Rationale and protocol of the CREFEL trial: a single-centre, single-arm pilot trial of cardiac radiotherapy for heart failure with reduced left ventricular ejection fraction

Abstract

Introduction

Heart failure with reduced ejection fraction (HFrEF) remains a significant cause of morbidity and mortality worldwide, particularly in patients who remain symptomatic despite guideline-directed medical therapy (GDMT). Preliminary data suggest that a single fraction low-dose whole-heart external beam radiotherapy (EBRT) may improve cardiac function by modulating inflammatory and fibrotic processes. This trial aims to evaluate the preliminary efficacy of a single fraction 5 Gy whole-heart EBRT to improve left ventricular ejection fraction (LVEF) in patients with HFrEF on GDMT. Secondary objectives will assess safety, cardiac biomarkers and patient-reported outcomes.

Methods and analysis

Single-centre, single-arm, prospective interventional trial aiming to enrol 40 patients with HFrEF (LVEF≤35%) on maximal GDMT, New York Heart Association (NYHA) classes II–IV, and stable for ≥6 months prior to enrolment, without recent heart failure admissions or GDMT changes. All participants are required to have cardiac implantable electronic devices. Recruitment will be balanced with 20 patients with ischaemic and 20 with non-ischaemic aetiology. Eligible patients will receive a single fraction of 5 Gy whole-heart EBRT guided by a non-contrast enhanced primary planning CT. Follow-up assessments will be conducted at baseline, 6 weeks, 12 weeks and 6 months. The primary outcome is an improvement in LVEF of ≥5% at 6 months, assessed by transthoracic echocardiography. This is an open-label trial with blinded ascertainment of the primary outcome. Secondary outcomes include acute and late toxicity, overall survival, hospital admission for heart failure, patient-reported quality of life, cardiac biomarkers and device-reported arrhythmia burden.

Ethics and dissemination

The trial has been approved by the Ethics Committee Research UZ/KU Leuven, Belgium (S69569). The study results will be shared through peer-reviewed journals and presentations at academic conferences.

Trial registration number

NCT06661876.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Prospective, single-arm pilot design with clearly defined eligibility criteria ensuring a stable heart failure with reduced ejection fraction population on maximal guideline-directed medical therapy.

• Use of standardised radiotherapy planning (5 Gy whole-heart external beam radiotherapy) and strict device dose constraints, enhancing reproducibility and safety.

• Blinded, independent assessment of the primary endpoint reduces measurement bias.

• Lack of a control group limits the ability to attribute observed changes solely to the intervention.

Background

Heart failure (HF) has been defined as a global pandemic affecting more than 64 million patients worldwide.^{1 2} It is a multifaceted and life-threatening syndrome with a major social and economic burden as it affects a patient’s functional capacity and quality of life.¹ Although the prognosis has slightly improved in the past decades, outcomes for patients with HF remain poor, especially in patients with HF with reduced ejection fraction (HFrEF).² Despite recent advances in guideline-directed medical therapy (GDMT), with sodium-glucose co-transporter 2 (SGLT2) inhibitors and angiotensin receptor-neprilysin inhibitors (ARNI), as well as cardiac resynchronisation therapy (CRT), mortality and symptom burden remain high in a substantial proportion of patients.2,5 Even on maximal 4-pillar GDMT, many patients progress to advanced HF, a clinical stage defined by persistent symptoms and poor functional capacity despite maximal treatment. Advanced HF is associated with a 1 year mortality rate ranging from 25% to 75%, demonstrating the need for novel treatment strategies.²

