14 Nasopharynx Cancer

SECOND EDITION

The GEC ESTRO Handbook of Brachytherapy

PART II: CLINICAL PRACTICE 14 Nasopharynx Cancer Warren Bacorro, Michael Mejia, Erik Van Limbergen, Joseph T.S. Wee, Melvin L.K. Chua

Editors Erik Van Limbergen Richard Pötter

Peter Hoskin Dimos Baltas

Nasopharynx Cancer

3

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

14 Nasopharynx Cancer

Warren Bacorro, Michael Mejia, Erik Van Limbergen, Joseph T.S. Wee, Melvin L.K. Chua

1. Summary 2. Introduction

3 3 4 4 4 4 6 6

9. Treatment planning

12 13 16 16 18 19 19

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

1. SUMMARY

The excellent outcomes achieved with the current management of nasopharyngeal cancer, which incorporates modern external beam radiotherapy techniques, and concurrent chemotherapy for locally-advanced disease, have resulted in the relegation of brachytherapy as a tool for salvage treatment for recurrent or persistent disease and possibly, for dose-escalation in primary locally- advanced cases where chemotherapy is contraindicated. Brachytherapy allows for delivery of higher doses to the nasopharynx and moderate doses to the proximal parapharyngeal spaces, clivus, inferior sphenoid, posterior maxillary sinuses, and posterior nasal cavity while sparing largely the pterygoid muscles, visual pathway, spinal cord, brainstem and pituitary gland. This dosimetry profile makes brachytherapy a useful tool in salvage re-irradiation where organ-at-risk tolerances could be significantly restrictive. Image-guidance and three-dimensional planning, together with the use of modern fractionation schemes, could improve further control and toxicity outcomes. Endoscopy-guidance and incorporation of interstitial techniques are currently under exploration.

2. INTRODUCTION

in Table 1), radiotherapy (RT) alone is the primary treatment modality. This is because the nasopharynx presents a challenging anatomical site for wide-field surgical resection. Moreover, compared to other squamous cell carcinomas of the head and neck, NPC is exquisitely radiosensitive [Chua, 2016]. It is therefore appreciable that factors relating to RT dosimetry and accuracy are crucial to achieving long-term tumour control [Ng, 2014]. On this note, the transition from two-dimensional (2D) external beamRT (EBRT) to contemporary RT techniques like intensity-modulated RT (IMRT) and image-guided RT (IGRT) have been pivotal in driving the improved survival, including reducing late RT-induced toxicities and better quality of life scores reported in NPC patients, even for those with advanced T3-4 and N2-3 disease [Lee, 2015; Peng, 2012; Au, 2017]. In patients with locoregionally advanced disease (Stage II to IV), radiotherapy with concurrent high dose cisplatin given every 3 weeks with or without adjuvant/neoadjuvant chemotherapy is the standard of care. In this regard, CCRT is an effective measure for targeting occult systemic metastases and for its radiosensitising effect. Altogether, these have resulted in superior local tumour control, with the majority of cases being now sufficiently treated with EBRT alone, with the shift in disease relapse patterns being currently dominated by distant metastasis [Chua, 2017].

Nasopharynx cancer (NPC) is a unique disease, with specific demographical and epidemiological patterns: men are more commonly affected thanwomen, with SouthernChinese and certain Southeast Asian populations being themost susceptible racial group. Overall global incidences are low, but there is preponderance for extremely high rates (age-standardised rates, ASR, of 10-30 cases per 100,000 person-years) in specific parts of the world, including Southern and Eastern China, Southeast Asia, and Northern and Eastern Africa [Wee, 2015; Chua, 2016]. Broadly, these regions can be grouped according to their reported ASRs. Southern and Eastern China, including Guangdong, Guangzhou, Guangxi, Fujian and Hainan report 15-30 cases per 100,000 person-years for males. This is followed by parts of Southeast Asia (Singapore, Malaysia, Indonesia, Vietnam and Philippines) and Northern/ Eastern Africa (Algeria, Kenya and Tunisia) that report 3-10 cases per 100,000 person-years for males. In these regions where NPC is endemic, this disease is invariably associated with exposure to the Epstein-Barr virus (EBV). ASRs for NPC in other parts of the world are typically less than 1 per 100,000 person-years.

In patients with localised or Stage I NPC (staging summarized

Nasopharynx Cancer

4

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

4. PATHOLOGY

In a small subset of cases, residual or recurrent disease persists in the nasopharynx and neck, either as a result of intrinsic radioresistance or suboptimal RT dose intensity due to the proximity to critical normal structures [Li, 2018]. Given the uniqueness of the nasopharynx anatomical topography, NPC has the potential to infiltrate critical neural structures intracranially via the porous base of skull and its foramens and laterally beyond ambiguous tissue planes to involve the masticator space. As a result, in cases where the tumour extends to these anatomical regions, optimal surgical resection with negative margins is not always feasible. When considering salvage RT, EBRT and brachytherapy techniques are possible options [Chua, 2016]. EBRT includes stereotactic radiosurgery (SRS) and IMRT. In principle, SRS is effective and safe in small recurrent tumours that are away fromcritical structures like the major neck vessels, cranial nerves and brainstem. The advent of IMRT has opened the possibility of re-irradiation in larger T3-4 recurrences, by its superior ability to conform radiation doses and derive sharp dose-gradients around critical normal tissues that had been previously irradiated. Several series have reported 3-year local control and survival rates of 60% and 50%, respectively, in recurrent NPC cases treated using this contemporary technique [Qiu, 2012, Tian, 2014, Li, 2018]. Nasopharyngeal brachytherapy, as described in literature, mostly entails intracavitary brachytherapy (ICBT) and 2D planning techniques. Invasive interstitial brachytherapy (ISBT) techniques have been described in older literature, but less invasive techniques incorporating endoscopic guidance, three- dimensional (3D) planning and high-dose-rate (HDR) delivery have been detailed in more recent publications. These advances could allow better coverage of deeper recurrences, as shall be discussed in subsequent sections. The nasopharynx is cuboidal in shape, and is bounded superiorly by the sphenoid sinus, posteriorly by the pharyngeal mucosa and longus colli muscles, laterally by the Eustachian cushion, in front of the fossa of Rosenmüller, and anteriorly and inferiorly by two arbitrary anatomical landmarks – the choanae, which links to the anterior nasal cavity and the caudal edge of the uvula, respectively (Figure 1). The nasopharynx is surrounded by critical normal organs. Superiorly, the porous base of skull and the clivus are barriers to the crucial cranial nerves (CN) in the cavernous sinus, including CN III, V, VI; the optic nerves (CN II) and the optic chiasm are also at risk in cases where the tumour involves the orbital apex and the sellar and parasellar spaces. Bulky tumours that extend laterally can also invade the pterygoid muscles and the masticator space, which would be associated with symptom of trismus. The nasopharyngeal region also harbours a rich lymphatic network, which would explain the frequent occurrence of nodal metastases (>60%); of note, the retropharyngeal and level II nodal stations are the first echelon of involvement [Ho, 2012]. 3. ANATOMICAL TOPOGRAPHY OF THE NASOPHARYNX

