30. Paediatric malignancies - The GEC-ESTRO Handbook of Brachytherapy

SECOND EDITION

The GEC ESTRO Handbook of Brachytherapy

PART II: CLINICAL PRACTICE 30 Paediatric malignancies Cyrus Chargari, Petra Kroon, Bradley R. Pieters

Editors Bradley Pieters Erik Van Limbergen Richard Pötter

Peter Hoskin Dimos Baltas

Paediatric malignancies

3

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

30 Paediatric malignancies

Cyrus Chargari, Petra Kroon, Bradley R. Pieters

1. Summary 2. Introduction

3 3 4 4 5 5 6 7

9. Treatment planning

10 11 12 12 13 13 14

10. Dose, dose rate and fractionation

3. Anatomical topography

11. Monitoring

4. Pathology 5. Work up

12. Results

13. Adverse Side Effects

6. Indications and contra-indications

14. Key messages 15. References

7. Target Volume 8. Technique

1. SUMMARY

Brachytherapy is a major tool for treatment of paediatric soft tissue tumours, as part of a multimodal approach. It allows dose escalation to the residual tumour while minimizing organs at risk dose exposure. Main indications are sarcoma of the head and neck area, urogenital tract, limbs and trunk. There are increasing data suggesting that the integration of image-guided brachytherapy will allow for more treatment personalization. Most of the clinical evidence comes from low-dose rate and pulsed-dose rate brachytherapy, but use of high-dose rate brachytherapy is developing, with retrospective series suggesting satisfactory efficacy and toxicity profile. Due to the rarity of paediatric cancer and the difficulty to acquire expertise in these treatments, one prerequisite for paediatric brachytherapy treatments is that the team is familiar with adult brachytherapy procedures. The indications and details for the techniques to be applied in paediatric RMS are mainly derived from the respective chapters in adults, with specific adaptation to children. Centralization to high volume centres is recommended.

2. INTRODUCTION

severe side effects associated with external beam radiotherapy when delivered in very young patients. The dosimetric advantages of BT are unequalled in terms of organs at risk sparing capacity. This superiority of BT over other irradiation modalities is particularly relevant to avoid irradiation of bone structures, especially to growth cartilages. It is also an excellent irradiation modality to perform focal dose escalation. In addition, the integral dose to the patient body is low, potentially minimizing the risk of second malignancy that is a significant concern among very young patients treated with ionizing radiation [2,10). Due to the rarity of paediatric cancer and the difficulty to acquire expertise in these treatments, one prerequisite for paediatric BT treatments is that the team is familiar with adult BT procedures. These include medical as well as paramedical expertise. In addition, it is recommended that paediatric cases are referred to expert centres, to increase patient volume and therefore quality of treatments through a high level of expertise [9, 33]. Few centres have expertise for paediatric BT worldwide and most of the knowledge in this setting comes from a few specialist centres, with retrospective experience mainly based on 2D-guided low-dose rate (LDR) BT procedures. Over the past decade, the developments of stepping source technology and 3D-image guidance concepts have been applied to paediatric BT procedures, enabling an increase in

The incidence of paediatric malignancies is low, with an annual frequency of approximately 15 per 100,000 children up to 15 years. It is however a major cause of death among children. Overall, the most frequent paediatric cancers are leukaemia, brain, and central nervous system tumours, and lymphoma. Altogether, these tumours account for approximately 65 % of paediatric tumours. Other tumour sites are less frequent and mainly represented by bone tumours (5%), soft tissue tumours (7%), neuroblastoma (8%), nephroblastoma (8%), and germ cell tumours (4%). There have been significant improvements in survival for paediatric tumours, in parallel with the development of multimodal strategies including active chemotherapy regimens. It is estimated that the probability of 5-year survival was 80-85% among children and adolescents diagnosed with cancer and treated in the 2010-2016 [32, 34, 50, 51]. Local treatments have a major role in the patient’s cure probability. Their impact in terms of long-term morbidity is however to be considered [55]. Together with surgery and external beam radiotherapy, brachytherapy (BT) has potentially a major role in patient treatment, especially in the context of organ-sparing strategies, to avoid mutilating surgery and also the long-term

