Angiogenesis is necessary for the growth and maintenance of preinvasive and invasive cervical lesions. Angiogenesis and neovascularization in cervical intraepithelial neoplasia (CIN) are reflected by the characteristic appearance of mosaic and punctation patterns on colposcopy. Widening of the intercapillary distance has been postulated to be associated with increasing severity of the lesion. Computerized colposcopy/image analysis is the non-invasive digital recording and processing of a colposcopic image. Computerized colposcopy has the capacity to enhance, process, magnify, quantify, store, retrieve, and sequentially compare colposcopic images. In addition, it has the ability to enhance the visualization of capillaries. The intercapillary distance can, thus, be quantified and intercapillary irregularities identified. The clinical usefulness of computerized colposcopy and image analysis in the conservative management of CIN in the non-pregnant (1) and pregnant (2) states have been previously reported. The purpose of this study was to objectively measure and compare the intercapillary distance in different grades of CIN using computerized colposcopy. Subjects were patients with abnormal Pap tests attending a colposcopy clinic at an inner-city hospital in the Bronx Borough of New York City. The computerized colposcope used was the DENVU 3100 (DENVU Inc, Tucson, Arizona, USA). The system consists of a standard photocolposcope, an optical interface for image digitization, a computer for recording and analyzing the image, a software package, and a charge-coupled devise camera. The software package includes image storage, file manipulations, image digitization, contrast and edge enhancement, and distance and area measurements. The measurement and graphics feature provides the ability to: draw and measure straight-line distances, draw irregular lines around regions of interest to measure perimeter (P) and surface area (SA), draw adjustable circles around fixed points, paint areas and apply text. Using computerized colposcopy and image analysis we studied 73 consecutive patients with a fully visualized squamocolumnar junction and a histopathologic diagnosis of CIN II (n = 42) and CIN III (n = 31). After annotating and storing the image for analysis, a colposcopically directed biopsy was obtained from the same studied area and was submitted for histopathologic examination. To obtain accurate measurements the system was calibrated to compliment the colposcope magnification power setting used for acquiring the images. All patients had computer-assisted measurements of the farthest two intercapillary points within each single mosaic as well as the mosaic P, and mosaic SA. Mosaic P represents the total length of the outer boundary of the mosaic while mosaic SA exemplifies the total area contained within the perimeter. To measure the intercapillary distance the mouse cursor was moved to the measurement start point and a marker was developed. The mouse cursor was, then, moved to the measurement end-point and, after clicking the marker, the distance was displayed and recorded. To measure mosaic P and SA the mouse cursor was directed to the desired start-point. The cursor was, then, moved to follow the outline of the mosaic. A line trailed the cursor and line segments were continued until the mosaic outline was completely traced. After reaching the end-point the mosaic P and SA were displayed and measured. All measurement values were expressed in millimeters. The mean intercapillary distance, P, and SA for each patient was calculated by adding all intercapillary distances, P's, and SA's of the whole mosaic pattern divided by the number of mosaics. The acquired digital images were analyzed blind of histopathologic findings. The Student's t-test was used for statistical analysis. The mean age of the patients was 31 ± 7.3; the majority were hispanics (49%), and African Americans (33%). The mean intercapillary distance in CIN II was 0.164 mm (SD ± 0.037), and in CIN III was 0.227 mm (SD ± 0.042). The difference was statistically significant (p < 0.01). The mean mosaic P in CIN II and III were 0.251 mm (SD ± 0.042), and 0.439 mm (SD ± 0.051), respectively, and the mean mosaic SA were 0.006 mm2 (SD = 0.001), and 0.013 mm2 (SD ± 0.002), respectively. The difference was statistically significant (p < 0.05). The results demonstrate increased intercapillary distance, mosaic perimeter and mosaic surface area with increasing severity of CIN. The increase in intercapillary distance may reflect insufficient vascular adaptation to an accelerated cellular growth rate. Measuring the intercapillary distance was previously reported using various techniques including enlarged colpophotographs (3), immunohistochemical staining for factor VIII antigen (4), histochemical staining for capillary endothelial cells (5) and a semiautomatic image analysis system (6). In the present study, computerized colposcopy allowed for the objective recording and measurement of changes in the intercapillary distance and neovascularization patterns in different grades of CIN. It has long been believed that invasive tumors depend on neovascularization for their continued growth, expansion, and possibly metastasis. Cell proliferation leads to capillary compression and consequently increased intercapillary distance. The intercapillary distance therefore is thought to reflect tumor oxygenation. Kolstad (3) assessed the oxygenation of a cervical cancer tumor surface, using the intercapillary distance as a parameter, and revealed a gradual decrease in tumor oxygenation with advancing stage of the disease. Kolstad also reported increased superficial tumor necrosis and increased local recurrence with intercapillary distances greater than 350 and 400 µ, respectively. Tumor intercapillary distance was also utilized as a useful tool to identify subgroups where hypoxia can interfere with the effectiveness of radiotherapy. The mean intercapillary distance was reported to decrease progressively during the course of external beam pelvic irradiation of cervical cancer (5). In preinvasive cervical disease neovascularization is confined to a narrow zone immediately underlying the dysplastic epithelium and it is suggested that neovascularization may be necessary for the growth, maintenance or progression of dysplastic lesions (4). The degree of neovasularization in cervical dysplasia does not correlate with the amount of inflammation in the lesion and the angiogenic response probably reflects signals generated by the dysplastic cells themselves. The degree of neovascularization also appears to be independent of the type of human papillomavirus present (4). Finally, computerized colposcopy may be useful in differentiating the degree of vascular atypicality, selecting the appropriate site for directed biopsy, and objectively evaluating neovascularization patterns of preinvasive and invasive cervical lesions. Because vascular changes may precede the histopathologic criteria of cervical neoplasia the use of computerized colposcopy may allow for the early detection of cervical lesions. In addition, computerized surveillance of the intercapillary distance may help to elucidate the pathophysiologic changes associated with progression or regression of cervical intraepithelial neoplasia.
No takes yet. Share an insight, caveat, or question.
Mikhail et al. (2004) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: