Cervical cancer continues to be a significant health issue and cause of cancer morbidity and mortality, with around 660,000 new cases and 350,000 deaths related to the disease globally each year and approximately 1,000 new cases and about 400 deaths in Hungary [1]. Of the many modern approaches, radiotherapy is one that may improve outcome. It is estimated that by implementing an image-guided adaptive radiotherapy (IGART) technique, 5-year disease-free survival will be more than 90% among patients with locoregional disease [2].

While early-stage cervical cancer can be cured with surgery or even with radiotherapy alone in certain clinical circumstances, the standard curative treatment for locally advanced cervical cancer is concomitant chemoradiotherapy (CRT) [3–8]. Intracavitary (IC) or combined intracavitary/interstitial (IC/IS) brachytherapy (BT) has been an essential component of the IGART approach in the last decade. No relevant clinical data indicate that BT boost could be replaced by any advanced external beam radiotherapy (EBRT) technique providing an equivalent outcome [9]. Due to excellent soft tissue resolution providing additional information regarding residual tumor extent, environmental propagation, and surrounding normal anatomy, magnetic resonance imaging (MRI)-based image-guided adaptive BT (IGABT) has become the gold standard curative treatment for locally advanced cervical cancer [2]. Since access to MRI has been limited in many clinical centers and its shortage is often aggravated by logistical and financial obstacles, computed tomography (CT)-based 3-dimensional (3D) image guidance is often used to replace MRI for BT planning.

Based on The Groupe Européen de Curiethérapie (GEC) and the European SocieTy for Radiotherapy & Oncology (ESTRO) recommendations, the widespread use of MRI-based image guidance has been receiving increasing clinical support in terms of both precision treatment planning and clinical outcome. The EMBRACE-I trial is considered to be the first prospective multicenter study offering a benchmark for the utilization of MRI-guided IGABT, with an excellent 92% overall 5-year local control rate [10]. Data from retroEMBRACE, a large retrospective multicenter analysis ongoing in parallel with EMBRACE-I, showed an 89% overall local control rate at 5 years [11]. Consequently, the EMBRACE-II was launched in 2016 as the first prospective multicenter interventional study aiming at uniform target concept and dose prescription protocol for MRI-guided IGABT to further improve efficiency and reduce treatment-related moderate and severe morbidity. A strict complex protocol of imaging, EBRT, and BT treatment planning and treatment delivery has been devoted to these aims. Since MRI-guided IGABT provides individually inhomogeneous dose distribution and the possibility of dose escalation in large and irregular volumes based on complex applicator arrangements and 3D treatment planning, this sophisticated technique ensures a superior outcome reinforced by the excellent clinical results from the previous EMBRACE studies [12].

We introduced the IGART technique in routine practice in our department at the end of 2015 for the definitive CRT of locally advanced cervical cancer. As a next step, we set out to introduce and adapt the recommendations of the EMBRACE-II protocol [2]. For various reasons, not all patients received MRI-based IGABT during the procedure. We intended to analyze our first findings on both the dosimetry and treatment-related toxicity aspects depending on the BT technique applied (MRI-based vs non-MRI based).

The clinical data of 50 locally advanced cervical cancer patients treated with definitive CRT and IGABT boost between February 2016 and August 2022 at the Department of Oncotherapy University of Szeged were evaluated in our retrospective study. The average (±SD) age of the patients overall was 52.7 ± 12.6 years, ranging between 26 and 83 years. The median (±SD) overall treatment time was 53.8 ± 9.9 (range: 42–81) days. All patients had FIGO IB-IIIB stage locally advanced cervical cancer, 22 (44.0%) of them had negative lymph node status, and 28 (56.0%) were diagnosed with N1 disease (Table 3). The vast majority of cases, 43 patients (86.0%), had squamous cell carcinoma, while seven patients (14.0%) had adenocarcinoma. Two-thirds of the study population was diagnosed with grade 2-3 tumors (Table 3). Of the 50 patients, 43 (86.0%) completed the scheduled five chemotherapy cycles, while seven (14.0%) received only four cycles of cisplatin. In the MRI-based IGABT group, among 24 patients, eight received combined the IC/IS IGABT treatment. Altogether, six patients had three brachytherapy sessions only due to dose limitations to the OARs: one in the MRI-based IGABT group and five in the CT-based treatment group. Histology, tumor grade, FIGO, and lymph node stage were well balanced between the two treatment groups. Relevant tumor characteristics at the time of EBRT are presented in Table 3.

