. Furthermore, preclinical models showed that immune cells were unable to get inside the tumour following TACE. Instead, they were present mainly at the tumour margin, suggesting a synergistic effect with immunotherapy may not be plausible.⁷⁰

On the contrary, radiotherapy kills cancer cells mainly by mitotic death and immunogenic cell death. This means that cancer cells do not die immediately but slowly after radiotherapy, resulting in a constant release of neoantigens. While the blood vessels might also be affected by radiotherapy, trafficking of immune cells into the tumour microenvironment is still observed in preclinical models, as well as the persistence of arterial enhancement of HCC following radiotherapy for up to 1 year.⁷¹ Therefore, as a combination partner, radiotherapy could potentially be more appealing than conventional TACE.

The majority of the studies that explored a combination of radiotherapy and immunotherapy came from the Asia-Pacific region, due to the higher utilization of radiotherapy to treat HCC in the region. One of the earliest prospective studies that explored the combination of radiotherapy and immunotherapy is the START-FIT study (Table 2).⁷² START-FIT is a single-arm, prospective, phase II study to report the efficacy and safety of the sequential combination of TACE, SBRT, and immunotherapy with avelumab, in patients with locally advanced unresectable HCC. The primary study endpoint was the proportion of patients deemed amenable to curative treatment, which was defined as those who had a sustained complete or partial treatment response for ≥2 months and whether curative treatment could be carried out. In START-FIT, 33 patients were enrolled. The median sum of the largest diameters of lesions was 15.1 cm, and 21 patients (64%) had macrovascular invasion. At a median follow-up of 17.2 months, 18 patients (55%) were deemed amenable to curative treatment. Radiological objective response by modified RECIST criteria was observed in 22 patients (67%) and by RECIST 1.1 was observed in 8 patients (24%). Median PFS was 20.7 months, and median OS was 30.3 months. In terms of TRAEs, 11 patients (33%) developed at least one grade ≥3 TRAE, most commonly hepatic impairments (n = 7).

Table 2.

Selected study on combining radiotherapy with systemic therapy

Author (year)	Study nature	N	Sequence	Systemic therapy	Median size	BCLC C (%)	Vascular invasion (%)	ORR (RECIST 1.1)	PFS (months)	OS (months)	TRAE
Dawson et al. (2025)63	Phase III	85	RT → TKI	Sorafenib	7.5 cm	80	75	38	9.2	15.8	Grade ≥3: 47%; liver failure: 1%; duodenal perforation: 1%; gastrointestinal haemorrhage: 4%
Chiang et al. (2026)74	Phase II	16	TACE → RT → IO	Durvalumab plus tremelimumab	11 cm	73	73	Not reported	18-month: 61.8%	18-month: 90.0%	Grade ≥3: 36%
Kim et al. (2024)85	Phase II	50	RT → IO	Nivolumab	7 cm	100	100	36	5.6	15.2	Grade ≥3: increased AST (6%); increased ALT (6%)
Chiang et al. (2023)72	Phase II	33	TACE → RT → IO	Avelumab	15.1 cm	64	64	24	20.7	30.3	Grade ≥3: increased ALT or AST (15%); increased bilirubin (6%)
Zhu et al. (2024)92	Phase II	46	IO → RT → IO	Sintilimab plus bevacizumab	>5 cm, 73.9%	100	100	45.7	13.8	24	Grade ≥3: Hypertension (8.7%); thrombocytopenia (8.7%)
Hu et al. (2023)91	Phase II	60	RT → IO plus TKI	Camrelizumab plus apatinib	≥10 cm, 57.5%	100	100	47.5	4.6	12.7	Grade ≥3: Hypertension (12.5%); hand-foot syndrome (10%); increased ALT (10%); neutropenia (10%)

ALT, alanine aminotransferase; AST, aspartate aminotransferase; IO, immunotherapy; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; RT, radiotherapy; TACE, transarterial chemoembolization; TKI, tyrosine kinase inhibitors; TRAE, treatment-related adverse events.

A similar study was conducted by the same group using the STRIDE regimen. The STRIDE regimen consists of one intravenous infusion of high dose anti-CTLA-4 tremelimumab and continuous infusion of anti-PD-L1 durvalumab in every 4-week interval. This regimen has demonstrated improved survival over sorafenib in the HIMALAYA trial, with a 5-year overall survival of ∼20%.^68,73 In the START-FIT study using STRIDE, 33 patients with locally advanced HCC were recruited. The objective response rate (ORR) by modified RECIST was at a remarkable 72.7% (42.4% complete response).⁷⁴ The 18-month PFS was 61.8% and 18-month OS was 90.0%. In terms of toxicity, 13 patients (36.3%) experienced grade ≥3 TRAEs, most commonly transient elevation of transaminase after TACE.⁷⁴

When combining systemic therapy with radiotherapy, two key issues are typically discussed: the optimal sequencing of the modalities and the safety of their combination. Sequencing is particularly relevant due to the hypothesized synergistic interaction between radiotherapy and immunotherapy, especially immune checkpoint inhibitors (ICIs).⁶² Yet, no head-to-head comparison exists to date that directly compares the sequential versus concurrent approaches, let alone in the specific context of HCC. However, a recent systematic review and network meta-analysis suggested sequential administration of radiotherapy followed by ICIs [predominantly anti-PD-(L)1] agents was associated with improved survival. The study examined 24 randomized controlled trials involving 9480 participants. Sequential administration of radiotherapy and ICI was shown to improve PFS (HR 0.81, 95% CI 0.67-0.97) compared with concurrent administration. To illustrate this concept, one example is the PACIFIC trial and PACIFIC-II trial, which explored the strategy of sequential administration of chemoradiotherapy and immunotherapy durvalumab (PACIFIC trial), and concurrent administration of chemoradiotherapy and durvalumab (PACIFIC-II), compared with the control arm with chemoradiotherapy alone, in unresectable, locally advanced non-small-cell lung cancer.^75,76 In the PACIFIC study, sequential administration of CCRT followed by durvalumab improved both PFS (16.9 months versus 5.6 months, HR 0.55, 95% CI 0.45-0.68) and OS (47.5 months versus 29.1 months, HR 0.72, 95% CI 0.59-0.89).⁷⁷ On the contrary, in the PACIFIC-II study, which explored the concurrent strategy, there was no improvement in PFS (HR 0.85, 95% CI 0.65-1.12) and OS (HR 1.03, 95% CI 0.78-1.39).⁷⁶ One possible explanation for these observed phenomena is that lymphocytes are exquisitely sensitive to radiation, and radiation-induced lymphopenia is known to adversely affect treatment outcome.⁷⁸ A concurrent approach may therefore deplete the lymphocytes and potentially negate the synergistic effects of radiotherapy and immunotherapy.

Regarding the safety of combining systemic therapy with radiotherapy, early studies reported significant risk of serious bowel toxicities when radiotherapy was given concurrently with bevacizumab (an anti-VEGF agent). These studies documented a 9% incidence of serious bowel injury that included grade 3-4 gastrointestinal ulceration or grade 4-5 gastrointestinal perforation, with associated fatalities.^79,80 As a result, the recently published European Society for Medical Oncology–European Society for Radiotherapy and Oncology (ESMO-ESTRO) consensus statements on the safety of combining radiotherapy with anti-VEGF inhibitors recommend that a major adaptation is required when concurrent anti-VEGF therapy and abdominal radiotherapy are planned. Specifically, a clinically meaningful interruption of the anti-VEGF agent, and/or a radiotherapy dose reduction, is advised.⁸¹ However, due to the long half-life of bevacizumab (∼20 days), it may not be clinically feasible to withhold the drug for 4-5 half-lives in order to fully eliminate its effects. A reasonable approach would be to stop bevacizumab ≥4 weeks before and after RT, as recommended by the ASTRO guideline.⁸²

