Hypofractionated radiotherapy with hyperthermia in radiation-associated and in-volume recurrent soft tissue sarcomas of the extremities and trunk wall: results of a proof-of-concept prospective trial

Abstract

Objective

Radiation-associated sarcomas (RAS) and in-volume recurrent sarcomas (IRS) are significant therapeutic challenges due to prior radiotherapy (RT) exposure and the risks associated with re-irradiation. The aim of this study was to evaluate the safety and efficacy of moderately hypofractionated RT combined with locoregional hyperthermia (HT) in patients with RAS/IRS.

Methods

A prospective, proof-of-concept, single-arm, phase II clinical trial (NCT04398095) was conducted in which 20 patients with locally advanced RAS/IRS of the extremities and trunk wall were enrolled. Treatment consisted of RT at 36 Gy in 12 fractions over three weeks, combined with at least weekly HT sessions. Unresectable tumors received an integrated boost to 42 Gy. The primary endpoint was grade 3 or higher late toxicity according to the Common Terminology Criteria for Adverse Events version 5.0 at 18 months, while secondary endpoints included local control, progression-free survival (PFS), and cancer-specific survival (CSS).

Results

Median follow-up was 48.9 months. No patient developed grade ≥ 3 late toxicity. Seven patients (35%) experienced local progression. No local progression was observed in the unresectable group treated with integrated boost. All tumors showed radiologic response (partial or stable disease). We observed up to 10% viable tumor cells in the postoperative specimens of five of the 12 patients who underwent surgery. One- and two-year PFS were 70% and 45%, respectively; CSS was 95% and 90%, respectively.

Conclusions

Moderately hypofractionated RT with HT is a safe approach for RAS/IRS reirradiation. Dose escalation may improve local control and warrants further investigation in larger studies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13014-025-02772-y.

Introduction

Radiation-associated sarcomas (RAS) are rare malignancies that may be related to previous radiotherapy (RT) [1]. RAS account for approximately 3% to 6% of all diagnosed sarcomas [2–5]. The risk of developing RAS over a 10-year period is 0.03% to 0.2% [2]. Another challenging clinical scenario is local recurrence after prior perioperative RT for localized soft tissue sarcomas. The combination of RT and surgery significantly improves local control compared to surgery alone to over 90% in extremity and trunk wall sarcomas [6, 7]. However, 5–10% of patients still experience local recurrence.

Due to the rarity of both RAS and recurrent sarcomas inside the treatment volume (IRS), there are no established guidelines or prospective clinical trials addressing their management. For non-metastatic cases, the only potentially curative approach is radical resection with wide negative margins [8]. While chemotherapy and targeted therapy are options for metastatic disease, their efficacy in the treatment of localized RAS/IRS remains uncertain, except for the benefit of taxanes, doxorubicin, liposomal doxorubicin, and ifosfamide in radiation-associated angiosarcoma [9].

The previous dose restricts the possibility of re-irradiation. Methods that enhance the local efficacy of RT can be used to overcome this issue and allow good local control despite a lower total RT dose. Hyperthermia (HT) may improve the effectiveness of RT and chemotherapy [10]. Regional HT (39–43 °C for ~ 60 min) is one of the most potent clinically validated radiosensitisers and can improve tumour control at lower equivalent dose in 2 Gy fractions (EQD2) by adding direct cytotoxicity and enhancing the effectiveness of each subsequent fraction. HT enhances radiosensitivity through three complementary mechanisms: (i) inhibition of DNA damage repair, most notably the transient degradation of BRCA2, which results in the blockade of homologous recombination and impaired RAD51 focus formation; (ii) functional reoxygenation, via increased tumour perfusion and a reduced hypoxic fraction; and (iii) protein denaturation and proteotoxic stress, which compromise cellular recovery after irradiation. These effects increase the lethal yield per Gy, making them particularly relevant for hypoxic, previously irradiated and intrinsically radioresistant sarcomas [11, 12]. Clinical studies, particularly those focusing on RAS, have demonstrated the feasibility of reirradiation combined with HT, achieving high response rates and durable local control [13]. This supports its use when further dose escalation is not possible. One study analyzed a cohort of patients with RAS who underwent moderately hypofractionated RT combined with locoregional HT [14]. The most common fractionation schedules were 32 Gy in eight fractions delivered twice weekly for four weeks and 36 Gy in twelve fractions delivered four times weekly for three weeks. RT was combined with HT once or twice weekly. Sixteen patients were treated between 1979 and 2009, most for unresectable disease. Median survival was 15.5 months, and among the twelve evaluable patients, the response rate was 75%, with seven patients achieving a complete response. Late toxicities were observed in seven patients (44%) and included hyperpigmentation, telangiectasia and fibrosis. The rationale for integrating HT with other modalities in sarcomas is also supported by the results of the randomized EORTC 62,961/ESHO RHT-95 trial led by Issels et al. In that study, patients with high-risk localized or locally advanced soft tissue sarcomas were randomized to receive neoadjuvant chemotherapy with or without regional hyperthermia. The addition of hyperthermia significantly improved local progression-free survival (hazard ratio 0.58; p = 0.003) and disease-free survival, with acceptable toxicity and no increase in surgical morbidity. Long-term follow-up confirmed a sustained benefit in local control and overall survival [15, 16].

