Quantitation of microvessel density in squamous cell carcinoma of the head and neck by computer-aided image analysis

OBJECTIVE: Angiogenesis is the formation of new blood vessels from a preexisting vascular network. In healthy individuals it is normally suppressed and observed only transiently during development, reproduction and wound healing. However, growth, invasion, and metastasis of most solid tumors are dependent on angiogenesis. Without formation of new blood vessels, also termed as neovascularization, tumors cannot exceed a size of about 1 mm3. Therefore, neovascularization is a basic requirement for nutrition and oxygenation of tumor cells. Numerous studies in different solid as well as non-solid tumors have evaluated the prognostic value of tumor neovascularization. In solid tumors the increased microvessel density, the pathological correlate to tumor neovascularization, has been linked to a worse prognosis of the disease. The aim of the current study was to assess the prognostic value of tumor neovascularization for recurrences in squamous cell carcinoma of the head and neck by determining microvessel density. Data was collected using an automated-computerized method and as well as a manual counting method. BASIC RESEARCH DESIGN: Immunohistochemistry was performed to detect intratumoral microvessels in tumor samples of 50 patients with squamous cell carcinoma of the head and neck. We used a monoclonal mouse antibody directed against the CD34 antigen. After immunostaining, the entire tumor section was scanned microscopically at low power (×40) to identify hot spots, which are the areas of highest neovascularization. Individual tumor microvessels were then counted under high power (×200) to obtain a vessel count in a defined area, and the average vessel count in 4 hot spots was taken as the microvessel density. Microvessel counting was performed twice by computerized method, as well as manually by two independent investigators without any previous knowledge of patients’ pertinent clinical data. Subsequently, both counting techniques were statistically compared with each other. RESULTS: On computer-aided image analysis an increased microvessel density was significantly correlated with recurrence of disease (p = 0.02). Repetitive computer counts yielded similar results (p = 0.08), whereas repeated manual counts by two investigators varied significantly (p = 0.04). However, no further statistical correlations between microvessel density and patients clinical data i.e. tumor status, lymph node status, overall survival, or disease free interval could be found. Furthermore, estimation of overall survival of patients with an increased microvessel density by Kaplan–Meier curves revealed non-significant results. CONCLUSION: There is mounting evidence that suggests, that assessment of tumor neovascularization might provide a novel approach of prognostication in patients with squamous cell carcinomas of the head and neck. In particular, in the present study, the degree of angiogenesis of a tumor, as assessed by microvessel density, was found to be correlated with recurrent disease in squamous cell carcinoma of the head and neck. Computer aided image analysis, an automated technique, constitutes a time-efficient and reproducible technique for quantification of tumor vascularization. We suggest that this computerized microvessel determination could be used as a reliable method for microvessel counts, which, furthermore, seems to be superior to manual counting. However, for a reliable and reproducible assessment of tumor neovascularization, validation procedures and quality control protocols are mandatory.