• Users Online: 138
  • Print this page
  • Email this page

Table of Contents
Year : 2003  |  Volume : 5  |  Issue : 1  |  Page : 21-29

Prognostic value of MRI in nasopharyngeal carcinoma

1 Diagnostic Radiology Department, Faculty of Medicine, King Abdulaziz University; Dr. Erfan and Bagedo General Hospital,Jeddah, Saudi Arabia
2 Diagnostic Radiology Department, Faculty of Medicine, Cairo University, Cairo, Egypt; Dr. Erfan and Bagedo General Hospital,Jeddah, Saudi Arabia
3 Kasr Al-Aini Center of Clinical Oncology and Nuclear Meicine (NEMROCK), Faculty of Medicine, Cairo University, Cairo, Egypt; Dr. Erfan and Bagedo General Hospital, Jeddah, Saudi Arabia

Date of Web Publication11-Jul-2020

Correspondence Address:
MD., FRCPC Sattam S Linjawi
P.O. Box 54403 Jcddah 21514
Saudi Arabia
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/1319-8491.289561

Rights and Permissions

Objective: The aim of this work was to evaluate the role of MRI in prediction of nasopharyngeal carcinoma (NPC) response to radiotherapy.
Materials and Methods: Fifty-two NPC patients had MRI before and after radiotherapy to the local disease and the nodal involvement. The MRI assessed the size, site, signal intensity, enhancement pattern and extension of the regional disease in twenty-five anatomical sites. Radiotherapy was delivered using X-rays of 4-6 MV Linac.
Results: Complete remission (CR) rates in the lateral, anterior, inferior, intracranial, orbital, skull base and paranasal sinuses extensions were 77% (177/229), 89%(23/28), 68%(13/19), 18%(5/28), 0%(0/5), 16%(28/17l) and 31%(9/29) respectively. The CR rate of the lateral, anterior, inferior extensions collcctivcly (Tl-2 lesions) was 77% (213/276) compared with 17.6% (42/238) for intracranial, orbital, skull base and paranasal sinuses extensions collectively (T3-4 lesions) (p<0.0000!). Among T2 lesions the CR rate was 89% (23/28) for nasal versus 68% (13/19) for oropharyngeal extension (p=0.04).
Conclusion: The intracranial, orbital, skull base and paranasal sinuses extensions of NPC were associated with significantly lower rates of CR.

Keywords: Nasopharyngeal MRI Staging Radiotherapy

How to cite this article:
Linjawi SS, Ragab YF, Mansour M. Prognostic value of MRI in nasopharyngeal carcinoma. Saudi J Otorhinolaryngol Head Neck Surg 2003;5:21-9

How to cite this URL:
Linjawi SS, Ragab YF, Mansour M. Prognostic value of MRI in nasopharyngeal carcinoma. Saudi J Otorhinolaryngol Head Neck Surg [serial online] 2003 [cited 2022 Dec 4];5:21-9. Available from: https://www.sjohns.org/text.asp?2003/5/1/21/289561

  Introduction Top

Although Nasopharyngeal Carcinoma (NPC) represents less than 1% of all malignancies in USA, it constitutes up to 20-30% in certain endemic areas like Hong Kong and Southern China [1]. In Saudi Arabia, NPC constitutes 5% of all malignant tumors and 37% of all head and neck cancers [2].

The current staging systems including the recent American Joint Committee of Cancer (AJCC)- 1997 [3] [Table 1],[Table 2], and the Union Internationale Contre le Cancer (UICC) [4], have their limitation in staging NPC and in the determination of the real loco-regional tumor cell burden which is the basis of successful management [2],[3],[4],[5]. Precise determination of tumor cell burden has its relevance in NPC especially in the Middle East area where stages III and IV represent more than 90% [2],[6].
Table 1: Tumor Characteristics among 52 NPS cases.

Click here to view
Table 2: Tumor Characteristics among 52 NPS cases.

