Despite an adequate diagnostic workup, 2% to 3% of neck node metastases from a primary head-and-neck tumor cannot be detected. There is no general consensus on the optimal diagnostic workup, and its timing, for carcinomas of an unknown primary (CUP). The purpose of our study is to describe the detection rate and identification of the location of the primary after a conventional workup and panendoscopy with or without a tonsillectomy and with or without an 18-F-FDG positron emission tomography (PET) or PET/computed tomography (CT) scan. We retrospectively reviewed the charts of 236 patients treated between June 1983 and December 2008 who were diagnosed as having CUP after a conventional workup that included physical examination (PE), mirror and fiber-optic examination, CT, and/or magnetic resonance imaging (MRI) of the head and neck, and for whom there was no identified primary tumor. Panendoscopy was subsequentely performed at least once in all patients. 18-F-FDG PET or PET/CT was obtained in 21 cases. A tonsillectomy was performed in 79 patients, 7 of which were bilateral. Patients with a low-neck or supraclavicular tumor location were not considered for this study since they had a primary most likely localized below the clavicles. Only patients with squamous cell carcinomas were included in the study. A primary lesion was found in 126 patients (53.3%), 6 of whom had 2 synchronous primaries. Locations were: tonsillar fossa, 59 patients (44.7%); base of tongue, 58 patients (43.9%); pyriform sinus, 10 patients (7.6%); pharyngeal wall, 3 patients (2.3%); nasopharynx, 1 patient (0.75%); and supraglottic larynx, 1 patient (0.75%). Detection rate was significantly lower in patients with no suspicious findings on workup and a negative physical exam CT and/or MRI (29.2%), as compared to those with at least 1 suspicious finding (64.0%). FDG-PET or PET/CT was performed in 21 patients and was positive in 5 of 21 patients (23.8%); in 3 patients a biopsy-proven primary was confirmed (14.3%). In 35 of 79 patients (44.3%) who underwent a tonsillectomy, a primary was found. Multivariate analysis revealed that clinical suspicion on PET and CT and/or MRI were significantly-related with the identification of the primary (p = 0.0234 and p < 0.0001, respectively). The most common primary sites for CUP are the tonsillar fossa (44.7%), base of tongue (43.9%), and pyriform sinus (7.6%). Panendoscopy with biopsy of suspicious sites remains the mainstay of the diagnostic workup. Metabolic studies do not appear to add valuable information to the detection rate of the primary site. An ipsilateral tonsillectomy should be performed, particularly if there are suspicious signs on PET and/or radiologic studies.

This report presents an incidental finding of a potentially life-threatening large fusiform aneurysm of the internal carotid artery at the level of the posterior communicating artery in a patient who reported with complaints relating to the presence of a radiolucent lesion in the mandible that was diagnosed as odontogenic keratocyst. This case highlights the need to study a case in detail in preparation for prescribing the appropriate advanced imaging modality examination, and also to carefully evaluate the entire imaged volume for incidental pathology and not just the region of interest. This report assumes significance in the light of widespread use of cone beam CT by dental clinicians for routine diagnostic tasks without a formal interpretation being carried out on all such studies.

Advanced imaging modalities are being increasingly used in dentistry for a variety of diagnostic tasks, including implant treatment planning. Interpretation of such images by radiologists is overlooked as the primary objective is to obtain diagnostic information pertinent to the task at hand. There is danger of overlooking critical lesions that may co-exist and appear on such studies as an incidental finding.