There exists in the nasopharynx an intermediate pseudostratified cuboidal type that is a transition between the pseudostratified columnar epithelium and the squamous epithelium. This intermediate epithelium is most susceptible to carcinogenesis, thus the propensity of NPC to grow in the lateral walls, posterior wall and anterior walls, in that order, corresponding to the predominance of this intermediate epithelium [Batsakis JG, 1980]. NPC is conventionally thought of as a tumour of squamous origin. Morphologically, the transformed cells are ovoid looking, with a distinct nucleoli and scanty chromatin, and often, there is an abundance of lymphoid cells that are admixed with the transformed epithelial cells. NPC is categorised into three pathological subtypes based on the degree of squamous differentiation. Type I – differentiated tumours with surface keratin; Type II – non-keratinising differentiated tumours; Type III – undifferentiated tumours. Of note, in regions where NPC is endemic, the non-keratinising subtypes (Types II and III) constitute the majority of cases (>95%). Interestingly, there is emerging evidence that Type II and III tumours are associated with EBV exposure, while some Type I tumours that are more common in the Western population may be associated with the human papilloma virus (HPV) [Lin 2013, Dogan, 2014]. All newly diagnosed NPC patients require a detailed clinical examination, which includes nasoendoscopic examination of the head and neck region, as well as a biopsy of the primary tumour. Laboratory investigations include routine complete blood count, a full renal and liver panel, and plasma cell-free EBV DNA titre. Of note, markers such as neutrophil-to-lymphocyte ratio (NLR), lactate dehydrogenase, and EBV DNA have been proposed as prognostic biomarkers in NPC. Given the proclivity for nodal and distant metastases (lungs, liver, and bones), an MRI of the head and neck region, a CT of the thorax and abdomen, and bone scan are necessary to accurately stage the patient. For distant metastasis staging, 18FDG-PET is a superior modality to conventional CT and bone scans. At present, NPC is staged according to the 8th edition of the UICC/AJCC stage classification system (Table 1) [Pan, 2016]. 5. WORK UP

6. INDICATIONS AND CONTRAINDICATIONS

Given the principles of brachytherapy dosimetry, only superficial mucosal or submucosal lesions are amenable to this RT technique. Satisfactory outcomes had been observed with early brachytherapy techniques involving the use of radioactive gold grain (Au198) implants or insertion of iridium (Ir192) wires through a customised mould to treat superficial lesions. However, with the advent of remote after-loading, HDR intracavitary brachytherapy is favoured

Nasopharynx Cancer

5

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Figure 2. Clinical tumour volumes, organs at risk and typical isodoses with intracavitary brachytherapy using the Rotterdam applicator. Shaded areas: Red, gross tumour volume; blue violet, nasopharynx; light orange, soft palate; dark orange, brainstem; atlanto-axial joint, yellow. Lines: Light pink, 250% isodose; dark pink, 150%; red, 100%; yellow, 90%; green, 75%; blue: 40%. [Bacorro, 2018]

Figure 1. Sagittal anatomical topography on co-registered radiological images; MRI (upper), CT, with applicator in place (lower). Nasopharyngeal tumour, red; Nasopharynx, blue green; Soft palate, light orange; Brainstem, dark orange; Spinal Cord, blue.

B

A

Figure 4. Orthogonal radiographs of inserted mould nasopharyngeal applicator. Left: lateral view; right: anteroposterior view. [Mazeron, 2002]

D

C

Figure 3. Retrograde insertion of the mould applicator. A, A personalized mould applicator. B, Insertion of the Nelaton tubes through the nostrils. C, Recuperation of the Nelaton tubes through the mouth and fixation onto the personalized mould applicator. D, Retrograde maneuvering of the applicator into the nasopharynx by traction on the Nelaton tubes. [Mazeron, 2010]

Figure 5. Computed Tomography with the mould applicator in place. Left: coronal view; right: transverse view. [Mazeron, 2002]

Figure 6. Customized mould applicators requiring recess dissection. Left: preloading mould; Middle: afterloading mould; Right: recess dissector. [Law, 2002]

