Extension of Local Disease in Nasopharyngeal Carcinoma Detected by Magnetic Resonance Imaging: Improvement of Clinical Target Volume Delineation

Int. J. Radiation Oncology Biol. Phys., Vol. 75, No. 3, pp. 742–750, 2009 Copyright Ó 2009 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/09/$–see front matter

doi:10.1016/j.ijrobp.2008.11.053

CLINICAL INVESTIGATION

Head and Neck

EXTENSION OF LOCAL DISEASE IN NASOPHARYNGEAL CARCINOMA DETECTED BY MAGNETIC RESONANCE IMAGING: IMPROVEMENT OF CLINICAL TARGET VOLUME DELINEATION SHAO-BO LIANG, M.D.,*y YING SUN, M.D., PH.D.,*y LI-ZHI LIU, M.D.,*z YONG CHEN, M.D.,*y LEI CHEN, M.D.,*y YAN-PING MAO, M.D.,*y LING-LONG TANG, M.D.,*y LI TIAN, M.D.,*z AI-HUA LIN, M.D., PH.D.,x MENG-ZHONG LIU, M.D.,*y LI LI, M.D., PH.D.,z AND JUN MA, M.D.*y * State Key Laboratory of Oncology in Southern China, Guangzhou, People’s Republic of China; y Department of Radiation Oncology, Cancer Center, Sun Yat-sen University, Guangzhou, People’s Republic of China; z Imaging Diagnosis and Interventional Center, Cancer Center, Sun Yat-sen University, Guangzhou, People’s Republic of China; and x Department of Medical Statistics and Epidemiology, School of Public Health, Sun Yat-sen University, Guangzhou, People’s Republic of China Purpose: To define by MRI the local extension patterns in patients presenting with nasopharyngeal carcinoma (NPC) and to improve clinical target volume delineation. Methods and Materials: Consecutive patients (N = 943) with newly diagnosed and untreated NPC were included in this study. All patients underwent MRI of the nasopharynx and neck, which was reviewed by two radiologists. Results: According to the incidence rates of tumor invasion, the anatomic sites surrounding the nasopharynx were initially classified into three risk grades: high risk ($ 35%), medium risk ($ 5–35%), and low risk (< 5%). Incidence rates of tumor invasion into anatomic sites at medium risk were increased, reaching 55.2%, when adjacent high-risk anatomic sites were involved. However, the rates were substantially lower, mostly < 10%, when adjacent high-risk sites were not involved. The incidence rates of concurrent tumor invasion into bilateral sites were < 10%, except in the case of prevertebral muscle involvement (13.1%). Among the 178 incidences of cavernous sinus invasion, there were often two or more simultaneous infiltration routes (60.6%); when only one route was involved, the foramen ovale was the most common (26.4%). Conclusions: In patients presenting with NPC, local disease spreads stepwise from proximal sites to more distal sites. Tumors extend quickly through privileged pathways such as neural foramina. The anatomic sites surrounding the nasopharynx are at low risk of concurrent bilateral tumor invasion. Selective radiotherapy of the local disease in NPC may be feasible. Ó 2009 Elsevier Inc. Nasopharyngeal carcinoma, Magnetic resonance imaging, Intensity-modulated radiotherapy, Local extension patterns, Clinical target volume delineation.

The highest rates of nasopharyngeal carcinoma (NPC) incidence in the world occur in southern China, with the yearly incidence rate varying between 15 and 50 cases per 105 people (1). Radiation therapy is the primary approach taken to treat locoregionally confined NPC. Clinical target volume (CTV) delineation is the key point in local tumor control and normal tissue protection. Intensity-modulated radiotherapy (IMRT), an important milestone in the development of radiation therapy, requires highly precise delineation of the CTV.

Most centers continue to use CTV definitions determined by two-dimensional (2D) radiotherapy, which leads to a significant intercenter CTV delineation variability (2–5). There are several disadvantages in this approach to CTV delineation. First, measured CTV generally does not differ according to clinical stage. In our opinion, the CTV should be delineated according to the gross tumor volume (GTV) rather than with respect to a fixed pattern. Second, the CTV includes most anatomic sites surrounding the nasopharynx bilaterally, without distinguishing the delineation of anatomic sites contralateral to the gross tumor from those ipsilateral to the

Reprint requests to: Jun Ma, M.D., Department of Radiation Oncology, Cancer Center, Sun Yat-sen University, 651 Dongfeng Road East, Guangzhou 510060, People’s Republic of China. Tel: (+86) 20-87343469; Fax: (+86) 20-87343295; E-mail: majun2@ mail.sysu.edu.cn This work was supported by the Science Foundation of Key Hospital Clinical Program of Ministry of Health P.R. China (Grant No.