Paediatric malignancies

4

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

treatment quality, and possibly minimizing the risk of long-term side effects. Historically, Iridium-192 wires were widely used for interstitial BT procedures in the paediatric population [25]. Since Iridium192 wires are no longer commercially available in most parts of the word, there has been a shift from LDR to either pulsed-dose rate BT (PDR) or high-dose rate (HDR) BT. Due to a theoretical superiority, PDR treatments are preferred, if possible, to minimize the risk of long-term normal tissue complication. An increasing number of recent publications however suggests that HDR treatments can also be delivered with good results, although the number of patients treated is still low. The vast majority of paediatric tumours that can be treated with BT are soft tissue sarcoma, mainly represented by rhabdomyosarcoma (RMS) and the ongoing European Paediatric Soft tissue sarcoma Study Group (EpSSG) FAR-RMS protocol has included a description on BT techniques and indications (clinicalTrial.gov reference NCT04625907). RMS are the most frequent paediatric soft tissue tumours, accounting for 4-6 % of all paediatric cancers. The main indications for BT are tumours arising in the gynaecological, urological, or head and neck area. Tumours of the limbs or the trunk are also good indications for BT, usually as peri-operative treatments. Prior to the 1990’s, there was a high incidence of gynaecological clear cell carcinoma related to post intrauterine diethylstilbestrol exposure and numerous clinical data show the place of BT in this situation, alone or in combination with surgery or external radiotherapy [27, 31]. Since the use of diethylstilbestrol has stopped, these tumours have become excessively rare and are not discussed here. BT may be also indicated in non-sarcoma tumours (ex: rare squamous cell carcinoma or adenocarcinoma, germ cell tumours), alone or in combination with external irradiation and/or surgery. Indications are the same as in adults, though technical aspects of paediatric treatments should be considered, because of obvious differences in organ dimensions. This chapter focusses on the place and treatment modalities for BT in paediatric RMS. For tumour sites suitable for BT, the anatomy and topography of tumours of children are usually comparable to those in adults. However, the dimensions of organs are obviously smaller. This must be considered at time of implant and at time of treatment planning. There is a large amount of clinical data showing that irradiated and treated volume is a major factor in developing complications. Such relationship between these volumes and the risk of long-term treatment-related morbidity was shown in paediatric tumours, especially for vaginal tumours treated with BT [27]. In addition, RMS are usually very large tumours at diagnosis, and therefore relationships between the target volume and the organs at risk may be closer than in adults. There are variations in frequency of the primary sites of the tumours related to the age of the patients. For example, vaginal rhabdomyosarcoma is seen during early childhood, while RMS of the uterine cervix are more frequently seen in teenagers. In the head and neck area there is typically a distinction between 3. ANATOMICAL TOPOGRAPHY

tumours arising from the orbit, non-parameningeal tumours and parameningeal tumours. Important in this regard is the difference in prognosis between these different sites. Parameningeal tumours are those located in the nasopharynx, nasal cavity, parapharyngeal area, paranasal sinuses, infratemporal and pterygopalatine fossa, middle ear, and mastoid. Non-parameningeal tumours are for example located in the oral cavity, ear, cheek, scalp, salivary glands, and neck.

4. PATHOLOGY

RMS are classified into embryonal subtype RMS (~80% of all rhabdomyosarcomas) or alveolar RMS (15-20% of all RMS). Other RMS subtypes are rare in children and adolescents. Alveolar rhabdomyosarcomas exhibit a poorer prognosis. Botryoid sarcoma is a variant form of embryonal RMS, mainly seen in gynaecological, bladder and urinary tract RMS. Embryonal and alveolar RMS show distinct molecular patterns and biology. Thus, balanced chromosomal translocations resulting in PAX3/FOXO1 and PAX7/FOXO1 fusion genes are evidenced in more than 80% of alveolar RMS. The outcome of patients with fusion-gene negative alveolar RMS seems to be like the one for patients with embryonal RMS [57, 69]. Soft tissue sarcoma other than RMS are rare (e.g., synovial sarcoma, extra-osseous Ewing and Ewing-like sarcoma, malignant peripheral nerve sheath tumour, fibrosarcoma, alveolar soft part sarcoma). Though few data are available for BT in these tumour types, mainly based on case reports, these are potential indications for BT and should be discussed on a case per case basis, considering tumour site, the possibility of alternative conservative treatments, technical feasibility of BT, and tumour volume. Various risk factors for paediatric soft tissue sarcoma have been identified. Those are mainly from inherited disorders and include: Li-Fraumeni syndrome, DICER1 syndrome (for gynaecological RMS), RB1 gene changes, neurofibromatosis type 1, SMARCB1 gene changes, intrauterine diethylstilbestrol exposure and a history of radiation exposure. When there is a syndrome associated with a high risk of second cancers (e.g., Li-Fraumeni syndrome), local treatments without irradiation should be prioritized. When irradiation cannot be avoided, BT allows minimizing patient body radiation dose. Patients < 10 years of age more frequently show embryonal RMS, while those ≥ 10 years old more frequently have alveolar RMS. In addition to this correlation, age was found to be an independent prognostic factor in RMS. Thus, patients aged ≥10 years have a poorer prognosis after adjusting on other risk factors, such as tumour stage, histology, and nodal status. It has also been shown that failure-free survival was lower for infants (≤ 1 year), due to a higher risk of local failure. This may be in relation to the fact that complete local treatment including radiotherapy is highly challenging to deliver in these patients and in this context BT has potentially a major role to achieve local control and avoid the long-term side effects of external irradiation.

Paediatric malignancies

5

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

5. WORK UP

Favourable tumour sites in EpSSG FAR RMS are defined as the orbit, the head and neck area (except for parameningeal sites), bladder-prostate, genital and bile ducts. Unfavourable sites include limbs, parameningeal sites, and other parts of the body. Recent data showed an excellent prognosis for patients with BP RMS treated with BT [12, 13]. Risk classification for paediatric RMS is different between European and North American groups. In general, it is based on primary site (favourable/unfavourable), tumour size (< or > 5 cm) and patient age (< or > 10 years), regional lymph node extension, distant metastasis, extent of initial surgical procedures performed prior to systemic treatment (IRS1: Localized tumour, completely removed with pathologically clear margins and no regional lymph node involvement ; IRS2 : localized tumour, grossly removed with (a) microscopically involved margins, (b) involved, grossly resected regional lymph nodes, or (c) both ; IRS3 : Localized tumour, with gross residual disease after incomplete removal, or biopsy only ; IRS4 : Distant metastases present at diagnosis), and biopathology findings (i.e. histology or FOXO1 fusion). According to data from the IRS (International Rhabdomyosarcoma Study)-IV study, the three-year failure free survival rate is 83-86% in IRS group I and II, 73% in group III, and less than 30% in IRS group IV [7]. Radiotherapy has a major role in the treatment of paediatric RMS, in combination with multi-agent chemotherapy regimens [3, 6, 59, 60). Various chemotherapy regimens do exist, according to risk group. BT is indicated for treatment of the tumour residuum, following chemotherapy. Usually, the treatment decision for BT is taken after 4 cycles. However, the local treatment can be delivered after additional cycles, to achieve optimal tumour regression, with regular imaging (every 2 cycles) to monitor tumour response. Main sites that are suitable for BT in paediatric tumours are: gynaecological tumours (vulva, vagina, cervix), urological tumours (bladder, prostate, urethra), perineal tumours, anus-rectum, tumours of the trunk and of the extremities, head and neck tumours (orbit, non-parameningeal tumours, parameningeal tumours). In most cases, local treatment is mandatory in patients with RMS. However, for gynaecological sites, up to 40% of patients can be cured without local treatment, if there is clinical and radiological complete response confirmed by biopsies after four cycles [36, 52]. On the contrary, all BP/RMS patients and parameningeal patients should receive local treatment, as only 10% of BP RMS can be cured without local treatment in historical cohorts [60, 61] BT is indicated as an alternative to external radiotherapy and mutilating surgery, in the context of organ sparing strategies. It is proposed schematically in the following situations: 1/ as exclusive treatment for in situ tumours (that have not been removed by surgery), to treat the post-chemotherapy residuum, to avoid the long-term morbidity associated with external beam radiotherapy in very young patients or with mutilating surgery (e.g., prostate RMS, gynaecological RMS) 6. INDICATIONS AND CONTRA-INDICATIONS

RMS are very chemo-sensitive tumours. It is therefore crucial to have tumour extent assessed prior to any chemotherapy, after surgery (if any), then after every 2-3 chemotherapy cycles, and prior to BT. In addition to clinical examination, which is an important component of workup, radiological assessment for soft tissue rhabdomyosarcoma mainly relies on MRI. Though radiological procedures are specific to each tumour site, MRI provides the most accurate morphological staging to accurately assess tumour extents. Computed Tomography (CT) scans can be also useful to preclude bone invasion, that would contra-indicate solely BT procedures (e.g., for floor of the mouth RMS). In case of bone invasion macroscopic radical surgery in combination with BT is advised; the AMORE procedure (see paragraph Technique). For gynaecological tumours, an examination under general anaesthesia should be performed prior to any chemotherapy if feasible. If tumour volume does not allow the gynaecological examination, for example in a vaginal botryoid RMS, it may be postponed. Colposcopy should be performed as far as tumour regression allows. At the time of colposcopy, a vaginal impression may be performed to properly see tumour extents at the level of the vaginal mucosa, and serve as a basis for customizing a personalized vaginal mould applicator (Figure 1). For bladder and prostate RMS (BP RMS), MRI should be performed with full bladder to better show tumour extent, and select patients that are good candidates for a conservative approach based on BT. Adequate nodal staging is also important at diagnosis (using conventional imaging and FDG PET/CT and if appropriate, sentinel lymph node biopsy), especially for alveolar RMS (and some specific sites such as limb RMS). When patients are referred for adjuvant BT, for example after surgical treatment of a limb RMS, it is crucial to have a complete report and description of the operative findings and to discuss thoroughly with the surgeon, to better define the area that should be treated.

Figure 1. Example of a vaginal mould applicator with four catheters for intracavitary treatment of a vaginal rhabdomyosarcoma

Paediatric malignancies

6

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

7. TARGET VOLUME

2/ in combination with conservative surgery, when radical surgery would lead to mutilation (e.g., bladder RMS, vaginal RMS, limbs RMS, head and neck RMS). For combined treatments, BT is delivered postoperatively or (preferably) as a perioperative approach to guide catheter placement according to operative findings, including extemporaneous analysis. If the BT decision is taken postoperatively, based on histopathological findings, it may be necessary to re-operate to place BT catheters in the tumour bed. 3/ as adjuvant treatment, in cases with adverse histopathological factors (e.g., involvement of resection margins, alveolar histology). 4/ in combination with external beam radiotherapy, to increase focally the dose to area at high risk for local relapse and minimize dose to organs at risk. For residual tumours > 5 cm in size, achieving a good dose distribution with exclusive BT may be complex, and combination of BT and external irradiation and/ or debulking surgery may be discussed. In case of lymph node extension, BT is usually associated with external radiotherapy to treat locoregional tumour extension. Based on adult data, BT is also usually not delivered as exclusive treatment and rather proposed as a boost following external irradiation for tumour sites at high risk for necrosis (e.g., anal canal) [56]. 5/ for salvage treatments. Local treatment is crucial in patients having a local relapse after primary treatment and radiotherapy and/or BT have a major role in this situation. Salvage BT may be indicated when the tumour recurrence occurs in an area that has not been previously irradiated and that is technically accessible to implantation. When the patient has a relapse in a previously irradiated area, this is a very complex situation, and no definitive recommendation can be made here. Treatment should be discussed on a case per case basis according to previous radiation dose, BT feasibility and the possibility of alternative salvage surgery. Only tumours which relapse with small volume disease can be treated with reirradiation through BT. Reports have shown the feasibility and satisfactory outcome of performing salvage treatment with limited surgery and BT for selected cases [68] Tumour bone extension is a contra-indication for solely BT. Selected patients with bone invasion however can be treated with combination of surgery and brachytherapy. Radiation hypersensitivity syndromes (e.g., Fanconi Anemia, Ataxia Telangiectasia) are also definitive contra-indications for BT. Fertility sparing approaches: Preserving fertility of the child is a major objective of treatment. The indication for BT should be carefully weighed against other possible local treatments. For example, patients with RMS of the uterine cervix may be preferably treated with trachelectomy, rather than with uterovaginal BT, in order to avoid sterility that will be caused by high uterine and ovarian radiation doses. The same approach should be discussed in patients with cervical and/ or upper vaginal tumours, where trachelectomy +/- upper partial colpectomy may lead to less fertility impairment than cervico vaginal BT. Patients requiring pelvic irradiation should undergo a temporary transposition of the gonads to limit radiation exposure, if it is expected that dose exposure will exceed usual dose/volume constraints and if oncologically feasible. This transposition can be performed by paediatric surgeons at the time of BT implantation [16, 17, 29].