We hypothesized that moderately hypofractionated RT combined with regional HT is a feasible treatment modality for patients with localized RAS/IRS. The aim was to develop a safe regimen that provides benefit to patients facing these challenging scenarios. Thus, based on a previously published retrospective study, we proposed a three-week regimen of 36 Gy in twelve fractions with HT at least once a week, preferably twice a week, with or without simultaneous integrated boost to 3.5 Gy per fraction depending on tumor resectability [14].

Therefore, we report the results of the first clinical trial designed for patients with RAS/IRS.

Materials and methods

We conducted a prospective proof-of-concept phase II, open-label, single-arm clinical trial (NCT04398095). Ethical approval for this study was obtained from the Institutional Ethics Committee of Maria Sklodowska-Curie National Research Institute of Oncology, approval number 74/2019.

Inclusion criteria

We recruited patients with locally advanced pathologically-proven RAS/IRS localized to the extremities or trunk wall. Indications for preoperative or definitive RT, resectability, and operability were discussed on a case-by-case basis at the Sarcoma Multidisciplinary Tumor Board (sMTB). Eligible patients were 18 years of age or older with Eastern Cooperative Oncology Group (ECOG) performance status 0 to 2. All patients provided written informed consent.

The criteria for diagnosis of RAS were based on adapted criteria from Huvos et al.: (1) prior exposure to RT; (2) tumor growth within the irradiated volume; (3) latency between treatment and tumor onset greater than one year; and (4) exclusion of cancer predisposition syndromes such as Li-Fraumeni [17]. IRS was defined as recurrence after prior perioperative RT and surgery within the previous planned target volume and 1 cm margin in all directions.

The major institutional criteria of resectability included extracompartmental extension of the tumor, involvement of the bone or major vessels that may require vascular reconstruction, an extension of the tumor through natural foramina, or technical difficulties due to the tumor volume or its anatomical localization. Previous surgeries and systemic treatment for RAS/IRS were allowed. The study used the FNCLCC (Fédération Nationale des Centres de Lutte Contre le Cancer) grading system.

Exclusion criteria

We excluded patients with distant metastases, patients with lymph node involvement, and patients with contraindications to RT or HT. Pathologic diagnoses of bone sarcoma, round cell sarcoma, embryonal or alveolar rhabdomyosarcoma, gastrointestinal stromal tumor, dermatofibrosarcoma protuberans, and aggressive fibromatosis were excluded. Contraindications to RT included significant late toxicity after prior radiation as judged by the investigator. Contraindications to HT included implanted medical devices or objects in the vicinity of the heated volume, significant tumor-related pain that was not controlled with medication, severe heart or lung disease, uncontrolled hypertension, and myocardial infarction or cerebrovascular event less than six months prior.

Treatment schedule

Baseline evaluation included a sMTB, physical examination, hematologic and serum biochemistry tests, and contrast-enhanced computed tomography (CT) scans of the chest, abdomen, and pelvis to exclude distant metastases. In addition, gadolinium-enhanced magnetic resonance imaging (MRI) of the RAS/IRS was performed or, if contraindicated, contrast-enhanced CT of the tumor was used as an alternative. If available, the previous electronic RT plan was obtained and uploaded into the treatment planning system for further evaluation and registration.