Nasopharynx Cancer

6

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

in current times, given its advantages in radiation protection and dosimetry. For ICBT, the thickness of the clinical target volume (CTV) should not exceed 10mm. Consequently, only superficial tumours without involvement of the parapharyngeal space or underlying bone are eligible. ISBT by endoscopic guidance can potentially allow for treatment of parapharyngeal extension. Tumours extending into the anterior nasal cavity or oropharynx, or closely adhere to the major carotid vessels would not be candidates for brachytherapy. Potential indications for brachytherapy may therefore include (1) upfront dose-escalation of T1 and T2 disease after EBRT alone, when concurrent chemotherapy is contraindicated, or even after CCRT, (2) boosting minimal residual local disease in an initially T1, as well as select T2 and T3 disease, after satisfactory response to EBRT or CCRT, and (3) salvage therapy for well-circumscribed and superficial local recurrences limited to the nasopharyngeal space. As basic principle, the gross tumour volume (GTV) at brachytherapy shouldbe delineatedusing information fromendoscopic examination and cross-sectional imaging (ideally MRI). There is not yet a consensus onCTVdelineation for nasopharyngeal brachytherapy.The high-risk CTV (HR-CTV) should cover at least the GTV plus a 5-mmmargin, with or without including the entire nasopharynx. Most tumours are located at the roof (i.e. the soft tissue situated between themucosa and the base of skull) and/or the lateral walls of the nasopharynx. In central tumours, the CTVwould include both roof and the lateral walls, however, in well-lateralised tumours, the CTV could be restricted to the roof and one lateral wall [Mazeron, 2002]. For nasopharyngeal boost, whether upfront or as salvage for persistent disease, an intermediate-risk CTV (IR- CTV) could be defined to include at least the GTV at diagnosis (GTV-D). With endocavitary brachytherapy, the 40% isodose (i.e., 1.4 Gy in a 3.5 Gy fraction) could adequately encompass the inferior third of the sphenoid, the entire clivus, the posterior third of the nasal cavity and maxillary sinus, and the proximal (pre-styloid) parapharyngeal spaces (Figure 2)[Bacorro, 2018; Mejia, 2018]. 7. TUMOUR AND TARGET VOLUMES

The brainstem, spinal cord, pituitary, optic chiasm, retina, clivus, atlanto-axial joint and soft palate are contoured as OARs using the MRI as aid for delineation. The atlanto-axial joint is drawn using the following borders: upper limit of the dens (superior), 2mm below the lower limit of the atlas (inferior), ventral limit of the dens and including the ligaments visualized on the MRI or corresponding area on CT (anterior), dorsal limit of the dens (posterior), and inner limits of the axis including the ligaments visualized on the MRI or corresponding area on CT (lateral). The soft palate is drawn from its junction with the hard palate down to its free border and along its junction with the tonsillar pillars, at least 6mm below the inferior limit of the constructed catheters.

8. BRACHYTHERAPY TECHNIQUES

8.1 Intracavitary Approaches Personalized ICBT applicators required important resources and expertise [Mazeron, 2002; Law, 2002]. Less invasive anterograde trans-nasal insertion techniques employing pediatric endotracheal tubes and commercially available nasopharyngeal balloon applicators [McLean, 1998, Chang 2001], and retrograde trans- oral insertion techniques using the Rotterdam nasopharyngeal applicator (RNA), have been since developed.The RNA, unlike the pediatric endotracheal tubes and balloon applicators, is compatible with multiple treatments and may remain in place for up to six days of treatment [Levendag, 1997 8.1.1 Imprint-Based Mould Technique: France In the customisedmould technique, two to four sagittally oriented plastic tubes are fixed on the surface of a rigid acrylic applicator, made from an individual impression of the nasopharyngeal cavity (Figure 3) [Chassagne, 1962]. The procedure is performed under neuroleptic analgesia, which allows some preservation of pharyngeal reflexes and muscular tonicity and does not require endotracheal intubation that could limit oropharyngeal access. Nasal secretions are suctioned and the mucous membranes anaesthetised with 5% xylocaine spray until it is possible to manipulate the oropharynx without provoking pain or triggering the gag reflex. Rubber Nelaton catheters are then passed through both nostrils and brought out the mouth. The oral end of the catheters is then tied to cords attached to a dummy applicator, which is much smaller than a normal nasopharyngeal cavity and is thickly coated with a quick-setting silicone paste. By pulling the nasal catheters, the dummy applicator is maneuvered retrogradely through the naso-oropharyngeal passage and into the nasopharyngeal cavity. The silicon paste is allowed to set up for a few minutes and is then extracted from the nasopharynx. An exact impression of the nasopharynx is obtained, demonstrating surface details of the tumour. If the impression appears incomplete, the procedure is repeated after adding more paste in the defective areas. Once an acceptable impression is achieved, the patient is kept under light anaesthesia, while the rigid applicator is fabricated from the impression. A bivalve plaster of Paris negative mould is prepared from the impression. The rigid applicator with walls about 2-5 mm thick is

7.1 Benavides Cancer Institute Approach Brachytherapy simulation

With the applicator in place, non-contrast 1-mm thick axial CT scans are obtained and co-registered with gadolinium-enhanced T1-weighted 1-mm thick axial MR scans if the latter are obtained.

7.2Delineation of clinical target volumes (CTV) and organs-at-risk (OAR) The residual gross tumour volume (GTV-R) is delineated as the residual tumour on the CT simulation scan and the co-registered MRI, the high-risk CTV (HR-CTV) as the GTV-R plus 5mm, and the intermediate-risk CTV (IR-CTV) as 5 mm around the HR- CTV plus the initial extent of disease (GTV at diagnosis, GTV-D), carving out bone and air.

Nasopharynx Cancer

7

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Table 1. UICC/AJCC 8 th Edition Stage Classification System for Nasopharyngeal Cancer a

T0 b No appreciable primary tumour T1 Tumour confined to the nasopharynx, or extension to oropharynx and/or nasal cavity without parapharyngeal involvement T2 Tumour extends to the parapharyngeal space , and/or adjacent soft tissue involvement (medial and lateral pterygoid muscles, prevertebral muscles) T3 Tumour invades bony structures and/or paranasal sinuses T4 Tumour with intracranial extension and/or involvement of cranial nerves, hypopharynx, orbit, parotid gland, and/or extensive soft tissue infiltration beyond the lateral surface of the lateral pterygoid muscle N0 No cervical or retropharyngeal nodal metastases N1 Unilateral cervical nodal metastasis/es and/or unilateral or bilateral retropharyngeal nodal metastasis/es 6 cm or smaller, above the caudal border of the cricoid cartilage

T

N

N2 Bilateral cervical nodal metastases, 6 cm or smaller, above the caudal border of the cricoid cartilage N3 Unilateral or bilateral cervical nodal metastasis/es larger than >6 cm in greatest dimension and/or and/or extension below the caudal border of the cricoid cartilage

M0 No distant metastases M1 Presence of distant metastases

M

a Changes to previous 7 th edition highlighted in bold italics b In cases of EBER-positive metastatic cervical carcinoma of unknown primary for which the nasopharynx is the presumed origin