2007-353), Hi-Tech Research and Development Program of China (Grant No. 2006AA02AA404), and the National Natural Science Foundation of China (Grant No. 30470505). Shao-Bo Liang and Ying Sun contributed equally to this study. Conflict of interest: none. Received July 16, 2008, and in revised form Nov 9, 2008. Accepted for publication Nov 13, 2008.

INTRODUCTION

742

Local extension patterns and CTV delineation of NPC d S.-B. LIANG et al.

invasion. Third, although the pattern of local extension is important for CTV delineation, there are few reports that have systematically examined the pattern of local extension at presentation. Furthermore, 2D radiotherapy cannot carry out precise delineation of CTV. However, IMRT could achieve a much more precise CTV delineation. Improvement in CT technique has permitted detailed examination of the extent of local disease in NPC (6–8). However, compared with CT, MRI has better soft tissue contrast resolution (9–10). Indeed, recent data suggest that MRI has several advantages over CT for studying local disease in NPC. Specifically, MRI provides more accurate definition of early invasion beyond the nasopharynx, improved differentiation of retropharyngeal nodes from the primary tumor, and more accurate assessment of the parapharyngeal space, skull base, paranasal sinus, and cranial invasion (11–14). Given these advantages, MRI is considered the optimal imaging technique for studying the extension of local disease in NPC. We initiated a study using a large sample of NPC patients to document in detail the patterns of local disease in NPC. The purpose of this study is to improve CTV delineation and thereby increase the effectiveness of treatment planning. Our data may improve understanding of the biological nature of NPC. METHODS AND MATERIALS Participants Between January 2003 and December 2004, 943 consecutive patients with newly diagnosed, untreated, and nondisseminated NPC were included into our study. There were 698 male patients and 245 female patients, with a male to female ratio of 2.9:1.0. The median age was 45 years (range, 11–78 years). Histologically, 98.6% of the patients had World Health Organization (WHO) Type II disease, 0.4% had WHO Type III disease, and the remainder (1.0%) had WHO type I disease. All patients underwent a pretreatment evaluation that included a complete history, physical and neurological examinations, hematology and biochemistry profiles, MRI scan of the nasopharynx and neck, chest radiography, and abdominal sonography. Medical records and imaging studies were analyzed retrospectively, and all patients were staged according to the American Joint Committee on Cancer (AJCC) classification, 6th edition (15). Both clinical examination and MRI findings were included in the clinical staging. Tumor extent (T stage) was diagnosed on the basis of the presence of MRI findings together with clinical signs and symptoms of cranial nerve palsy revealed by physical examination. Lymph node involvement (N stage) was based on clinical examination; MRI data were consulted when metastatic lymphadenopathy was not revealed by clinical examinations. The stage distribution for all patients was 19.7% Stage T1, 24.3% Stage T2, 34.3% Stage T3, and 21.7% Stage T4; 34.1% Stage N0, 35.3% Stage N1, 21.1% Stage N2, and 9.4% Stage N3; 8.4% Stage I, 23.1% Stage IIA–B, 39.4% Stage III, and 29.1% Stage IVA–B.

743

in the axial, coronal, and sagittal planes (repetition time of 500–600 ms and echo time of 10–20 ms, two excitations, a 22-cm field of view (FOV), and a 320  224 frequency matrix) and T2-weighted fast spin-echo images in the axial plane (repetition time, 4000– 6000 ms and echo time, 95–110 ms, one excitation, a 22-cm FOV, and a 320  224 frequency matrix) were obtained before injecting the contrast material. After intravenous administration of gadopentetate dimeglumine (Gd-DTPA; Magnevist, Schering, Berlin, Germany) at a dose of 0.1 mmol per kg body weight, spin-echo T1-weighted axial and sagittal sequences and spin-echo T1-weighted fat-suppressed coronal sequences were performed sequentially, with parameters that were used before Gd-DTPA injection. We used a section thickness of 5 mm and a matrix size of 512  512. Image assessment. All MRI materials and clinical records were reviewed to minimize heterogeneity in restaging. Two radiologists with a clinical focus on head and neck cancers and certifications of professional diagnostic imaging in China and who have been on staff for 10 years evaluated the MR images separately. Any disagreements were resolved by consensus. Tumor soft tissue had low signal intensity on T1-weighted images; intermediate signal intensity, superior to the muscle signal, on T2-weighted images; and moderate enhancement on contrast-enhanced T1-weighted images, replacing the normal anatomy of the structure. Skull-base invasion was suspected by the presence of low-intensity tissue in the high signal bone marrow on the T1-weighted image and the Gd-DTPA enhancement of the abnormal tissue.

