25 Oesophageal Cancer

SECOND EDITION

The GEC ESTRO Handbook of Brachytherapy

PART II: CLINICAL PRACTICE 25 Oesophageal Cancer Brachytherapy Razvan Galalae, Richard Pötter, Erik Van Limbergen

Editors Erik Van Limbergen Richard Pötter

Peter Hoskin Dimos Baltas

Oesophageal Cancer Brachytherapy

3

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

25 Oesophageal Cancer Brachytherapy v Razvan Galalae, Richard Pötter, Erik Van Limbergen

1. Summary 2. Introduction

3 4 5 5 6 7 8

9. Treatment planning

12 14 16 16 20 20 21

10. Dose, Dose Rate, Fractionation

3. Anatomical topography

11. Monitoring

4. Pathology 5. Work up

12. Results

13. Adverse side effects 14. Key messages

6. Indications, contra-indications 7. Tumour and Target Volume

15. References

8. Technique

11

1. SUMMARY

Oesophageal brachytherapy is very effective in terms of local tumour control and more efficient than external beam radiotherapy. Thus, oesophageal brachytherapy may be used in curative indications as a boost procedure (e.g. 2 x 4-5 Gy once-weekly schedule) following downsizing with external radiotherapy giving 50 (to 60) Gy, particularly in tumours with less than 5 cm length. However, oesophageal brachytherapy may lead to substantial long-term morbidity in curative settings (ulceration, stricture, fistula), in particular if small diameter applicators are used (2-6 mm) and the dose prescription is at 10 mm from the source axis. Mucosal doses above 65-70 Gy EQD2 3 are associated with a significant increase of ulceration and stricture risk (≥G2) based on recent Japanese experience. In squamous cell carcinoma with 5 cm or less tumour length, no deep invasion (T1 or T2) and good response after external beam radiotherapy (complete or partial remission), however, level 1 evidence showed superior cause-specific survival outcomes for higher tumour EQD2 10 doses at 5 mm mucosal depth of 72.5 Gy despite consequently higher total mucosal EQD2 3 doses of about 83 Gy (see table 24.1). The main goal in avoiding severe morbidity in curative settings is to prevent the mucosal dose becoming more than twice the reference dose at 5 mm depth. This can be best achieved by using the largest possible applicator with diameter ≥10 mm; preferably 15 or 20 mm. Oesophageal brachytherapy boost used in curative indications will also reduce substantially the lung and heart dose decreasing the radiation pneumonitis risk and chronic cardiac morbidity. Delivering the oesophageal brachytherapy boost upfront followed by external radiotherapy and concomitant chemotherapy may be useful in severe obstruction to alleviate dysphagia. Oesophageal brachytherapy is the standard treatment in palliative indications with a long-lasting beneficial effect on dysphagia and a more pronounced health-related quality of life benefit than after stent placement. Combination treatments with stents are possible but may be challenging. High-dose-rate brachytherapy dominates oesophageal brachytherapy with doses per fraction of 4-6 Gy using various prescriptions. The diameter of the applicator is the key factor and has to be adjusted to the oesophageal lumen diameter at the time of brachytherapy. The applicator diameter should be as wide as possible, in particular in curative settings (>10 mm; preferably 15-20 mm). Small diameter applicators should only be used in highly obstructive lesions (palliative intent). Prescribing the dose at 5 mm from the applicator surface is important in curative settings taking into account the mucosal dose which should be significantly less than 200% of the prescribed dose. Recent Japanese experience even suggests prescribing at the mucosal surface (e.g. 2x6 Gy after 50 Gy EBRT; 72 Gy EQD2 3 ) to minimize the risk of long-termmorbidity. The corresponding tumour dose at 5 mmmucosal depth would be approximately 55 Gy EQD2 10 . In palliative settings the dose is usually prescribed at 10 mm from the source axis (e.g. 5 Gy) as only small diameter applicators (2-6 mm) can be used in obstructive lesions. Mucosal doses (e.g. 5.6 x 5 Gy in a 4 mm diameter applicator) in palliative settings should be recorded but not taken into account for the prescription when chronic morbidity is not of interest. Interdisciplinary cooperation with experienced interventional gastroenterologists is strongly recommended. In addition, treatment decisions should be taken in interdisciplinary tumour boards after careful and detailed work-up.

Oesophageal Cancer Brachytherapy

4

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

Main indications for oesophageal brachytherapy include:

1. Single (boost) procedures in superficial tumours following non-radical surgery - curative indications in the definitive treatment of superficial tumours (T1). 2. Boost procedure following combination treatment of external beam radiotherapy and chemotherapy or as an upfront procedure in selected cases – curative indications in the definitive treatment of inoperable tumours (T2 and T3). 3. Boost procedure following combination treatments of external beam radiotherapy and chemotherapy or as an upfront procedure in appropriate cases – selected curative indications in the neo-adjuvant treatment of operable tumours in order to reduce radiation pneumonitis risk and chronic cardiac morbidity and alleviate dysphagia with the goal of improving detrimental nutritional status and operability (T2 and T3). 4. Combination treatment with external radiotherapy with or without chemotherapy - palliative indications in advanced tumour stages after excluding a fistula by radiological work up. 5. Single procedure or combination treatment with stents to ameliorate severe dysphagia symptoms – palliative indications in advanced tumour stages after excluding a fistula by radiological work up.