Recently, stereotactic body radiotherapy (SBRT) has emerged as a novel approach to treat patients with ventricular tachycardia (VT) refractory to antiarrhythmic drugs and catheter ablation.⁶ In the ENCORE-VT phase I/II trial and subsequent studies, SBRT was shown to significantly reduce VT burden in the majority of patients.^{7 8} Besides the reduction in VT burden, recent data are showing other potentially beneficial cardiac effects of radiotherapy. SBRT for refractory VT is delivered using a single-fraction high dose ≥25 Gy and is accurately targeted to the critical substrate in the myocardium. However, given the dose distribution and decay, the rest of the heart receives a low-dose bath. Pedersen et al showed, both through analysis of the ENCORE-VT cohort and in murine models, that low-dose radiation may improve left ventricular function in patients with various aetiologies of HF, with the benefit persisting at least up to 3 months.⁹ Notably, the improvement was observed days after SBRT and therefore deemed unlikely to result from the reduction in VT burden which usually manifests after several weeks. This preliminary data suggest that low-dose whole-heart radiotherapy can modulate the inflammatory and fibrotic processes that contribute to adverse cardiac remodelling and may represent a novel therapeutic strategy for improving outcomes in patients with HFrEF.⁹

Methods

Aims

The primary aim of the Cardiac RadiothErapy For hEart faiLure (CREFEL) trial is to assess the impact of a single-fraction, low-dose, whole-heart EBRT of 5 Gy in patients with symptomatic, yet stable HFrEF. Based on recent preclinical and clinical data, this intervention is hypothesised to result in a clinically meaningful improvement in left ventricular ejection fraction (LVEF) at 6 months follow-up.

Secondary objectives are as follows: (1) safety, including both acute toxicity within 30 days after intervention and late toxicities beyond 30 days; (2) efficacy, including overall survival, heart failure admissions and cardiac biomarkers; (3) patient-reported outcomes with quality of life questionnaires. Because the study is not powered for these outcomes, they will be assessed descriptively and considered exploratory.

Study design and setting

The CREFEL trial is a prospective, single-centre, single arm, open-label study with blinded ascertainment of the primary outcome, conducted in an academic hospital in Belgium. The primary endpoint will be assessed through a blinded evaluation of the echocardiographic images to minimise bias and ensure objective outcome measurement. Two independent assessors will be blinded to both patient identity and the imaging time point.

Study population and recruitment

Eligible participants are adults (≥18 years) diagnosed with ischaemic or non-ischaemic cardiomyopathy with LVEF≤35% and presenting with NYHA classes II–IV HF while on maximal GDMT. Maximal GDMT is defined as the use of all four pillars of contemporary guideline-directed therapy for HFrEF, including a β-blocker, a renin–angiotensin system inhibitor preferably an ARNI, a mineralocorticoid receptor antagonist and a SGLT2 inhibitor, at the maximum tolerated dose. All participants have cardiac implantable electronic devices (CIEDs) capable of recording arrhythmic events and have been free of acute HF events for at least 6 months. The latter implies no hospital admissions for HF, nor any changes in GDMT in the past 6 months before recruitment. Key exclusion criteria include eligibility for heart transplantation, pregnancy or breastfeeding, prior thoracic radiotherapy involving the heart.

Patients are identified and recruited during routine ambulatory clinic visits at the cardiology department of UZ Leuven. A local investigator informs potential participants about the study in detail. Written informed consent is obtained prior to the required investigations. An example of the informed consent form is provided as online supplemental file 1. The target sample size is 40 patients, accounting for 20% dropout and stratified evenly between ischaemic and non-ischaemic cardiomyopathy.