Diagnostic criteria Nasal involvement was defined as tumor involvement beyond the posterior line of the pterygopalatine fossa. Oropharynx involvement was defined as tumor involvement below C1/C2 (15, 16). Tumor detection below the lower margin of C3 or the free edge of the epiglottis was regarded as laryngopharynx involvement (13). Parapharyngeal involvement was defined as posterolateral infiltration of tumor beyond the pharyngobasilar fascia (12, 15). Infratemporal fossa involvement was defined as extension beyond the anterior surface of the lateral pterygoid muscle or lateral extension beyond the posterolateral wall of the maxillary sinus or pterygomaxillary fissure (15). The parapharyngeal space was subdivided into prestyloid and poststyloid compartments by a layer of fascia that was an extension from and contained the fibers of the tensor veli palatini muscle. Anterior and predominantly lateral to the tensor veli palatini fascial layer was the prestyloid compartment. Posterior and medial to the tensor veli palatini fascial layer was the poststyloid parapharyngeal space, which has also been called the carotid space (17).

Statistical analysis All statistical analyses were performed using the Statistical Package for Social Sciences, version 11.0 (SPSS, Chicago, IL). The chi-square test (or Fisher’s exact test, if indicated) was used to compare the relative frequencies of tumor invasion into various anatomic sites. The criterion for statistical significance was set at a = 0.05, and p values were two-sided.

RESULTS Procedures Imaging protocol. All patients underwent MRI with a 1.5-T system (Signa CV/I, General Electric Healthcare, Chalfont St. Giles, United Kingdom). The region from the suprasellar cistern to the inferior margin at the sternal end of clavicle was examined with a head-and-neck combined coil. T1-weighted fast spin-echo images

The risk of tumor invasion into various anatomic sites Cumulative incidence rates of tumor invasion into the anatomic sites surrounding the nasopharynx ranged from 0.2% to 66.7% (Table 1). According to the cumulative incidence rates of tumor invasion, we initially classified anatomic sites

I. J. Radiation Oncology d Biology d Physics

744

Table 1. Incidence rates of tumor invasion into anatomic sites surrounding the nasopharynx Anatomic sites High risk Parapharyngeal space Levator veli palatine muscle Prestyloid compartment Tensor veli palatine muscle Poststyloid compartment Nasal cavity Pterygoid process Basis of sphenoid bone Petrous apex Prevertebral muscle Clivus Foramen lacerum Medium risk Foramen ovale Great wing of sphenoid bone Medial pterygoid muscle Oropharynx Cavernous sinus Sphenoidal sinus Pterygopalatine fossa Lateral pterygoid muscle Hypoglossal canal Foramen rotundum Ethmoid sinus Jugular foramen Low risk Inferior orbital fissure Cervical vertebrae Infratemporal fossa Maxillary sinus Cistern Temporal lobe Meninges Orbital apex Superior orbital fissure Hypopharynx Frontal sinus

No. of patients (%) 638 (67.7) 618 (65.5) 605 (64.2) 539 (57.2) 477 (50.6) 451 (47.8) 437 (46.3) 418 (44.3) 365 (38.7) 363 (38.5) 361 (38.3) 339 (35.9) 219 (23.2) 210 (22.3) 188 (19.9) 187 (19.8) 164 (17.4) 163 (17.3) 162 (17.2) 100 (10.6) 96 (10.2) 87 (9.2) 50 (5.3) 48 (5.1) 35 (3.7) 31 (3.3) 27 (2.9) 27 (2.6) 20 (2.1) 17 (1.8) 13 (1.4) 11 (1.1) 6 (0.6) 5 (0.5) 2 (0.2)

surrounding the nasopharynx into three risk grades: high risk ($ 35%), medium risk ($ 5–35%), and low risk (< 5%), as shown in Fig. 1. It was concluded that the anatomic sites at high risk of tumor invasion were adjacent to the nasopharynx, and the anatomic sites at medium or low risk of tumor invasion were separated from the nasopharynx. When high-risk anatomic sites were involved, the adjacent sites at medium risk had high rates of tumor invasion, with the highest incidence rate reaching 55.2% (Table 2). In contrast, when anatomic sites at high risk were not involved, the adjacent sites at medium risk had low rates of tumor invasion, mostly < 10% (Table 3). An exception to this pattern was found for oropharynx involvement, which had an incidence rate of 13.1% without prevertebral muscle involvement; this exception can likely be attributed to the fact that the main infiltration route to oropharynx was not through prevertebral muscle. These results demonstrate that tumor invasion is relative to the involvement of adjacent anatomic sites. Therefore, we further analyzed the incidence rate of tumor invasion into