The patient then underwent RT planning and began treatment within two weeks. The treatment regimen is presented in Fig. 1. Surgery for resectable RAS/IRS was scheduled between six and eight weeks after completion of RT. Patients with unresectable tumors were referred for follow-up.

Fig. 1.

Please change as dot

Radiotherapy

Previous clinical trials of moderate hypofractionation have shown promising local efficacy and good tolerability. Preclinical radiobiological data by Haas et al. and Thames et al. suggest that soft tissue sarcomas may have a relatively low α/β ratio, typically estimated between 2 and 5 Gy, implying a greater sensitivity to fraction size compared to many carcinomas [18–20]. Consistent with previous studies, we used an alpha/beta ratio of 4 Gy for STS [21–23]. Therefore, we decided to use one of the regimens evaluated in the study published by Jong et al., namely 36 Gy in 12 fractions delivered over three weeks (four fractions per week) with preferably six HT sessions. This corresponds to EQD2 of 42 Gy [14]. In the case of unresectable tumors, we used a simultaneous integrated tumor volume boost of 3.5 Gy per fraction. This allowed an EQD2 of up to 52.5 Gy. The addition of HT should radiosensitize tumor cells and compensate for the lower EQD2 than is typically used in primary tumors [14, 24].

Immobilization and bolus use were determined on an individual basis [25, 26]. The gross tumor volume (GTV) was delineated by two sarcoma-experienced radiation oncologists on the planning CT co-registered with either contrast-enhanced MRI or diagnostic contrast-enhanced CT. As a standard approach, the clinical target volume (CTV) was defined by expanding the GTV by at least 2 cm in all directions for deep-seated tumors and by at least 3 cm longitudinally for extremity STS. For superficial RAS/IRS, the recommended CTV margins were at least 3 cm circumferential and 0.5 cm deep. The CTV was further refined to correspond to anatomic boundaries relevant to tumor spread, including bone, fascia, and critical organs. However, given the variability of clinical presentations, the final determination of CTV was based on a risk-benefit assessment based on PTV coverage and a risk of toxicity and required the consensus of at least two radiation oncologists during a common case discussion. All delineated target volumes and organs at risk were independently reviewed by the third radiation oncologist. The planning target volume (PTV) was generated by expanding the CTV with an additional safety margin of 0.5–1.0 cm. The simultaneous PTV boost volume was generated by expanding the GTV by the same safety margin without adding an elective volume margin.

We used highly conformal RT techniques with dose intensity modulation and daily image guidance (cone beam CT) to deliver the prescribed dose. The dose was prescribed on mean PTV. If a prior electronic RT plan was available, we performed deformable registration, dose mapping and dose calculation using Varian Velocity 4.0 software (Varian Medical Systems). The acceptance of treatment plans was at the discretion of the radiation oncologist on a case-by-case basis rather than based on pre-specified constraints.

Hyperthermia

We treated patients using a Celsius TCS locoregional deep HT system. Heat energy is delivered via electromagnetic waves at a frequency of 13.56 MHz using capacitive coupling for efficient energy transfer [27]. A major limitation of the Celsius TCS system is the lack of direct tumor temperature monitoring; however, this has been mitigated using predefined treatment protocols tailored to the anatomical location and individual patient parameters. The technical details regarding the device and description of protocols were provided in the Supplement A. A dedicated team of radiation therapists administered HT, ensuring strict adherence to protocols while closely monitoring hyperthermia-related adverse events and any deviations in each session. In the case of poor tolerance or any other deviations, treating physician may decide to reduce heating power or interrupt session. All such events were recorded. Sessions were performed at least once a week with the goal of treating patients twice a week, with a minimum of 48 h between treatments to avoid theoretical thermotolerance. To maximize therapeutic efficacy, each HT session was followed by an RT fraction within one hour of completion of heating. Delivering radiotherapy (RT) within ~ 1 h after hyperthermia (HT) therapy leverages (1) transient DNA repair blockade, (2) improved oxygenation and (3) maximized tumor-cell killing at a given RT dose, while keeping toxicity acceptable. Furthermore, a similar timing strategy was employed in another trial, yielding favorable results and tolerance [24].