Purpose. To analyze parameters that may influence the likelihood of local control after definitive radiotherapy for head and neck cancer.Methods. Between April 1980 and January 2000, 404 patients were treated with definitive RT alone (358 patients) or combined with adjuvant chemotherapy (46 patients) at our institution and were followed up for 0.25 to 20.25 years (median, 3.5 years.) All living patients were followed up for at least 2 years. All patients had the primary tumor volume calculated on pretreatment CT. End points were local control after RT and local control after RT without a severe late complication. Parameters evaluated in multivariate analyses of these end points included primary site, T stage, primary tumor volume, N stage, histologic differentiation, fractionation schedule, adjuvant chemotherapy, and gender.Results. The rates of local control and local control without a severe late complication after RT were significantly influenced by primary tumor volume for patients with cancer of the supraglottic larynx and true vocal cord. In contrast, the rates of local control and local control without severe complications for patients with tumors of the oropharynx and hypopharynx were less influenced by tumor volume. Multivariate analysis of the overall population revealed that the only parameter that was significantly related to the probability of local control after RT was T stage. Multivariate analyses stratified by primary site revealed that tumor volume significantly influenced local control for patients with cancers of the supraglottis (p = .0220) and glottis (p = .0042) but not for those with lesions of the tonsillar fossa/posterior tonsillar pillar (p = .0892), base of tongue (p = .9493), anterior tonsillar pillar/soft palate (p = .5909), and hypopharynx (p = .2282).Conclusions. The most important parameter that has an impact on local control after RT is T stage. Primary tumor volume also significantly influences the probability of local control in cancers of the supraglottis and glottis. (C) 2003 Wiley Periodicals, Inc.

Background. To define the role of planned neck dissection after definitive radiotherapy for patients with nodepositive squamous cell carcinoma of the head and neck.Methods. Review of the pertinent literature.Results. Radiotherapy alone produces a relatively high likelihood of regional control for patients with early-stage neck disease. Patients with more advanced neck disease have a higher probability of regional control if a planned neck dissection follows radiotherapy. However, for patients who have a complete response to radiotherapy, the likelihood of an isolated recurrence in the neck is low. Radiographic evaluation of the response to radiotherapy may better define the subset of patients who are likely to benefit from a neck dissection.Conclusions. Neck dissection after definitive radiotherapy improves regional control for patients with advanced neck disease. Patients who have a complete clinical and radiographic regional response to radiotherapy have a low probability of an isolated recurrence in the neck. It is advisable to proceed with a neck dissection for patients who have an equivocal response to radiotherapy, because the likelihood of salvage of an isolated recurrence in the neck is remote. (C) 2002 Wiley Periodicals, Inc.

Background. Perineural invasion is observed in a small subset of patients with carcinomas of the skin of the head and neck.Methods. Review of the patient literature highlighting the University of Florida experience.Results. Patients with early perineural invasion are asymptomatic, and the phenomenon is discovered only on pathologic examination of the excised lesion. These patients are defined as having "incidental" perineural invasion, and treatment with surgery followed by postoperative irradiation results in a cure rate of approximately 80%. Undiagnosed, the perineural carcinoma slowly progresses and eventually results in symptoms, usually facial weakness or numbness. The disease eventually extends to the skull base and becomes incompletely resectable. Symptomatic patients are defined as having "clinical" perineural invasion, and aggressive treatment results in a cure rate of approximately 45%.Conclusions. Perineural invasion is an uncommon spread pattern observed in patients with skin cancer and is associated with a relatively poor prognosis. The likelihood of cure is inversely related to the proximal extent of the cancer and is lower for symptomatic compared with asymptomatic patients. (C) 2002 John Wiley & Sons, Inc.

PURPOSE: To establish the normal variations of the postcricoid portion of the hypopharynx, esophageal verge, and cervical esophagus, as seen on computed tomographic (CT) and magnetic resonance (MR) images.MATERIALS AND METHODS: One hundred twenty-one CT and 92 Md images were reviewed. Diameter and wall thickness were measured at multiple levels. Depiction of the layers of the musculature and adjacent fat planes was evaluated. The frequency and size of the tracheoesophageal lymph nodes were noted.RESULTS: An esophageal anteroposterior diameter greater than 16 mm and lateral diameter greater than 24 mm were considered abnormal. The average wall thickness was 4.8 mm laterally and 3.8 mm posteriorly. Demonstration of the intramural fat planes of the postcricoid region decreased from the upper to the lower region of the cricoid cartilage. The ability to separate the esophageal wall from the trachea was highest at the esophageal verge and declined markedly more distally. The tracheoesophageal groove nodes were seen more often on the right (mean size [+/- SD], 4.5 mm +/- 2.2).CONCLUSION: Knowledge of the normal appearance-and variations of the postcricoid region and cervical esophagus is essential in detecting abnormalities in these areas.