The features of the target volumes did not differ between the two groups; using the same EBRT technique, no significant difference was found between the coverage of PTV45, PTV-N, CTV-N_PIL, or CTV-N_PAO of the two groups, as shown in Table 4.

While the CTVHR volumes were similar, the D90 dose to them at BT was significantly higher in the MRI-based IGABT group than in the non-MRI-based group; the CTVHR D98 doses showed a strong trend in favor of the MRI-based technique (Tables 5, 6). Likewise, the cumulative doses to the CTVHR by means of both D90 and D98 were significantly higher in the MRI-based treatment group than in the other group (Table 7). The Point A target doses, however, did not differ between the two groups (Table 6). In the MRI-based IGABT group, an average EQD2 D98 dose of 8.01 ± 1.15 (5.3–10.2) Gy per BT session and 90.08 ± 11.99 Gy cumulative dose was delivered to the GTVres (Tables 6, 7). When the doses to the CTVHR and GTVres within the MRI-based IGABT group were compared according to the BT technique, no difference was found between the IC only and IC/IS cases.

Next, the same dosimetry parameters of the MRI-based IGABT cohort were studied in chronological order; an apparent increase of doses by time was found (Figure 3). The prespecified OAR doses (the EQD2 doses to 0.1 cm³ and 2.0 cm³ volumes of the bladder, rectum, sigmoid, and bowel) are demonstrated in Tables 8, 9. The doses to the bladder, rectum, and sigmoid did not differ between the two groups. Interestingly, significantly less dose to the bowel was experienced in Group 2, both if the dose per BT fraction and the cumulative dose after EBRT and BT was considered (Tables 8, 9). OAR doses in the MR-based IGABT cohort did not fluctuate over time.

Regarding treatment-related side effects, the most frequent acute toxicity of any grade was nausea, leucopenia, and neutropenia in both groups. Interestingly, thrombocytopenia of grade 1 occurred significantly more in Group 2 (Table 10). There were no serious grade 3–4 adverse events registered (Table 10). No cessation of the curative treatment course was necessary due to severe toxicity.

In this exploratory analysis of our first clinical experience, we found that the MRI-based IGABT technique is appropriate both for individual dose escalation and control of OAR doses and toxicities in locally advanced cervical cancer. This should be the preferred approach, especially in cases with large and/or irregular extent post-teletherapy residual tumors.

The studied IGABT technique based on multimodality imaging, modern radiobiology knowledge, and novel teletherapy and BT technologies is a rational approach but requires effort from the whole medical staff to adopt special skills and make a series of labor-intensive key steps in patient-management. For this new strategy, invaluable help is provided by the published protocol of the EMBRACE-II international clinical study [16]. This book of precision on CRT and IGABT provides information on all challenges, including complex patient management, 3D image guidance, delineation and treatment planning, and proper BT applicator selection and arrangement. Furthermore, the EMBRACE team and Gyn GEC ESTRO collaborative group have been making efforts in the last two decades to disseminate the fundamentals of the aforementioned comprehensive concept through special training programs and within the frame of the ongoing EMBRACE-II international multi-center clinical study [17].