Surgery

An experienced sarcoma surgical team performed tumor resection with the primary goals of achieving negative margins, preserving limb function when applicable, and ensuring an acceptable functional outcome. Reconstructive surgery was permitted after resection.

Follow-up

Recommended imaging included contrast-enhanced CT of the chest, abdomen, and pelvis, as well as MRI or CT of the tumor site. Follow-up visits were scheduled every three months for the first two years, every six months for the subsequent three years, and annually thereafter. Due to the outbreak of COVID-19 pandemics in 2020, we also allowed teleconsultations, but the preferred way was still a physical visit to our center.

Primary endpoint

The primary endpoint of the study was the rate of Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 grade 3 or higher late toxicity related to reirradiation or HT at 18 months after completion of treatment [28]. Late toxicity was defined as occurring at any time after 90 days from the start of RT with HT. The primary endpoint was considered clinically reliable only in patients who completed 18 months of follow-up (physical visit) or who developed such an event earlier.

Secondary endpoints

Secondary endpoints calculated for all enrolled patients included one-year and two-year time to local progression (TTLP), one-year and two-year progression-free survival (PFS), one-year and two-year cancer-specific survival (CSS), and local control rate at the moment of analysis.

Local progression was defined as the radiographic evidence of disease progression in the irradiated volume, confirmed by a board-certified musculoskeletal radiologist. Additional endpoints included radiologic response, pathologic response and acute toxicity. Radiologic response was assessed in all patients according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 [29]. Pathologic response was assessed according to the recommendations of the European Organization for Research and Treatment of Cancer-Soft Tissue and Bone Sarcoma Group (EORTC-STBSG) [30]. It was expressed as a percentage of the total tumor area stained with hematoxylin in representative tissue slabs. Response scores were classified into five different categories: (A) no stained tumor cells; (B) single stained tumor cells or small clusters of cells or small clusters (< 1% of the total sample); (C) 1% to < 10% stained tumor cells; (D) 10% to < 50% stained tumor cells; and (E) 50% stained tumor cells. Grades A, B, and C were considered good pathologic response.

Statistical methods

Due to the rarity of RAS/IRS and the innovative nature of the study, the sample size was pragmatically set at 20 patients based on institutional historical data. Median follow-up was estimated by Kaplan-Meier analysis using the inverse of the status indicator. TTLP was calculated from enrollment to last follow-up (censored), death (censored), or local progression (event). PFS was calculated from enrollment to last follow-up (censored), progression in any location (event), or death (event). CSS was calculated from enrollment to last follow-up (censored), information that a patient was alive (censored), death from other causes (censored), or cancer-related death (event). For cancer-specific survival (CSS), deaths from other causes were treated as competing events. Death from unknown causes was treated as death from disease progression. Missing data on date of death were obtained from the National Cancer Registry when available. To calculate the above parameters, we used the date of the last known status as the date of the last follow-up, regardless of the physical visit to our center. Data analysis was performed using the R software environment, version 4.3.2 (R Foundation for Statistical Computing, Vienna, Austria) and the jamovi project, version 2.5.4.0 (obtained from https://www.jamovi.org, Sydney, Australia). Analyses were performed using the following packages: survival (v3.5-7) for Kaplan–Meier estimates, survminer (v0.4.9) for visualization, and cmprsk (v2.2-11) for competing risk analysis.

Results

Patient and treatment data

Twenty patients were enrolled between November 2019 and August 2021. Complete patient data are available in Supplement B. Patient characteristics are shown in Tables 1 and 2. Twelve of the enrolled patients were diagnosed with IRS, while the remaining eight patients were diagnosed with RAS. The most common pathologic diagnoses were synovial sarcoma (n = 4), undifferentiated pleomorphic sarcoma (n = 4), and angiosarcoma (n = 3). All tumors were intermediate- or high-grade. At the sMTB, 15 tumors were deemed resectable or marginally resectable and five were deemed unresectable and referred for integrated boost. In 14 patients, the current RT plan could be co-registered with previous RT plans and relevant deformable dose calculations could be performed. The summary of registrations is available in Supplement B. All patients completed RT without interruptions and underwent at least four HT sessions. Details regarding HT parameters and hyperthermia-related adverse events are also available in Supplement B.

