Detection of cervical metastasis: A meta-analysis comparing computed tomography with physical examination

Background: Despite extensive coverage in recent literature, controversy continues with regard to the relative sensitivities of computed tomography (CT) and physical examination (PE). Objective: To identify a statistically significant consensus. Data Sources: Initially, data were reviewed on 47 consecutive patients with head and neck cancer on whom a total of 53 neck dissections were performed. These data were combined with findings from a 15- year MEDLINE review of the English-language literature, including references. Study Selection: All publications that contained a direct comparison of CT with PE, with appropriate data availability, were included. Data Extraction: Multiple-observer independent extraction was used. A total of 647 neck dissections were included in the meta-analysis. The definition of metastasis varied minimally among studies as follows: (1) nodal size, greater than 10 to 15 mm; (2) multiplicity of 8- to 10-mm nodes; or (3) evidence of necrosis. Necks were compared for positivity or negativity rather than for the actual nodal staging. In all cases, a final determination was made by results of histopathologic examination of surgical specimens. Data Syntheses: The results in this review favored CT over PE but were not statistically significant by use of the Fisher exact test. A combination of the present study's data with those of the literature review yielded the following meta- analysis results: sensitivity, 83% (CT) vs 74% (PE) (P=.002); specificity, 83% (CT) vs 81% (PE) (P=.7); and accuracy, 83% (CT) vs 77% (PE) (P=.006). Overall, PE identified 75% of pathologic cervical adenopathy; this detection rate increased to 91% with the addition of CT. The results of sensitivity analysis confirmed homogeneity across study designs. Conclusions: Computed tomography is a more sensitive indicator of cervical metastasis than PE. More importantly, these diagnostic modalities were additive, with CT significantly enhancing the detection rates of PE alone. All patients who are at risk for cervical metastasis should have CT or equivalent radiographic imaging performed prior to therapeutic intervention. Future studies correlating CT detection rates to the primary site and staging are needed before more specific conclusions can be drawn.