Volume 75, Number 3, 2009

anatomic sites at low risk when adjacent sites were involved (Fig. 2). The incidence rates of tumor invasion into anatomic sites at low risk increased significantly when adjacent sites were involved, with the highest incidence rate reaching 34%. Privileged pathways in local extension To demonstrate characteristics of privileged pathways in local extension, we analyzed the infiltration routes to cavernous sinus. Five well-known infiltration routes were determined: foramen ovale, foramen lacerum, foramen rotundum, sphenoidal sinus, and clivus. Analysis of the infiltration routes of tumor invasion into the 178 sides of the cavernous sinus was performed (Table 4). The incidence of tumor invasion into the cavernous sinus through a single infiltration route accounted for 39.4% (70 of 178), among which the most common was the foramen ovale (Fig. 3A, 47 sides, 26.4%), followed by the foramen lacerum (Fig. 3B and 3C, 18 sides, 10.1%). The incidence rate of tumor invasion into the cavernous sinus through two or more infiltration routes accounted for 60.7% (107 of 178), among which the most common routes were through the foramen ovale and foramen lacerum (86 sides, 48.3%). From these data, there were always several simultaneous infiltration routes to the same anatomic site, and thus it was difficult to identify which infiltration route occurred first and which routes occurred later. Therefore, we defined a pathway as being involved when all anatomic sites along the pathway were involved (Fig. 4). It is apparent that there are some privileged pathways in local extension. Furthermore, neural foramina, such as the pterygopalatine fossa, foramen rotundum, foramen ovale, jugular foramen, and hypoglossal canal, are important in tumor extension. Tumor invasion into bilateral anatomical sites With the exception of anatomic sites on the midline, such as the base of the sphenoid bone and clivus, the bilateral nasopharyngeal carcinoma was defined by MRI as a tumor extending across the midline of the nasopharynx. Nine hundred and four patients (904/943, 95.9%) had bilateral tumor invasion into the mucous membrane of the nasopharynx. However, most anatomic sites were at low risk of concurrent bilateral tumor invasion (< 10%), except prevertebral muscle (13.1%), shown in Table 5. DISCUSSION The emergence of IMRT has not only brought about a significant survival benefit to NPC patients, it has also reduced suffering caused by radiotherapy side effects (18). However, some problems in defining CTV during IMRT remain, such as determining the origin in 2D radiotherapy and a lack of individualization. The definition of individual CTV should be based on the GTV, the pattern of local extension, and the biological nature of NPC. This study has shown patterns of local disease in NPC by concrete data to improve CTV delineation.

Local extension patterns and CTV delineation of NPC d S.-B. LIANG et al.

745

Fig. 1. Summary of tumor invasion risk. The anatomic sites shown in red were at high risk of tumor invasion. Sites shown in yellow were at medium risk of tumor invasion, and those in blue were at low risk of tumor invasion.

The pattern of local extension in NPC Tumor invasion risk gradually reduces moving away from the nasopharynx. Our data show that all anatomic sites at high risk of tumor invasion lack any barrier and are a short

distance from the nasopharynx, such as parapharyngeal space (67.7%). However, anatomic sites at medium or low risk of tumor invasion are either structurally distant from the nasopharynx or separated from it by other anatomic sites.

Table 2. Tumor invasion into anatomic sites at medium risk with invading adjacent high-risk anatomic sites

Foramen ovale Great wing of sphenoid bone Medial pterygoid muscle Oropharynx Cavernous sinus Sphenoidal sinus Pterygopalatine fossa Lateral pterygoid muscle Hypoglossal canal Foramen rotundum Ethmoid sinus Jugular foramen

Parapharyn-geal space (n = 638)

Nasal cavity (n = 451)

Pterygoid process (n = 437)

Basis of Prevertebral sphenoid bone Petrous apex muscle (n = 418) (n = 365) (n = 363)

208 (32.6) 198 (31.0%)

— —

— —

— 204 (48.8%)

191 (52.3%) 190 (52.1%)

— —

186 (29.2%)

—

169 (38.7%)

—

—

—

159 (24.9%) — —

— — 129 (28.6%)

— — —

— — 162 (38.8%)

137 (21.5%)

143 (31.7%) 152 (34.8%)

Clivus (n = 361)