A

Figure 7. The afterloading mold is flexible in the longitudinal and axial planes. [Law, 2002]

B

Figure 8. The afterloading mould in-situ. A, Lateral view; B, Submento-occipital view. [Law, 2002]

Figure9.Paediatricendotrachealtubetechnique. 1A,paediatricendotracheal tubewithballoon cuff inflated and 1B, deflated; 2A, source carrier with cesium 137 slugs loaded and 2B, unloaded; 3, cesium 137 slugs; 4, pediatric rubber catheter occasionally used to guide endotracheal tube insertion; 5, 5-cc syringe; 6, local anaesthesia oropharyngeal spray; 7, cotton swabs; 8, local anaesthesia and nasal decongestant spray; 9, contrast for balloon filling. [Wang, 1987]

Nasopharynx Cancer

8

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Figure 10. Typical isodoses associated with the paediatric endotracheal tube technique with cesium 137 slugs. Left, lateral view; Right, anteroposterior view. [Wang, 1987]

Figure 11. Balloon applicator with 3-cc balloon. Modified cuffed paedriatic endotracheal tube with more slender catheters to allow for greater patient comfort. [Slevin, 1997]

Figure 12. Sagittal MR T1-weighted image of balloon applicator in-situ , inflated with 4mL diluted Gadolinium contrast. [Slevin, 1997]

Figure 14. Retrograde naso-oropharyngeal insertion of a pre-fabricated nasopharyngeal applicator. Top: Silicon feeding tubes inserted transnasally, through the naso-oropharyngeal passage, into the mouth and attached and secured into the nasal limbs of the applicator. Bottom: The applicator maneuvered into the nasopharyngeal cavity by traction onto the feeding tubes.

Figure 13. Rotterdam nasopharyngeal applicator. A, Old-type Rotterdam nasopharyngeal applicator (RNA); B, New-type RNA, with flanges of both catheters tilted more sideways to allow more lateral dosimetric coverage of the parapharyngeal space. This newer version was never released commercially for clinical use. [Levendag, 2013]

Nasopharynx Cancer

9

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

thenmade in this mould using an acrylic compound.The applicator surface is carefully smoothedwith a grinder and a double nylon cord is attached to its lower posterior aspect to be used later for applicator removal from the nasopharynx. Two rubber strips are attached at its upper anterior aspect at the level of the small projections, which extends into the posterior choanae, where the paste overflowed. These strips would hold the applicator in positionwhen brought out through the nostrils and tied together anterior to the nasal septum. At the time themold technique was employed, iridium (Ir192) wire source afterloading was used. Based on the tumour topography imprinted on the applicator, the radiation oncologist specifies the configuration of the blind-ended plastic tubes upon which wire sources are to be loaded.The tubes are oriented sagittally and spaced 10-15 mm. The radiation oncologist determines the appropriate source lengths and may require 2-4 tubes in an averaged-sized nasopharynx, depending on the tumour extent. Once the applicator has been assembled, the endoluminal trajectory is checkedwith dummy sources.The applicator is inserted in a similar fashion as for the dummy applicator and fixed using immobilization buttons applied on the anterior rubber strips. The positioning in the nasopharyngeal cavity is documented with orthogonal films and the projected dosimetry is verified. The Ir192 wires are then loaded as prescribed. (Figures 4 and 5). 8.1.2 Endoscopy-Based Mould Technique: Hong Kong In contrast to the French approach, this approach only entailed customization of the size of the applicator relative to the dimensions of the nasopharynx of the patient as measured during endoscopy but not of the configuration of the source paths relative to the tumour configuration (Figures 6-8). The endoscopy procedure may entail tumour debulking for better visualization andmeasurement of the nasopharyngeal cavity, as well as to allow better fitting of the applicator and eventual dosimetry. The earlier, preloading version is built on a foam base, while the later, afterloading version is built on a two-jointed silicon and plastic base that allowed for greater flexibility. The insertion procedure is similar to that described for the French approach. 8.1.3 Massachusetts General Hospital Technique The procedure uses two paediatric endotracheal tubes with inner and outer diameters of 5 and 6.8 mm, respectively. After topical anaesthesia of the nasal and the nasopharyngeal mucosa, endotracheal tubes are introduced into the nasopharynx through the nostrils. Under fluoroscopic control, the distal tip of each dummy slugs is placed at the free edge of the soft palate posteriorly and at the posterior wall of the maxillary sinus anteriorly. The inflated balloon, which is attached to the distal end of the endotracheal tube, is used to anchor the tubes and to create a distance between the radiation sources and the nasopharyngeal vault to obtain a better depth dose. The entire implant treatment can be performed as an outpatient procedure (Figures 9 and 10) [Wang, 1987; Wang 1991]. 8.1.4 Nasopharyngeal Balloon Applicators The balloon applicator, inserted transnasally [Slevin, 1997, McLean, 1998] is fixed in place by inflating a balloon concentrically around the catheter (Figure 11).This does not allowpreferential displacement of the catheter away from the soft palate and towards the nasopharynx (Figure 12). Eccentrically placed balloons have been proposed to allow better apposition of the catheter against the nasopharynx [Chang, 2001].These applicators are not compatible withmultiple treatments over several days as the relation of the balloon catheters is not stable or fixed.