Intervention

Patients receive a single-fraction, 5 Gy EBRT treatment targeting the entire heart. This dose was selected by analogy to the dosimetric analysis of Pedersen et al, in which 5 Gy approximated the mean whole-heart exposure outside the ablation target in SBRT-treated patients (ENCORE-VT cohort).⁹ The intervention is planned and delivered by a radiation oncologist, in collaboration with the cardiology team. Treatment planning includes non-contrast enhanced primary planning computed tomography (4D-CT) imaging, with additional cardiac MRI or echocardiography used for guidance if available. EBRT is delivered using either intensity-modulated radiotherapy or volumetric modulated arc therapy (VMAT) with a photon beam of at least 6 MV energy. Both techniques enable a highly conformal dose distribution to be delivered, with a sharp dose fall-off outside the target, thereby decreasing the dose to the surrounding organs at risk (OAR). Both techniques differ only in how the radiation is delivered to the patient. The first technique involves the linear accelerator delivering radiation from various static beam angles, while the latter technique involves the linac rotating around the patient while delivering the radiation. VMAT will generally be used unless this is not possible for technical reasons. The planning target volume (PTV) encompasses the whole heart including a respiratory and cardiac motion treatment range and a technical treatment margin, usually 5 mm. The near-maximum dose is limited to 5.35 Gy (PTV D2%≤5.35 Gy). Figure 1 shows a dose distribution map of 5 Gy EBRT delivered to the whole heart. Given that half of the enrolled patients will have ischaemic cardiomyopathy, a subset will have undergone prior percutaneous coronary interventions (PCIs). Organ-dose estimates for fluoroscopically guided coronary procedures indicate median heart doses in the mGy range.¹⁰ Therefore, even with multiple previous PCIs, the cumulative myocardial dose is negligible compared with the prescribed dose of 5 Gy and is not considered a safety concern in this protocol. The surrounding OAR around the heart has a dose tolerance well above the prescribed 5 Gy in this trial; for example, the maximum dose (Dmax) to the left coronary artery is ≤20.0 Gy. Therefore, no hard constraints to the OAR are implemented. The only relevant dose constraint is the CIED where Dmax is ≤0.5 Gy.¹¹ Treatment planning excludes the CIED from primary beam irradiation. Despite the 5 Gy myocardial prescription, the steep dose fall-off ensures that the CIED dose constraint is met. The rapid isodose decay outside the PTV is shown in figure 1. The treatment will be performed under close cardiological monitoring with continuous ECG monitoring. Patient (re)positioning, image guidance and target localisation will be performed by either using a time-resolved or triggered volumetric cone-beam CT or by using stereoscopic X-ray images as landmarks near the target (eg, transvenous leads if present) for motion compensation. A wall-mounted surface tracking device will be used for initial positioning and to monitor patient motion during treatment (intrafraction motion). Directly after the treatment, the CIED device will be evaluated for functionality and will be reprogrammed, if necessary. Participants will continue the baseline GDMT unchanged throughout the study. Diuretic adjustments are permitted based on clinical need. All changes to concomitant therapy will be documented.

Figure 1. Dose distribution map of whole-heart EBRT. The figure illustrates the dose distribution of whole-heart irradiation in the axial (A), sagittal (B) and coronal (C) planes. (D) The dose distributions are indicated by a colour wash ranging from 1 Gy (20% of the prescribed dose) to 5.5 Gy (110% of the prescribed dose). EBRT, external beam radiotherapy.

Outcomes

The primary outcome is based on efficacy and defined as an improvement in LVEF of ≥5% from baseline at 6 months after the intervention. LVEF will be assessed by transthoracic echocardiography using Simpson’s biplane method. The echocardiographic images are obtained by an experienced cardiologist. LVEF measurements will be performed only after completion of the trial by exporting all images to an off-line workstation. To ensure a blinded assessment of the primary endpoint, all echocardiographic studies will be pseudonymised and presented in random order to two independent assessors who are therefore blinded to participant information and the timepoint in the trial. This approach ensures that the blinded assessors can objectively evaluate LVEF without knowing whether the imaging study was conducted before or after radiotherapy. Analysis will be performed based on the average measurements of the two assessors. A subgroup analysis comparing the primary endpoint between participants with ischaemic and non-ischaemic cardiomyopathy will be performed.