Click here to view

Recent MRI data on NPC suggests that MR is probably the modality of choice in studying the patterns of loco-regional disease in NPC [5]. It provides accurate detection of early invasion outside the confines of the nasopharynx (NPX), differentiation of lateral retropharyngeal nodes from primary tumor [7], and also assessment of cranial, orbital, skull base invasion as well as cervical lymph nodal involvement [8],[9]. MRI is also valuable in accurate evaluation of the response to radiotherapy [10].

The aim of this work was to evaluate the role of MRI in predicting the response to radiotherapy in NPC patients through the assessment of local extent and regional lymph nodal involvement before and after radiotherapy. Although this will unlikely change patients management drastically, however, it helps in the development of a clearer idea about patient prognosis. It allows both the physician and the patient to design the plan of therapy, particularly since most of the NPC patients receive their treatment on out-patient bases.

  Materials and Methods Top

This was a retrospective study in which all patients with newly diagnosed (biopsy-proven) NPC who presented to our hospital between January 1996 and December 1999 were included (52 patients). All patients underwent Mfyl examination before and at 2, 6, 12 and 24 months after the completion of radiotherapy (internal hospital protocol). The mean age was 42 years (range of 15-80 years) with 40 males and 12 females (3:1).

Undifferentiated carcinoma (WHO Type III) cases represented 54% (28/52) of the whole group followed by non-keratinized squamous cell carcinoma (WHO Type II) which constituted 33% (17/52). Keratinized squamous cell carcinoma (WHO Type I) accounted for less than 5% (2/52) of patients. NPC not otherwise specified constituted 9% (5/52) of patients. T3 and T4 lesions represented 60% (31/52) of cases whereas N2 and N3 lesions were reported in 65% (34/52) of patients. More than 90% (48/52) of the present study population were AJCC stage III and IV [Table 1].

The mean dose of radiotherapy for the primary nasopharyngeal lesion was 65.8 Grays over 63 days over 36 fractions. The mean dose to the neck was 56.60 Grays over 57 days over 32 fractions.

MR imaging:

MR imaging was performed on a 1.0-T magnet (Magnetome Expert, Siemens Medical Systems, Erlangen, Germany) with maximum gradient strength of 20 mT/m; and rise time of 1 msec. A Circularly Polarized phased array head coil was used in all cases to cover the region from the frontal sinuses down to C7 level. The study protocol included T1 spin echo weighed images (T1WI) (TR=500-800ms; TE= I4-20ms) with and without the application of fat saturation pulses in the sagittal, axial and coronal planes, T2 turbo spin echo weighed images (T2WI) were obtained in the axial plane (TR = 4000ms; TE =90ms). The slice thickness was 5mm with no inter-section gap in all planes and the field of view was 20-25 cm. The post contrast T1WI were obtained after a bolus injection of 0,lmmol/kg of gadolinium dimeglu- mine (Magnivest, Schering Pharmaceutical, Germany). The MR imaging was performed prior to therapy for staging purposes. All patients had the first three MR follow up (at 2,6 &12 months), while only 23 patients had the fourth MR follow up (24 months).

Each study was retrospectively assessed by two of the authors in a consensus reading fashion for sites of primary disease and regional nodal spread. The readers were not blinded to the pathology or the treatment.

Twenty five anatomical sites of tumor involvement on MRI were classified into seven categories according to the involved structures: 1- Lateral extension included levator palatini muscle, tensor palatini muscle, prevertebral muscle, parapharyngeal fat space, medial pterygoid muscle, lateral pterygoid muscle and infratemporal fossa; 2- The anterior extension included the nasal cavity; 3- The inferior extension included oropharynx and hypopharynx; 4- The intracranial extension included cavernous sinus and cranial fossae; 5- The orbital extension; 6- The skull base extension that included sphenoid wing, clivus, pterygoid process, foramen lacerum, petrous temporal bone, jugular foramen, foramen ovale, pterygopalatine fossa and foramen rotundum; and 7- The paranasal sinus extension that included sphenoid, ethmoid and maxillary sinuses [Table 2].

