Keywords

1 Introduction

Nasopharyngeal carcinoma (NPC) is relatively uncommon in most parts of the world but is endemic to certain regions such as Southern China [1]. NPC is rare in the United States, with an incidence of less than 1/100,000 person-years compared with 27/100,000 person-years in Southern China.

NPC is unique histologically from other head and neck cancers. In the most recent World Health Organization classification in 2005, NPC comprises three main types, namely, keratinizing squamous cell carcinoma (type 1), non-keratinizing carcinoma (type 2), and basaloid squamous cell carcinoma [2]. Non-keratinizing carcinoma (type 2) is subdivided into differentiated (type 2a) and undifferentiated (type 2b). Type 2 is also strongly associated with Epstein-Barr virus (EBV) and is the most common histologic type found in endemic regions.

Intergroup 0099 established concurrent radiation therapy (RT) and chemotherapy as the standard of care for locally advanced NPC [3]. Although surgical resection is often an option for tumors at other head and neck sites, successful resection of NPC is nearly impossible given its location and frequent involvement of the lateral retropharyngeal lymph nodes. Thus, surgery is mostly limited to radical or selective neck dissections for persistent or recurrent disease after RT.

Toxicity is an issue with the use of conventional RT given the proximity of the nasopharynx to critical normal structures. Intensity-modulated radiation therapy (IMRT) offers advantages over conventional RT by optimizing the delivery of radiation to irregularly shaped volumes so as to spare organs at risk. Further, different doses can be delivered simultaneously to select regions by dose painting or a simultaneous integrated boost. These techniques allow increased sparing of nearby critical normal structures by simultaneously delivering higher radiation doses to gross disease and lower doses to regions suspected of harboring microscopic disease. In the next section, we evaluate the clinical evidence that has established IMRT as the standard of care for definitive RT in NPC.

2 Clinical Evidence for Intensity-Modulated Radiation Therapy

2.1 Dosimetry

Dosimetrically, IMRT improves coverage of disease while reducing the dose to the numerous critical adjacent structures relative to conventional RT (Fig. 8.1) [4, 5]. Dosimetric comparisons of IMRT versus two-dimensional (2D) RT and three-dimensional (3D) RT plans showed that IMRT led to lower doses to the spinal cord, mandible, temporal lobe, parotid glands, optic chiasm, and brainstem.

Fig. 8.1
figure 1

Dosimetric comparison of treatment plans for intensity-modulated radiation therapy (IMRT) vs. 3D-conformal RT vs. traditional RT. Axial dose distributions through the center of the nasopharynx and neck for IMRT (left), 3D-conformal RT (middle), and traditional treatment plans (right). Note the relatively poor coverage of the skull base and medial nodal regions in the traditional plan and the improved dose conformality of the IMRT plan (From: Hunt et al. [4], with permission from Elsevier)

Full size image

2.2 Salivary Function and Treatment Compliance

The most common complication of RT for NPC is a decline in salivary function, known as xerostomia, due to a damage of nearby salivary structures. The symptoms of xerostomia can significantly affect a patient’s quality of life [6]. The severity of xerostomia depends mostly on the dose and volume of salivary gland within the radiation field. Dosimetric comparisons have revealed that a mean dose of 26 Gy or less to the parotid glands is necessary to preserve salivary function [7, 8].

The main benefit of IMRT over conventional RT for NPC is the ability to spare the parotid glands. Two phase III randomized controlled trials assigned patients to receive either 2D RT or IMRT with parotid-sparing techniques and evaluated outcomes at 1 year after treatment. The first trial found that IMRT was associated with superior quality of life outcomes [9]. The second study also found benefits in observer-rated xerostomia outcomes and preservation of parotid function (measured by parotid flow rate) with the use of IMRT [10], as well as a trend towards improvement in patient-reported xerostomia outcomes. The same study revealed the somewhat surprising finding that xerostomia quality of life scores only correlated weakly with both salivary flow rates and observer-rated xerostomia outcomes. Therefore, evaluation of both patient-reported and physician-reported outcomes remains important. Regardless, both phase III studies showed improved xerostomia outcomes with the use of IMRT compared with conventional 2D RT.

