Background and Rationale:

Locally advanced cervical cancer remains a major public health concern worldwide and in France. Standard-of-care management involves concurrent chemoradiotherapy followed by brachytherapy, which is critical for local tumour control and survival. Conventional high-dose-rate (HDR) brachytherapy protocols often require multiple implantations under anaesthesia, entailing increased logistical demands, operating room use, and patient discomfort.

The UNICURE-HD study investigates a simplified brachytherapy approach - a single intraoperative implantation delivering all planned fractions - combined with image guidance (MRI and/or CT) and 3D dosimetry, in order to assess whether it can maintain oncological outcomes while reducing treatment complexity and resource consumption.

Objectives:

Primary objective: To evaluate local control at 3 and 5 years after completion of chemoradiotherapy and image-guided HDR uterovaginal brachytherapy with single implantation.

Secondary objectives: To assess pelvic nodal control, para-aortic nodal control, distant control, progression-free survival, overall survival, and acute/late severe toxicities (CTCAE v5.0 ≥ grade 3). Prognostic factors will be analysed, and exploratory predictive models will be developed.

Study Design:

This is a retrospective, observational, multicentre study involving six French cancer centres (Pitié-Salpêtrière, Saint-Louis, Tenon, Institut de Cancérologie de Lorraine, Centre Oscar Lambret, CHU de La Réunion). Eligible patients are women ≥18 years, diagnosed with FIGO 2018 stage IB3-IV cervical cancer, treated between January 2014 and June 2024 with concurrent chemoradiotherapy followed by image-guided HDR uterovaginal brachytherapy using a single implantation.

Methods:

Data will be extracted from medical records (electronic or paper), pseudonymised locally, and entered into a secure AP-HP-hosted database compliant with GDPR and the MR004 framework. Variables collected include demographics, tumour characteristics, EBRT details, brachytherapy technique (endocavitary vs hybrid/interstitial), dosimetric parameters (EQD2 for HR-CTV, IR-CTV, OARs), and outcomes. Imaging protocols, applicator types, and treatment times will also be recorded.

Statistical analysis will use descriptive statistics, Kaplan-Meier survival estimation, log-rank tests, and multivariable Cox regression. Predictive models (LASSO, random forests) will be explored. Analyses will be conducted in R software.

Sample Size:

Approximately 400 patients are expected, representing all eligible cases treated over the 10-year period in the participating centres.

Ethics and Regulatory Compliance:

The study follows GDPR and French data protection regulations (MR004). Ethics opinion has been requested from the Sorbonne Université Ethics Committee. Each centre will retain the correspondence table linking identifiers to pseudonyms locally; no directly identifying data will be shared.

Potential Impact:

If the single-implantation HDR brachytherapy strategy achieves equivalent tumour control and toxicity profiles compared to multi-implant protocols, it could streamline cervical cancer management, reduce anaesthetic risk, improve patient comfort, and optimise resource use in high-volume oncology departments. This could have significant implications for accessibility and cost-effectiveness of cervical cancer care, particularly in settings with limited resources.

Background and Rationale Cervical cancer remains a major global health concern, ranking among the leading causes of cancer-related mortality in women in low- and middle-income countries, but also persisting as a significant public health problem in high-income countries such as France. The prognosis of locally advanced cervical cancer (LACC) has markedly improved over the past two decades thanks to the standardisation of concurrent chemoradiotherapy followed by brachytherapy, in accordance with international guidelines (e.g., GEC-ESTRO, ABS, ESGO).

Brachytherapy is an essential component of curative treatment, enabling the delivery of high doses to the tumour volume while sparing surrounding organs at risk (OARs) through the steep dose fall-off inherent to the inverse square law. Multiple large-scale studies have demonstrated that omission or suboptimal delivery of brachytherapy significantly compromises local control and survival. However, conventional high-dose-rate (HDR) image-guided adaptive brachytherapy (IGABT) for cervical cancer often requires multiple applicator implantations (usually 2-3) under general or spinal anaesthesia, spread over separate hospitalisations or days.

This traditional multi-implantation schedule, although effective, poses several challenges:

1. Logistical complexity: multiple operating room slots, anaesthetic procedures, and inpatient stays are required, which increases healthcare system burden.
2. Patient comfort and safety: repeated invasive procedures are associated with increased pain, anxiety, risk of anaesthesia-related complications, and overall reduced quality of life.
3. Resource constraints**: in high-volume oncology departments or in resource-limited settings, multiple implantations can delay treatment or limit access for other patients.
4. Overall treatment time: prolongation of the total treatment time beyond 50 days is associated with a \~1% decrease in local control per day of delay, underlining the importance of optimising schedules.

To address these challenges, some teams have adopted a *single-implantation* strategy, whereby all HDR brachytherapy fractions are delivered from a single applicator placement, either over consecutive days (with in-patient applicator maintenance) or on the same day with multiple fractions. This approach is less common but potentially offers major advantages in terms of logistics, patient comfort, and cost-efficiency, provided that it maintains oncological efficacy and safety.

Study rationale The UNICURE-HD study aims to comprehensively evaluate the outcomes of image-guided HDR uterovaginal brachytherapy delivered with a single implantation in patients with LACC treated in French reference centres. This large, multicentre retrospective analysis will address a gap in the literature, as most published data on single-implant strategies come from small single-centre series with heterogeneous protocols and limited follow-up.

Given that France benefits from a strong brachytherapy infrastructure and experienced multidisciplinary teams, but also faces increasing demands on operating theatres and inpatient beds, demonstrating the safety and efficacy of single-implant HDR brachytherapy could have significant organisational and economic impact without compromising patient outcomes.