PURPOSE:To determine the utility of pretreatment computed tomography (CT) for predicting primary site control in patients with supraglottic squamous cell carcinoma (SCC) treated with definitive radiotherapy (RT).MATERIALS AND METHODS:Pretreatment CT studies in 63 patients were reviewed. Minimum length of follow-up was 2 years. Local recurrence and treatment complications resulting in permanent loss of laryngeal function were documented. Tumor volume was calculated using a computer digitizer, and pre-epiglottic space (PES) spread was estimated. The data were analyzed using a combination of Fisher's exact test, logistic regression modeling, and multivariate analyses. Five-year local control rates were calculated using the product-limit method.RESULTS:Local control rates were inversely and roughly linearly related to tumor volume, although there seemed to be a threshold volume at which primary site prognosis diminished. Local control was 89% in tumors less than 6 cm3 and 52% when volumes were > or =6 cm3 (P = .0012). The likelihood of maintaining laryngeal function also varied with tumor volume: 89% for tumors less than 6 cm3 and 40% for tumors > or =6 cm3 (P = .00004). Pre-epiglottic space involvement by tumor of > or =25% was associated with a reduced chance of saving the larynx (P = .0076). Multivariate analyses revealed that only tumor volume independently altered these end points.CONCLUSION:Pretreatment CT measurements of tumor volume permits stratification of patients with supraglottic SCC treated with RT alone (which allows preservation of laryngeal function) into groups in which local control is more likely and less likely. Pre-epiglottic space spread is not a contraindication to using RT as the primary treatment for supraglottic SCC.

Over the past 18 years, numerous classifications have been proposed to distinguish among the diverse nodal levels. Some classifications have used surgical landmarks, others physical assessment criteria. These classifications do not agree precisely and exhibit sufficient variation that competent physicians could arrive at slightly different staging of the patient's nodal disease. In the past 2 decades, computed tomography and magnetic resonance imaging have offered progressively more refined anatomical precision, reproducibility, and visualization of deep, clinically inaccessible structures. Because the majority of patients with head and neck malignancies presently undergo sectional imaging prior to treatment planning, we felt a need to integrate anatomical imaging criteria with the 2 most commonly used nodal classifications: those of the American Joint Committee on Cancer and those of the American Academy of Otolaryngology-Head and Neck Surgery. The imaging-based nodal classification proposed herein has been developed in consultation with surgeons interested in such classifications in the hope that the resultant classification would find ready acceptance by both clinicians and imagers. It is our desire that the best attributes of imaging, combined with those of the physical assessment, can result in a better and more consistently reproducible nodal staging than is possible by either

PURPOSE: To describe the computed tomographic (CT) measurements and features that enable accurate diagnosis of congenital nasal piriform aperture stenosis (CNPAS).MATERIALS AND METHODS: The clinical and Imaging features of six patients (age range,0-11 months) with CNPAS were retrospectively evaluated and compared with those of 61 control subjects younger than 24 months.RESULTS: The average widths of the piriform aperture in patients aged 0-3, 4-6, and 10-12 months were 4.8, 7.0, and 6.0 mm, respectively, compared with 13.4 (P <.001), 14.9, and 15.6 mm, respectively, in the age-matched control subjects. The average areas of the piriform aperture in patients in the three age groups measured 0.24, 0.46, and 0.36 cm(2), respectively, compared with 0.67 (P <.001), 0.86, and 1.11 cm(2), respectively, in the age-matched control subjects. The width of the nasal cavity in patients with CNPAS was in the less than 5th percentile compared with that in the control subjects. Abnormal dentition and a midline bone ridge projecting from the inferior palate were present in all six patients.CONCLUSION: CT measurement of the piriform aperture width is a simple and accurate method for diagnosing CNPAS; a width less than 11 mm in a term infant is considered to be diagnostic. An abnormal dentition and a bone ridge along the underside of the palate are confirmatory imaging findings.

