Y.F. Fan

Nasopharyngeal carcinoma with intracranial spread: CT and MR characteristics.

PURPOSE Nasopharyngeal carcinoma (NPC) frequently spreads intracranially. We compare CT and MRI in identifying intracranial spread and reexamine the route of infiltration. METHOD One hundred fourteen consecutive patients with proven NPC were evaluated prospectively with T1-, T2-weighted, contrast-enhanced MRI and CT. RESULTS MRI showed 35 (31%) patients with middle cranial fossa involvement. Twenty-nine (25%) patients had cavernous sinus infiltration, while six (5%) showed only dural thickening. The most common route of spread is through the foramen ovale (FO) (12/35 patients, 34%), followed by skull base destruction (6/35 patients, 17%), foramen lacerum (FL) (6/35 patients, 17%), sphenoid sinus (6/35 patients, 17%), and combined FO and FL (5/35 patients, 14%). Using MRI as a standard, CT demonstrated the following involvement: cavernous sinus in 26 of 29 (90%) patients, FO in 9 of 12 patients, skull base in 6 of 6 patients, FO and FL in 3 of 5 patients, FL in 6 of 6 patients, sphenoid sinus in 6 of 6 patients and dura in 0 of 18 patients. CONCLUSION It is believed that NPC most commonly spreads intracranially via the FL or by direct erosion. Perineural spread through the FO is an important route, which explains why with CT evidence of cavernous sinus involvement there may be no skull base erosion. These findings are best seen on MRI.

Skull base erosion in nasopharyngeal carcinoma: Detection by CT and MRI

It is generally accepted that computed tomography (CT) is superior to magnetic resonance imaging (MRI) in demonstrating bony erosion while MRI is better in delineating soft tissue abnormality. The ability to detect skull base erosion by CT and MRI was compared in a retrospective study of 114 patients with nasopharyngeal carcinoma (NPC). Involvement of the following structures was demonstrated on CT and MRI: pterygoid plates [CT--10 (9%) patients, MRI--8 (7%) patients]; pterygoid process [CT--22 (19%) patients, MRI--22 (19%) patients]; clivus [CT--17 (15%) patients, MRI--26 (23%) patients]; petrous apex [CT--20 (18%) patients, MRI--34 (30%) patients]; sphenoid body/sinus [CT--31 (27%) patients, MRI--32 (28%) patients]; sphenoid wing [CT--12 (11%) patients, MRI--16 (14%) patients]. Erosion of the foramen ovale could be seen on CT in 19 (17%) patients but tumour was noted in the foramen in 28 (25%) patients using MRI. Contrary to common belief, MRI appears to be more sensitive in detecting bony involvement in the petrous apex, the clivus and the sphenoid wing. MRI is, therefore, the preferred technique in demonstrating skull base involvement.

Retropharyngeal lymphadenopathy in nasopharyngeal carcinoma

PURPOSE This study reports the incidence, numer, size and distribution of enlarged lateral retropharyngeal (LRP) nodes in patients with nasopharyngeal carcinoma (NPC). MATERIALS AND METHODS One-hundred and fourteen patients with histopathologically proven NPC were staged prospectively with magnetic resonance imaging (MRI) and computed tomography (CT). T1-weighted, gadolinium-DTPA enhanced, T2-weighted images and CT were obtained. RESULTS Ninety-one (80%) patients showed enlargement of both LRP and cervical nodes. LRP lymphadenopathy was detected in 59 (65%) of these 91 patients. A total of 347 neck nodes were identified. There were 72 enlarged LRP nodes (average 13 mm, range 6-34 mm). Sixteen groups of fused nodes were also seen (average 24 mm, range 15-40 m). Thirty-two patients (35%) with cervical lymphadenopathy showed no LRP lymphadenopathy. LRP nodes were seen at the following levels: C1, 35 (49%) nodes; C1/C2, 23 (32%); C2, 13 (18%) and C2/C3, one (1%). It is not known if all these nodes are due to metastasis or hyperplasia as no histopathological proof is available. CONCLUSION LRP lymphadenopathy may be seen in up to 65% of NPC patients with neck node enlargement. All the retropharyngeal nodes were located in the lateral group. Although medial retropharyngeal nodes are mentioned in the literature, they were not seen in this study. LRP nodes should be documented when surgical resection is contemplated in malignancy of the head and neck.