The treatment volume was calculated by the sum of the measurements of longest perpendicular dimensions of the lesion in the three orthogonal planes. Local invasion was considered to be present when abnormal signal intensity and/or enhancement similar to that of the primary tumor has directly extended to the surrounding tissues. Skull base marrow infiltration was considered positive when areas of T1WI low signal intensities and contrast enhancement were detected particularly in the fat suppression sequence. T2WI hyperintensities of the same sites were also considered as additional evidence of marrow replacement [11]. All cervical lymph nodes were evaluated. Lymph nodes were considered positive when a size more than 1.0 cm in long axis was encountered. They were classified into discrete or amalgamated and the presence of central necrosis was noted.

Follow-up MRI studies were performed at 2, 6, 12 and 24 months after completion of radiotherapy using the same sequences and parameters as in the initial examination. These studies were reviewed by the same authors in a similar fashion to evaluate the tumor response to radiotherapy in all specified anatomic sites. AJCC (1997) was adopted for tumor Classification and Staging.

Patients who were radiologically classified as complete remission were followed up with imaging while those who were thought to have partial remission under went biopsy for pathological confirmation.

The criteria of local tumor response on MRI were: 1- reduction in size, 2- decreased in the signal intensity onT2 weighted sequence and 3- decrease of tumor enhancement. The criteria of response for marrow lesions were 1- fatty replacement of the marrow and 2- regression of the enhancement of the marrow infiltrative lesions.

Tumor response was classified into:

A- Complete remission (CR): defined as the disappearance of all known disease, with or without residual post radiation changes. If post radiation changes develop, then they should be negative for malignant cells on biopsy or remain stationary over the follow up period.

B- Partial remission (PR): defined as at least 50% reduction of the sum of the products of the largest perpendicular diameters of measurable lesion(s). It was not necessary for all lesions to have regressed to qualify for a partial response, but no lesion should have progressed and no new lesion should have appeared.

Statistical Analysis:

The “P” value was calculated using the software “Statistica” version 5, 1997. Two-sided method was employed.

  Results Top

The maximum local tumor size in the NPX as measured by MRI ranged from 5mm to 95mm (mean 67mm). It is measured by the sum of the products of the largest perpendicular diameters of measurable lesion(s). The tumors appeared in all patients as nasopharyngeal masses with homogenous, intermediate signal intensity on theTl WI similar to that of skull base muscles. These lesions were of bright signal intensity on T2WI. Contrast enhancement was of moderate to intense degree and mostly homogenous. This was best appreciated in fat suppressed images. For bone invasion T1WI showed well defined low signal intensity patches replacing the normal high signal intensity bone marrow. On T2WI the tumor infiltrations within the bone marrow showed variable signal intensity while on the enhanced T1WI with fat suppression they displayed moderate to intense enhancement.

The most common extension was the lateral. The second common extension was the skull base. The third common extension was the intracranial extension. The least site of involvement was the orbit [Table 2].

According to the specified criteria of local tumor response on MRI (materials and methods): 1- reduction in tumor size was noted in all patients, 2- decreased in the tumor signal intensity on T2 weighted sequence was observed in 85% (437/514) and 3- decrease of tumor enhancement was reported in 71% (365/514). Similarly applying the criteria of response for marrow lesions revealed 1- fatty replacement of the marrow in 60% (103/171) and 2- regression of the enhancement of the marrow infiltrative lesions in 87% (149/171).

The best CR rate after radiotherapy was reported with the anterior extension where 89% (23/28) of the reported sites showed complete disappearance of the local lesions on MRI. This was followed by the lateral extension where 77.3% (177/229) of the recorded sites achieved CR on MRI, followed by the inferior extension (oropharynx) in which the CR rate was 68% (13/19). On the other hand the CR rate was 0% (0/5), 15% (5/35), 16% (28/171) and 31% (9/29) in extensions of the orbit, the intracranial, the skull base and paranasal sinuses, respectively [Table 2]. Collectively, the CR rate of the lateral extension, anterior extension and inferior extension was 77% (213/276) compared with only 17.6% (42/238) for the orbital, intracranial, skull base and paranasal sinuses all in all (p<0.0000l).[Figure 1][Figure 2]
Figure 1: 29-year-old man with T4 NPC

Click here to view
Figure 2: 33-year-old man with T4 NPC

Click here to view

The PR rate among the lateral, anterior, inferior, intracranial, orbital, skull base and paranasal sinuses extensions were 22%, 11%, 32%, 75%, 100%, 595, and 66% respectively. [Figure 3]
Figure 3: 60-year-old man with T4 NPC

Click here to view

Various groups of lymph nodes were noted to be enlarged in our study. This included different levels of cervical lymph adonopathy. The prevalence of involvement of each level is beyond the scope of this study and was not specifically evaluated.