PURPOSE Our goal was to describe the CT findings before and after radiation therapy in a series of patients with laryngeal chondroradionecrosis. METHODS The CT studies obtained before and after radiation therapy in nine patients with the diagnosis of laryngeal chondroradionecrosis were reviewed retrospectively. RESULTS CT scans revealed abnormalities in all patients. A variable degree of laryngeal soft-tissue swelling was seen in eight of the patients. In four patients, cartilaginous abnormalities were visible initially, and appeared in three of four other patients who had further follow-up CT studies. Six patients had involvement of the thyroid cartilage; collapse of the thyroid cartilage was seen in two cases and gas bubbles were visible adjacent to the thyroid cartilage in three cases. Four patients with involvement of the thyroid cartilage eventually underwent total laryngectomy, and one died suddenly in severe respiratory distress. In all three patients with arytenoidal involvement, anterior dislocation of this cartilage was seen; in two of these patients, the adjacent part of the cricoid cartilage showed some sclerosis. Two patients with arytenoidal necrosis (both with cricoidal sclerosis) kept a functional larynx. In one case, cricoidal sclerosis was seen in association with lysis of the thyroid cartilage. CONCLUSION The CT appearance of laryngeal chondroradionecrosis is nonspecific, but the diagnosis can be strongly suggested in cases of sloughing of the arytenoid cartilage, fragmentation and collapse of the thyroid cartilage, and/or in the presence of gas bubbles around the cartilage.

PURPOSE: Our purpose was to determine the variation in location, distribution, and sex predilection of arytenoid cartilage sclerosis in a population of patients without laryngeal cancer as an aid to understanding the significance of this entity when seen in patients with laryngeal cancer. METHODS: One hundred CT examinations of patients of different ages and with no history of laryngeal cancer or radiation therapy were evaluated for the presence of arytenoid cartilage sclerosis. The arytenoid cartilage was subdivided into three regions: superior process, body, and vocal process. Each region was assessed on each side separately on bone window scans. RESULTS: Arytenoid cartilage sclerosis was seen in 16% (n = 16) of the patients. Sclerosis was most common in the body (n = 12) and least common in the true vocal process (n = 2); the left side was the preferred location for all three regions. In 50% of the patients, at least two regions were affected. Eighty-one percent of sclerotic arytenoid cartilage was seen in women. CONCLUSION: Knowledge of the frequency and distribution of arytenoid cartilage sclerosis as a normal variant can aid in determining the risk of arytenoid cartilage involvement in patients with laryngeal cancer.

OBJECTIVE Defects of the cochlear modiolus have been found to be associated with most cases of large vestibular aqueduct. The clinical significance of these modiolar defects has not been studied previously. The purpose of this article is to correlate clinical (functional) parameters, such as hearing outcomes, with the severity of the radiographic findings in these dysplastic inner ears. STUDY DESIGN The study design was a retrospective chart review, supplemented with telephone interviews and clinic visits. SETTING The study was conducted at an academic, tertiary care center. PATIENTS Thirty consecutive patients with large vestibular aqueducts participated. RESULTS Scores of modiolar deficiencies yielded inconsistent correlations with hearing loss. Vestibular aqueduct morphology and thickness correlated very strongly with the severity of hearing loss. CONCLUSIONS These observations support the hypothesis that large vestibular aqueduct-related hearing loss may be caused by transmission of subarachnoid pressure forces into the inner ear. However, the thickness and morphology of the vestibular aqueduct may simply be markers for more subtle cochlear dysplasia manifest by modiolar deficiency.

PURPOSE:To determine the computed tomographic (CT) appearance of the normal modiolus and the pathologic alteration in patients with a large vestibular aqueduct and an otherwise normal-appearing cochlea.MATERIALS AND METHODS:Temporal bone CT studies obtained before and after a major upgrade of CT capability in 1992 were reviewed in four groups: Group A (1.5-mm section thickness) comprised 50 normal ears in 43 patients, group B (1-mm section thickness) comprised 75 normal ears in 50 patients, group C (1.5-mm section thickness) comprised 16 ears with a large vestibular aqueduct in 10 patients, and group D (1-mm section thickness) comprised 23 ears with a large vestibular aqueduct in 12 patients. All groups comprised adult and pediatric patients.RESULTS:In groups A and B, the normal modiolus was visualized in 90% and 100% of ears, respectively. In groups C and D, with a total of 39 ears with a large vestibular aqueduct and an otherwise normal cochlea, modiolar deficiency was demonstrated in 100% of ears.CONCLUSION:CT is an excellent technique for depicting the cochlear modiolus. Results suggest that all ears with a large vestibular aqueduct have associated cochlear modiolar deficiencies. Thus, a large vestibular aqueduct may be only occasionally, if ever, an isolated developmental anomaly of the inner ear.