2. INTRODUCTION

in operable OC since its publication [Sjoquist et al.]. A head-to- head comparison of the two multimodal therapeutic principles (perioperative chemotherapy versus neoadjuvant chemoradiation) has not been performed yet. The initiated prospective randomized controlled trial ESOPEC (Trial registration NCT02509286), however, aims to compare FLOT against CROSS protocols and to definitely answer this open question [Hoeppner et al.]. Sanford and co-workers demonstrated in 215 patients treated with neoadjuvant chemoradiation impressive pathologic complete responses (pCR) in one third and a downstaging in two-thirds of the analysed patients [Sanford et al.]. The complete response rate was significantly higher (60.7% versus 18.4%) in patients with OSCC than in OAC [Sanford et al.]. Definitive external beam radiation therapy combined with chemotherapy is the better alternative in inoperable patients for whom surgery is not possible, e.g. because of surgically inaccessible locoregional tumour burden or clinical conditions with an unacceptable anaesthetic risk. Cooper et al. demonstrated in a randomized controlled trial the superiority (26% versus 0%) of chemoradiotherapy (Herskovic protocol: 50 Gy plus 5-fluorouracil and cisplatin in weeks 1, 5, 8 and 11) versus radiotherapy alone in terms of overall survival with acceptable toxicity [Cooper et al.]. A local brachytherapy boost may be added with caution to the combination of external beam locoregional radiotherapy and chemotherapy [Okawa et al., Ishikawa et al., Gaspar et al.]. Very advanced stages have a poor prognosis and are usually treated with palliative (chemo-)radiation in patients with good performance status (e.g. Herskovic protocol, see above). Radiation therapy alone is given externally, or in combination with brachytherapy in patients with poor performance status not likely to tolerate an aggressive chemoradiation regimen. In addition, oesophageal brachytherapy alone may be used for advanced or recurrent disease and in patients with bleeding, ulcerating or obstructive tumours [Rosenblatt et al.]. Alternatively, different types of local palliative treatments are available such as: non-radical surgery, thermal ablative treatment, metal stents, self-expandable metal stents (SEMS), plastic stents, and self-expandable plastic stents (SEPS). Several randomized controlled trials have been conducted

Oesophageal cancer (OC) is the sixth leading cause of cancer-related mortality worldwide. OC generally occurs in elderly patients and is associated with a prolonged history of tobacco smoking and alcohol dependence, especially in oesophageal squamous cell carcinoma subtype (OSCC) [Yang et al.]. In developed countries, the oesophageal adenocarcinoma (OAC) subtype is increasing and is aetiologically linked with Barrett’s oesophagus, gastro- oesophageal-reflux-disease (GERD) and abdominal obesity as GERD’s most common cause [Yang et al.]. Dietary factors, including carcinogens and low level of fruits and vegetables, are a common aetiology for both subtypes [1, Yang et al.]. Radical surgery remains the standard in early disease (cT1-2, cN0, cM0). In early superficial OC (cT1) non-radical surgery and brachytherapy may be considered as an alternative. Meta-analyses have revealed inferior outcomes in patients with squamous cell carcinomas compared to adenocarcinomas following surgery alone [Pennathur et al., Graham et al.]. For more advanced operable locoregional disease (cT2-3, cN0-2, cM0) the treatment strategy has evolved from classical monotherapy with radical surgery to multimodality therapy. Usually, oesophageal cancer is diagnosed in advanced inoperable stages (cT3-4, any cN+) [Krug et al.]. In operable patients, perioperative chemotherapy [Al-Batran et al.] or multimodality protocols including neoadjuvant preoperative chemoradiation followed by radical surgery compare favourably with previous reported results [Agranovich et al.]. In randomized phase 3 trials, the FLOT protocol (docetaxel, oxaliplatin, and fluorouracil/ leucovorin) was proven superior to previous used perioperative chemotherapy schemes [Al-Batran et al.], and perioperative chemotherapy compared superiorly in terms of overall survival versus surgery alone in resectable OC [Ychou et al.]. Van Hagen et al. have proven in a randomized controlled trial the superiority of neoadjuvant chemoradiation to surgery alone as well [Van Hagen et al.].This neoadjuvant chemoradiation protocol with 41.4 Gy plus carboplatin/paclitaxel (CROSS protocol) has provided the clearest evidence for survival benefit and has become the clinical standard

Oesophageal Cancer Brachytherapy

5

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

in order to assess the effectiveness of local palliative treatments [Homs et al., Amdal et al. Bergquist et al.]. There is no consensus regarding the optimal palliative local treatment. In general, the use of stents achieves an immediate effect on dysphagia [Homs et al.]. Endoluminal brachytherapy, in contrast, has a longer-lasting effect [Homs et al., Amdal et al., Bergquist et al.], and is superior in terms of health-related quality of life [Bergquist et al.].

the oesophageal lumen is filled the mucosa and submucosa are smoothed, pressed against the elastic muscle layers and thus the whole oesophageal wall is thinned, with the thickness usually less than 3 mm. Xia et al. found in a study on CT images an average oesophagus wall thickness between 1.87 and 2.70 mm when the oesophagus was dilated [Xia et al.] (fig. 27.1).

4. PATHOLOGY

3. ANATOMICAL TOPOGRAPHY

Tumours of the upper two thirds are mostly squamous cell carcinoma, while adenocarcinoma is the predominant type occurring in the lower third [Yang et al.]. In Western countries, the incidence of squamous cell carcinoma subtype is declining, while the incidence of OAC subtype is on the rise [Cook et al.]. Submucosal microscopic tumour spread is common and may extend longitudinally and radially in up to two thirds of patients [Landau et al.]. Depending on tumour volume, tumour extension in the oesophageal wall in the transverse plane may vary from a fewmm up to >20 - 30 mm. Lymph node involvement for several centimetres in the proximal and distal directions is also very common and is strongly linked to submucosal invasion [Raja et al.]. Raja et al. showed that with an increase of submucosal invasion

The oesophagus extends approximately from the level of the 6th cervical vertebra to the 11th thoracic vertebra and is an elastic tube about 25 cm long, which is - if empty - in the transverse cross section similar to a thumb, with a diameter of up to 2 cm in the latero- lateral and about 1-1.5 cm in the anteroposterior direction. The mucosa is folded in the longitudinal direction giving the transverse section a star-like appearance. The submucosa is loose allowing for folding of the mucosa and may reach a thickness of about 5 millimetres, if the lumen is not filled. The muscle layers are thin. The shape in the transverse section is very dependent on the state of filling: this may be with food or a tube which is introduced. When

Fig. 27.1: Cross sectional anatomy of the oesophagus with its different layers: mucosa, muscularis mucosae, submucosa, circular and longitudinal muscle (magnification factor 4). The thickness of these layers is highly dependent on the filling status.