8.1.5 Rotterdam Nasopharyngeal Applicator The Rotterdamnasopharyngeal applicator (RNA) is a commercial applicator, made out of soft silicone, which is well-tolerated by the patient (Figure 13)[Levendag, 1997]. It is suitable for applications with stepping-source afterloaders for PDR- or HDR-brachytherapy, as well as for classical low-dose-rate (LDR) techniques. The applicator can remain in situ for the duration of the treatment, which varies from2 to 6 days and can be performed on an outpatient basis in case of HDR brachytherapy. The RNA consists of two silicone tubes fixed on a silicone base that maintains a curved path following the curvature of the nasopharyngeal recess and displaces the catheters away from the soft palate. After local anaesthesia of the nasal cavities and the oropharynx with a 2% xylocaine spray, a flexible guide wire is introduced into one nasal cavity. The end is recuperated with forceps through the oropharynx and brought outside the mouth. The procedure is repeated at the other side. The applicator is then advanced over the guide wires, fixed to them by clamps, and pulled gently through the mouth and the oropharynx into the nasopharynx. The legs of the applicator exit through the nostrils and are fixed with a silicone bridge, pushed against the nasal septum. The applicator tubes are then brought into the applicator and fixed. The applicator placement is secured externally using a silicone flange (Figure 14). A newer design features displacement of the two catheters laterally and away from each other allowing a better coverage of the nasopharyngeal recess [Levendag, 1997]. The tubes have an outer diameter of 15 French (5 mm) and an inner diameter of 9 French (3 mm) and can accommodate standard 6 French afterloading catheters. In both designs, the catheters are inserted into the tubes but are not optimally displaced towards the recess as they are fixed upon the base, the maximum height of which is limited by the diameter of the naso-oropharyngeal passage, that is, the passage bounded by the posterior and lateral walls of the pharynx and the soft palate. Development of newer applicator designs should ideally allow for deployable catheters that can be advanced closer to the nasopharynx and away from the soft palate [Bacorro, 2018]. The procedure is performed under general anesthesia and entails splitting the palate to access the nasopharynx, interstitial implantation radioactive gold grains, and closure of the palate in three layers (Figure 15). This technique has been associated with brisk nasopharyngeal ulceration and necrosis whichmay persist for two to three weeks, and velopalatal insufficiency due to fistulisation or atrophy in 11-19% of the cases [Choy, 1992; Kwong, 2000] 8.2.2 Transnasal Permanent Interstitial Implants The procedure is performed under general anaesthesia [Vikram, 1984]. The soft palate is retracted forward with a retractor, and the nasopharynx is visualised, by means of a fiber optic nasopharyngoscope. After satisfactory visualisation of the tumour site has been obtained, hollow afterloading needles are introduced into the nasal passages and are advanced through the posterior choanae, and inserted into the mucosal surface. Radioactive I-125 seeds are then introduced submucosally through these needles, and the needles are withdrawn. (Figure 16) 8.2.3 Transnasal, Endoscopy-Guided HDR Interstitial Implants The procedure is performed using endoscopic guidance via the 8.2 Interstitial Approaches 8.2.1 Split-palate Interstitial Implantation

Nasopharynx Cancer

10

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Figure 15. Transpalatal flap approach for nasopharyngeal brachytherapy. AU-shaped incision of the palate is performed with preservation of the greater palatine vessels bilaterally. The sutures on the palate flap help provide the retraction for direct visualization of the nasopharynx. The tongue is retracted with a Dingman mouth gag. A trocar would then be inserted for permanent implantation into the nasopharyngeal tumor. [Han, et al, in Devlin, 2007]

Figure 16. Transnasal permanent interstitial seed implantation. A, tongue depressor; B, nasopharyngoscope; C, palate retractor; D. Implant needles; E, Implantation gun. Orotracheal tube employed for general anesthesia not shown. [Vikram, 1984]

Figure 17. Endoscopy-guided interstitial brachytherapy. A, The instruments and applicators used in the procedure: nasal endoscopes, fixator buttons, interstitial sharp plastic needles, plastic needles obturators, needle holder, needle and suture; B, The nasal outward view of applicators sewed to the nose wings.; C, CT scan acquisition with the applicator and dummy sources in-situ. D, Axial CT image of the residual tumor, interstitial needle and isodose line. The red arrow indicates the 100% isodose curve covering the whole GTV. E, The 3D reconstruction image of two needles encompassing the residual tumor (outlined by red fine grid lines) in coronal section. F, Treatment delivery by afterloading system. [Wan, 2014]

A

B

C

D

E

F

Nasopharynx Cancer

11

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Table 2. The classic Levendag system [Levendag, 1997] Point

Designation

Definition

Targets Nasopharynx, right and left

Na (R)(L)

Intersection of the Pa-BOS line with the bony outline of the base of skull Intersection of the line drawn between the anterior clinoid process and point R and the line drawn from the contralateral outer canthus and tragus Ventral part of corpus of the atlas

Base of skull, right and left

BOS (R)(L)

Node of Rouviére

R

Organs at Risk Pituitary Optic chiasm Retina, right and left Cord Palate, right and left

P OC Re (R)(L) C Pa (R)(L)

0.5 cm from center of sella 1.5 cm ventrally from P 1 cm posterior to the line drawn from the contralateral outer canthus and tragus Posterior to R at the posterior border of the atlas Junction of soft and hard palate

Figure 18. Orthogonal radiographs of the Rotterdam applicator in-situ. Left, lateral view; Right, anteroposterior view. A modified Rotterdam applicator, to which incorporated a thin lead shield at the base, is used, to decrease soft palate dose. [Courtesy of Dr. Duc Hoang Lam, Ho Chi Minh Oncology Center].

Table 3. Dose Prescription and Constraints at the Ho Chi Minh Oncology Center a [Courtesy of Dr. Duc Hoang Lam] Point Designation Dose Constraints Targets Nasopharynx, right and left Node of Rouviére Na (R)(L) R 95-105% PD 95-105% PD

Organs at Risk Pituitary Optic chiasm Retina, right and left Cord Palate, right and left

P OC Re (R)(L) C Pa (R)(L)

≤30% PD ≤25% PD ≤20% PD ≤30% PD ≤110% PD

a A modified Rotterdam applicator, to which incorporated a thin lead shield at the base, is used, to decrease soft palate dose. PD, prescription dose. At the Ho Chi Minh Oncology Center, boost for definitive RT is given as four 3-4 Gy fractions given twice daily, and palliative brachytherapy as three to four 5-7 Gy fractions given once weekly.