The secondary outcomes include safety, overall survival, heart failure hospitalisations, patient-reported outcomes, cardiac biomarkers, arrhythmic burden and LVEF improvement at 6 and 12 weeks. Safety is evaluated by the incidence of acute (≤30 days) and late (31 days to 6 months) toxicity. Toxicity is assessed using the Common Terminology Criteria for Adverse Events (CTCAE) V.5.0,¹² and will be reported as prevalence. Overall survival is defined as death from any cause occurring within 6 months following treatment. The study will additionally assess HF-related hospitalisations, defined by the number and duration of admissions exceeding 24 hours that are attributed to HF and necessitate intensification of diuretic therapy.¹³ Patient-reported outcomes will be evaluated using two validated instruments. First, the 36-item Short Form Healthy Survey (SF-36) and the Minnesota Living with Heart Failure Questionnaire (MLWHFQ).^{14 15} High scores on the SF-36 indicate better health status, whereas higher scores on the MLWHFQ indicate a greater negative impact of heart failure on quality of life. Other heart function indicators that will be assessed include end-systolic and end-diastolic left ventricular volumes as obtained by echocardiography using the Simpson’s biplane method, NT-proBNP levels, high-sensitive troponin levels, high-sensitive CRP levels and the arrhythmic burden measured on the implanted device. Echocardiographic assessments at 6 and 12 weeks will be analysed as secondary, exploratory outcomes (LVEF change from baseline), using the same blinded reading procedures as for the primary endpoint.

Data collection and follow-up

An overview of the trial timeline and follow-up schedule is presented in table 1. Study assessments will be conducted at baseline, 6 weeks, 12 weeks and 6 months after intervention. Each assessment will include transthoracic echocardiography, blood analysis, patient-reported outcome questionnaires, electrocardiography, toxicity assessment and CIED interrogation. At 12 weeks postintervention, a thoracic CT is performed to evaluate for subclinical radiation pneumonitis and pericarditis. Remote CIED monitoring will be utilised when available to enhance data collection between visits. To promote complete follow-up, study visits are aligned with routine clinical appointments where possible. Participants will receive appointment reminders and flexible scheduling will be offered to accommodate individual needs. In the event of a missed follow-up, the study team will contact the participant to reschedule the visit or arrange remote data collection when feasible. For participants who withdraw or deviate from the protocol, efforts will be made to collect available outcome data up to the point of discontinuation.

Table 1. Follow-up schedule.

	Enrolment	EBRT treatment	Week 6	Week 12	6 months
Eligibility screening	x
Informed consent	x
Medical history and medication profile	x
CIED analysis	x	x	x	x	x
Planning CT	x
Transthoracic echocardiography and electrocardiography	x		x	x	x
Blood work	x		x	x	x
Thoracic CT				x
Registration toxicity (CTCAE V.5.0)	x	x	x	x	x
Patient-reported outcomes	x		x	x	x

CIED, cardiac implantable electronic device; CTCAE, Common Terminology Criteria for Adverse Events; EBRT, external beam radiotherapy.

Safety reporting

As defined above, safety is a secondary outcome of the trial. All adverse events (AEs), whether solicited, spontaneously reported by the participant or observed by the study team, will be documented from the time of intervention until the end of the 6 month follow-up period. At each scheduled study visit, participants will be actively monitored for symptoms of toxicity using the CTCAE V.5.0. All AEs will be recorded in the electronic case report form (eCRF) and will be assessed for seriousness, severity and potential relationship to the intervention. Adverse events of special interest (AESI) include upper gastrointestinal tract symptoms and any coronary event requiring intervention. All serious adverse events or AESIs will be promptly reported to the institutional review board in accordance with institutional and regulatory requirements. Participants experiencing AEs will be followed clinically until resolution or stabilisation. All deaths will be reported without delay to the institutional review board (irrespective of whether the death is related to disease progression, study procedure or is an unrelated event). Because follow-up is limited to 6 months, and some radiation-related AEs have longer latency, this study may not capture very late events. However, given the single 5 Gy prescription to the myocardium and OAR doses well below established constraints, delayed serious, radiation-related AEs beyond 6 months are not anticipated. A formal Data and Safety Monitoring Board has not been established given the study’s sample size and anticipated risk.

Statistical analysis

A sample size of 40 patients is expected to be recruited in a 24 month accrual period. Based on the expected baseline of LVEF of 20% with a 5% SD, this sample size is based on detection of a clinically meaningful improvement of ≥5% in LVEF, with 90% power and an alpha of 0.05. After correction for 20% drop-outs, a minimum of 16 patients is required for each subgroup. Given the objective of subgroup analysis, establishing preliminary efficacy in both ischaemic and non-ischaemic cardiomyopathy, the overall sample size was set at 40 patients.

Descriptive statistics for demographics, clinical variables and outcomes described above will be conducted. The primary efficacy outcome, defined as change in LVEF, will be assessed using repeated measures analysis of variance (ANOVA), including a subgroup analysis comparing participants with ischaemic and non-ischaemic cardiomyopathy. Secondary outcomes are analysed descriptively, using time-to-event analysis or repeated measures ANOVA where appropriate. As the study is not powered to formally assess these outcomes, analyses will be considered exploratory. All results will be presented with the corresponding 95% CI. A p-value<0.05 will be considered significant.

Ethical considerations and dissemination

The study protocol has been approved by the Ethics Committee Research UZ/KU Leuven under registration S69569 on 18 October 2024. Since two amendments have been made. The first was approved on 6 February2025 and concerns approval of patient documents in English and French. The second amendment was approved on 10 April 2025 and concerns the NYHA class as inclusion criterion. Any future protocol amendments will be submitted for further review and approval. The trial adheres to the Declaration of Helsinki and the Belgian Law of 7 May 2024 concerning human experiments. Insurance is provided in accordance with Belgian Law and provides cover for damage to research subjects through injury or death caused by the study. Dissemination of the study outcomes will occur through academic conference presentations and submissions to scientific journals. Authorship of publications will be determined in accordance with the requirements published by the International Committee of Medical Journal Editors and in accordance with the requirements of the respective journal. No professional medical writers will be employed in the preparation of manuscripts or other study-related publications.

Confidentiality and data management

This study will be conducted according to the principles of the Declaration of Helsinki (version 7, October 2013) and in accordance with the ICH-GCP guidelines of 17 January 1997. The data will be handled confidentially and anonymously. The handling of personal data will comply with the Dutch Personal Protection Act and with the Belgian law of 8 December 1992 on the protection of privacy regarding the processing of personal data. A subject identification code list will be used to link the data to the subject. These codes will not be based on the patient initials or birth date. The principal investigator will safeguard the key to this code. All study data will be entered into a secure, password-protected eCRF. Data entry will be performed by trained study personnel, and access will be restricted based on user roles. To promote data quality, automatic validation checks will be embedded within the eCRF system, including range checks, logic checks and completeness checks to ensure all required fields are accurately filled. The data will be kept for 15 years. Given the early-phase, single-arm design of this study and the relatively low anticipated risk of the intervention, a formal data monitoring committee has not been established. Instead, safety and study conduct will be monitored internally by the study team and communicated with the institutional review board, in accordance with institutional and regulatory requirements. Deidentified individual participant data underlying the published results, along with the study protocol, statistical analysis plan, informed consent form and analytic code, will be made available beginning 6 months after publication of the primary results, for at least 5 years. Access will be granted to researchers with a methodologically sound proposal, subject to approval of a data use agreement. Requests can be submitted to the corresponding author.

Patient and public involvement

Patients or the public were not involved in the design, or conduct, or reporting, or dissemination plans of our research.

Discussion

The CREFEL trial is a first clinical pilot study investigating the use of whole-heart low-dose EBRT as a novel therapeutic strategy for patients with HFrEF on maximal GDMT. The intervention is based on recent preclinical and clinical data suggesting that low-dose cardiac radiation may modulate cardiac function, resulting in improved LVEF and reduced myocardial fibrosis.⁹

Unlike previous cardiac radiotherapy trials that focused on VT suppression through targeted high-dose radio-ablation of arrhythmogenic substrates, this study explores the potential benefit of a ‘whole-heart’ low-dose approach on myocardial remodelling. EBRT can be used for a wide variety of benign conditions based on its anti-inflammatory or antiproliferative effects.16,18 Recent translational research by Pedersen et al demonstrated that low-dose cardiac radiation was associated with significant improvements in LVEF in both murine models and in patients with cardiomyopathy who underwent high-dose cardiac radiotherapy in the setting of therapy-refractory VT. In the retrospective analysis of the ENCORE-VT cohort, these functional benefits were observed shortly after treatment, suggesting a mechanism independent of VT suppression. The rationale for selecting a dose of 5 Gy was based on dosimetric analysis of this cohort, which showed that 5 Gy approximated the mean whole-heart dose received outside the targeted scar during VT radiotherapy. They observed depletion of inflammatory and fibrotic cells, specifically macrophages and fibroblasts, which are more radiosensitive than non-proliferative cardiomyocytes. These findings provide a biological rationale for applying whole-heart radiation in heart failure beyond its initial use in arrhythmia control.