Table 1.

Characteristics of patients with radiation-associated sarcomas

Patient number	Sex	Age at diagnosis [years]	Site	Description of previous RT	Time interval from prior RT [years]	Pathologic diagnosis	Radiological maximum tumor diameter [cm]	Group
RAS1	F	33	Chest wall	15 × 2.67 Gy + boost 8 × 2 Gy 2017 breast cancer	3	UPS G2	3.1	No boost
RAS2	F	70	Chest wall	25 × 2 Gy + boost 8 × 2 Gy 2011 breast cancer	9	Angiosarcoma G2	NA	No boost
RAS3	F	68	Chest wall	17 × 2.5 Gy + boost 6 × 2.5 Gy 2012 breast cancer	8	Angiosarcoma G2	NA	No boost
RAS4	F	64	Chest wall	18 × 2.5 Gy + boost 4 × 2.5 Gy 2016 breast cancer	5	Angiosarcoma G2	NA	Boost
RAS5	F	67	Chest wall	20 × 2.75 Gy 2014 breast cancer	7	UPS G3	5	No boost
RAS6	F	85	Neck and supraclavicular area	22 × 1.8 Gy 2011 thyroid DLBCL	10	UPS G3	6.7	Boost
RAS7	F	58	Arm	25 × 2 Gy + boost 5 × 2 Gy 2011 breast cancer	10	MPNST G3	5	No boost
RAS8	M	60	Chest wall and arm	25 × 2 Gy 2013 aggressive fibromatosis	8	UPS G2	13	No boost

DLBCL - diffuse large B-cell lymphoma, F - female, G - grade, M - male, MPNST - malignant peripheral nerve sheath tumor, NA - not applicable (in angiosarcomas due to the character of the disease), RT - radiotherapy, UPS - undifferentiated pleomorphic sarcoma

Table 2.

Characteristics of patients with in-volume recurrent sarcomas

Patient number	Sex	Age at diagnosis [years]	Site	Description of previous RT	Pathologic diagnosis	Radiological aximum tumor diameter [cm]	Group
IRS1	M	69	Groin	25 × 2 Gy 2017 postoperative	Myxoid liposarcoma G3	5	No boost
IRS2	F	67	Arm	5 × 5 Gy 2015 preoperative	Synovial sarcoma	4.3	No boost
IRS3	F	34	Foot	28 × 2 Gy 2018 postoperative	Clear cell sarcoma	2.2	No boost
IRS4	M	73	Arm (stump)	5 × 5 Gy 2014 preoperative	Myxofibrosarcoma G2	7.2	No boost
IRS5	F	74	Forearm	5 × 5 Gy 2019 preoperative	Myxofibrosarcoma G2	7	Boost
IRS6	F	73	Groin, pelvis	5 × 5 Gy 2008 preoperative	Synovial sarcoma	16	Boost
IRS7	M	55	Knee	5 × 5 Gy 2015 preoperative	Synovial sarcoma	3.5	No boost
IRS8	M	53	Forearm	5 × 5 Gy 2019 preoperative	Myxofibrosarcoma G2	1	No boost
IRS9	F	42	Arm, axilla	5 × 5 Gy 2017 preoperative	Epithelioid sarcoma	1.5	No boost
IRS10	F	71	Forearm	33 × 2 Gy 2014 postoperative	Sarcoma NOS G3	8	No boost
IRS11	M	75	Knee	5 × 5 Gy 2014 preoperative	MFS G3	3.6	No boost
IRS12	F	73	Shoulder	23 × 2 Gy + boost 5 × 2 Gy 2017 postoperative	Synovial sarcoma	5	Boost

F - female, G - grade, M - male, MFS - myxoinflammatory fibroblastic sarcoma, NOS - not otherwise specified, RT - radiotherapy

Twelve of 15 patients underwent scheduled surgery. Two did not undergo surgery due to COVID-19 outbreak and one refused the operation due to complete clinical response after RT + HT.