Fig. 27.1A. Anatomy with an intraluminal applicator 15 mm in diameter: flattened layers of the wall, in particular the sub-mucosa: wall thickness varying from 1.87-2.7 mm (mean 2.1 mm; Xia et al. 2009)

Fig. 27.1B. Normal anatomy with a star like irregular mucosa and submucosa: wall thickness varying from 4.44-4.95 mm (mean 4.7 mm; Xia et al. 2009)

Oesophageal Cancer Brachytherapy

6

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

(double contrast), demonstrating both the tumour length and the resting lumen in obstructive lesions in AP and in lateral views. In addition, deviation of the oesophagus may be apparent indicating extra-oesophageal spread. On the radiograph, tumour extension must be reproducibly related to certain anatomical landmarks, e.g. carina of the trachea, and bony landmarks (e.g. thoracic vertebral bodies). A fistulamust be carefully excluded.The posterior-anterior and lateral chest X-ray are useful additional tools - if combined with barium swallow - to document tumour extension related to anatomical landmarks. For assessment of local tumour extension beyond the oesophageal mucosa, sectional imaging methods are necessary because barium swallow can only accurately estimate endo-oesophageal tumour growth. The gold standard to detect and document distant metastases is Fluorodeoxyglucose-positron emission tomography (FDG-PET) and computed tomography (CT) [Tangoku et al.]. Ameta-analysis on staging investigations showed that distant metastatic disease is best detected by FDG-PET with higher performance than CT and para-oesophageal lymph node metastases are detectable with the highest sensitivity by EUS [vanVliet et al.]. In addition, preoperative EUS is indispensable in T-staging providing the best cross-sectional method accuracy [Choi et al.]. FDG-PET is superior over CT in evaluating tumour response after neoadjuvant chemotherapy [Makino et al.] or chemoradiotherapy [Wieder et al.]. An SUV max reduction >70% after neoadjuvant therapy was associated with better survival outcomes. In addition, contrast enhanced computed tomography gives a relatively precise definition of transmural growth and of tumour infiltration into the surrounding fat and organs (aorta, trachea, major bronchi). MRI gives similar results with some advantages due to the coronal and sagittal imaging and the better soft tissue discrimination adding more information to assess the local and locoregional tumour extent. TNM clinical staging 8th edition [Brierly et al.] has changed the systematic tumour assessment mostly to depth of invasion, circumferential involvement and extra-oesophageal spread in contrast to the preceding classifications which addressed tumour length and amount of obstruction as well. Location categories (lesion epicentres) - assessed ideally during oesophagoscopy - are the following: upper location (cervical oesophagus to lower border at the azygos vein), middle location (lower border of azygos vein to lower border of inferior pulmonary vein), and lower location (lower border of inferior pulmonary vein to stomach including oesophagogastric junction). Pathologic stages correspond to clinical stages. However, pN0 is considered when a minimumof seven nodes are examined in the resection specimen. As neoadjuvant therapy replaces surgery alone pathologic staging is losing its clinical relevance with the exception of early stages (pT1-2). For those patients undergoing neoadjuvant therapy post- neoadjuvant staging includes adding yp to the TNM categories. As discussed before the role of ypTNM postneoadjuvant staging is limited. For adenocarcinoma subtypes only, the finding of residual nodal disease (ypN+) might be used for decision-making in adding adjuvant chemotherapy with a possible survival benefit. Burt et al. demonstrated a 40% lower risk of death for patients with adenocarcinoma and residual nodal disease after neoadjuvant chemoradiation and radical surgery [Burt et al.]. Dysphagia is a cause of significant morbidity in patients with OC which requires often nutritional interventions and specific therapies. Thus, the severity of patient-reported obstruction

the maximal cancer length also increased [Raja et al.]. The largest cancer width increased also according to the depth of invasion [Raja et al.]. Other established predictors for nodal metastatic disease are higher grade [Barbour et al.], higher T- category [Leers et al.], and lymphovascular invasion [Buskens et al.].

5. WORK UP

Screening, in general, is not recommended due to the lack of evidence for reduced cancer mortality. However, patients with Barrett’s oesophagus (BE) as a complication of GERD are considered at risk to develop oesophageal cancer especially OAC subtype [Dubecz et al.] and may benefit from screening upper gastrointestinal endoscopy [DE Carli et al.]. A detailed medical history including all risk factors (alcohol and tobacco abuse, Barrett’s disease, overweight and abdominal obesity etc.) is important. The work up then includes a clinical examination, that particularly addresses the patient's general condition and performance status, nutritional status, weight loss during the last six months, degree of dysphagia (i.e. dysphagia to solids, semi-solids, liquids, total dysphagia – indicating the degree of luminal obstruction) and associated symptoms like cough (due to fistula or aspiration), chest pain or back pain. The physical examination includes the chest, palpable lymph node drainage sites (supraclavicular and cervical nodes), and an abdominal examination (liver). Investigations which aim to define the extent of tumour spread so that appropriate treatment may be planned include endoscopic and radiological investigations. Endoscopic investigations include an endoscopy of the upper gastrointestinal tract and a bronchoscopy especially in carcinomas extending proximally above the carina.The gastroenterologist should take several biopsies and measure the precise distance between the macroscopic start of the tumour and its distal end, measured from the teeth (fig. 27.2). Furthermore, a detailed description of the tumour surface should be available with regard to the pattern of tumour growth (exophytic, ulcerative) and to vulnerability and risk of bleeding. The resting lumen in obstructive lesions should be given in mm (e.g. by comparison with the diameter of the oesophagoscope). In addition, endoscopic ultrasonography (EUS) is a significant advance for imaging the gastrointestinal (GI) tract wall and enables guidance for fine needle aspiration (FNA) biopsy of suspicious paraoesophageal nodes [Tangoku et al.]. EUS may be also used to place clips or other markers to visualize the cranial and caudal edges of the tumour lesions on cross sectional imaging or fluoroscopy during oesophagoscopy. Moreover, endosonography combinedwith oesophagoscopy can lead to amore accurate estimate of submucosal and intramural tumour spread and staging. EUS determines ideally the pattern of local tumour growth, in particular for limited tumours distinguishing between T1a (invasion lamina propria/muscularis mucosa) and deeper T2 lesions [Thosani et al.]. Thus, EUS plays an important role in selecting patients for endoscopic non-radical surgery. However, these excellent diagnostic and staging methods may be technically impossible in very advanced obstruction. In advanced disease (T3, T4) then, the pattern of intraluminal local tumour growth may be documented by classical barium swallow radiography