Due to the progression attributes of cervical carcinoma, sufficient loco-regional dose delivery is essential for tumor control. The review of Tanderup et al based on a large amount of clinical data demonstrated a strong correlation between the delivered dose to the target volumes and clinical outcome [18]. The importance of dose escalation is also supported by the results of the EBRACE-I study, achieving an excellent 5-year local control rate of 92% for the overall patient cohort [10]. Meanwhile, the complete remission rate was 98%, and only 98/1,318 patients (7.4%) had local failure in the entire study population. Notably, 90% of these cases showed relapse within the BT target volume only. According to the report of Schmid et al, higher local failure rates were associated with the presence of tumor necrosis, uterine corpus or mesorectal space infiltration, or a larger initial tumor extent (CTVHR > 45 cm³) [19]. Moreover, an adenocarcinoma or adenosquamos carcinoma histology subtype was associated with a significantly higher risk of local failure. Delivering 85 Gy D90 dose to the CTVHR resulted in a 95% local control for squamous cell carcinomas vs. only 86% for adenocarcinoma/adenosquamous carcinoma cases at 3 years. In accordance with the retroEMBRACE data, overall treatment time (OTT) was also found to be an independent risk factor on clinical outcome [19]. It is envisioned that the prospective follow-up data of EMBRACE II will not only point to the role of radiation dose in improved outcome but will demonstrate the feasibility of individually chosen dosing according to stage and tumor-related biomarkers in line with the importance of long-term quality of life (QOL) consideration [2]. In other words, dose escalation and dose de-escalation will become individually chosen strategies in modern practice.

Our first series of patients and clinical data in this paper represent the efforts and learning period we went through during the integration of the modern IGART technique into our institutional practice. First, we systematically introduced the bladder protocol and rigid image fusion of the multiple series of treatment planning CT and diagnostic pelvic MRI and PET/CT scans for EBRT contouring, including the recommended complex volume of interest template, into the planning system. Accordingly, all patients received risk-adapted elective pelvic irradiation and SIB to pathological lymph nodes if applicable using IMRT planning aims; daily cone-beam CT verification with correction to bony anatomy allowed for halving former PTV-CTV margins of 10 mm. Great attention was given to both teletherapy and chemotherapy as the basis of high-dose delivery at IGABT due to down-staging. Regarding the implementation of MRI-guided IGABT, the greatest challenges were ensuring the infrastructure of anesthesia, appropriate patient transport, adequate MRI imaging, and applicator reconstruction during the first couple of combined IC/IS treatment sessions. Handling the early target delineation and planning optimization uncertainties, we could gradually improve the BT target dose utilizing MRI guidance, resulting in a mean dose to the D90 CTVHR higher than 85 Gy. Although in the first few cases our main goal was to safely remain within the limits of OAR hard-dose constraints, later on we simultaneously aimed at the fulfillment of OAR soft-dose constraints and the planning aims of D90 CTVHR (90–95 Gy) and D98 GTVres (>95 Gy). The only explanation why the target dose was higher in the MRI-guided IGABT group could be that by having had identified the residual tumor and delivered a prespecified radiation dose to it, the dose to the CTVHR was unintentionally increased too. Our experience on eight cases (16,0%) having received BT with interstitial needles was favorable, indicating that the combined IC/IS technique is the adequate solution to serve this dual goal. The analysis in the retroEMBRACE study suggested that the simple IC BT technique has limitations in CTVHR dose coverage, especially in cases with larger residual tumor volumes (CTVHR > 30 cm³), hence resulting in suboptimal local control [20]. Another retroEMBRACE analysis by Fokdal et al. demonstrated that by using the combined IC/IS technique, the D90 CVTHR dose could be elevated by 9 Gy EQD₂ and consequently significantly higher local control rates were found in patients with residual CTVHR > 30 cm³ volumes without increasing toxicity as compared to the use of the sole IC BT technique [21]. In fact, dedicated IC/IS applicators represent an excellent tool for the implementation of personalized treatment plans with unique complexity, allowing appropriate target coverage even if large residual tumors or parametrial infiltration are present [22, 23].