The target volume in RMS depends on various factors. These include: tumour topography, tumour characteristics (infiltrative or expansive with well-defined tumour borders), relationships to organs at risk and possible involvement, tumour stage at diagnosis (organ confined tumour, regional or metastatic extent), amount of residual disease after surgery (presence of a microscopic or macroscopic residuum), tumour histology (embryonal or alveolar subtype), response to chemotherapy (complete or partial response, stable disease) and age of the patient. When defining the clinical target volume (CTV) for BT, the tumour volume at diagnosis and after induction chemotherapy must be taken into consideration. A precise clinical examination, if necessary, under general anaesthesia, should be done. The findings from this examination are integrated with information about tumour volume and topography from the different imaging procedures. Post-implant 3D imaging is mandatory to guide target definition and include any residual tumour as part of the CTV. T2 weighted MRI is the best modality to define the residual tumour site, because of higher soft tissue contrast in comparison with CT scans. A discussion with expert radiologists may be necessary to guide target delineation. Residual hyper-T2 regions are parts of the residual gross tumour volume (GTV). In addition to MRI, a CT scan is usually necessary to see with high accuracy the catheters and the seed markers. If a surgical procedure is associated with BT the target volume is defined by the radiation oncologist together with the surgeon and the pathologist. After complete excision, the benefit of MRI is lower than for intact tumours, and a CT scan may be enough to delineate tumour bed. With an intraoperative procedure, the target volume is more easily defined because of the direct view of the structures to be treated [20]. Proper identification of the target volume requires a close collaboration between surgeons and brachytherapists, considering peri-operative findings, including frozen sections, and surgical seed markers. The target volume must be as small as possible because of the high risk of radiation morbidity, in particular the impairment of soft tissue growth and its possible adverse impact on cosmetic and functional outcome. However, for reasons of local tumour control, it is also necessary to consider the initial tumour volume. In fact, the clinical target volume is a compromise between initial and residual tumour volume considering the different variables listed below. For multimodal treatment (induction chemotherapy, surgical resection) there are in principle three situations that are most important for clinical target volume definition: • Residual macroscopic gross tumour: CTV is defined including at least the gross tumour volume (GTV) after induction chemotherapy plus a minimum safety margin of 5 mm. In any case the initial GTV must also be considered. • Residual microscopic disease (confirmed by pathology): For CTV the region of microscopic disease is included with some safety margin (at least 5 mm), also considering the dimensions of GTV at diagnosis, and considering postoperative anatomical changes. • No residual tumour as determined by clinical examination, imaging, or biopsy: if BT is considered, a CTV is defined, which will most likely prevent local recurrence without inducing major morbidity (e.g., prostate, cervix, tongue). This CTV is defined

Paediatric malignancies

7

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

Figure 2. Example of treatment of a vaginal rhabdomyosarcoma Fig. 2a, sagittal magnetic resonance imaging (T2-weighted) showing a bulky vaginal tumour. Fig. 2b, vaginal mould applicator in situ, inserted for treatment of residual disease after 6 cycles of chemotherapy. Fig. 2c, dose distribution for a treatment delivering 60 Gy through 143 hourly pulses of 0.42 Gy per pulse. In yellow: isodose 100%, in purple: isodose 90%, in blue: isodose 50%. The target volume is indicated as a red broken line structure, encompassed by the 90% isodose.

the young child. All BT procedures are performed under general anaesthesia. The following examples for BT techniques are described according to major locations which have been treated: cervix/ vagina/vulva, bladder prostate, limbs and trunks, head and neck, anus-rectum. These examples include different methods of BT (interstitial, endocavitary). Some procedures are or can be performed intraoperatively (bladder, prostate, trunk, limbs). Only a short description is given here; for detailed description we refer to the respective chapters in this book. Gynaecological RMS For vaginal and cervical RMS, treatment is usually based on an intracavitary procedure, with a vaginal applicator being inserted under general anaesthesia. It is recommended that a personalized mould applicator or a small cylindric applicator (the vaginal diameter is usually less than 15-20 mm) is used, following the same procedure as described in adults. Nevertheless, in children the cervical-vaginal impression is made under general anaesthesia often using a condom introduced into the vaginal cavity for easy removal of the impression from the vagina, which is too small to receive liquid paste and strips. The following steps are identical to making a mould in adult patients: impression into plaster, rough mould, positioning of the catheters according to the anatomy and the CTV, making appropriate perforations including four small holes to suture the applicator with silk thread to the inferior lateral parts of the vagina A washing catheter is also fixed to the vaginal mould applicator, to allow vaginal irrigation during treatment. The mould applicator is inserted under general anaesthesia, together with a Foley urinary catheter (Figure 2). Gold fiducial markers may be placed to mark the tumour landmarks. For patients with significant residual disease, a conservative tumour debulking is performed, without resection of the vaginal wall, to allow vaginal mould insertion. Para-vaginal interstitial catheters are rarely useful in vaginal RMS but may be required when there is significant residual para-vaginal disease (> 5mm thickness), following a free-hand technique. For tumours involving the upper part of the vagina and/or the cervix, an intrauterine catheter may be inserted and sutured to the vaginal mould applicator.