Oesophageal Cancer Brachytherapy

7

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

Fig. 27.3. Staging of oesophagal cancer ( Korean Journal of Thoracic and Cardiovascular Surgery 2015 ) - Tis: high-grade dysplasia - T1a: invasion of lamina propria or muscularis mucosae - T1b: invasion of submucosa - T2: invasion of muscularis propria - T3: invasion of adventitia - T4a: invasion of pleura, pericardium, azygos vein, diaphragm, or peritoneum - T4b: invasion of other surrounding structures/organs as aorta, trachea or a vertebral body

Fig. 27.2. Basic anatomy of the oesophagus with topographic landmarks for the radiological determination of the longitudinal dimensions of the GVT: bony landmark, tracheal bifurcation and aortic notch. The length which is given by oesophagoscopy for the determination of the proximal and distal ends of the length of the GTV is also shown. The black dots indicate potential lymphatic spread.

- N0: no regional lymph node metastases - N1: metastases in 1-2 regional lymph nodes - N2: metastases in 3-6 regional lymph nodes - N3: metastases in 7 or more regional lymph nodes - M0: no distant metastases - M1: distant metastases

6.A. Curative indications in superficial OC (T1a or T1b-SM1) Endoluminal brachytherapy has been used successfully to treat superficial oesophageal cancer in combination with external beam radiotherapy, however, distinguishing clearly between cT1a indications with excellent 5-year outcomes and submucosal tumours cT1b where the results are less favourable [Murakami et al]. This study byMurakami et al. revealed inmultivariate analyses that tumour depth was the only significant predictor for overall survival (p=0.002). Tamaki et al. showed good 5-year outcomes of endoluminal brachytherapy boost in superficial OC in terms of cause-specific survival and locoregional control as well, and could not identify a significant difference between HDR and LDR brachytherapy boost after external beam radiotherapy (50-60 Gy) in terms of efficacy and safety [Tamaki et al.]. The endoscopic classification of the Japan Esophageal Society (JES) based on invasion depth of superficial OC may help to identify the best candidates for endoluminal brachytherapy following non-radical surgery (e.g. superficial submucosal dissection) with excellent cause specific survival rates for SM1 lesions with a submucosal invasion ≤ 200µm [Nagami et al.]. Thus, the indications for endoluminal brachytherapy as a single procedure in superficial OC are T1a N0 M0 and T1b-SM1 N0M0 without risk factors for para-oesophageal lymph node involvement such as G3 and angio-lymphatic invasion. When those risk factors are present in T1a / T1b-SM1 lesions external beam radiotherapy should be added to endoluminal brachytherapy. T1 with significant submucosal invasion (>SM1) and T2 lesions should be considered contraindications for any non-radical surgical procedure.

should be measured in an objective way at baseline and regularly after the therapy (e.g. after 6 weeks, 3 months and 6 months etc.). Standardised means of measuring the respective endpoints should be used. Degree of dysphagia and amount of obstruction is recorded as: 0=no dysphagia, 1=dysphagia to solids, 2=dysphagia to semisolids, 3=dysphagia to liquids, 4=total dysphagia (or aphagia) [Francis et al.]. The performance status (e.g Karnofsky index or WHO score) is documented to evaluate the general condition. The nutritional status (weight, body-mass-index, and nutritional risk score) should be recorded. In order to exclude synchronous malignancies complete examinations of theGI tract and a head&neck examination including panendoscopy are recommended. For patients older than 70 years at diagnosis a Comprehensive Geriatric Assessment (CGE) might be useful. Apart from blood counts, blood coagulation tests, renal tests, liver tests, andCEAno further specific laboratory examinations are necessary. In general, identification of comorbidities and their specific medications is mandatory.

6. INDICATIONS, CONTRA-INDICATIONS

Themain indications for endoluminal brachytherapy in oesophageal cancer are:

Oesophageal Cancer Brachytherapy

8

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

6.B. Curative indications in inoperable OC (cT2-3 cN0-3 cM0) – boost procedure In these indications, as most oesophageal tumours are large at diagnosis (T3/4, stage III/IV, extensive disease), endoluminal brachytherapy alone is not sufficient to achieve complete tumour remission, especially for disease deep in the oesophageal wall and plexus (cT2-3). The combination of external radiotherapy (with or without chemotherapy) and endoluminal brachytherapy is therefore the recommended treatment, starting with external beam radiotherapy combined with chemotherapy in order to induce tumour remission and adding brachytherapy as a boost treatment to a reduced volume.This strategy can follow the response adaptive target approach which was first introduced in cervical cancer with definition of initial and residual GTV and based on this to define the intermediate-risk and high-risk CTVs (CTV IR and CTVHR) (see section 7 on target volume). In selected cases, brachytherapy may be used at the start of radiotherapy to achieve rapid symptom relief from advanced oesophageal stenosis and bleeding, if the cancer lesions are passable by the endoscope and the endoluminal brachytherapy applicator can be safely introduced prior to the downsizing effect of external beam radiotherapy.The use of oesophageal stenting - as an alternative to brachytherapy - in patients with cancer-related dysphagia undergoing combined chemotherapy and external beam radiotherapy (CRT) with curative intent is associated with a significant increase of grade 3 and higher toxicity (71%versus 27%, p<0.01) and inferior survival (p=0.026) and therefore cannot be recommended [Francis et al.]. Finally, endoluminal brachytherapy may also be added if not initially planned, when tumour remission achieved by external radiotherapy has not been sufficient, to increase the local effect in curative situations. In stage I-IIIA tumours - in particular in tumours which are inoperable for technical reasons (cervical oesophagus, upper third of thoracic oesophagus) or medical reasons - a combined treatment using external radiotherapy and HDR-brachytherapy with curative intent is indicated. Endoluminal brachytherapy as a boost procedure was introduced in the late 1980s and proved feasible in phase II trials [Calais et al.]. In randomized controlled trials endoluminal brachytherapy boost has been tested against external beam boost after chemoradiation and proved superior in patients with tumour length <5 cm [Okawa et al.]. This study by Okawa et al. of the Japanese Society of Therapeutic Radiology andOncology (JASTRO) also demonstrated the feasibility of HDR or LDR brachytherapy with two fractions of 5 Gy as a once-weekly procedure after 60 Gy of external beam irradiation (EBRT). Cancer specific survival was improved in lesions smaller than 5 cm in patients receiving a brachytherapy boost [Okawa et al.] (see table 27.1). However, the continuation of chemotherapy (combination of cisplatin and 5-fluorouracil) during the brachytherapy boost phase after 50 Gy chemoradiotherapy and the use of higher brachytherapy boost doses of 15-20 Gy may be very toxic with life-threatening toxicity [Gaspar et al.]. 6.C. Curative indications in operableOC (cT2-3 cN+ cM0) – boost procedure Since alleviating dysphagia is important to maintain nutritional status [Homs et al., Amdal et al., Bergquist et al.] endoluminal brachytherapymay be used as a boost procedure prior to combination treatment with neoadjuvant external beam radiotherapy and chemotherapy as an upfront procedure in selected cases with the goal of improving detrimental nutritional status and operability

(cT2 and cT3). Another specific indication for endoluminal brachytherapy as a boost within neoadjuvant treatment in operable oesophageal cancer is to reduce the radiation pneumonitis risk and chronic cardiac morbidity by lowering the mean lung [Abou Yehia et al.] and heart doses [Darby et al.] or in cases of inadequate downsizing (<70% SUVmax drop in FDG PET pre-operative restaging) following neoadjuvant radio-chemotherapy. 6.D. Palliative indications in inoperable OC (cT3-4 cN0-3 cM0-1) – boost procedure In advanced locoregional tumours the use of external beam radiotherapy for initial downsizing with or without chemotherapy is important. The addition of endoluminal brachytherapy may be useful when significant downstaging is achieved after excluding a fistula by radiological work up and the residual oesophageal lumen becomes passable by an endoscope. Again, the response adaptive boosting strategy should be applied based on the residual GTV, defining a CTVHR and PTV after EBRT (see section 7 on tumour and target volumes). Chemotherapy may be added in palliative cases only if patients present with ECOG status 0-2. Brachytherapy alone may be indicated in specific situations: for palliation to improve dysphagia in obstructive lesions or for tumour recurrence after external beam radiotherapy to prevent tumour regrowth [Homs et al., Amdal et al., Bergquist et al.]. If the obstruction is so advanced that the oesophageal lumen has become impassable for the brachytherapy applicator, treatment to clear the lumenmust be performed first by a specialized gastro-enterologist.Thismay consist of gradually widening the obstruction with a bougie dilator, or by laser resection.This reopening therapy should quickly be followed by endoluminal brachytherapy to prevent tumour regrowth. If a tube or a stent has been inserted, additional brachytherapy to prevent re-obliteration is also indicated.The primary aimof brachytherapy in obstructive lesions in a palliative setting is to achieve symptom relief from dysphagia for the resting life span. 6.E. Palliative indications to treat tumour local relapse and/or alleviate cancer-related dysphagia – brachytherapy alone 7.A. Curative indications Curative indications require high precision treatment. Planning should be 3D based, using endoscopy, EUS and CT/MRI (see section 9) in order to cover the CTVwith an adequate prescription isodose (fig. 27. 5) 7.A.1. Superficial tumours Superficial tumours that can be treated with curative intent may be de novo lesions as well as small recurrences in previously irradiated areas. In these superficial targets, the extent of the GTV has to be carefully documented by the available diagnostic information: endoscopy, endoluminal ultra sound, CT and MRI. Around the GTV a CTV margin of 1 cm in longitudinal direction is taken to cover microscopic spread. In the radial dimension the thickness 7. TUMOUR AND TARGET VOLUMES

Oesophageal Cancer Brachytherapy

9

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

Fig. 27.4 A. Focal, polypoid thickening of the oesophageal wall (<5mm) non-transmural CM-enhancement, smooth outer wall contours: T1.

B

Fig. 27.4 B. symmetrical, circumferential wall thickening (5- 15mm), smooth outer wall contour: T2.

C

Fig. 27.4 C. asymmetrical, circumferential wall thickening (>10mm) with stenosis, irregular outer wall contours and para- oesophageal fat tissue infiltration: T3.