The lesser toxicity associated with IMRT may also improve treatment compliance or the ability of patients to tolerate the prescribed therapy. A multi-institutional trial of IMRT by the Radiation Therapy Oncology Group, RTOG 0225, showed that 90 % of patients were able to receive the full 70-Gy prescribed dose and that 88 % of the patients with T2b or higher or N+ disease were able to receive the full three cycles of concurrent cisplatin [11]. These findings compare favorably to previous studies that used conventional RT techniques, for example, chemotherapy compliance rates were 63 % in the Intergroup 0099 trial, 71 % in a Singapore randomized trial, and 52 % in the Hong Kong NPC-9901 trial [3, 12, 13].

2.3 Disease Control

In addition to improving toxicity outcomes, excellent disease control outcomes have been reported by several institutions. Lee et al. reported findings from an initial series of patients with NPC treated with IMRT, with an incredible 4-year local control rate of 97 %, despite 70 % of patients in that study having locally advanced disease [14]. Kwong et al. reported the first prospective series, with 3-year outcomes of 100 % local control (LC), 92.3 % regional control, and 100 % overall survival (OS) rates [15]. These excellent outcomes are supported by additional published series from many individual institutions, comprehensively reviewed by Wong et al. (Table 8.1) [16]. The RTOG 0225 trial further demonstrated the feasibility of implementing IMRT techniques across the multiple US institutions [11]. That phase II study reported excellent 2-year outcomes of 93 % LC, 89 % local-regional control (LRC), and 80 % OS rates (Table 8.1).

Table 8.1 Published studies of IMRT for nasopharyngeal cancer
Full size table

Notably, other factors may contribute to improvements in LRC associated with IMRT, including the use of chemotherapy, better supportive care, and technologic advances in imaging that provide better tumor delineation. Other limitations associated with historical comparisons include changes in the criteria for disease staging over time as well as improved staging with the use of magnetic resonance imaging (MRI) and positron emission tomography (PET) [17]. For example, because MRI is more sensitive than computed tomography (CT) for detecting minimal skull base involvement or intracranial extension, the T category tends to be upstaged when MRI is used rather than CT.

3 Techniques

3.1 Diagnostic Work-Up for Target Volume Delineation

Disease staging should include both CT and MRI of the head and neck. CT is important for assessing cortical bone involvement; MRI provides superior visualization of skull base involvement and tumor invasion into soft tissue structures compared with CT [18]. Infiltration of disease into the bone marrow is best seen as hypodense regions relative to normal marrow on T1-weighted non-contrast MRI scans. Fusion of the skull base portion of the CT scan with the MRI scan should aid in delineating the gross tumor volume (GTV). MRI also allows retropharyngeal lymph nodes to be distinguished from primary tumor, whereas CT may not.

Enlarged retropharyngeal lymph nodes should be considered a gross disease. Involvement of other lymph node regions is suggested by the presence of central necrosis, extracapsular spread, or nodal diameters of 1 cm or more. PET/CT may help to clarify involvement of borderline lymph nodes. Generally, because NPC has a high likelihood of nodal spread, any nodes suspected of harboring disease should be considered a gross disease.

3.2 Simulation and Daily Localization

The patient should be set up for treatment simulation supine, with the neck extended. The immobilization device to be used should include at least the head and neck; if possible, shoulders should also be immobilized to ensure the reproducibility of patient setup from day to day, especially when an extended-field IMRT plan is to be used. A bite block can be placed during treatment simulation and throughout treatment to move the tongue away from the high-dose regions in the nasopharynx.

CT-based treatment simulation should involve 3-mm-thick scan slices with intravenous contrast to help delineate the GTV, particularly the lymph nodes. The isocenter is typically placed immediately above the arytenoids. Image registration and fusion applications with MRI and PET should be used to help delineate target volumes, especially regions of interest that encompass the GTV, skull base, brainstem, and optic chiasm.

3.3 Target Volume Delineation and Treatment Planning

Several IMRT dose-fractionation regimens have been used for NPC (Table 8.2). Excellent LRC rates in excess of 90 % have been reported with the use of these regimens.