Foramen lacerum (n = 339)

— 187 (55.2%) 196 (54.3%) 184 (54.3%) —

—

— 112 (30.9%) — — 150 (41.1%) — — 150 (44.2%) — — 148 (41.0%) —

—

—

—

—

—

100 (15.7%)

—

95 (21.7%)

—

—

—

—

—

86 (13.5%)

—

—

—

—

90 (24.8%)

89 (24.7%)

—

80 (12.5%)

—

—

—

70 (19.2%)

—

—

68 (20.1%)

— 46 (7.2%)

46 (10.2%) —

46 (10.5%) —

— —

— —

— 43 (11.8%)

— 45 (12.5%)

— —

746

I. J. Radiation Oncology d Biology d Physics

Volume 75, Number 3, 2009

Table 3. Tumor invasion into anatomic sites at medium risk with noninvasive high-risk adjacent sites

Foramen ovale Great wing of sphenoid bone Medial pterygoid muscle Oropharynx Cavernous sinus Sphenoidal sinus Pterygopalatine fossa Lateral pterygoid muscle Hypoglossal canal Foramen rotundum Ethmoid sinus Jugular foramen

Parapharyn-geal space (n = 296)

Nasal cavity (n = 492)

Pterygoid process (n = 506)

Basis of sphenoid bone (n = 525)

Petrous apex (n = 578)

Prevertebral muscle (n = 580)

Clivus (n = 582)

Foramen lacerum (n = 604)

11 (3.7%) 12 (4.1%)

— —

— —

— 6 (1.1%)

28 (4.8%) 20 (3.5%)

— —

— 14 (2.4%)

32 (5.3%) 26 (4.3%)

2 (0.7%) 29 (9.8%) — — 27 (9.1%) 0 (0.0%) 10 (3.4%) 7 (2.4%) — 3 (1.0%)

— — — 34 (6.9%) 19 (3.9%) — — — 4 (0.8%) —

19 (3.8%) — — — 10 (2.0%) 5 (1.0%) — — 4 (0.8%) —

— — — 0 (0.0%) — — — — — —

— — 14 (2.4%) — — — — 17 (2.9%) — —

— 76 (13.1%) — — — — 6 (1.0%) — — 5 (0.9%)

— — — 15 (2.6%) — — 10 (1.7%) — — 3 (0.5%)

— — 14 (2.3%) — — — — 19 (3.1%) — —

For example, the inferior orbital fissure is separated from the nasopharynx by the nasal cavity and pterygopalatine fossa, which results in low risk of tumor invasion in that structure (3.7%). The risk of tumor invasion into various anatomic sites is relative to the distance of the anatomical structures from the nasopharynx cavity. Our data also show that tumor invasion into various anatomic sites is relative to the involvement of adjacent sites. Tumors invade first into sites adjacent to the nasopharynx. When these sites are involved, adjacent sites may also be invaded. Anatomic sites at low risk had a low incidence rate of tumor invasion (0.2%–3.7%); however, tumor invasion into adjacent sites increased the rate of tumor invasion (Fig. 2). It could be concluded that local disease spreads

stepwise from proximal sites to more distal sites and that a skip pattern of local extension was unusual. Privileged pathways in local extension. Several investigators have reported that tumors are likely to spread through neural foramina in head and neck cancers (19, 20). Su et al. (21) reported that the most common infiltration route to the cavernous sinus was through foramen ovale. This route was the most important route to the cavernous sinus in our study, confirming the findings of Su et al. with a larger patient sample. However, in locally advanced disease, the tumor invasion was so extensive that all infiltration routes could be involved, such that the incidence rate of tumor invasion into the cavernous sinus through two or more infiltration routes accounted for 60.7% of all extensions to cavernous sinus in our study.

Fig. 2. The incidence rate of tumor invasion into various anatomic sites (A-H) at low risk when adjacent sites were involved.

Local extension patterns and CTV delineation of NPC d S.-B. LIANG et al.