For patients receiving tyrosine kinase inhibitors (TKIs) with anti-VEGF activity, several studies have reported high rates of serious bowel and liver toxicities when these agents were combined with radiotherapy, particularly with sorafenib.^83,84 Therefore, a reasonable approach is to suspend TKIs with anti-VEGF properties for 5-10 days before and after radiotherapy.^81,82 Of note, in the RTOG-1112 study, patients in the SBRT plus sorafenib arm were required to start sorafenib at half of the standard dose (200 mg twice daily) at 1-5 days after the completion of SBRT. Grade ≥3 adverse events were similar between sorafenib (42%) and SBRT plus sorafenib arm (47%) Importantly, no cases of gastrointestinal perforation occurred in the SBRT plus sorafenib arm, and only two patients (2%) developed grade 3 gastric or upper gastrointestinal bleeding.⁶³ In contrast, for patients who are to receive concurrent ICIs such as anti-PD-(L)1 and radiotherapy, no treatment modifications are recommended.⁸¹ Indeed, prospective phase II studies evaluating the concurrent administration of radiotherapy and immunotherapy have not demonstrated increased toxicity.^85,86

Tackling portal vein tumour thrombosis using combination of radiotherapy and immunotherapy

Combining radiation with immunotherapy has also been explored in patients with macrovascular invasion, in particular for those with major portal vein tumour thrombosis (PVTT). These patients carried a grave prognosis compared with those without PVTT; without treatments their survival was only 2-4 months.⁸⁷ As a result, these patients are often excluded from pivotal clinical trials.^65,68,69 Although multiple locoregional therapies are available for the treatment of HCC, none of them are unsuitable for patients with major PVTT. On the contrary, retrospective studies have suggested that PVTT is exquisitely sensitive to radiotherapy, and a substantial number of patients with PVTT treated with radiotherapy achieved high ORR.88, 89, 90 These encouraging results prompted exploration of radiotherapy in combination with immunotherapy.

The NEXTRAH study was the first to evaluate the combination of radiotherapy with immunotherapy in patients with macrovascular invasion⁸⁵ (Table 2). The study was conducted in Korea and recruited 50 patients to receive nivolumab and external beam radiotherapy dose between 30 and 50 Gy in 10 fractions concurrently, followed by nivolumab maintenance. Of note, the study enrolled patients predominantly with a single tumour (60%) and a median tumour size of 7 cm. There were 26 patients (52%) with vascular invasion of the portal vein (Vp)3 or Vp4, and 6 patients (12%) with extrahepatic spread. The overall ORR was 36.0% and disease control rate (DCR) was 74.0%. The median PFS was 5.6 months and the median OS was 15.2 months. Among the patients with major PVTT, the PFS was 3.6 months and median OS was 10.8 months. Another randomized study explored the combination of SBRT with camrelizumab and apatinib, in patients with hepatitis B virus-related HCC and PVTT (Table 2). Of the patients, 67% had major PVTT. The addition of SBRT improved PFS and OS, compared with camrelizumab and apatinib alone (median PFS 4.6 months versus 2.5 months; median OS 12.7 months versus 8.6 months).⁹¹ A single-arm prospective phase II study conducted in China evaluated the combination of SBRT with sintilimab and bevacizumab in patients with PVTT (Table 2).⁹² Among the 46 patients enrolled, nearly half of them had Vp4 (45.7%). ORR according to modified RECIST was 59% and DCR was 100%. At a median follow-up time of 26 months, the median PFS was 13.8 months and the median OS was 24 months. For TRAEs, the most common were decrease in white cell count and platelet count. The most common high-grade TRAEs were decreased platelet count (n = 4, 8.7%) and hypertension (n = 4, 8.7%).

Taken together, these studies highlighted that combining radiotherapy with immunotherapy is generally safe and effective in patients with major PVTT. Notably, in a similar group of patients enrolled in the IMBrave150 study, patients treated with atezolizumab plus bevacizumab had a median PFS of 5.4 months and a median OS of 7.6 months only.⁹³ Thus, the addition of radiotherapy to immunotherapy, when feasible, should be discussed in a multidisciplinary team as a strategy to prolong survival in patients with PVTT. The ongoing NRG-GI012 study (NCT07166406) is recruiting to examine the efficacy of the combination between immunotherapy and radiotherapy in patients with HCC and macrovascular invasion.

Management of oligometastatic and oligoprogressive disease

The concept of oligometastatic disease (OMD), which consists of a maximum of three to five metastases, describes an intermediate state of locally advanced disease to polymetastatic, widespread disease.⁹⁴ Clinical research in OMD has garnered significant interest recently because these patients could be potentially cured if the OMD could be aggressively treated by local therapies.^95,96 This has motivated the establishment of a consensus statement from ESTRO and the American Society for Radiation Oncology (ASTRO) to provide formal definition and classification of OMD, namely, synchronous OMD, metachronous OMD (e.g. oligorecurrence), oligoprogression, and oligopersistence disease.⁹⁷ Studies have shown the benefits of radiotherapy for OMD in multiple cancers, including lung,^11,98 kidney,¹² and prostate cancers.⁹⁹

Emerging evidence suggests potential benefit of SBRT in de novo and recurrent OMD for HCC. The first prospective study on the role of SBRT in OMD came from China, and was a single-arm phase II study, enrolling 25 patients with de novo and recurrent OMD.¹⁰⁰ These patients were treated with SBRT to OMD sites at a median dose of 54 Gy in six fractions, followed by anti-PD-1 therapy sintilimab. The majority of OMD (76%) was intrahepatic. The 1- and 2-year local control rates were 100% and 91%, respectively. The median PFS was 19.7 months and the median OS was not reached. The most common TRAE was decreased lymphocyte count (56%), but the lymphocyte counts gradually returned to baseline in the third month of treatment. Grade ≥3 toxicity occurred in three patients, namely thrombocytopenia (n = 1), increased γ-glutamyltransferase (n = 1), and myositis (n = 1). Another prospective study conducted in Korea evaluated the role of SBRT in 40 patients with de novo or recurrent OMD.¹⁰¹ These patients had a primary tumour in the liver definitively treated ≥3 months before enrolment. As a result, majority of the OMD sites were extrahepatic. Among the 62 lesions that were treated, the predominant OMD locations were in the lung (48.4%), lymph node (22.6%), and bone (17.7%). Following SBRT to OMD, the 2-year time to local progression was 91% and ORR was 76%. However, the high local response did not translate into longer PFS, with a median PFS reaching at 5.3 months. In terms of safety, low grade acute (10%) and late (7.5%) toxicities were uncommon, and none reported high-grade toxicities.¹⁰¹ It is worthwhile to note the stark PFS difference despite similar high local control rates between the Korean and Chinese study. The key difference between the two studies is the utilization of systemic therapy. In the Chinese study, all the patients received systemic immunotherapy. In contrast, in the Korean study, 40% of patients did not receive any systemic therapy, and 45% of patients received concurrent or sequential TKI therapy. These findings suggest that, in the setting of OMD, the use of systemic therapy, in particular immunotherapy, to control micrometastases, is important for systemic control to improve PFS. Ablative treatments like SBRT to OMD could improve response but are insufficient to prolong survival.

In contrast to OMD, the evidence supporting the use of radiotherapy in managing oligoprogressive HCC is less reported but emerging. The introduction of effective systemic therapy such as immunotherapy combination has prolonged survival and enabled opportunities to explore the use of radiotherapy in oligoprogressive disease.¹⁰² Specifically, based on the latest results from first-line systemic therapy, it is expected that the OS for patients with unresectable HCC will reach ∼2 years. Approximately 30%-40% of patients would develop oligoprogression and they are potential candidates for ablative radiotherapy and maintenance of systemic treatment.