considering pre-treatment description of the tumour site, as per clinical and radiological evaluation. Specific points: In vaginal RMS, it has been shown that inclusion of the initial sites (extent) of disease was associated with more complications, without any benefit in terms of local control [37, 42, 43]. Therefore, only residual disease at time of BT should be included in the target volume. For a gynaecological implant, e.g., considering the CTV to be treated, MRI should be performed before and during the implant with the moulded vaginal applicator in place [41]. Tumour thickness and the exact topography of the residual disease can be evaluated and can be compared to the initial findings including the vaginal imprint. In BP RMS, it is recommended to systematically include both the prostate and the bladder neck as parts of the CTV when tumours extended to both organs at diagnosis, even if residual tumour only involves the prostate or the bladder. The prostate and bladder neck have the same embryogenesis and only very selected cases of patients with prostate only disease, and no initial extension to the bladder neck can be treated with prostate only BT. For multimodal strategies combining macroscopic tumour resection and intracavitary BT such as in the AMORE protocol (orbits, head, and neck tumours), the CTV encompasses the microscopically residual tumour volume and is defined as being 5 mm of tissue as measured from the surface of the mould.

8. TECHNIQUE

The details for the techniques to be applied in paediatric RMS are mainly derived from the techniques developed for adult BT as outlined in the respective chapters in this book for adult BT, but adaptation is required because of the small dimensions of

Paediatric malignancies

8

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

buttons. Additional interstitial catheters may be required, especially for patients with bulky residual tumour (> 3-4 cm), or involving the perineum. The patient is transferred to the brachytherapy department a few days after the implant procedure (usually 5 days later), to allow for the management of postoperative pain and of any early postoperative complication by a specialized team of anaesthesiologists and paediatric surgeons. Anus, rectum RMS For ano-rectal tumours, the BT technique follows the principles used in the adult patient. However, the dimensions of the applicator must be especially adapted. One layer of plastic or steel needles (3 - 5) is implanted depending on the target volume. Specific care must be taken to keep the contralateral side of the rectal wall as far away from the needles as possible with a rectal cylinder. Perineal RMS For perineal tumours, the prognosis is poorer than for urogenital RMS and achieving good local control may warrant combining conservative surgery and BT. When BT is associated with surgery, catheters are usually placed postoperatively, after complete surgical healing. Catheters are placed through a transperineal approach, which may involve intraoperative imaging procedures, such as ultrasound. The catheters should be fixed to the skin, possibly by means of buttons or through a perineal template. Soft tissue sarcoma of the extremities and the trunk The technique applied for paediatric soft tissue sarcoma of the extremities and the trunk is schematically the same as for adults. The most common technique is the intraoperative placing of flexible catheters, with buttons at the level of the skin and a template to keep the implant geometry. Usually, this treatment is associated with gross tumour resection, and an interstitial single-plane implant with catheters placed according to Paris system rules (parallel and equidistant) which allows an adequate dose distribution to treat the tumour bed. In cases of unresectable tumours, deeply seated tumour, large operative cavity, or microscopic residual disease, it may be necessary to use a multiple plane implant. Catheters are usually inserted orthogonal to the axis of surgical incision (Figure 4). In some cases, with wide excision, placement can be done parallel to the surgical bed. Attention should be paid at the time of surgery to avoid placing the catheters close to the nerves and vascular structures. Close collaboration between surgeons and the brachytherapist is therefore mandatory. Orbits BT has a place as local treatment for orbital RMS. It may be proposed as part of primary treatment in patients who have a residual disease after induction chemotherapy or as part of a salvage treatment after initial radiotherapy (external beam radiotherapy or brachytherapy). As described in detail by Blank and colleagues, the BT procedure is part of a multimodal approach to treat the operative bed after macroscopic tumour resection. The technique relies on an individual mould of silicone material that fits into the surgical defect, with flexible catheters inserted into the mould and adapted for remote afterloading treatment (Figure 5). The wound is closed over the mould, with only the catheters protruding through the closed incision [4]. A contraindication is deeply seated tumours close or in the orbital apex. These tumours cannot be reached for macroscopic radical excision and the chance of permanent damage with BT to the optic nerve is very high. Only in the case of radiorecurrent disease a mutilating orbital exenteration is an option followed by BT of the whole orbital surface (Figure 6).