The involved lymph nodes showed low signal intensities inTlWI and slightly to moderately high signal intensities on T2WI with a uniform mild to moderate enhancement or peripheral pattern of enhancement with central necrosis. Discrete nodal masses represented 27% (11/40) while amulgumat- ed lymph nodes were 73% (29/40). Thirty-five percent (14/40) of the involved lymph nodes showed evidence of central necrosis that was low signal intensity on T1WI with non-enhancing internal components and moderate hyperintensity on T2WI. The response to radiotherapy in the neck nodal disease was dependent on the clinical stage. CR rate was 100% (6/6) in N1 patients compared with 64% (22/34) for N 2-3 patients (p=0.03).

  Discussion: Top

Due to the anatomical location of the NPC it is traditionally treated by radiotherapy which is considered the standard treatment of almost all the patients rather than surgery [1],[12],[13],[14],[15],[16],[17],[18],[19],[20],[21]. Hence, proper assessment of the primary lesion as well as the neck disease is of utmost importance prior to and after radiotherapy.

Ng et al and Sakata et al [9],[10] reported the results of a comparative study between MRI and CT scan in 67 NPC patients to estimate the capacity of each modality in assessing the extent of the disease. Their study concluded that MRI allowed more accurate evaluation of the extent of NPC than CT and should be the primary mode of evaluation.

Other investigators supported the superiority of MRI over CT in detection of the NPC extensions. This is vital to optimize the radiotherapy planning and to avoid geographical miss. It also facilitates the assessment of response to radiotherapy through comparison of serial MRI studies [5],[6],[7],[8].

Patients with AJCC T1 disease constituted 13% (7/52) of our study. Eighty six percent [6/7] of T1 lesions were located in the lateral recesses of the nasopharynx. This was in accordance to what was reported by King et al 1999 [5]. These T1 tumors showed excellent response to radiotherapy, as detected in follow-up MRI studies, leaving only post-irradiation mural effacements at the site of tumor involvement. In three cases with T1 lesions we could not precisely differentiate between partial response and post-radiotherapy focal mural thickening. Endoscopies and biopsies were performed and revealed no evidence of viable malignant cells in all three cases.

In 23 patients, MRI precisely detected the extensions to the nasal fossae. The CR rate in cases with nasal extension was 89% (23/28). This good response could be attributed to its early detection since that usually present with epistaxis.

Patients with T2a NPC constituted 11% (6/52) whereas T2b lesions represented 15%(8/52) of our study population. MRI could demonstrate direct tumor invasion of the parapharyngeal fat space and its compression by bulging tumors. In our series bulging tumors was encountered in 35% (12/33) of the parapharyngeal involvement whereas the infiltration was reported in 65% (21/35). The CR rate was 90% (13/12) in the bulging parapharyngeal tumors compared with 62% (13/21) for the parapharyngeal infiltrative lesions (p=0.047). This response was demonstrated on follow-up MRI as regression in the size, less contrast enhancement and restoration of the fat plane [5].