Table 4. Incidence rates of tumor invasion routes to 178 sides of cavernous sinus Routes One route Foramen ovale Foramen lacerum Foramen rotundum Sphenoidal sinus Clivus Two routes Foramen ovale + foramen lacerum Foramen ovale + foramen rotundum Foramen ovale + sphenoidal sinus Foramen ovale + clivus Foramen lacerum + foramen rotundum Foramen lacerum + sphenoidal sinus Foramen lacerum + clivus Foramen rotundum + sphenoidal sinus Foramen rotundum + clivus Sphenoidal sinus + clivus $ Three routes Including foramen ovale or foramen lacerum Without foramen ovale or foramen lacerum Total

No. of Sides (%) 47 (26.4%) 18 (10.1%) 1 (0.6%) 1 (0.6%) 3 (1.7%) 50 (28.1%) 2 (1.1%) 7 (3.9%) 1 (0.6%) 2 (1.1%) 3 (1.7%) 4 (2.2%) 0 (0.0%) 0 (0.0%) 2 (1.1%) 1 (0.6%) 36 (20.2%) 178 (100.0%)

Dubrulle et al. (22) summarized the well-known routes of local disease extension of NPC (22). On the basis of his study and our data, we have outlined the overall pathways of local extension in NPC (Fig. 4). This is the first figure to show systematically the incidence rate of tumor invasion in each infiltration route of local extension in NPC, which could be useful in the delineation of target volume during treatment planning and in understanding the biological nature of NPC. Tumor invasion into bilateral anatomic sites. The main treatment for NPC is radiotherapy, which means there is an unavoidable lack of pathological confirmation of imaging findings concerning tumor extension (23, 24). Therefore,

747

bilateral nasopharyngeal carcinoma was defined according to MR imaging. Studies by Sham et al. (25) showed that the tumor was confined to the unilateral nasopharynx in only 2.8% of patients (7 of 247) by pathology of multiple sites (25). There are two explanations for these findings: first, tumor spreads easily across the midline of the nasopharynx following the mucous membrane; second, the tumor may originate in multiple sites in the mucous membrane of the nasopharynx. In our study, 95.9% of patients had bilateral tumor invasion in the nasopharynx as detected by MRI. Therefore, we believe that this ratio would increase by a histopathological examination. However, the anatomic sites, with the exception of the prevertebral muscle, are at low risk of concurrent bilateral tumor invasion (< 10%). This may be because the prevertebral muscles are adjacent to each other, so when one site is invaded, the tumor easily extends to the other site. Improvement in clinical target volume delineation Two-dimensional radiotherapy, which was widely used in the 20th century, was based on CT and clinical examination in NPC. The target volume (including the nasopharynx, the posterior ethmoid sinus, the sphenoidal sinus and base of the sphenoid bone, the posterior nasal cavity and maxillary sinus, the skull base, and the lateral and posterior pharyngeal wall to the lower pole of the tonsil) were all given a definitive-intent radiation dosage (26). However, the target volume seemed to lack individualization. Further, the question remains as to whether the clinical target volume from 2D radiotherapy is suitable in the era of MRI and IMRT. In recent years, local tumor control for NPC has been greatly improved with the development of IMRT. However, delineation of CTV originated from 2D radiotherapy, but with a different dose gradient (18, 27, 28). Our data suggest some improvement in CTV delineation for three reasons. First, the anatomic sites at high risk should be defined as CTV, because their incidence rates of tumor

Fig. 3. (A) Enhanced coronal T1-weighted coronal fat suppression image (repetition time [TR] = 600 ms, echo time [TE] = 15 ms) at the level of the foramen ovale shows the tumor extending into the cavernous sinus through the enlarged left foramen ovale. (B) Enhanced coronal T1-weighted coronal fat suppression image (TR = 600 ms, TE = 15 ms) at the level of the foramen rotundum shows that the left foramen rotundum was invaded. (C) Enhanced coronal T1-weighted coronal fat suppression image (TR = 600 ms, TE = 15 ms) at the level of the cavernous sinus shows that the tumor extending into the cavernous sinus through the enlarged left foramen rotundum.

I. J. Radiation Oncology d Biology d Physics

748

Volume 75, Number 3, 2009

20.5 % Great wing of sphenoid bone 64.2 % Prestyloid compartment

19.8%

10.5% Medial pterygoid muscle

21.3%

Infratemporal fossa

12.8% Foramen ovale

Cavernous sinus

1.2 %

9.3 % Hypoglossal canal 50.6 %

2.3% Lateral pterygoid muscle

4.7%

Cistern segment 0.5 %

Jugular foramen

Poststyloid compartment

Cistern segment

31.4% Petrous apex

2.4 %

2.8% Maxillary sinus 47.8%

Infratemporal fossa

4.9% Nasal cavity

Ethmoid sinus

6.3 %

4.3 % Foramen rotundum

Cavernous sinus

15.2% Pterygopalatine fossa

3.2 %

1.2 % Inferior orbital fissure

1.2 %

Superior orbital fissure

Orbital apex

2.1 % 0.6 % Maxillary sinus

Nasopharynx 16.4%

38.3% Clivus

Cavernous sinus

17.2 %

44.3 % Basis of sphenoid bone

9.3% Sphenoidal sinus

Cavernous sinus

33.3% Petrous apex 32.0% 47.8%

Clivus Foramen lacerum 16.1% Cavernous sinus

3.0% Inferior orbital fissure 17.4%

1.2% Cavernous sinus Orbital apex 6.0% Foramen rotundum

Fig. 4. The overall pathways of local extension in 943 nasopharyngeal carcinoma patients.