Table 8.2 IMRT dose and fractionation schemes
Full size table

Several acceptable definitions of target volumes, including the GTV, the clinical target volume (CTV), and planning target volume (PTV), have been used at different institutions, as reviewed by Wong et al. [16]. The RTOG established a guideline for target volume delineation with RTOG 0225, which was successfully implemented in that multi-institutional study [11]. Suggested target volumes for the GTV and high-risk CTV are described in the following sections (Tables 8.3 and 8.4 and Figs. 8.2, 8.3, and 8.4). In a recent RTOG 0615 trial, the lower-than-expected 2-year LRC rate of 84 % was attributed to an increased incidence of major deviations in target volume [19]. Thus, attention must be paid to accurate target delineation to avoid marginal misses when using IMRT.

Table 8.3 Definitions of target volumes for gross disease
Full size table
Table 8.4 Definition of target volumes for high-risk subclinical region
Full size table
Fig. 8.2
figure 2

Delineation of target volumes in a case of T1N1 nasopharyngeal carcinoma (NPC). GTV70 (inner contour, red) and CTV59.4 (green) contours in a patient with T1N1 NPC with coverage of the retropharyngeal and level II nodes (Figure 1.2 from: Lee NY, Le QT, O’Sullivan B, Lu JJ (2003) Chapter 1 Nasopharyngeal Carcinoma. Target Volume Delineation and Field Setup: A Practical Guide for Conformal and Intensity-Modulated Radiation Therapy, with kind permission from Springer Science + Business Media)

Full size image
Fig. 8.3
figure 3

Delineation of target volumes in a case of T3N2 nasopharyngeal carcinoma (NPC). PTV70 (red) and PTV59.4 (green) in a patient with T3N2 nasopharyngeal carcinoma (Figure 1.4 from: Lee NY, Le QT, O’Sullivan B, Lu JJ (2003) Chapter 1. Nasopharyngeal carcinoma. Target Volume Delineation and Field Setup: A Practical Guide for Conformal and Intensity-Modulated Radiation Therapy, with kind permission from Springer Science + Business Media)

Full size image
Fig. 8.4
figure 4

Delineation of target volumes in a case of T3N2 nasopharyngeal carcinoma (NPC) with the use of different CT window settings. GTV70 (green) and CTV59.4 (red) in bone window (left) and soft tissue window (right) (Figure 1.5 from: Lee NY, Le QT, O’Sullivan B, Lu JJ (2003) Chapter 1. Nasopharyngeal carcinoma. Target Volume Delineation and Field Setup: A Practical Guide for Conformal and Intensity-Modulated Radiation Therapy, with kind permission from Springer Science + Business Media)

Full size image

Reductions in high-risk subclinical volumes with IMRT have also been described. Lin et al. reported a prospective, single-institution study involving 323 patients with NPC; that study reduced the CTV suggested in the RTOG guidelines and resulted in excellent LCR outcomes [20]. One reduction involved the exclusion of upper deep jugular lymph nodes (level IIa above the C1 vertebrae) in the CTV. Treatment volume reductions may be important for reducing toxicity and even secondary primary tumors, the rate of which has been reported to be as high as 1 % among patients with NPC receiving definitive RT [21]. The next section reviews the guidelines used at the authors’ institutions and some variations in those guidelines used at other institutions.

3.3.1 Gross Tumor Volume

Generally, the GTV is defined as the primary tumor and any involved lymph nodes. Involved lymph nodes are typically defined as any lymph node larger than 1 cm in diameter or those that show avidity on PET scanning.

Expansions around the GTV have included those for both a CTV and a PTV or a single, larger PTV expansion alone. The RTOG studies recommended the use of a CTV70, defined as a 0.5-cm margin with an optional posterior margin reduction of 0.1–0.5 cm (Table 8.3) as well as a PTV70 expansion of 0.5 cm. Variations on these expansions have included a larger CTV expansion of 1 cm [15, 22, 23] or the elimination of a CTV and the use of a larger PTV of 1 cm [12, 24]. The use of the latter method may avoid confusion with the CTVs described below for high-risk and low-risk subclinical regions.

3.3.2 High-Risk and Low-Risk Subclinical Regions

The CTV is generally defined as regions at high risk of harboring microscopic disease (Table 8.4). This volume can be treated to a lower dose of 59.4 Gy (CTV59.4), which includes all potential routes of spread for primary and nodal disease. Specifically, CTV59.4 typically covers the clivus, skull base, inferior sphenoid sinus, cavernous sinus, pterygoid fossae, parapharyngeal space, posterior nasal cavity and maxillary sinus, retropharyngeal lymph nodes, and neck levels II through V. The bilateral level IB can be spared in carefully selected patients without compromising LRC [20, 25]. Whether the inferior orbital fissure or the anterior arch of C1 can be spared remains unclear owing to a lack of data [16]. Variations also exist for the inferior border of the retropharyngeal lymph nodes. A consensus guideline published by Gregoire et al. defines the border as the cranial edge of the hyoid bone [26], but others have described it as the inferior border of the hyoid bone [22] and the cranial edge of the second cervical vertebrae [20, 25].