Our group is one of the earliest groups to report the potential of using SBRT to treat oligoprogressive disease in patients progressed on first-line immunotherapy.¹⁰³ In this small case series, five patients had advanced disease, including portal vein invasion and extrahepatic metastases. Despite the adverse features, this small group of patients reached a median PFS of 10 months. Recently, a retrospective study from China compared the treatment outcomes of patients with oligoprogression treated with radiotherapy (majority external beam radiotherapy), with switching to second-line systemic therapy, with or without radiotherapy. A total of 154 patients with oligoprogression were included in the study.¹⁰⁴ Forty-two patients remained on first-line systemic therapy and received radiotherapy, 24 patients switched to second-line systemic therapy and received radiotherapy, and 88 patients switched to second-line therapy.¹⁰⁴ The median PFS was notably longer for patients who were maintained on first-line systemic therapy and who had oligoprogression treated with radiotherapy, at 8.6 months, compared with second-line systemic therapy alone at 3.1 months. The addition of radiotherapy to oligoprogressive sites together with switching to second-line therapy resulted in a median PFS of 5.8 months. Highlighting the maintenance of first-line treatment is important even though ablative treatments were given to oligoprogressive sites. Similarly, in another study consisting of 28 patients with oligoprogressive disease during systemic therapy, the median PFS was 11.6 months following switching to next-line of systemic therapy plus radiotherapy, compared with 16.5 months for patients who were maintained on first-line systemic therapy and who received radiotherapy.¹⁰⁵ The addition of radiotherapy appears to be safe, and the most common TRAE was lymphopenia. No patients developed grade 4 or 5 TRAEs.

Given these promising signals, radiotherapy in combination with immunotherapy in OMD, in particular oligoprogressive disease, is expected to be an active area of research, especially given that a substantial number of patients with advanced HCC develop OMD. At the moment, a number of prospective studies (NCT06434480, NCT06870942, NCT06841172) are underway evaluating the role of radiotherapy in the setting of oligoprogressive disease, with maintenance immunotherapy. These studies will provide further data on the optimal treatment strategy for HCC following progression on immunotherapy.

Conclusion and future direction

Technological advancement has enabled the safe and effective delivery of radiotherapy to intra-abdominal targets, such as HCC. Although many studies have shown that the use of MR-guided radiotherapy and particle radiotherapy to treat HCC is generally safe with minimal high-grade toxicities, these technologies have not been compared with SBRT in a large, randomized study to demonstrate a clinically meaningful superior local control and survival. Furthermore, compared with conventional linacs, the preparation and treatment time for each patient are significantly longer, which will limit the number of patients who can be treated with these technologies.^106,107 The limited number of machines and cost involved in installing and operating these advanced technologies also posts challenges to the widespread application of these technologies. In particular, multiple health economic analyses have indicated that PBT is currently not cost effective for all patients compared with conventional photon-based therapy, but it might be cost effective in patients with high risk of developing severe treatment-related toxicities.108, 109, 110 This highlights that patient selection is key to fully utilizing these technologies.^111,112

Apart from technological advancements, the incorporation of radiotherapy as part of a combination treatment strategy is emerging. However, prospective evidence is lacking in this space, and the majority of the studies have originated in Asia, where there is a high utilization of radiotherapy in the treatment of HCC. The emergence of oligometastatic HCC opens an arena for future research. Given the promising results of combination therapy, the role of radiotherapy in the management of HCC will become crucial. However, despite these opportunities, the underlying complexity in the management of HCC and geographical differences in treatment strategies highlight the need for multidisciplinary team discussion to provide the best treatment options available for patients in their respective geographical regions.