PURPOSE: To establish the range of normal variation in the CT appearance of the middle ear ligaments and the stapedius tendon as an aid in detecting abnormal changes in these structures. METHODS: CT scans of the temporal bone in 75 normal middle ears, obtained with 1-mm-thick sections, were reviewed by two observers, who rated the visibility of the structures of interest on a scale of 1 to 5. RESULTS: The anterior, superior, and lateral malleal ligaments and the medial and lateral parts of the posterior incudal ligament were seen in 68%, 46%, 95%, 26%, and 34% of the ears, respectively. The stapedius tendon was seen in 27% of the cases. When visible, the ligaments were judged to be complete in 90% to 100% of the ears and the stapedius tendon was complete in 65% of cases. Their width varied considerably. Interobserver variability was high for most observations. CONCLUSION: CT scans are more likely to show the malleal than the incudal ligaments. Although the interobserver agreement was statistically significant for most study parameters, the percentage of agreement above that expected by chance was low. When seen, the ligaments usually appeared complete. Understanding the normal range of appearance may help identify abnormalities of the ligaments and tendons of the middle ear.

Joubert syndrome is characterized by episodic hyperpnea and apnea, developmental delay, hypotonia, truncal ataxia, ophthalmologic abnormalities, and vermian dysgenesis. We studied 15 patients with the diagnosis of Joubert syndrome to (1) more fully define the syndrome's clinical features, and (2) correlate the clinical features with magnetic resonance imaging (MRI) findings. Eight of 15 patients had a history of episodic hyperpnea and apnea. All patients had developmental delay and hypotonia. Of the 13 patients receiving detailed neuro-ophthalmologic evaluations, three had optic nerve dysplasia, pendular nystagmus, and gaze-holding nystagmus. All 13 patients had a normal vestibulo-ocular reflex based on head thrust, but had absent to poor ability to cancel the vestibulo-ocular reflex horizontally and vertically. Twelve of 13 patients had impaired smooth pursuit. Twelve of 13 patients had defects in initiation of saccades and quick phases. Two of the most consistent radiologic features were absent or hypoplastic posterior cerebellar vermis, and deformed midbrain and pontomesencephalic junction, which based on ocular motor physiology correlate with the vestibulo-ocular reflex cancellation/pursuit defect and saccade initiation defect, respectively As a result of midbrain, vermian, and superior cerebellar peduncle abnormalities, axial neuroimaging showed a unique "molar tooth" appearance of these structures. These results indicate that Joubert syndrome results from maldevelopment of the midbrain and cerebellar vermis, producing a pathognomonic sign on MRI. (J Child Neurol 1997;12:423-430).

PURPOSE: To establish the computed tomographic (CT) visibility of the incudostapedial joint and of the stapes superstructure in normal and opacified middle ears. MATERIALS AND METHODS: Two independent observers reviewed CT studies with 1-mm section thickness of the temporal bone in 75 normal ears and in 33 opacified middle ears (13 of which had definitive proof of disease) to establish the rate of visualization of the stapes and incudostapedial joint. RESULTS: In normal ears, both the stapes crura and the continuity between the incus and stapes were seen in almost 100% of cases, whereas the actual incudostapedial joint was identified in 86% (86 observations) and 67% (67 observations) of cases in the axial and coronal planes, respectively. The position of the incudostapedial joint was below footplate level on the axial images and mostly at or anterior to the midportion of the foot-plate on the coronal images. In 13 clinically confirmed diseased middle ears, the status of the stapes superstructure in all cases and that of the incus in 11 cases was correctly predicted with CT. CONCLUSION: It is now possible to visualize routinely the incudostapedial joint and stapes superstructure at CT. Absence of these structures in an opacified middle ear strongly indicates abnormality.