Oesophageal Cancer Brachytherapy

10

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

of the oesophageal wall is measured. This thickness can be derived from the endoluminal ultrasonography or from the planning CT or MRI in 3D planning. The thickness of the wall is dependent on tumour thickness and the filling condition. Without thickening by tumour infiltration, the wall measures under physiologic resting conditions about 1.87-4.95 mm (Xia et al. 2009). In order to eliminate localization uncertainties and geographic mismatch it is important to implant radiopaque markers (e.g. clips or contrast media) at the cranial and caudal edges of the CTV (GTV plus 1 cm safety margins cranially and caudally) into the oesophageal wall during the oesophagoscopy and document by fluoroscopy and digital radiographs. The oesophagoscopy is best performed with the specialized gastro-enterologist and will also be used to place the applicator for endoluminal brachytherapy. To allow for applicator location uncertainties (fig. 27.5) a PTV margin of 1 cm is defined at both longitudinal ends of the CTV (GTV to PTV total margin 2 cm).

7. A.2. Boost after external beam (chemo)radiotherapy The traditional target approach takes into account the extension and depth of the GTV at diagnosis and adds margins as appropriate for longitudinal and circumferential extension (eg. 1 cm). Another PTV margin of 1 cm may be added to account for applicator location uncertainties. The Adaptive Image Guided Brachytherapy concept as developed and already applied for IGABT in cervix [ICRU Report 89, Haie- Meder et al.], anal [Van Limbergen et al.], rectal [Dresen et al.], nasopharynx [Bacorro et al.], and vaginal cancer [Schmid et al.], is also applicable to curative intent oesophageal brachytherapy after initial chemoradiotherapy. High precision localisation with endoscopy and EUS and 3D planning on CT and MRI is in this setting mandatory as well. The residual tumour volume (GTV) and residual oedematous (grey) zones at the time of brachytherapy represent the high-risk clinical target volume (HR-CTV). However, it is also necessary to take into consideration the initial size of the gross tumour volume

D

Fig. 27.4 D. large oesophageal tumour with stenosis, transmural growth and infiltration into adjacent organs (left main bronchus, pericardium, diaphragm, pleura): T4

Fig. 27.4 E. Barium Swallow.

E

From fig 27.4 A to fig. 27.4 E. T-staging of oesophageal cancer (T1-T4) based on Computed Tomography and Barium swallow X-ray for brachytherapy treatment planning Courtesy to Ahmed Ba-Ssalamah and Sarah Pötter-Lang, Department of Biomedical Imaging and Image-guided Therapy, Medical University of Vienna, Austria

Oesophageal Cancer Brachytherapy

11

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

and define it as an intermediate risk CTV zone (IR CTV) as this area may still contain significant microscopic disease.

(Create Medic Ltd, Yokohama, Japan; http://www.createmedic. co.jp/english/) (fig. 27.8). In general, the larger applicators are used preferentially, whereas applicators with smaller diameters are reserved for highly obstructing lesions. It is always advisable to use the largest possible diameter to improve the dose gradient at the mucosa and the depth dose (see figure 27.9 and table 27.1). Themost serious problem is an obstructive tumour, where difficulties may occur even in placing an applicator with a small diameter. As these tumours usually are rather rigid, it is difficult to dilate the lumen mechanically. If no dilation is possible - e.g. in rigid recurrent lesions -, a small diameter applicator is introduced and the high tumour surface dose with a significant overdosage to the mucosa and submucosa are considered in the dose prescription. The chosen dose per fraction is in this context a trade-off to achieve the maximal acceptable mucosal dose. A shrinking volume strategy can be used in multiple fraction regimens in order to minimize mucosal doses and in consequence therapy-related toxicity. In these highly obstructive cases, laser surgery may be an alternative to mechanical dilatation to clear the lumen prior endoluminal brachytherapy. If there is a wide oesophageal lumen, the diameter of even a large diameter (bougie) applicator (14-16 mm) may not be sufficient to fill the entire lumen. In certain situations, balloon applicators with at least two channels may be used, which allow for a better fit with larger lumen diameters. These applicators are commercially available (CreateMedic Ltd, Yokohama, Japan).These large diameter applicators may be used for the lower, medium or upper third of the oesophagus. The balloon (usually about at least 6 cm in length) can be filled with a dilute contrast medium up to a diameter of 20 mm leading to a large distance up to 10 mm between the source and the surface of the balloon (tumour/oesophageal mucosa). The outer surface of applicators usually carries a scale in cm similar to endoscopes. 8.2. Patient Preparation The application should preferably be performed using an endoscope. Fluoroscopic guidance alone should be used only when endoscopic guidance is not available. HDR treatments may be performed on an outpatient basis. However, the risk of post-procedural serious complications (e.g. bleeding, perforation, mediastinitis) with immediate necessity for emergency treatment should be considered and discussed with the patient. The patient must be fasting on the morning of treatment and intravenous access should be secured. Application is facilitated by sedation and a vagolytic agent. Local anaesthesia of the oral cavity and oro-hypopharynx is needed to introduce the endoscope and applicator. In fluoroscopy assisted applications, a recent barium swallow should be available - the barium swallow should not be performed more than 2 - 3 days before the application. This documents the actual pathologic situation precisely. A recent barium swallow is only unnecessary in recently stented patients or in patients with a wide lumen. However, barium swallow excludes an existing pre-treatment fistula. 8.3. Application Methods Depending on whether endoscopy or fluoroscopy is used, there are different methods for application appropriate for the individual pathologic situation. The goal is always to match the applicator