The low anterior neck can also be treated to a lower dose than the GTV because it is at low risk of harboring disease. This low-risk region can be treated separately with a dose of 50.4 Gy in 1.8 Gy per fraction using conventional anteroposterior (AP) or posteroanterior (PA) portals or with a dose of 54 Gy (CTV54) in 1.64 Gy per fraction in a single IMRT plan.

Finally, an additional CTV (CTV63) can be used at the discretion of the treating physician. A lower dose (63 Gy) can be used for a small-volume lymph node disease. Examples of the appropriate application of this intermediate dose would include the presence of small lymph nodes near the mandible or in the lower neck and close to the brachial plexus.

3.3.3 Planning Target Volume

The margin for the PTV also varies between institutions [16]. Most institutions have described the PTV as 0.2–0.5 cm beyond the CTV. The use of a PTV margin of 0.3–0.5 cm would be reasonable, as many published studies have shown an efficacy using these limits. Daily image guidance with kV imaging can facilitate margin reduction.

3.4 Plan Assessment and Dose Constraints

For NPC, the organs at risk include the brainstem, spinal cord, optic nerves, chiasm, parotid glands, pituitary, temporomandibular (TM) joints, middle and inner ears, skin (in the region of the target volumes), oral cavity, mandible, eyes, lens, temporal lobe, brachial plexus, esophagus (including postcricoid pharynx), and glottic larynx (Table 8.5). In cases of advanced disease, we typically prioritize normal structure constraints, specifically the brainstem, spinal cord, and optic chiasm, over full coverage of the tumor. Ideally, at least 95 % of the PTV70 should receive 70 Gy. In addition, the minimum dose to 99 % of the CTV70 should be >65.1 Gy. The maximum dose received by 0.03 cm3 of the PTV70 should be <80.5 Gy.

Table 8.5 Normal tissue dose constraints
Full size table

For the PTV59.4, 95 % of the volume should receive the prescription dose. The minimum dose to 99 % of the CTV59.4 should be >55.2 Gy. The maximum dose to 0.03 cm3 of PTV59.4 should be 69.3 Gy.

4 Future Directions

Overall, LRC with IMRT is excellent, with rates generally exceeding 90 % in the current era when chemotherapy is included as part of the treatment. Future directions in therapy are now focusing on identifying patients with NPC who are more likely to experience local regional or, more commonly, distant failure after RT. These high-risk patients are likely to benefit from treatment intensification.

Monitoring levels of EBV DNA in plasma samples is one way to stratify patients in terms of risk, as this biomarker is showing great potential in the clinical setting. Many studies, including prospective and phase II studies, have established that pretreatment and posttreatment levels of EBV DNA are reliable indicators of tumor burden, predictors of recurrence and distant failure, and independent prognostic factors in EBV-related NPC [27–32]. Quantification of plasma EBV DNA has also been shown to be useful for monitoring patients with NPC and predicting the outcome of treatment [33]. A recent four-center study sought to harmonize EBV DNA assay methods, to bring us step closer to using EBV in biomarker-driven trials [34]. Indeed, we anticipate that an upcoming phase III study by the RTOG will incorporate plasma EBV DNA levels in treatment stratification.

Several imaging methods are also being used to identify patients with high-risk NPC that is more aggressive and more likely progress despite treatment. On the basis of evidence linking hypoxia with radioresistance [35], Chao et al. tested a PET-based technique to measure hypoxia with a Cu-ATSM [Cu(II)-diacetyl-bis(N(4)-methylthiosemicarbazone] tracer and considered the results promising [36]. Lee et al. also demonstrated the feasibility of using 18F-labeled fluoromisonidazole (18F-FMISO) PET/CT for guiding IMRT so as to allow the dose to radioresistant hypoxic regions to be escalated to 84 Gy (Fig. 8.5) [37]. Findings of an ongoing trial, NCT00606294, are expected to show whether FMISO PET-guided visualization of hypoxia can be used to stratify patients in terms of risk.

Fig. 8.5
figure 5

Multimodality image acquisition, processing, and registration for 18F-FMISO PET/CT-guided IMRT. Shown are computed tomography (CT) (top left), fluorodeoxyglucose (FDG) (top right), 18F-fluoromisonidazole (18F-FMISO) (bottom left), and fused FDG-18F-FMISO (bottom right) images. Also shown are three enlarged areas from each scan type (From Lee et al. [37], with permission from Elsevier)

Full size image

Treatment intensification in the form of dose escalation is now an option with IMRT. Previous attempts at dose intensification with conventional RT were limited by toxicity [38, 39]. However, at least two dose-escalation studies have shown that IMRT can allow a safe dose escalation in NPC [22, 23]. In one of those studies, Kwong et al. used a prescribed dose of 76 Gy given in 35 fractions for T3-T4 NPC and found an excellent 2-year LCR rate of 96 % and acceptable toxicity [23]. In another such study, Kam et al. used a boost technique to provide a total dose of 74 Gy and also reported excellent LRC [22]. Although current dose levels have resulted in excellent LRC, dose escalation in selected patients with high-risk NPC may confer further benefits.

The use of adjuvant chemotherapy is another potential form of treatment intensification. Findings from the INT0099 trial indicated that the current standard of care should include adjuvant chemotherapy in addition to concurrent chemoradiation. However, results of a more recent phase III trial found no benefit from the use of adjuvant chemotherapy [40]; in that trial involving 508 patients, the 2-year failure-free survival rate was 86 % in the group with adjuvant chemotherapy and 84 % in the group without adjuvant chemotherapy (P = 0.13). Additional follow-up is needed, however, as the failure-free survival Kaplan-Meier curves may well separate over time. Moreover, that study was not designed to directly compare this therapy with that of INT0099. Nevertheless, patients with high-risk NPC may be more likely to benefit from adjuvant chemotherapy.

Interest has also been growing in the use of proton therapy in the form of intensity-modulated proton therapy (IMPT). IMPT plans have been shown to provide additional dosimetric advantage over IMRT by improving tumor coverage and reducing the mean dose to organs at risk (Fig. 8.6) [41]. We look forward to identifying potential benefit from protons in the clinical setting. Currently, an ongoing phase II trial at Massachusetts General Hospital is evaluating the potential for reduction in toxicity from the use of proton beam therapy (NCT00592501).

Fig. 8.6
figure 6

Dosimetric comparison of treatment plans for intensity-modulated radiation therapy (IMRT) vs. intensity-modulated proton therapy (IMPT). Dose distributions are shown for IMRT plans (left) and IMPT plans (right) for a patient with T4N1N0 nasopharyngeal carcinoma. Dotted lines denote 95 % of the prescribed dose to the gross tumor volume. Figure 2 from Taheri-Kadkhoda et al. [41] (License accessible at: http://creativecommons.org/licenses/by/2.0/legalcode)

Full size image

Adaptive RT is also being investigated for its potential to improve clinical outcomes. The rationale for this therapy is that significant anatomic changes during therapy, such as those resulting from loss of body weight or shrinkage in tumor volume (reportedly most severe after the first 2 weeks of treatment [42]), can lead to movement of the organs at risk into the planned radiation field. Conversely, marginal misses may occur if the tumor becomes displaced out of the treatment field, especially given the current efforts to reduce margins and treatment volumes to the greatest possible extent. The potential value of repeated treatment simulations is being considered and has shown some potential [42].

5 Conclusions

Clinical outcomes with IMRT have demonstrated clear dosimetric advantages, excellent LRC rates of more than 90 %, and lesser toxicity (specifically by improving salivary function) compared with conventional RT. The use of IMRT with specific target volume guidelines has been replicated successfully in a multi-institutional setting in the United States.

Further improvements in toxicity after IMRT will rely on either further reductions in margins within treatment volumes or the use of adaptive RT. Proton therapy also shows promise in terms of further sparing critical structures. Regarding approaches to improve disease control, dose escalation with IMRT is now feasible and could be considered for cases of particularly aggressive NPC. The use of imaging parameters and biomarkers, such as EBV DNA levels, also shows promise for risk stratification and consequent treatment intensification for high-risk NPC.