The rationale for the patient eligibility criteria has been well considered to ensure the feasibility of detecting a consistent effect of cardiac radiotherapy on LVEF. A population of patients with stable HFrEF for a minimum of 6 months on maximal GDMT was chosen to exclude variation in LVEF based on recent episodes of congestion or effects secondary to changes in GDMT. This approach does not strictly follow the European Society of Cardiology (ESC) and American Heart Association (AHA) definitions of advanced heart failure (NYHA III–IV with recurrent hospitalisations), but was deliberately chosen to ensure a stable population in whom treatment effects can be evaluated more reliably. However, given the described anti-inflammatory properties of low-dose radiotherapy, future studies may consider expanding this approach to patients with acute HF as this is characterised by systemic inflammation, with elevated levels of inflammatory mediators and upregulated cytokines associated with worse prognosis.¹⁹

Interim echocardiograms at 6 and 12 weeks will permit detection of early effects, in analogy with Pederson et al,⁹ but the primary endpoint at 6 months focuses on persistence of a potential functional benefit. A transient effect limited to the first weeks after treatment would not be therapeutically useful, as repeated radiation sessions would not be acceptable given safety considerations. Although Pedersen et al reported a mean absolute increase of 13.7% in LVEF in the ENCORE-VT cohort,⁹ this trial is designed to detect a more modest absolute increase of 5% in LVEF. Given that patients have a baseline LVEF of ≤35%, even a 5% improvement would be clinically meaningful and could provide a foundation for larger trials that may apply higher thresholds.

Several limitations and practical challenges of this pilot study should be acknowledged. First, successful implementation of the CREFEL trial requires close coordination between cardiology, radiation oncology and medical physics, an interdisciplinary collaboration not typically seen in standard HF care. Nevertheless, this collaborative model has already been successfully established at our centre through the ongoing CREVET and CREVET II trials (Cardiac RadiothErapy for VEntricularTachycardia—ClinicalTrials.gov NCT05973578 and NCT06744530). Second, the assessment of LVEF is based on transthoracic echocardiography, which is inherently subject to interobserver and intraobserver variability. To mitigate this, all echocardiographic images are acquired by the same experienced imaging cardiologist and subsequently evaluated by two independent, blinded cardiologists, which helps to minimise measurement bias and enhance consistency. Third, the lack of a control group rules out formal conclusions about efficacy. However, as the first clinical trial to investigate whole-heart low-dose radiotherapy in HF, the CREFEL trial is designed to establish preliminary efficacy and provide a foundation for larger-scale studies. At this early stage, inclusion of a control group would not only require a substantially larger sample size but also the use of a sham radiotherapy intervention, which poses ethical and logistical challenges in this vulnerable patient population. Finally, the relatively short duration of follow-up may not capture the full trajectory of cardiac remodelling or potential late radiation-related toxicities.

While many of these limitations may be addressed in future larger, controlled studies with extended follow-up, these studies will first depend on pilot studies such as the CREFEL trial. This pilot will provide preliminary data on efficacy and safety, together with practical insights into patient recruitment, treatment planning, follow-up and data collection to guide the design of future larger scale studies. Effect sizes from this early-phase trial will be interpreted with caution and used primarily to plan subsequent multicentre, randomised controlled trials. Our study group intends to use the results of the CREFEL pilot to inform and design future trials that will further evaluate the safety and preliminary efficacy of this therapeutic approach. If successful, cardiac radiotherapy could extend from a niche antiarrhythmic intervention to a novel, disease-modifying therapy for HF, especially in cases refractory to pharmacological management.