7.B. Palliative Indications In palliative situations, it is usually not possible to entirely cover the gross tumour volume with endoluminal brachytherapy. The historical approach in these cases, as used in palliative endobronchial brachytherapy, is to prescribe the dose at 1 cm distance from the source axis, deep into the tumour mass, accepting that the cancer lesion is usually not covered in depth by an adequate tumour sterilizing dose (fig. 27.6). The downsizing effect of the previously irradiated brachytherapy fractions, however, may be used to achieve a better coverage of the actual tumour lesion in multiple-fraction procedures. Although in palliative settings depth dose delivery can be suboptimal, it is always important to treat the malignant stricture over an adequate length. Around the GTV visible on barium swallow or at endoscopy a composite CTV-PTV margin of 2 cm is taken. It is very helpful to indicate the CTV/PTV borders with radiopaque (clip) marking. 8.1. Applicator Types For patient comfort and logistic reasons, HDR-brachytherapy should be preferred. However, LDR(PDR) brachytherapy may be also used [Vuong et al. Okawa et al.]. The applicators are single catheters, allowing the treatment of cylindrical shaped PTVs. There are several variations in the construction of the applicators: pure cylindrical, with an inflatable double lumen, or bougie type with a gradually increasing diameter to facilitate the mechanical dilation of stenotic disease. Importantly, dilation by bougies must be carried out by experienced gastro-entorologists, because there is a significant danger of perforation and bleeding. The diameters of these applicators vary greatly. Bougie-type oesophageal applicators between 18F (Ø 6 mm) and 45F (Ø 15 mm) for both afterloaders Flexitron and microSelectron along with an applicator fixation mask for use in a wide oesophageal lumen are available by Nucletron (an Elekta company, Elekta AB, Stockholm, Sweden; https://www.elekta.com/brachytherapy/) as well as small-diameter applicators between 6 and 8 mm for palliative treatments. Single-use bougies between 24F (Ø 8 mm) and 36F (Ø 12 mm) along with a universal 2 mm-diameter applicator are provided by Eckert & Ziegler (BEBIG GmbH, Berlin, Germany; https://www.bebig.com/home/products/hdr_brachytherapy/) for the afterloaders SagiNova and MultiSource. Varian offers bougie oesophageal applicators between 8 mm to 14 mm diameter for the afterloaders VariSource and GammaMed plus (VarianMedical Systems Inc, Palo Alto, USA; https://www.varian.com/oncology/ solutions/brachytherapy). All applicators available on the market are CT- compatible. See also fig. 27.7. In addition, Buzurovic et al. have successfully tested in collaboration with Ancer Medical (Hialeah, FL, USA) a novel multi-balloon self-centring applicator [50, Buzurovic et al.]. This central source catheter surrounded by five equally spaced, 2cm (individually inflatable up to 2.3 cm) balloons allows the treatment of an active length of 10 cm. Balloon applicators are also well established and used in Japan 8. TECHNIQUE

Oesophageal Cancer Brachytherapy

12

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

diameter to the oesophageal lumen diameter in its maximum extension.This avoids eccentric applicator positions, under-dosage to the tumour and over-dosage to the healthy tissue and treats as far as possible beyond the steep slope of the depth dose curve. The application may be performed in a dedicated brachytherapy room. Artificial teeth are removed. The patient lies on the left side as usual for oesophageal endoscopy. A sedative drug and a vagolytic agent are given intravenously. Extensive local anaesthesia is applied to the oral cavity, the oro- and hypopharynx (nomeals are allowed for 2 - 3 hours afterwards). The oro- and hypopharynx are then digitally palpated to determine the individual anatomical situation. In endoscopy assisted applications, the surface of the endoscope is covered with local anaesthetic gel and then introduced between two fingers placed into the oropharynx to lead the tip of the endoscope towards the proximal aperture of the oesophagus. With the patient swallowing, the endoscope is then introduced into the oesophagus. The proximal and distal end of the tumour are inspected by the endoscopist and the radiotherapist. The distances in cm from the teeth are noted and compared to the results of pre-therapeutic examinations. If localisation is difficult because of changes since pre-therapeutic imaging, the proximal and distal tumour end may be documented by two anterior-posterior radiographs with the tip of the oesophagoscope at the respective tumour margin. Radio-opaque markers should be used to document the distal and the cranial ends of the actual tumour at first brachytherapy. A flexible guide wire (>twice the length of the oesophagoscope) is advanced through the biopsy channel and placed far beyond the distal tumour margin, preferably within the stomach (fig. 27.12). The oesophagoscope is then withdrawn over the guide wire which remains in place. The applicator (open end or with perforated guiding tip) with a diameter larger than the biopsy channel (> 2mm) is advanced using the guide wire as glide path to the correct position. If an applicator with a small diameter (<= 2 mm) is to be used, it can be directly placed under endoscopy through the biopsy channel (as in bronchus). If the applicator with a diameter larger than 2 mm has a closed end, it has to be introduced alongside the oesophagoscope, which may be difficult in obstructive lesions. In fluoroscopy assisted applications, a semiflexible applicator with a scale is covered with local anaesthetic gel. The applicator is introduced between two fingers which are placed in the oropharynx to guide the tip of the applicator towards the proximal aperture of the oesophagus. With the patient swallowing, the applicator is then introduced into the oesophagus. It is gently advanced avoiding any significant pressure against resistant tissue. As the tumour region is reached - according to the known distance from the teeth - it is advanced even more cautiously. In obstructive lesions the resistance increases and must be gently overcome. After passing the obstruction, further advancement becomes easy again. To avoid perforation, the radiation therapist performing the application without the support of a gastro-enterologist should never try to overcome resistance by force. After the applicator has been placed correctly, the distance between the tip of the applicator and the teeth is noted andmarked. If an inner tube is needed, this can now be introduced into the outer tube ("applicator"). If there are difficulties with introducing the applicator, the Seldinger technique with a guide wire can be used. In such cases the guide wire (> twice the length of the tube) is put into a semi-flexible gastric tube with an open distal end. The gastric tube is gently introduced into the oesophagus with the patient actively swallowing and then

carefully advanced well beyond the distal end of the tumour, preferably into the stomach. The gastric tube is removed over the guide wire leaving the guide wire in place. The applicator can then be carefully introduced into the patient over the guide wire, which is finally removed (fig. 27.12). The guide wire based Seldinger technique offers in addition to localization accuracy, procedural safety. Ideally this will be performed with interdisciplinary expertise of the specialized cooperating gastroenterologist. The application can also be done through the nose with the patient sitting upright. Local anaesthesia is applied.The guide wire and later the applicator is then introduced through the nose.This procedure is only possible for small diameter applicators up to 6-8 mm.