Cavernous

Local extension patterns and CTV delineation of NPC d S.-B. LIANG et al.

Table 5. Concurrent bilateral tumor invasion into anatomic sites surrounding of the nasopharynx Anatomical sites

No. of patients (%)

Prevertebral muscle Levator veli palatini Parapharyngeal space Prestyloid compartment Pterygoid process Poststyloid compartment Foramen lacerum Tensor veli palatini Petrous apex Vidian canal Pterygopalatine fossa Cavernous sinus Hypoglossal canal Great wing of sphenoid bone Oropharynx Foramen ovale Foramen rotundum Masticator space Medial pterygoid muscle Maxillary sinus Inferior orbital fissure Lateral pterygoid muscle Jugular foramen Infratemporal fossa Orbital apex Superior orbital fissure Hypopharynx

124 (13.1%) 89 (9.4%) 82 (8.7%) 70 (7.4%) 60 (6.4%) 56 (5.9%) 54 (5.8%) 52 (5.5%) 47 (5.0%) 24 (2.5%) 23 (2.4%) 14 (1.5%) 14 (1.5%) 13 (1.4%) 13 (1.4%) 9 (1.0%) 9 (1.0%) 7 (0.7%) 7 (0.7%) 3 (0.3%) 2 (0.2%) 1 (0.1%) 1 (0.1%) 1 (0.1%) 1 (0.1%) 0 (0%) 0 (0%)

invasion were > 35%. Further, whether the anatomic sites at medium or low risk could be defined as CTV was based on tumor invasion into the ipsilateral adjacent anatomic sites. It might be reasonable to define anatomic sites at medium risk as within the CTV when adjacent sites at high risk are

749

involved and to define sites at low risk as the CTV only when the adjacent sites at high or medium risk are involved. It may be incorrect to define the ethmoid sinus and maxillary sinus as part of the CTV because they are at low risk of tumor invasion (2–5). Second, the critical role played by neural foramina in tumor extension indicates that they should be given greater consideration in CTV definitions and that their adjacent anatomic sites should be defined in the local extension pathway as part of the CTV when neural foramina have been invaded. Third, when the tumor invades primarily on one side of the nasopharynx, the bilateral anatomic sites at high risk should be included in the CTV, whereas the sites at medium or low risk and contralateral to the tumor invasion area should be excluded from the CTV. The information presented here, which includes a large number of patients at a single institution receiving sensitive and specific imaging technologies, offer valuable data for defining the local extension patterns in NPC. However, it should be noted that it lacks pathological confirmation of imaging findings concerning tumor extension. The main treatment for NPC is radiotherapy, and thus a lack of pathology concerning tumor extension. Further, whether the recommended CTV delineation is valid requires further study.

CONCLUSIONS In patients presenting with NPC, local disease spreads stepwise from proximal sites to more distal sites. Tumors extend quickly through privileged pathways such as neural foramina. The anatomic sites surrounding the nasopharynx are at low risk of concurrent bilateral tumor invasion. Selective radiotherapy of local disease in NPC may be feasible.

REFERENCES 1. Chan AT, Teo PM, Johnson PJ. Nasopharyngeal carcinoma. Ann Oncol 2002;13:1007–1015. 2. Xia P, Fu KK, Wong GW, Akazawa C, Verhey LJ. Comparison of treatment plans involving intensity-modulated radiotherapy for nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 2000;48:329–337. 3. Kam MK, Teo PM, Chau RM, et al. Treatment of nasopharyngeal carcinoma with intensity-modulated radiotherapy: The Hong Kong experience. Int J Radiat Oncol Biol Phys 2004;60:1440–1450. 4. Hunt MA, Zelefsky MJ, Wolden S, et al. Treatment planning and delivery of intensity-modulated radiation therapy for primary nasopharynx cancer. Int J Radiat Oncol Biol Phys 2001;49:623–632. 5. Luo Wei, Deng Xiao-wu, Lu Tai-xiang. Dosimetric evaluation for three dimensional radiotherapy plans for patients with early nasopharygeal carcinoma. Ai Zheng 2004;23:605–608. 6. Cheung YK, Sham J, Cheung YL, Chan FL. Evaluation of skull base erosion in nasopharyngeal carcinoma: Comparison of plain radiography and computed tomography. Oncology 1994;51:42–46. 7. Sham JS, Cheung YK, Choy D, et al. Nasopharyngeal carcinoma: CT evaluation of patterns of tumor spread. Am J Neuroradiol 1991;12:265–270. 8. Yamashita S, Kondo M, Hashimoto S. Conversion of T-stages of nasopharyngeal carcinoma by computed tomography. Int J Radiat Oncol Biol Phys 1985;11:1017–1021.