Nasopharynx Cancer

12

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Table 4. Sample Dosimetry for a T2 Case

Case 1 Prescription dose: 3.5 Gy Rotterdam applicator

Levendag 2D pointsa

Physical Dose (Gy)

IGBT volumes

Physical Dose (Gy)

EQD2 b (Gy)

Targets

HR-CTV D95 HR-CTV D90 HR-CTV D85 IR-CTV D95 IR-CTV D90 IR-CTV D85

3.2 3.5 3.8 2.4 2.7 3.0 1.1 0 0 0 1.7 1.1 2.3 1.9 2.0 2.1 6.4 5.2

3.5 3.9 4.4 2.5 2.9 3.3 0.9 0 0 0

Na(R)/Na(L) BOS(R)/BOS(L) R

3.8/4.0 2.7/2.6 6.3

Organs-at-risk

Brainstem D2cc Spinal cord D2cc Pituitary D2cc

C P OC

2.5 0.84 0.59 0.54 0.52

Optic chiasm D2cc Retina Right D2cc Retina Left D2cc Clivus D2cc Clivus DMean Atlantoaxial joint D2cc Atlantoaxial joint DMean Soft palate D2cc Soft palate DMean

Re(R) Re(L)

1.6 0.9 2.4 1.9 2.0 2.1 12.0c 8.5

Pa(R) /Pa(L)

4.5/4.9

a Na – nasopharynx (intersection of the Pa-BOS line with the bony outline of the base of skull), P – pituitary (0.5cm from center of sella), OC – optic chiasm (1.5cm ventrally from P), Re – retina (1 cm posterior to the line drawn from the contralateral outer canthus and tragus), C – cord (posterior to R at the posterior border of corpus C1), R – node of Rouviére (ventral part of corpus C1), Pa – palate (junction of soft and hard palate), BOS – base of skull (intersection of the line drawn between the anterior clinoid process and point R and the line drawn from the contralateral outer canthus and tragus),(Levendag,1997). b Per fraction, equivalent dose in 2Gy using αβ ratio of 10 for the tumour, and αβ ratio of 3 for organs-at-risk. c Soft palate doses could be further decreased by incorporation of a lead shield onto the base of the Rotterdam applicator. The lead shield should be insulated to prevent leaching. Care should be taken during insertion to avoid injury to the uvula.

Three-dimensional planning Related to the CTV, dummy sources are loaded into the plastic tubes and their correct position verified with fluoroscopy. Orthogonal radiographs are taken to document the applicator placement and source position, if multiple treatments are intended for a single insertion. One- to three-millimeter CT scan cuts are carried out through the central plane of the sources. MR imaging at diagnosis or prior to external radiotherapy and/or after external radiotherapy may be co-registered in order to aid delineation of target volumes and organs at risk.

inferior meatus. The interstitial technique is used to complement intracavitary techniques to improve coverage of parapharyngeal involvement and entails insertion of sharp plastic needles. Three- dimensional treatment planning is then employed followed byHDR treatment delivery (Figure 17) [Ren 2014, Wan 2014].

9. TREATMENT PLANNING

Benavides Cancer Institute Approach Prescription, optimization and dosimetry

Two-dimensional planning Two-dimensional planning entails acquisition of orthogonal radiographs (anteroposterior and lateral) taken with the applicator and dummy wire markers in-situ, and wire markers taped onto the lateral canthi (Figure 18). The most commonly used system is the Levendag system which defines the following prescription and monitoring points outlined in Table 2. At the Ho Chi Minh Oncology Center, nasopharyngeal ICBT, either as boost in the primary setting or as exclusive salvage treatment for recurrences, is performed using a modified Rotterdam applicator, using 2D planning based on the Levendag system (Table 3).

Dose prescription is to the HR-CTV. Graphical optimization is performed with the following planning objectives HR-CTVD90% ≥100% of prescribed dose (PD)(high priority); IR-CTV D90% ≥75% PD (intermediate priority); and soft palate D2cc <120% PD or as low as possible (intermediate priority). Treatment planning is conducted using Oncentra Brachy version 4.5.3, 2018, Elekta AB, Stockholm, Sweden.

Nasopharynx Cancer

13

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Table 5. Brachytherapy boost protocols for primary disease Levendag, 1997

BCI protocol

Dose (EQD2, αβ =10)

Dose (EQD2, αβ =10)

T-stage

EBRT

BT

SEQD2

EBRT

BT

SEQD2

60Gy/30F (60.0Gy 10 ) 70Gy/35F (70.0Gy 10 )

18Gy/6F (19.5Gy 10 ) 12Gy/4F (13.0Gy 10 )

66Gy/33F (66.0Gy 10 ) 70Gy/33F (70.7Gy 10 )

12-14Gy/4F (13.0-15.8Gy 10 ) 12-14Gy/4F (13.0-15.8Gy 10 )

T1

79.5Gy

79.0-81.8Gy

10

10

T2

83.0Gy

83.7-86.5Gy

10

10

F, fractions; EQD2, dose equivalent in 2Gy; Σ, cumulative

Table 6. Brachytherapy monotherapy regimens for persistent and recurrent disease Previous EBRT GEC-ESTRO (Mazeron, 2009)

BCI protocol

EQD2, αβ =10

Dose (EQD2, αβ =10)

Dose (EQD2, αβ =10)

BT

ΣEQD2

BT

ΣEQD2

12-18Gy/4-6F (13.0-19.5Gy 10 ) ≥60Gy LDR-PDR (≥60.0Gy 10 )

12Gy-14/4F (13.0-15.8Gy 10 ) 50/10F; 56/14-16F

Persistence

66.0-70.0Gy

73.0-79.5Gy

79.0-86.5Gy

10

10

10

0.0Gy 10 (Ignored)

Recurrence

≥60.0Gy

62.4-65.8Gy

(62.6Gy

; 62.4-65.8Gy

)

10

10

10

10

Persistent disease would pertain to that treated to 66-70Gy, completed within the last 2-3 months. Recurrent disease would pertain to that with previ- ously documented complete response to definitive radiotherapy, now with documented recurrence. Recurrent disease considered for full brachytherapy monotherapy dose shall ideally be at least 6 months from previous radiotherapy course. F, fractions; EQD2, dose equivalent in 2Gy; Σ, cumulative