Toxicity

The median follow-up for the whole group reached 48.9 months (44.8–58.6 months, 95% confidence interval, CI). Of all the patients enrolled (n = 20), four did not reach 18 months of physical follow-up in our institute. Two of them died of the disease and two of them decided to continue follow-up in local oncology departments because of the risk of COVID-19 infection. However, one of the deceased patients developed grade 1 superficial soft tissue fibrosis. Therefore, we do not have a long enough observation period for three patients. Among the remaining 17 patients who reached 18 months of physical follow-up, none developed grade 3 or higher late toxicity. Two patients with chest wall angiosarcoma developed persistent grade 2 skin ulceration that were managed with topical therapy and dressings. Four patients, including one described above, developed grade 1 superficial tissue fibrosis. Two patients had decreased joint range of motion, grade 1 and 2, respectively. Acute toxicities were mild (see Supplement B). The notable perioperative acute toxicity was one severe wound complication, namely persistent infection followed by arm amputation.

Radiologic and pathologic response

All irradiated and heated tumors showed positive radiological response to the treatment, namely stable disease in 11 cases, partial response in five cases and complete response in four cases. Among patients who underwent surgery (n = 12), we observed good pathological responses in five patients (grade A, n = 2; grade C, n = 3; grade D, n = 1; grade E, n = 6).

Local control and time to local progression

Three of 15 patients scheduled for surgery did not undergo the procedure, and all developed local progression (see Table 3). Four additional patients from the twelve who underwent surgery also experienced local progression. In a group of patients with unresectable tumors who underwent RT + HT with simultaneous integrated tumor volume boost up to 42 Gy, none experienced local progression. One and two-year TTLP were 94.7% and 61.8%, respectively.

Table 3.

Local recurrences and their management

Patient number	Treatment group	Surgery	Physical follow-up time* [months]	Time to local progression [months]	Management of local progression	Synchronous metastases
H1	No boost	No (refused due to complete response)	35.5	18.5	Conservative surgery	No
H2	No boost	Yes	61	23	Amputation	No
H3	No boost	Yes	58.7	23.2	Conservative surgery	No
H5	No boost	No (prolonged COVID-19 infection)	24.9	13.5	Best supportive care	No
H9	No boost	No (COVID-19 outbreak, patient refused hospitalization)	18.9	15.3	Refused amputation, chemotherapy and further follow-ups	Yes
H10	No boost	Yes	25.3	6.9	Chemotherapy	No
H11	No boost	Yes	51	20.5	Surgery	No

* The last time the patient was physically present in our center

Survival

Five of the 20 patients enrolled in the study died of disease progression. Status and missing data on deaths of patients lost to follow-up were obtained from the National Cancer Registry. One and two-year PFS were 70% and 45%, respectively (Fig. 2). One and two-year CSS were 95% and 90%, respectively (Fig. 3).

Fig. 2.

Please change as dott

Fig. 3.

Please change as dottt

Discussion

Our study prospectively evaluated the safety of moderately hypofractionated RT combined with locoregional HT in patients with RAS/IRS localized to extremities or trunk wall. We found that the proposed approach is safe in terms of late toxicity. To the best of our knowledge, no prospective studies on this topic have been published in the literature. Therefore, direct comparison of results with other studies is obviously not possible.

The only study on which we base our fractionation regimen reported similar toxicity results to those mentioned in the introduction [14]. One may question whether arm amputation due to infection is an acceptable toxicity. However, this type of serious wound complication could also occur after initial RT [31, 32]. Moreover, when considering re-irradiation, we assume that the worst toxicity is the tumor itself, and any kind of RAS/IRS local recurrence after surgery could be much more destructive. Finally, RT may allow an attempt at surgery in marginally resectable tumors. Interestingly, RAS presented the largest reirradiated volume among all patients included in the study (see Supplement B). This may explain the significant perioperative toxicity. It is indeed noteworthy that patients with previously heavily irradiated areas (often exceeding 50 Gy) did not develop significant late effects even after receiving an additional 36 Gy in 12 fractions, in some cases with a boost. It may be explained by the proper selection of patients, namely the exclusion of patients with significant late toxicity from previous RT. Moreover, further follow-up is warranted due to the occurrence of late adverse events even many years after redirradiation.