9. Casselman JW. The value of MRI in the diagnosis and staging of nasopharynx tumors. J Belge Radiol 1994;77:67–71. 10. Ng SH, Chang TC, Ko SF, et al. Nasopharyngeal carcinoma: MRI and CT assessment. Neuroradiology 1997;39:741–746. 11. Olmi P, Fallai C, Colagrande S, Giannardi G. Staging and follow up of nasopharyngeal carcinoma: Magnetic resonance imaging verses computed tomography. Int J Radiat Oncol Biol Phys 1995;32:795–800. 12. King AD, Teo P, Lam WW, et al. Paranasopharyngeal space involvement in nasopharyngeal cancer: detection by CT and MRI. Clin Oncol 2000;12:397–402. 13. Chong VF, Fan YF. Skull base erosion in nasopharyngeal carcinoma: Detection by CT and MRI. Clin Radiol 1996;51: 625–631. 14. Chong VF, Fan YF, Khoo BK. Nasopharyngeal carcinoma with intracranial spread: CT and MRI characteristics. J Comput Assist Tomo 1996;20:563–569. 15. Cooper J, Flemming ID, Henson DE, et al. The American Joint Committee on Cancer. Manual for staging of cancer. 6th ed. Philadelphia: Lippincott; 2002. 16. Heng DM, Wee J, Fong KW, et al. Prognostic factors in 677 patients in Singapore with nondisseminated nasopharyngeal carcinoma. Cancer 1999;86:1912–1920. 17. Hugh D. Curtin. Separation of the masticator space from the parapharyngeal space. Radiology 1987;163:195–204.

750

I. J. Radiation Oncology d Biology d Physics

18. Nacy Lee, Ping Xia, Jeanne M, et al. Intensity-Modulated radiotherapy in the treatment of nasopharyngeal carcinoma: An update of the UCSF experience. Int J Radiat Oncol Biol Phys 2002;53:12–19. 19. Chong VFH, Fan YF. Hypoglossal nerve palsy in nasopharyngeal carcinoma. Eur. Radiol 1998;8:939–945. 20. Caldemeyer KS, Mathews VP, Righi PD, Smith RR. Imaging features and clinical significance of perineural spread or extension of head and neck tumors. RadioGraphics 1998;18:97–110. 21. Chih-Ying Su, Chun-Chung Lui. Perineural invasion of the trigeminal nerve in patients with nasopharyngeal carcinoma: Imaging and clinical correlations. Cancer 1996;78:2063–2069. 22. Dubrulle F, Souillard R, Hermans R. Extension patterns of nasopharyngeal carcinoma. Eur Radiol 2007;17:2622–2630. 23. Chua DT, Sham JST, Kwong DL, Au GK. Treatment outcome after radiotherapy alone for patients with Stage I–II nasopharyngeal carcinoma. Cancer 2003;98:74–80.

Volume 75, Number 3, 2009

24. Gearaa FB, Sanguineti G, Tuckerb SL, et al. Carcinoma of the nasopharynx treated by radiotherapy alone: Determinants of distant metastasis and survival. Radiother Oncol 1997;43: 53–61. 25. Sham JS, Wei WI, Kwan WH, et al. Fiberoptic endoscopic examination and biopsy in determining the extent of nasopharyngeal carcinoma. Cancer 1989;64:1838–1842. 26. Carlos A. Perez. Nasopharynx. In: Halperin EC, Perez CA, Brady LW, editors. Principles and practice of radiation oncology. 3rd ed. Philadelphia: Lippincott; p. 897–939. 27. Kam MK, Chau RM, Suen J, et al. Intensity-modulated radiotherapy in nasopharyngeal carcinoma: Dosimetric advantage over conventional plans and feasibility of dose escalation. Int J Radiat Oncol Biol Phys 2003;56:145–157. 28. Zhao C, Han F, Lu LX, et al. Intensity modulated radiotherapy for local-regional advanced nasopharyngeal carcinoma. Ai Zheng 2004;23:1532–1537.