MRI features of cervical nodal necrosis in metastatic disease

Magnetic resonance imaging (MRI) is increasingly used to evaluate head and neck tumours but its place in the assessment of nodal metastasis is still unclear. This study compares the MRI findings with the computed tomography (CT) sign of nodal necrosis in 114 patients with confirmed nasopharyngeal carcinoma (NPC). Ninety-one (80%) patients showed neck node enlargement on both MRI and CT. Thirty-one (34%) of these patient with lymphadenopathy showed nodal necrosis. Seventy-two necrotic nodes were seen on CT. Using CT as the standard, the sensitivity and specificity of individual MRI sequences are as follows: T1-weighted (36%, 100%), T2-weighted (47%, 98%) and enhanced T1 (67%, 100%). In combination (when one or more sequences are positive), the sensitivity and specificity are as follows: T1 and T2-weighted (60%, 99%), T1-weighted and enhanced scan (67%, 100%) and T1-, T2-weighted and enhanced scan (78%, 99%). Only 26 (36%) nodes with nodal necrosis could be recognised on T1-weighted MRI as an area of low signal intensity. Paradoxically, eleven (15%) nodes, showed slightly hyperintense necrotic areas. Contrast enhanced MRI showed 48 (67%) NN thus improving detection rate. High signal intensity with or without a rim of lower signal intensity could be seen in 39 (54%) nodes on T2-weighted images. Five nodes (7%), curiously showed low signal necrotic centres. Ideally, staging of the primary tumour and nodes should be performed using a single modality. The tumour is often more advantageously evaluated by MRI. However, CT should be considered if MRI findings of nodal involvement are equivocal.

Temporal lobe changes following radiation therapy: imaging and proton MR spectroscopic findings

Abstract Radiation therapy for nasopharyngeal carcinoma affects the temporal lobes. This paper characterizes proton MR spectroscopic findings of the temporal lobes and correlates them with imaging changes. Single-voxel proton MR spectroscopic examinations were acquired from 13 healthy adult volunteers (25 spectra) and 18 patients (28 spectra). All patients had biopsy-confirmed nasopharyngeal carcinoma and were previously treated with radiation therapy. Six patients (33 %) had a single treatment and12 (67 %) patients had two treatments. Point resolved spectroscopy (PRESS) method was used (TR = 3000 ms, TE = 135 ms) and data processed automatically using the LCModel software package for metabolite quantification. Voxel size and geometry were adapted to the lesion to reduce skull-base lipid contamination. The metabolites were quantitated relative to water signal. For each location, an additional non-water-suppressed reference scan in fully relaxed conditions was performed. The imaging findings were divided into four categories: I, normal; II, edema only; III, contrast-enhancing lesions; and IV, cystic encephalomalacia. The N-acetyl-aspartate levels were reduced in 27 (96 %) spectra. Choline was increased in 3 (11 %), normal in 4 (14 %), and reduced in 21 (75 %) spectra. The creatine level was normal in 8 (29 %) spectra and reduced in 20 (71 %) spectra. Imaging showed 4 (14 %) spectra with category-I imaging findings; 5 (18 %) spectra with category-II findings; 15 (54 %) spectra with category-III findings; and 4 (14 %) spectra with category-IV findings. Magnetic resonance spectroscopy showed reduced N-acetyl-aspartate in radiation-induced temporal lobe changes. Creatine levels were relatively more stable. Choline levels may be increased, normal, or reduced. Imaging findings ranged from normal to contrast-enhancing lesions and cystic encephalomalacia.

Temporal lobe necrosis following radiation therapy for nasopharyngeal carcinoma: 1H MR spectroscopic findings.

