Detection of metastases from head and neck cancers.

Before treatment for head and neck malignancies is begun, a search for distant metastases (DM) is performed. The first objective of this review was to determine the accuracy of liver function tests (LFT), alkaline phosphatase (AP) tests, and chest radiographs (CXR) in detection of DM. Second, an effort was made to identify tumor characteristics which are associated with a higher incidence of DM and therefore justify the use of more precise screening tools. An analysis of 97 patients with noncutaneous squamous cell carcinomas presenting to the Stanford Head and Neck Tumor Board in 1991 revealed 17 DM in 14 patients. There were 10 pulmonary metastases, 5 bone metastases, and 2 hepatic metastases. CXR had a sensitivity of 50% and a specificity of 94% for detection of pulmonary DM. AP tests showed a sensitivity of 20% and a specificity of 98% for detection of bone DM. LFT had a sensitivity of 50% and an 81% specificity for demonstration of hepatic DM. A separate analysis of 79 patients with known DM from two hospitals showed the incidence of DM to be increased in patients who had tumors of advanced stage, advanced T status, and poor histologic differentiation and to also be increased in the presence of local-regional recurrence. There was little association of DM with N status. The sensitivity of CXR and laboratory tests, which are currently used in evaluation for DM at most cancer centers, is disappointing; these tests should be viewed as gross screening examinations. We recommend a chest computed tomography scan in the event of an abnormal CXR, a bone scan in the event of an elevated AP, and either an ultrasound or computed tomography/magnetic resonance imaging scan of the liver when elevated LFT levels are present, depending on tumor stage and differentiation.