10. DOSE, DOSE RATE, FRACTIONATION

Dosimetric outcomes The following parameters are monitored and noted: HR-CTV D95%, D90% and D85%; IR-CTV D95%, D90% and D85%; D2cc of the brainstem, spinal cord, pituitary, optic chiasm, and retina; and D2cc and mean dose of the clivus, atlanto-axial joint and soft palate. Sample Case Table 4 summarises the comparison of the classic Levendag target andOAR point doses with dose-volume parameters on 3Dplanning for a rT2 case.The Na and BOS point doses correlate poorly with the HR-CTVD90. Specifically, the former overestimates the HR-CTV D90 by 9-14%, which could lead to clinically significant HR-CTV underdosage. The C points overestimate the D2cc for the spinal cord, which receive negligible doses. Similarly, the pituitary and optic chiasm receive negligible doses.The retina D2cc is significant but is underestimated by the Re point dose; routine monitoring of the retina D2cc is recommended. Significant volumes of the soft palate receive important doses (bothD2cc andDmean >3Gy) which are underestimated by the Pa point. Similarly, significant volumes of the clivus and the atlanto-axial joint receive significant doses (D2cc and Dmean in the 1.3-2.3 Gy range), but are not monitored in OARs in the Levendag system. For these OARs that are in close proximity to the applicator, routine D2cc and Dmeanmonitoring and optimizing to keep the doses to minimum are recommended. In cases of re-irradiation, routine monitoring and minimization of doses to all of the above OARs are recommended.

Some literature data are strongly suggestive for the existence of a dose-response relationship above 65 Gy in NPC. However, dose escalation is limited by OAR tolerances. Brachytherapy can be used to deliver an additional dose to a small volume after a full course of EBRT. Wang delivers with LDR afterloading intracavitary implant 7-12 Gy at 5 mm below the mucosa [Wang, 1997]. When brachytherapy is carried out for a recurrent NPC in a previously irradiated area, 60 Gy are delivered in roughly 6 days with a LDR or PDR technique. Early HDR brachytherapy regimens, which entailed delivery of 5-8 Gy fractions, were associated with acceptable toxicity rates. Levendag et al. deliver after a rest period of 1-2 weeks, a boost dose to the primary site with HDR brachytherapy, with two 3-Gy fractions daily with a 6-hour interfraction interval. Total of fractions is 6 (after 60 Gy EBRT) for T1 - 3 tumours, and 4 (after 70 Gy EBRT) for T4 tumours. In order to further risks of late toxicity, especially with re-irradiation, it is now recommended to give HDR treatments in 3.0-4.0 Gy twice-daily fractions with at least a 6-hour interfraction interval [Mazeron, 2009; Kovacs, 2017]. With the emergence of cross-sectional imaging and 3D brachytherapy planning techniques, combined with the use of stepping-source techniques, better evaluation of dose-volume parameters and thus optimization are now possible. Improvement of intracavitary applicator designs and combination with interstitial techniques can potentially improve dosimetry and outcomes.

Nasopharynx Cancer

14

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Table 7. Brachytherapy (BT) as boost to External Beam Radiotherapy (EBRT): Pre-Chemoradiation Era Study Number of patients (Boost: No boost) Patient Age T- and N- Stage RT course ΣEQD2

Local Control and Survival

Toxicity

Retrospective cohorts Chang (1996) 183 (133:50)

>40 (64%)

T1-T2, 100% EBRT (2D): Boost:

EBRT: Boost: 64.8-68.4Gy No boost: 68.4-72Gy

5-year LRFS Boost: 86.8% No boost: 73.8%

64.8-68.4Gy/36-38F No boost: 68.4-72Gy/38-40F BT (2D ICBT): 5-5.5Gy/F, 1-3F Rx pt: 2cm from CAX

10

10

BT: 6.3-7.1Gy

/F, 1-3F

10

Teo (2000)

509 (163:346)

Boost: >40 (60%) No boost: >40 (66%)

Boost: T1, 45% T2, 55%

EBRT (2D): 60Gy/24-30F

EBRT: 60-62.5Gy

5-year LC Boost: 94.6% No boost: 90% P=0.016 5-year CSS Boost: 88.1% No boost: 83.6% P=0.043 3-year LRFS Boost: 86% No boost 94% P=0.23 3-year CSS Boost: 76% No boost: 89% P=0.29 5-year LRFS Boost: 95.8% No boost: 88.3% P=0.020

NPUN Boost:

6%

10

No boost: <1% P=<0.001

N0-1, 74% N2-3, 26% No boost: T1, 62% T2, 38% N0-1, 59% N2-3, 40% Boost: T1-2, 73% T3-4, 27% N0-1, 60% N2-3, 40% No boost: T1-2, 56% T3-4, 44% N0-1, 47% N2-3 53%

BT (2D ICBT): 24Gy/3F (n = 94) 18Gy/3F ( n = 58)

BT: 36Gy 24Gy

(n = 94) ( n = 58)

10

10

8-46Gy/1-7F (n = 11) Rx pt: 1cm from CAX

(n = 11)

8-63Gy

10

Ozyar (2002)

144 (106:38)

Boost: >40 (60%) No boost: >40 (40%)

EBRT (2D): 66Gy/33F

Boost: 80.1Gy

10

No boost: 66Gy 10

BT (2D ICBT): 12Gy/3F Rx pt: 1cm from CAX

Leung (2008)

287 (145:142)

Boost: 48.5 (22-78) No boost: 43.5 (24-76)

Boost: T1-T2a, 94% T2b, 6% N0-1, 89% N2-3, 11% No boost: T1-2a, 93% T2b, 7% N0-1, 86% N2-3, 14%

EBRT (2D): 66Gy

Boost: T1-2a: 78.6Gy 10 T2b: 82Gy 10 No boost: 66Gy 10

CNP Boost: 4% No boost: 8% TLN Boost: 0 No boost: 2% NPUN Boost: <1% No boost: 0 Endocrine Boost: 4% No boost: 10%

BT (2D ICBT): T1-2a: 10Gy/2F T2b: 12Gy/2F Rx pt: 1cm from CAX

5-year OS Boost: 91.1% No boost: 79.6% P=0.006

5-year major complication- free rate

Nasopharynx Cancer

15

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Boost: 89.5% No boost: 85.6% P=0.23 NPUN Boost: 18% No boost: 19%

Ren (2010)

141 (40:101)