PURPOSE To observe the patterns of radiation-induced temporal lobe necrosis (TLN) following radiation therapy for nasopharyngeal carcinoma (NPC). METHODS AND MATERIALS Twenty-five proton magnetic resonance spectroscopic (1H MRS) examinations were acquired from 13 healthy adult volunteers for comparison with data from the patient population. There were 18 patients (28 spectra) with radiologic evidence of TLN and all patients were confirmed cases of NPC treated with radiation therapy. Six patients (33%) had a single treatment while 12 (67%) patients had two treatments. All 1H MRS examinations were performed on a 2-T whole body system (Bruker) using the point-resolved spectroscopy (PRESS) method with TE = 135 ms, TR = 3000 ms, and data processed automatically using the LCModel software package for metabolite quantification. RESULTS The N-acetyl-aspartate (NAA) levels were reduced in all except one spectrum (96%). Choline (Cho) was increased in 3 (11%), normal in 4 (14%), and reduced in 21 (75%) spectra. The creatine (Cr) level was normal in 8 (29%) spectra and reduced in 20 (71%) spectra. In four patients with normal imaging findings 1H MRS was abnormal. CONCLUSION 1H MRS can characterize radiation-induced TLN. Spectra with increased Cho can be mistaken for neoplasm. Spectroscopy can also identify metabolic derangement before imaging.

Pterygopalatine fossa and maxillary nerve infiltration in nasopharyngeal carcinoma

Nasopharyngeal carcinoma (NPC) may infiltrate the pterygopalatine fossa (PPF) and the maxillary nerve. This study illustrates involvement of the naxillary nerve in the PPF with perineural spread to the cavernous sinus.

Functional endoscopic sinus surgery (FESS): What radiologists need to know

The place of coronal computed tomography (CT) in the assessment of patients prior to functional endoscopic sinus surgery (FESS) is well established. The ability to accurately correlate radiological and surgical anatomy enhances precision and safety during FESS. This pictorial essay reviews the conceptual anatomical framework that forms the basis of FESS.

Parotid Gland Involvement in Nasopharyngeal Carcinoma

PURPOSE This study documents the imaging findings of parotid gland involvement in nasopharyngeal carcinoma (NPC). METHOD This study reviewed the film records of 1,916 patients with NPC seen over a 5 year period. There were 27 (1.4%) patients with parotid involvement. There were 27 CT and 4 MR studies, and the pattern of involvement was categorized into the following groups: (1) parotid nodal involvement; (2) parotid invasion from extracapsular cervical nodal spread; (3) direct invasion by NPC; and (4) diffusely infiltrated gland not related to (2) or (3). RESULTS In the subset of 27 patients, parotid nodal involvement was seen in 11 (41%) patients. Ten (37%) patients had cervical extranodal spread into the parotid gland. Three patients (11%) showed direct invasion by NPC, and another three patients (11%) showed an isolated and diffusely infiltrated parotid gland. CONCLUSION Metastatic NPC may involve parotid nodes. The parotid gland may also be involved by extracapsular spread from enlarged cervical nodes or infiltrated directly by the primary lesion in the nasopharynx.

Comparison of CT and MRI features in sinusitis

OBJECTIVE To correlate the features of inflammatory changes in the paranasal sinuses on magnetic resonance imaging (MRI) with computed tomography (CT). METHODS AND PATIENTS One hundred and fourteen patients with histologically proven nasopharyngeal carcinoma (NPC) were staged with both CT and MRI. All CT and MRI images of patients with mucosal thickening but no tumour involvement of the sinuses were retrospectively analysed. RESULTS There were inflammatory changes in 36 maxillary, 21 sphenoid and 16 ethmoid sinuses. These changes include mucosal thickening, retention cysts, retained secretions, inspissated secretions and dystrophic calcification. MRI is superior to CT in separating thickened mucosa, retained secretions and retentions cysts. CONCLUSION It is important to appreciate CT changes of sinusitis and the corresponding spectrum of MRI features.