Accuracy of magnetic resonance imaging in predicting absence of fixation of head and neck cancer to the prevertebral space.

BACKGROUND: The purpose of this study was to determine the preoperative accuracy of preservation of the retropharyngeal fat plane on magnetic resonance (MR) images in predicting the absence of fixation or extension of head and neck carcinomas to the prevertebral space. METHODS: The MR images of 75 patients with T3 or T4 primary pharyngeal or laryngeal cancers seen over a 5-year period and treated at our Head and Neck Cancer Center were retrospectively reviewed. The MR images were independently and blindly evaluated by two head and neck radiologists for preservation of the retropharyngeal fat plane between the tumor and prevertebral musculature. In cases in which the fat was preserved, the prevertebral muscle complex was assessed for the presence of T2 hyperintensity and enhancement. All patients underwent panendoscopy, surgery, or both. RESULTS: Forty of 75 patients had preservation of the retropharyngeal fat plane between the tumor and the prevertebral compartment on T1-weighted images. In all 40 cases, the prevertebral muscles had a normal appearance on T2-weighted and enhanced MR images. Intraoperative assessment revealed absence of fixation of tumor to the prevertebral fascia in 39 of 40 cases, and these tumors were resectable. CONCLUSION: In patients with advanced head and neck carcinomas, preservation of the fat between the tumor and the prevertebral musculature on unenhanced T1-weighted images reliably predicts absence of prevertebral space fixation.

MR imaging characteristics of cranial bone marrow in adult patients with underlying systemic disorders compared with healthy control subjects.

BACKGROUND AND PURPOSE: In young adults, hematopoietic bone marrow has usually converted to fatty marrow. Fat hyperintensity on T1-weighted MR images facilitates the evaluation of marrow abnormalities. Our purpose was to compare cranial marrow signal intensity patterns in adults with systemic disorders and in healthy subjects. METHODS: MR images in 25 adults with underlying systemic disorders (chronic anemia, lymphoma, leukemia, or other infiltrative processes) and 44 healthy aged-matched individuals were retrospectively reviewed. Calvarial and clival marrow signal intensity on sagittal T1-weighted images was graded relative to that of orbital fat, white matter (WM), and gray matter (GM). Marrow was classified as homogeneous (uniformly isointense), diffusely heterogeneous (mottled), or focally heterogeneous (generally isointense with a focal lesion). RESULTS: In 84% of the control subjects, bone marrow was iso- or hyperintense relative to WM. Patients had abnormal diploic (n = 22) or clival (n = 17) marrow; 22 had calvarial marrow that was hypointense relative to WM compared with that in seven healthy subjects (P <.001). Marrow hypointensity relative to WM was a sensitive (93%) and specific (86%) marker of pathologic abnormality. Although marrow hypointensity relative to GM was specific (96%), it was not sensitive (67%). Calvarial and clival marrow patterns, respectively, were homogeneous in 81% and 64% of control subjects and 76% and 60% of patients. Clival marrow intensity varied more than did calvarial intensity; therefore, clival criteria were less sensitive and accurate in systemic disease detection. CONCLUSION: Homogeneous diploic marrow hypointense relative to WM on non-contrast-enhanced T1-weighted images suggests an underlying systemic or hematologic disorder and requires appropriate clinical correlation and evaluation.