The important point is the high frequency of local progression after 36 Gy in twelve fractions without simultaneous integrated boost, namely in seven of fifteen patients scheduled for surgery. Notably, none of the patients with unresectable tumors who underwent a regimen with boost developed local progression despite more locally advanced tumors that precluded surgery. This may suggest a threshold dose of 52.5 Gy for GTV, assuming an alpha/beta ratio of 4 Gy. However, one patient with extensive non-metastatic unresectable recurrent synovial sarcoma with severe invasion of the femoral and external iliac veins died six months after treatment with boost despite a partial response at first follow-up (Supplement B). The most likely cause of death was an embolic event. However, local progression cannot be excluded. The problem of radiosensitivity of RAS/IRS has not been well studied and our study may provide valuable suggestions, especially considering the unexpectedly high rate of complete responses after reirradiation [33].

The study has limitations. The first limitation is the small sample size and the heterogeneity of the group. Nevertheless, the main objective of the study was to evaluate the safety of reirradiation using moderately hypofractionated RT with HT in a group of very rare entities, and we achieved this objective. Second, the COVID-19 outbreak might have influenced the quality of the data due to loss to follow-up, lack of surgery, and telephone rather than physical visits. However, this was unavoidable, and we tried to minimize the impact of the pandemic on the quality of the trial. Finally, data on the optimal use of HT in STS remains very limited. Treatment protocols, energy settings, and other heating parameters were determined by the HT team based on clinical experience rather than standardized objective criteria, as no formal guidelines currently exist. In addition, intratumoral temperature during HT could only be estimated rather than accurately measured. We followed predefined Celsius TCS protocols without direct temperature control, as the necessary temperature sensors were not available for this device. A potential improvement for future studies could be the integration of magnetic resonance thermometry or alternative temperature measurement techniques. Finally, in our study, late toxicities were evaluated using the current international standard for systemic and mixed-modality oncology trials, namely CTCAE version 5.0. However, most studies assessing late local toxicity after RT for extremity or truncal sarcoma have traditionally used the RTOG/EORTC Late Radiation Morbidity Scoring Scheme and the Stern scale for lymphoedema. These scales emphasise functional and site-specific outcomes (fibrosis, joint stiffness, oedema and skin induration) and may be therefore better suited to comparing local toxicity profiles between RT regimens. Consequently, our CTCAE-based assessment may underestimate or classify differently certain late effects, particularly low-grade fibrosis or limb swelling. This could lead to limited comparability with earlier studies. Future studies should therefore consider using the RTOG and CTCAE criteria in parallel to ensure alignment with the existing sarcoma literature.

Given that regional HT can potentiate both radiotherapy and chemotherapy, combining HT with active agents in chemosensitive histologies (e.g., angiosarcoma) is biologically plausible and may be considered on a case-by-case basis within a multidisciplinary team. However, our trial evaluated RT + HT without chemotherapy; therefore, evidence for adding chemotherapy to HT in RAS/IRS remains limited and warrants prospective study.

This study does not establish a new standard of care for RAS/IRS or confirm the long-term efficacy of the proposed regimen. Rather, it demonstrates that moderately hypofractionated RT with HT has a favorable toxicity profile and suggests the dose threshold required to achieve better local control.

Our results provide hypothesis-generating data suggesting that preoperative and definitive hypofractionated re-irradiation with HT is a safe and effective treatment for RAS/IRS. However, the local control without dose escalation is unsatisfactory, suggesting its importance in the design of further clinical trials and clinical practice.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Konrad Zasadziński: Data curation; Investigation; Formal analysis; Visualization; Writing – original draft. Aneta Borkowska: Data curation; Resources; Writing – review & editing. Dorota Kopeć: Data curation; Resources; Writing – review & editing. Maria Telejko: Resources; Writing – review & editing. Piotr Rutkowski: Validation; Writing – original draft; Writing – review & editing. Mateusz Jacek Spałek: Conceptualization; Data curation; Investigation; Methodology; Resources; Supervision; Validation; Writing – original draft; Writing – review & editing; Visualization.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Data availability

Data is provided within the supplementary information files.

Declarations

Ethical approval

The study was conducted in accordance with the tenets of the Declaration of Helsinki. It does not report the use of experimental or novel protocols. Ethical approval for this study was obtained from the Institutional Ethics Committee of Maria Sklodowska-Curie National Research Institute of Oncology, approval number 74/2019.

Competing interests

The authors declare no competing interests.