For vulval sites, catheters are implanted according to the Paris system rules (parallelism and equidistance of the catheters). The same interstitial-BT technique used in adult patients can be

employed for children. Bladder prostate RMS

BT implant for BP RMS is usually performed at the time of open surgery, and most data are derived from the Institut Gustave Roussy (IGR) experience [12, 13, 46, 45]. The possibility for a conservative procedure is determined according to the response after chemotherapy, based on imaging (MRI with full bladder) with or without cystoscopy. A conservative procedure is not indicated in patients who would require irradiation of the whole bladder height. In IGR experience, the main criterion to allow the conservative procedure is the absence of tumour extension more than 1 cm above the level of the trigone in the posterior bladder wall. When tumour response is insufficient, additional chemotherapy courses can be delivered to increase tumour shrinkage. For tumours located in the anterior part of the bladder wall, a partial cystectomy is usually performed with free upper margins, respecting the muscular layer of the bladder neck. When the trigone is involved, a conservative tumour resection is performed. A bilateral ureteral transposition may be indicated to avoid stenosis of the ureteral orifices. For bulky residual tumours involving the prostate, partial prostatectomy with urethral preservation is considered. As this is a conservative approach, the surgery never aims at being microscopically complete at the level of the prostate and/or bladder neck, as BT will treat residual tumour cells. Usually, the implantation is carried out during the surgical procedure. Plastic catheters are inserted through a transperineal approach, to encompass the prostate/urethra and the bladder neck (Figure 3). Catheters are sutured to the bladder wall to avoid topographical modification and to the perineum through

Figure 3. Example of treatment of a bladder prostate rhabdomyosarcoma Four catheters were placed through transperineal approach to encompass the prostate and bladded neck for a pulse dose rate treatment delivering 60 Gy in pulses of 0.42 Gy per pulse. The 100% isodose is shown in yellow. The target volume is indicated as a red broken line structure.

Paediatric malignancies

9

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

Figure 4. Example of treatment of the forearm alveolar rhabdomyosarcoma in a 10 months old girl. She was treated with second look surgery and perioperative placement of catheters within the surgical bed. Dose was 60 Gy in pulses of 0.42 Gy per pulse.

Figure 5. Silicone mould with brachytherapy catheters placed at the medioinferior part of the orbita.

Figure 6. Exenterated orbita filled with a thermoplastic mould and catheters in a conal arrangement.

Head and neck tumours A large experience in treatment of head and neck sarcoma comes from Amsterdam, with the AMORE protocol (‘‘Ablative surgery, MOuld technique with afterloading brachytherapy, and RE constructive surgery’’). The AMORE protocol is a treatment for non-para-meningeal and selected para-meningeal tumours that includes two surgical sessions within 1 week, with brachytherapy in between. It consists of ablative surgery, brachytherapy, and surgical reconstruction in three steps [4, 65]: The first step is complete macroscopic surgical tumour ablation, preserving vessels, muscles and nerves. An individual mould made of thermoplastic material in

the shape of the surgical defect is placed in the operative bed in the same operative procedure. Catheters are imbedded in the mould, with a distance to the surface of the mould of approximately 4 mm. Then, subcutaneous tissues and skin are closed over the mould, and the catheters protrude through the surgical incision (Figure 7); The second step is irradiation of the operative bed to treat microscopic tumour residuum, with 3D treatment planning; The third step is reconstructive surgery, directly after removal of the mould with a free-vascularised or pedicled muscle or musculocutaneous flap (Figure 8). One objective of reconstruction is to supply an adequate and well-oxygenated tissue volume in the irradiated defect. If osseous

Paediatric malignancies

10

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

Figure 9. Recurrent para-pharyngeal tumour, excised and implanted with a mould for the AMORE technique. Prescribed dose 40 Gy 5 mm from the mould surface. Planning Target Volume is indicated with a red broken line structure

Figure 7. Parapharyngeal space filled with thermoplastic mould pieces containing brachytherapy catheters.

A

B

Figure 8. (A) Excisional area. (B) Muscle flap to cover surgical and irradiated cavity

structures need to be removed because of bone infiltration, e.g., in case of the mandible, the defect is reconstructed with a bone graft. For tumours of the oral cavity, oropharynx, or oral mucosa, interstitial treatments as per adult patients are performed, following Paris system rules (see chapters 11, 12, 13, and 14) [18].

term adverse side effects, in particular minimizing functional and cosmetic impairment. Therefore, computer assisted dose calculation is always performed and treatment planning is systematically based on sectional images, preferably MRI. Specific issues on dose and fractionation are detailed in the next paragraph. Each treatment should be individualized, and computerized treatment planning is mandatory. After implantation, CT is performed, with scan acquisition in the supine position with applicator and/or interstitial catheters in place, slice thickness usually 1.5 mm. A T2-weighted MRI is useful to guide target delineation, especially for patients treated with in situ tumour (e.g., BP RMS, gynaecological RMS, perineal RMS). Catheters are localized and active source dwell positions are decided according to the target volume, that is delineated taking into account clinical findings,

9. TREATMENT PLANNING

General requirements The main goals of paediatric BT must be carefully considered in treatment planning. These are to achieve local control and preserve the function of the organ involved at the same time as reducing long