9. TREATMENT PLANNING

After placement of the /applicator and before 2D or 3D imaging a calibrated marker wire is inserted to enable definition of the correct stepping source positioning in the applicator. This is not necessary when using applicators with inner markers at their distal end (see section 8). 9.A. Curative indications Curative indications require high precision treatment.The treatment planning is preferably based on 3D imaging. Planning on CT is possible, but should be supplemented by endoscopic placement of radio-opaque clips to delineate the CTV. MRI allows discrimination of the GTV from the non-invaded oesophageal wall, or the residual GTV and grey zones after initial (chemo)radiotherapy. If PET-CT is available, it may be of additional value to accurately determine the GTV. 9.A.1. Radical brachytherapy for superficial tumours After delineation of the GTV on CT or MRI a CTV margin of 1 cm is delineated in the oesophageal wall, as well as a PTVmargin of an additional 1 cm in the longitudinal direction to take account of possible displacement of the applicator in the oesophageal lumen. Dose prescription should be defined by the D90 of the PTV. For comparison with other cases and other centres dose recording and reporting should contain the prescription depth from the surface of the catheter, as well as the reference dose at 5 mm mucosal depth (fig. 27.13). 9.A.2. Boost after (chemo)radiotherapy 2D planning (see in detail GEC-ESTRO Handbook of BT first edition chapter 24) Two orthogonal X-rays are takenwith the applicator and themarker wire in place. The position of the applicator on the radiographs is carefully checked and compared to the pre-treatment barium swallow, looking at anatomical landmarks (carina, vertebral bodies, aortic notch) or clips.The tip of the applicator should reach several centimetres beyond the distal end of the visible tumour to adequately cover the length of the PTV. The Length of the Planning Target Volume (PTL) is determined considering tumor length (GTL), safety margins for subclinical extension (CTL) and positional uncertainties e.g. catheter

Oesophageal Cancer Brachytherapy

13

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/06/2019

Curative Treatment: Radial dose distribution Larger diameter catheter ASL 80 mm

Palliative treatment: radial dose distribution endo-oesophageal catheter ASL 80 mm

Curative Treatment: Radial dose distribution Larger diameter catheter ASL 80 mm

Palliative treatment: radial dose distribution endo-oesophageal catheter ASL 80 mm

Endo-oesophageal Catheter Diameter 12mm

Endo-oesophagealCatheter Diameter 4mm

100 % Prescribed dose at 6 + 3 = 9 mm (Curative intent)

100 % Prescribed Dose at 10 mm from the source axis at 10 mm Reference depth dose (5 mm from applicator) 148 % PD

Reference depth dose (5 mm from applicator) 79 % in tissue Maximum dose at applicator surface 155 % PD on surface 200 % PD in applicator 400 % PD in applicator

Maximum dose at applicator surface 540 % PD on surface 400 % PD in tissue 200 % PD in tissue

Oesophageal wall

Fig 24.5. Radial dose distribution of endoluminal brachytherapy with curative intent for oesophageal cancer using a large diameter (12 mm) applicator. ASL is 80mm with 2mm step size and geometrical optimisation on distance. The prescription isodose covers the CTV. The high dose isodoses (200%, 400%) are inside the applicator. The maximum surface dose is 155% of the Prescribed Dose. The Reference Isodose at 5mm from the catheter is 79% of the Prescribed Dose. distribution of endoluminal brach therapy with curative intent for oesophageal cancer using a large diameter (12 mm) applicator. ASL is 80mm with 2mm step size and geometrical optimisation on distance. The prescription isodose covers the CTV. The high dose isodoses (200%, 400%) are inside the applicator. The maximum surface dose is 155% of the Prescribed Dose. The Reference Isodose at 5mm from the catheter is 79% of the Prescribed Dose. Fig. 27.5. Radial

Fig 24.6 Radial Dose distribution around a small (4mm) endoluminal applicator for a large obstructive oesophageal cancer in a palliative setting. The ASL is 80mmm with 2mm step size and geometrical optimization on distance. Dose prescription is at 10 mm from the source axis. The higher doses (200-400%) are in tissue and not in the applicator as is the case in curative treatment with a large size applicator. The reference depth dose at 5 mm from the applicator is at 148% Fig. 27.6. Radial dose distribution around a small applicator (4 mm). The ASL is 80 mm with 2 mm step size and geometrical optimisation on distance. The maxi um surfa e dose is 540% of the prescribed dose at 10 mm. The 200% and 400% isodoses are in the tissues (tumour and / or oesophageal wall).

Fig. 27.8. Japanese double balloon applicators, inflatable to 15mm diameter and 20mm (Create Medic Ltd, Yokohama, Japan; http://www.createmedic.co.jp/english/).

Fig. 27.7. Example of bougie type intraluminal oesophageal applicators, between 18F (Ø 6 mm) and 45F (Ø 15 mm) with an applicator fixation mask.

Fig. 27.9. Surface doses in relation to applicator diameter for an endo-oesophageal source of 80 mm ASL with 2 mm step size and geometrical optimization on distance. With dose prescription at 3 mm (yellow line) from the applicator surface doses decrease from 394% to128 % for diameters ranging from 2 to 26 mm. The gradient is steeper when dose is prescribed at 5 mm from 604% to 140% (blue line), and range from 1057% to 74% in case of prescription at 1 cm. See also table 27.1. It is shown that 12 mm is the minimum diameter of the applicator to avoid >200% of the prescribed dose to the mucosal surface.