Objectives

Primary objective:

To determine the 3 and 5-year local control rate following chemoradiotherapy and single-implantation image-guided HDR uterovaginal brachytherapy in women with LACC.

Secondary objectives:

1. To evaluate progression-free survival (PFS) and overall survival (OS).
2. To assess pelvic nodal control, para-aortic nodal control, and distant metastasis control rates.
3. To document the incidence and patterns of severe (grade ≥3, CTCAE v5.0) acute and late toxicities affecting urinary, rectal, gastrointestinal, and vaginal functions.
4. To identify clinical, radiological, and dosimetric prognostic factors for local control, PFS, and OS.
5. To develop exploratory predictive models for relapse and survival using advanced statistical and machine learning methods.

Study Design:

This is a retrospective, observational, national multicentre study involving six French centres with expertise in gynaecological brachytherapy:

• Pitié-Salpêtrière Hospital, Paris (Coordinating centre)
• Saint-Louis Hospital, Paris
• Tenon Hospital, Paris
• Institut de Cancérologie de Lorraine, Nancy
• Centre Oscar Lambret, Lille
• CHU de La Réunion

Population:

Inclusion criteria encompass adult women (≥18 years) with histologically confirmed LACC (FIGO 2018 IB3-IV), treated with concurrent chemoradiotherapy followed by image-guided HDR uterovaginal brachytherapy using a single implantation, between January 1, 2014, and June 15, 2024.

Non-inclusion criteria include absence of brachytherapy, treatment without concurrent chemoradiotherapy, or missing follow-up data.

Sample size:

An estimated 400 patients are expected, representing all eligible cases during the study period in participating centres.

Treatment Details:

All patients received EBRT to the pelvis (± para-aortic fields if indicated) with a total dose typically ranging from 45-50 Gy in 1.8-2 Gy fractions, delivered using 3D conformal radiotherapy or IMRT/VMAT. Nodal boosts were administered based on disease extent. Concurrent weekly cisplatin (40 mg/m²) or alternative regimens (e.g., carboplatin) were used unless contraindicated.

Brachytherapy was performed using an endocavitary applicator or hybrid/interstitial technique depending on tumour response and anatomy. Imaging for applicator reconstruction included MRI (preferred) and/or CT, with MRI often fused to CT for optimal target delineation. Dose prescription and planning followed GEC-ESTRO guidelines, aiming for HR-CTV D90 ≥85 Gy EQD2 (α/β = 10) while respecting OAR constraints (e.g., bladder D2cc <90 Gy EQD2, rectum/sigmoid D2cc <75 Gy EQD2, α/β = 3).

Data Collection and Variables

Collected data include:

Demographics: age Tumour characteristics: histology, FIGO stage, initial MRI findings, nodal status Treatment parameters: EBRT technique, fields, dose, chemotherapy regimen, brachytherapy technique, applicator type, imaging modality, fractionation scheme, dosimetric parameters (HR-CTV, IR-CTV, OAR doses), total treatment time Outcomes: local, pelvic nodal, para-aortic nodal, distant relapse (dates and locations), vital status, last follow-up Toxicities: maximum grade acute and late toxicities, date of onset, management strategies

Statistical Analysis:

Survival endpoints (local control, PFS, OS, nodal and distant control) will be estimated using the Kaplan-Meier method. Comparisons will be performed with log-rank tests. Prognostic factor analysis will use univariate and multivariable Cox proportional hazards models. Predictive modelling with machine learning algorithms (LASSO regression, random forests) will be explored for relapse and survival prediction. All analyses will be performed using R software (v4.4.2 or later).

Ethics and Regulatory Compliance:

The study is conducted under the MR004 framework for retrospective research without informed consent, in compliance with GDPR. The AP-HP Data Protection Officer has been notified, and the Sorbonne Université Ethics Committee opinion has been sought. Pseudonymisation is performed locally at each centre, and no directly identifying data will be transferred to the coordinating centre.

Potential Impact:

If single-implant HDR brachytherapy is shown to provide equivalent local control and toxicity outcomes compared with traditional multi-implant schedules, it could:

1. Streamline treatment delivery, freeing up operating rooms and staff.
2. Reduce anaesthesia-related morbidity and patient discomfort.
3. Shorten treatment times, potentially improving tumour control.
4. Optimise cost-effectiveness and scalability of brachytherapy programmes in both high-resource and resource-constrained settings.

The UNICURE-HD study will therefore provide robust, multicentre evidence to inform clinical practice guidelines and health policy regarding brachytherapy delivery for cervical cancer.

Inclusion Criteria:

1. Women aged 18 years or older
2. Diagnosis of locally advanced cervical cancer (squamous cell carcinoma, adenocarcinoma, or other)
3. FIGO stage IB3-IV
4. Treatment with chemoradiotherapy followed by uterovaginal brachytherapy with implantation in the operating room
5. Informed and not opposed to the collection of her data for research purposes

Exclusion Criteria:

1. Lack of information regarding follow-up
2. Patient under legal protection measures -

Locally advanced cervical cancer (FIGO 2018: IB3-IV) treated with concurrent chemoradiotherapy followed by HDR uterovaginal brachytherapy

The procedures carried out with the French data privacy authority (CNIL, Commission nationale de l'informatique et des libertés) do not provide for the transmission of the database, nor do the information and consent documents signed by the patients.

Consultation by the editorial board or interested researchers of individual participant data that underlie the results reported in the article after deidentification may nevertheless be considered, subject to prior determination of the terms and conditions of such consultation and in respect for compliance with the applicable regulations.

Beginning 3 months and ending 3 years following article publication. Requests out of these time frame can also be submitted to the sponsor

Researchers who provide a methodologically sound proposal.