Paediatric malignancies

11

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

HDR treatments, especially for sensitive sites such as urogenital, perineal, anal canal or head and neck tumours. However, emerging data suggest that HDR treatments are feasible, though the number of series is still limited. In most cases treated with PDR BT, BT is performed in one session; for other patients - for bulky tumour or where it is difficult to adapt the treated volume to the CTV - two, or sometimes even three sessions may be necessary. In the French experience of PDR BT, the physical prescription ranges from 50 to 60 Gy using continuous hourly pulses of 0.42 Gy per pulse to keep the daily dose at 10 Gy. For patients with macroscopic residuum, the objective is to deliver at least 50 GyEQD2 to 90% of the CTV (CTV D 90 ) and a minimal dose > 55-60 GyEQD2 to 95% of the GTV (α/β = 10 Gy for tumour). For microscopic only disease, the objective is to deliver at least 45-50 GyEQD2 to 90% of the CTV. In the Dutch experience of PDR BT, irradiation typically delivers 32-36 pulses of 1.25 Gy every 2.1 hours up to 42.2-47.5 GyEQD2 (α/β value = 10 Gy for tumour). For HDR treatments, physical doses of 27.5 Gy to 36 Gy are delivered through fractions of 3 to 5.5 Gy (two fractions per day), resulting in EQD2 doses of 35-40 GyEQD2 (α/β-ratio = 10 Gy for tumour). Hyperdose sleeves should be as thin as possible, with minimal confluence of the 150% isodose. The V150/V100 should be ideally less than 25-30%. For interstitial procedures, the hyperdose sleeve (200% idosose) should be less than 8-9 mm width. Few data are available in terms of dose/constraints for paediatric BT and adult data cannot be applied to the paediatric population. Because of smaller organs at risk volumes, dose/volume parameters should be considered very carefully. In particular, dose volume parameters must be adjusted for young children for whom a D 2 cm 3 may not be appropriate due to smaller size. Indeed, D 2 cm 3 refers to excessively large volumes, compared to the patient anatomy in particular for young children and therefore organs at risk D 2 cm 3 doses will likely not adequately predict the risk of complications with regard to the radiation dose exposure. In children, smaller volumes should be considered, such as the D 1cm 3 , the D 0.5 cm 3 or even the D 0.1 cm 3 . Only one series so far has studied dose/volume effects among 78 patients treated with PDR BT for urogenital tumours, median age 2.9 years. It was shown that the rectal D 0.5cm 3 and D 1cm 3 were significant for probability of grade 1 to 3 gastrointestinal complication (p = .009 and p = .017, respectively) and grade 2 to 3 anorectal morbidity (p = .007 and p = .049, respectively). The D 2cm 3 was not significant for the risk of rectal complications. A 10% probability (95% confidence interval, 4%-20%) for anorectal toxicity of grade 2 or greater was reached for a D 0.5cm 3 = 52 Gy. These data suggest higher radiation sensitivity in children, as compared to adults and the need to carefully look at smaller volumes (D 0.5cm 3 ) than in adults [62]. The use of vaginal spacers in female paediatric patients with bladder neck RMS resulted in significantly decreased doses to the rectum and the posterior vaginal wall [38]. More cooperative international research is required to provide evidence-based dose constraints.

as well as data from pre-implant imaging (usually MRI) and data from imaging with the applicator in place (CT scan +/- MRI fused with CT scan). For intracavitary BT, the dose is prescribed to the most external part of the CTV, at mid-level of the implant (usually at 2-3 mm from the applicator, depending on the depth of vaginal wall). Optimization is possible to avoid hyperdose sleeves extent into the vaginal wall. For interstitial procedures, the dose is usually prescribed to the 85% isodose, according to the Paris system rules, as a starting point for treatment planning. The homogeneity criteria are carefully to be considered and no large high dose volumes should be allowed. Dose-volume relations for organs at risk (including soft tissue and bone) are accurately assessed. Manual optimization can be done to improve dose distribution, for example in BP RMS where catheters are usually more distant each at the level of bladder neck than at the level of prostate. In the AMORE procedure CT-based treatment planning with the mould in situ is performed. Preoperative MR-images can be registered with the CT scan, but geometric deviations because of surgery and mould positioning should be considered. The reference isodose (40-45 Gy) typically surrounds the mould surface by 5 mm with limited optimization beyond 5 mm if necessary. Also, a high-risk CTV volume can be identified (tumour site and probable microscopic irradicality) which receives the above mentioned 40-45 Gy and an intermediate risk volume (no visible tumour on MRI, but microscopic tumour extension not excluded) for a lower dose (e.g., 30-36 Gy) (Figure 9). Dose-volume relations for organs at risk (including soft tissue and bone) are accurately assessed. Total dose varies according to the aim of radiotherapy within the specific treatment protocol applied. For interstitial or intracavitary BT alone, the total dose expressed as EQD2 (α/β-ratio = 10 Gy for tumour) usually varies from 45 to 60 GyEQD2: 45 GyEQD2 for standard prognosis (e.g., residual microscopic disease), > 50 up to 60 GyEQD2 for poor prognosis (e.g., gross residual disease) or for recurrent disease. Decisions about dose depend on the individual prognosis of the child according to the specific risk group. The following factors also influence the decision about total dose: tumour site, tumour stage at diagnosis and after surgery (amount of residual disease), histological subtype (good, intermediate, poor prognosis), response to chemotherapy (complete/partial remission, no response), dose volume relations in organs at risk, age of the child, place of BT within the treatment programme. When given as a boost in combination with external beam radiotherapy, one possible scheme is to deliver through PDR (hourly pulses of 0.5 Gy) a dose of 15-20 G EQD2, depending on the dose of external beam radiotherapy. Most of the available literature in paediatric BT comes from LDR, or more recently, from PDR. Because of the theoretical radiobiological superiority, those treatments may be preferred over 10. DOSE, DOSE RATE AND FRACTIONATION

Paediatric malignancies

12

THE GEC ESTRO HANDBOOK OF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 01/09/2023