We have prospectively collected a set of acute toxicity measures possibly related to CRT; these were easily controlled and no serious side effects resulting in the cessation of therapy occurred. Due to the short follow-up time, the analysis of late toxicity is out of scope of this paper. Regarding long-term toxicity related to the described modern approach, the association between the delivered dose and the risk of moderate to severe chronic toxicity is highly supported by the results of the EMBRACEh-I study. In terms of urinary morbidity, the dose to the bladder of D2cm³ was correlated with the risk of grade 2 or higher fistula, cystitis, or bleeding; an increase of that dose from 75 Gy to 80 Gy almost doubled the risk of grade ≥2 cystitis [24]. Beyond rectal and bowel D2cm³ BT doses, the incidence of grade ≥2 diarrhea was associated with higher prescribed EBRT doses (45 vs. 50 Gy), total body V43 Gy values, and pathological lymph node SIB volumes [24]. Attention to vaginal toxicity has recently come into view [25]. Vaginal doses show extreme diversity depending on both the teletherapy and BT techniques. Vaginal dose de-escalation is crucial for the prevention of grade ≥2 vaginal stenosis; the EMBRACE II protocol puts great emphasis on the documentation and control of the vaginal dose, assuming a maximum dose of 65 Gy EBRT + BT EQD2 to the ICRU recto-vaginal point. In addition, the doses to the mid and lower parts of the vagina must be assessed using the Posterior-Inferior Border of Symphysis (PIBS) dose points [26]. It seems advantageous to use a ring and tandem applicator instead of a tandem and ovoids for this vaginal dose de-escalation approach [27] Interestingly, Mohamed et al. demonstrated that significant vaginal dose de-escalation may be achieved by reducing the dwell times in the ring or ovoids while increasing that in the tandem/interstitial needles without compromising the D90 CTVHR ≥85 Gy target dose [28].

Despite the excellent local control rates with the systematic utilization of IGART, with cases of advanced-stage disease, distant recurrence has still been a significant problem in light of the modest overall survival (OS) data. As a result, the need for progress has emerged recently in the field of novel systemic therapies. Currently, many clinical trials are ongoing using immunotherapy in addition to the standard CRT approach. Human papillomavirus (HPV)-associated cervical cancers are considered to be immunogenic tumors with increased neoantigen formation, high somatic mutation rate, and enhanced programmed death-ligand 1 (PD-L1) expression [29, 30]. Indeed, immunotherapy is now a part of standard care of metastatic/recurrent cervical carcinoma. Interestingly, the first phase III CALLA study failed to meet its primary endpoint when adding durvalumab to standard CRT in locoregional cervical carcinoma; progression-free survival (PFS) was not improved in the investigational arm [31]. In contrast, the phase III ENGOT-cx11/KEYNOTE-A18 study - beyond PFS advantage - resulted in a significant increase in OS: at 3 years OS was 82.6% for pembrolizumab plus CRT vs 74.8% for CRT alone [32]. Another approach, the use of induction chemotherapy prior to CRT, was tested in the phase III GCIG INTERLACE trial; a significant 9% PFS and 7% OS absolute benefit was demonstrated at 5 years of follow-up in favor of induction chemotherapy followed by CRT compared to standard CRT alone [33]. The results of two later studies are encouraging and will probably exert a great impact on the standard of care of locally advanced cervical cancer in the near future. In addition, beyond traditional risk evaluation, the integration of biomarkers serving tailored management is awaited; notably, molecular biomarkers are being investigated in a Translational Research sub-study by EMBRACE as well [34].

The protocol used by us has been widely used within and outside of the EMBRACE network. In accordance with our modest experience, a report on a larger experience of 392 patients over 10 years has been published [35].

In summary, we found the IGABT approach as a key element of IGART feasible and adequate for tumor dose-escalation and OAR protection at the same time in the management of cervical carcinoma.