Oropharyngeal involvement could be also well delineated on pre- and post-radiotherapy MRI that showed actual enhancing tumor tissue in relation to involved oropharyngeal structures. Follow-up MRI studies demonstrated CR in 68% (13/19) of the involved sites. Although oropharyngeal and/or nasal extension are considered as T2 lesions according to AJJC, the CR rate in our study was better in the nasal extension than oropharyngeal. The CR rate in the former was 68% (13/19) whereas it was 89% (23/28) in the latter (p=0.04). This is consistent with the reported literature of poor local control of NPC invading oropharynx due to greater number of concurrently involved sites and large volume disease. It is also reported that over 90% of patients with oropharyngeal extension had evidence of skull base invasion and 35% had intracranial extension [5],[22],

Paranasal sinuses Invasion could be identified on basis of MRI as an enhancing soft tissue mass within the sinus cavity, however, after radiotherapy assessment it is rather difficult due to associated hypertrophy and enhancement of the inflamed mucosal lining. In the present study 3 patients with maxillary sinus invasion interpreted as partial response to radiotherapy, one case was proved to have complete response after sinus endoscopy and biopsy. Similar scenario occurred in one case of sphenoid sinus involvement that showed MRI partial response after treatment. Biopsy from the suspected MRI residual showed no viable malignant cells or the so-called “sterile tumor”.

Patients with T3 lesions represented around one third (31%, 16/52) of the current series. Nasopharyngeal carcinoma tends to invade skull base bones rather than to spread through foramina, although both events could co-exist.

In the present series the CR rate in the 9 anatomical sites of skull base extension varied from 6% (2/31) and (1/18) in foramen lacerum and petrous temporal bone respectively, to 29% (6/21) and 27% (3/11) in foramen ovale and pterygopalatine fossa respectively. The difficulty of using MRI to differentiate post-radiation tissue changes from local tumor recurrence was described by Ng et al [9], In their study they identified local invasion by the presence of abnormal signal intensity and/or enhancement in the soft tissue and/or bone marrow. They also utilized tumor size and signal reduction, and absence of enhancement as signs of tumor response to therapy. In this study we used similar criteria to attempt differentiation between tumor residual and post radiation change (see materials and methods).

Patients of T4 local disease constituted 29% (15/52) of our patients. After radiotherapy, the CR rate for all T4 lesions was 18% (5/28) while 71% (20/28) showed PR. The CR rate was higher for the anterior and middle cranial fossae extension 33% (2/6) compared with the extension to the cavernous sinus 14% (3/22), however, the difference was not statistically significant (p=0.15).

In the present study 39% (21/52) of our patients were classified as N3 and 65% (34/52) as N2-3 neck lesions. The CR rate after radiotherapy was 100% (6/6) in N1 patients compared with 64% (22/34) for N2-3 patients. The difference was statistically significant (p=0.03).

Although post radiotherapy re-evaluation can be performed based on MRI of the nasopharynx and neck to detect the regression of tumor size and degrees of contrast enhancement, however, differentiation between post-irradiation changes and possible residual disease might be difficult. Follow up and/or histopathologic evaluation are crucial in these situations. Non-invasive procedure such as positron emission tomography and MR spectroscopy might replace endoscopy and biopsy in the future when they are used on a wide scale.

In conclusion, different sites of involvement in the same tumor-staging category have different CR rate after radiotherapy. The intracranial, orbital, skull base and paranasal sinuses extensions of NPC were associated with significantly lower rates of CR.

Therefore, MRI has a prognostic value in cases of NPC and can be used as guidance to the response rates.