11. MONITORING

RMS patients, 125/275 (45.6%) had tumour size > 5 cm, 31/275 (11.3%) had alveolar histology, 12/275 (4.4%) had regional lymph node involvement and 15/175 (5.5%) had distant lymph node or visceral metastatic sites. The main tumour sites were urogenital sites, with 172 BP RMS and 87 gynaecological tumours. In this series, BT was planned on 2D radiographs in 180 (59.0%) patients and on 3D imaging in 125 patients (41.0%). Among patients with RMS, median follow-up was 55 months (range: 1 month – 48 years). At five years, local control probability was 92.4% (95% CI = 88.1-96.9) for BP RMS and 94.2% (95% CI = 88.1-100) for gynaecological RMS. The local control probability was poorer for perineal RMS (62.5% (95% CI = 38.2-100)). Overall survival probability in the whole cohort was 93.3% (CI95%:90.1-96.5) [11]. Another large experience comes from India, with a publication on 105 children (median age 10 years) treated for soft tissue sarcoma [35]. Treatment included wide local excision and BT with or without external beam radiotherapy (19% of patients). Synovial sarcoma (22%) was the most frequent histology. Eighty-five (81%) received BT alone. After a median follow-up of 65 months, local control, disease-free survival, and overall survival at 10 years were 83, 66, and 73%, respectively. On multivariate analysis, authors found higher local control probability for patients with tumours <5 cm versus >5 cm (p = 0.03) and trunk/extremity versus head and neck/genitourinary sites (P = 0.002). Wound complications were reported in 6%. Subcutaneous fibrosis (25%) and limb oedema (6%) were the most frequent late complications [35]. For head and neck and orbital tumours, the most mature data are available from the Amsterdam experience [8, 63, 68]. In a series of 42 with non-parameningeal and parameningeal RMS patients published in 2009 and treated at doses ranging from 40 to 50 Gy (including 11 patients referred at time of relapse), the authors reported 3 patients with local recurrences and six having both local and distant recurrence. Overall, the 5-year survival rate was 70% in the primary treatment group and 82% for the salvage group. In an analysis of failure patterns, it was reported that five of six patients with relapse in the residual area had gross total or debulking (incomplete) surgery, suboptimal position of the mould for BT, or both. These data show the importance of both high-quality surgery and brachytherapy to achieve high local control [8]. Factors found to be a contraindication to perform the AMORE procedure are: • Intracranial extension • Invasion of the nasopharynx • Involvement of the orbit leading to exenteration (except in radiorecurrent disease) • Encasement of the carotid artery (more than 50% surface contact) • Invasion into the pterygopalatine fossa (contraindicated if adequate positioning of the mould cannot be accomplished) For orbital brachytherapy in combination with local surgery (mostly eye preservation) the 10-year event free survival and overall survival was found to be 65% and 89%, respectively [63]. Promising results were also reported from smaller series (usually involving less than 20-30 patients) treated with HDR treatments, mainly for limb tumours or bladder prostate RMS [23, 39].

During irradiation, regular checks must be carried out as for adult patients, but specific checks are mandatory dependent on the age of the child and the support of the family. The treatment room is equipped with Tv monitoring, WiFi, and viewing facilities. To preserve the quality of the implant, it is necessary to check clinically twice a day that there is no displacement of the plastic tubes or needles in an interstitial implant or of the applicator (mould) in endocavitary BT. X ray or sectional image control is mandatory when there is any suspicion of movement of the material [12, 13]. For HDR treatments, a minimal interval of 6 hours should be followed between two fractions and cross-sectional imaging is necessary prior to each fraction. The experience with PDR BT shows that compliance issues can be managed successfully in expert institutions, and they seem to be of relatively minor importance when compared with the long term results, in particular in terms of minimizing late side effects [13]. The compliance during continuous or pulsed irradiation for a period of several days is usually satisfactory, even in young children. The whole team including the brachytherapist, nurses, technologists, anaesthesiologist, psychosocial staff, family, and paediatric oncologist must however collaborate closely to support the child in coping with the various situations creating discomfort. General anaesthesia during the whole treatment course is not required. To minimize pain and psychological problems related to isolation, “preventive” soft sedation and pain treatment can be prescribed, thus reducing discomfort. The continuous support of the family and the care of the nurses is crucial throughout the day and night and in particular during meals. With regard to other forms of supportive and “preventive” care (e.g., anti-inflammatory treatment, antibiotics), these depend on the site, the procedure and the specific risks. In general, they are similar to those in adults (see the respective organ chapters). Local control, survival Numerous series report high to excellent local control rates after BT for paediatric tumours [5, 12, 13, 14, 15, 21, 22, 23, 24, 26, 28, 30, 35, 39, 45, 46, 47, 48, 49, 53, 54, 58, 66, 67]. Most clinical data for paediatric BT comes from LDR or PDR treatments and for treatment of urogenital sites [19]. The largest study for paediatric BT was published from Gustave Roussy including 305 patients (no retinoblastoma or CNS primary malignancy) with a mean age of 2.2 years (range: 1.4 months–17.2 years) at diagnosis. In this cohort, 270 (88.5%) patients had localised disease, 16 (5.2%) had regional lymph node involvement, and 19 (6.3%) had distant lymph node or visceral metastases. This series included 42 (13.8%) patients referred for local relapse/ progression. Primary tumour sites were mainly represented by genitourinary tumours, with 172 (56.4%) patients treated for a BP-RMS and 87 (28.5%) having a gynaecological cancer. Among 12. RESULTS