Boost: >45 (43%) No boost: >45 (46%)

Boost: T2b, 100% N0-1, 53% N2-3, 48% No boost: T2b, 100% N0-1, 57% N2-3, 43%

EBRT (2D): Boost: 60Gy No boost: 68Gy

Boost: 77Gy

5-year LRFS Boost: 97.5% No boost: 80.2% P=0.012 5-year OS Boost: 98.3% No boost: 91.9% P=0.231 3-year LC Boost: 99.2% No boost: 94.1% P=0.030 3-year OS Boost: 91.3% No boost: 80.4% P=≤0.001

10

No boost: 68Gy 10

BT (3D ISBT): 16Gy/6F to 100% of GTV

Wu (2013)

348 (175:173)

Boost: 44 (22-69) No boost: 44 (18-74)

Boost: T1, 18% T2, 82%

EBRT (2D): Boost: 56-60Gy/28-30F No boost: 70-72Gy/35-36F

EBRT: Boost: 56-60Gy 10 No boost: 70-72Gy 10 BT: 10.4-26Gy

G ≥2 trismus Boost: 7% No boost: 21% P=<0.001

N0-1, 71% N2-3, 29% No boost: T1, 27% T2, 73% N0-1, 37% N2-3, 63%

BT (2D ICBT): 10-25Gy/4-10F BID ≥6 hours apart Rx pt: 1cm from CAX

G ≥2 neck fibrosis Boost: 13% No boost: 27% P=0.002 CNP Boost: 11% No boost: 23% P=0.003 TLN Boost: 1% No boost: 8% P=0.008 NPUN Boost: 2% No boost: 0% P=0.123

10

F, fractions; 2D, two-dimensional; BID, twice daily; ICBT, intracavitary brachytherapy; ISBT, interstitial brachytherapy; Rx pt, prescription point; CAX, central axis; ΣEQD2, cumulative dose equivalent in 2Gy; LRFS, local recurrence free survival; OS, overall survival; CSS, cancer-specific survival; NPUN, nasopharyngeal ulceration or necrosis; CNP, cranial neuropathy; TLN, temporal lobe necrosis.

Nasopharynx Cancer

16

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 10/09/2019

Table 8. Brachytherapy (BT) as boost to External Beam Radiotherapy (EBRT): Chemoradiation Era Study Number of patients (Boost: No boost) Patient Age T- and N- Stage RT course ΣEQD2

Local Control and Survival

Toxicity

Randomized Controlled Trial Rosenblatt (2014) 274 (135:139)

Boost: 40 ±14.8 No boost: 43.5 ±13.6

Boost T3-4N2-3: 26.7% No boost T3-4N2-3: 24.5%

EBRT (2D): 70Gy/35F

Boost: HDR: 112.8Gy 10 LDR: 81Gy 10 No boost: 70Gy 10

3-year LRFS Boost: 54.4% No boost: 60.5% P=0.647

Any G 3-4 late toxicity Boost: 24% No boost: 22%

BT (2D ICBT): HDR: 27Gy/3F LDR: 11Gy

Rx pt: Tumor tissue point, TT; Levendag system

3-year OS Boost: 63.3% No boost: 62.9% P=0.742

Retrospective cohort Chao (2017)

232 (124:108)

Boost: >50 (49%) No boost: >50 (51%)

Boost: T1–2, 88% T3, 12% N0-1, 65% N2-3, 35% No boost: T1-2, 83% T3, 17% N0-1, 57% N2-3, 31%

EBRT (IMRT): 70Gy/35F

Boost: 86Gy

5-year LC Boost: 94.3% No boost: 88.7% P=0.228

10

No boost: 70Gy 10

BT (2D/3D ICBT): 12Gy/2F Rx pt: 3-5mm beneath balloon surface (2D); 90% of CTV (3D)

F, fractions; 2D, two-dimensional; ICBT, intracavitary brachytherapy; Rx pt, prescription point; CAX, central axis; ΣEQD2, cumulative dose equivalent in 2Gy; IMRT, intensity-modulated radiotherapy; LRFS, local recurrence free survival; LC, local control; OS, overall survival.

12. RESULTS

Benavides Cancer Institute Approach Given the scarcity of literature on nasopharyngeal IGBT, we have devised our protocols based on our current IMRT protocols and on published 2D and LDR BT protocols, and modifying the BT fractionation based on GEC-ESTRO guidelines for HDR BT for head-and-neck cancers [Mazeron, 2009], resulting in cumulative doses as detailed in Tables 5 and 6.

A significant RT dose-response relationship was observed in retrospective studies among patients treated with 2D EBRT techniques, with better local control (LC) among patients treated to ≥70 Gy [Mesic, 1981; Perez, 1992]. A total dose of 77-81 Gy has been recommended if treating with RT alone [Levendag, 2002]. In the pre-chemoradiation era, several retrospective cohorts showed that for T1-T2 disease, dose escalation by ICBT [Chang, 1996; Teo, 2000; Leung, 2008; Wu, 2013], as well as ISBT [Ren, 2010], after EBRT has led not only to improved LC, but also survival, with similar or decreased toxicity (Table 5). On the other hand, a retrospective cohort that included both T1-2 and T3-4 disease found similar LC and survival rates with the addition of ICBT boost [Ozyar, 2002], suggesting the ineffectiveness of ICBT in more advanced T disease. In the chemoradiation era, excellent local control rates (up to 85.8% overall 8-year local failure free survival; T1: 91.7%, T2: 88.2%, T3: 87.2%; T4, 71.6%) have been achieved leading to a decline in ICBT use [Au, 2018; Lee, 2015]. For stage III and IV disease, a multi-center trial examining dose escalation by ICBT after neoadjuvant chemotherapy and concurrent CRT did not demonstrate significantly improved outcomes, including LC for T1-T2 tumours, compared to the latter regimen alone [Rosenblatt,

11. MONITORING

During the few days of the application, the patients receive sedatives and a liquid or pureed diet. Headache may occur but a correctly made applicator remains well positioned and does not interfere with sleep. It is very important to check the applicator position, clinically or radiologically, and to suction the secretions, which accumulate at the nares, several times daily.