  References Top

Schantz SP, Harrison LB, Forastiere A A. Tumors of the nasal cavity and paranasal sinuses, nasopharynx, oral cavity, and oropharynx. In: DeVita VT, Hellman S. and Rosenberg SA, ed. Cancer principles and practice of oncology. 5th ed. Philadelphia, Pa: J.B. Lippincott Company, 1997; 765-771.  Back to cited text no. 1
Mansour M, Atef H , Saber M. Nasopharyngeal Carcinoma (NPC): Pattern of failure. Med J Cairo Univ 2000; 68(3): 205-215.  Back to cited text no. 2
American joint committee on cancer. Cancer staging manual. 5th ed. Philadelphia, Pa: Lippincott-Raven, 1997;25:171-178.  Back to cited text no. 3
Union International Contre Cancer. TNM Classification of Malignant Tumor. Geneva, Switzerland, 1968.  Back to cited text no. 4
King AD, Lam WWM, Leung SF, et al. MRI of local disease in nasopharyngeal carcinoma: tumor extent versus tumor stage. BJR 1999; 72:734-741.  Back to cited text no. 5
Kuten A, Cohen Y, Lavie R, Dale J, Ben Arush MB, Goldberg H, Stein M. Update on nasopharyngeal carcinoma in northern Israel. Strahlenther Onkol 1994; 170 (10): 565-70.  Back to cited text no. 6
Olmi P, Fallai C, Colangrande, et al. Staging and follow up of the nasopharyngeal carcinoma: Magnetic Resonance Imaging versus computer tomography. Int J Radiat Oncol Biol Phys 1995; 32:795-800.  Back to cited text no. 7
Chang VFH, Fan YF. Skull base erosion in nasopharyngeal carcinoma. Detection by CT and MRI. Clin Radiol 1996;51:625-631.  Back to cited text no. 8
NG SH, Chang TC, KO SF, Yen PS, Wan YL, Tang LM, Tsai MH. Nasopharyngeal carcinoma: MRI and CT assessment. Neuroradiology (Germany) 1997; 39 (10): 741-746.  Back to cited text no. 9
Sakata K, Maryama M, et al. Prognostic factors in nasopharyngeal tumors investigated by MRI and value of MR imaging in the newly published TNM staging. Int J Radiat Oncol Biol Phys 1999; 43(2): 273-278.  Back to cited text no. 10
Graiger AJ, Elliot JM, Genant HK. MRI imaging of bone marrow. In: McCarthy SM, ed. Categorical course in diagnostic radiology: Body MRI, Radiologic Society of North America Inc, 1999; 167-178.  Back to cited text no. 11
Hayward JL, Carbone PP and Heuson JC. Assessment of response to therapy in advanced breast cancer. Eur J Cancer 1997; 13:89-94.  Back to cited text no. 12
Qin D, Hu Y, Yan J, et al. Analysis of 1379 patients with nasopharyngeal carcinoma treated by radiation. Cancer 1988; 61:1117-1124.  Back to cited text no. 13
Flores AD, Dickson RI, et al. Cancer of the nasopharynx in British Columbia. Am. J Clin Oncol 1986; 9:281-291.  Back to cited text no. 14
Rahima M, Rakowsky E, Barzilay J, et al. Carcinoma of the nasopharynx. An Analysis of 91 cases and a compari son of different treatment approaches. Cancer 1986; 58:843-849.  Back to cited text no. 15
Wang DC, Cai WM, Hu VC. Long-term survival of 1035 cases of nasopharyngeal carcinoma. Cancer 1988;61:2338-2341.  Back to cited text no. 16
Hoppe RT, Goffinet DR, Bagshaw MA. Carcinoma of the nasopharynx. Eighteen years experience with mega voltage radiation therapy. Cancer 1976; 37:2605-2612.  Back to cited text no. 17
Mesic JB, Fletcher GH, Goepfert H. Mega voltage irradiation of epithelial tumors of the nasopharynx. Int. J Radiat Oncol Biol Phys 1981; 7:453.  Back to cited text no. 18
Perez CA, Ackerman LV, Mill WB, et al. Cancer of the nasopharynx. Factors influencing prognosis. Cancer 1969; 24:1-17.  Back to cited text no. 19
Wang CC, Meyer JE. Radiotherapeutic management of carcinoma of the nasopharynx. Analysis of 170 patients. Cancer 1971; 28:566-570.  Back to cited text no. 20
Marks JE, Bedwinek JM, Lee F, et al. Dose response analysis for nasopharyngeal carcinoma. Cancer 1982; 50:1042-1050.  Back to cited text no. 21
Teo P, Shiu W, Leung SF, et al. Prognostic factors in nasopharyngeal tumors investigated by computer tomography- an analysis of 659 patients. Radiother Oncol 1992; 23:79-93.  Back to cited text no. 22


  [Figure 1], [Figure 2], [Figure 3]

  [Table 1], [Table 2]


    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

  In this article
Materials and Me...
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded46    
    Comments [Add]    

Recommend this journal