Picrosirius red staining

For visualization of the collagen fibres, 4-μm thick sections were stained with picrosirius red. In brief, after deparaffinization in xylene and ethanol, sections were hydrated in distilled water, followed by incubation in sirius red (0.1% in saturated picric acid, Electron Microscopy Sciences, USA) for 1 h at room temperature, rinsed with distilled water, and counter stained with haematoxylin. Sections were examined by polarizing microscopy (BX51, Olympus, Japan). Collagen fibres were illustrated as orange-red and/or green colour.

Examination of the polarized colours of stained collagen fibres

To correct potential variability in the staining intensity of sections from different staining batches, all image-acquisition parameters were fixed during the process of image-capture, and the intensity of acquisition illumination was calibrated by adjusting only the microscope condenser aperture. The evaluation process was as follows. Mature and immature collagen fibres were differentiated by the polarization colours. Against a black background, thick yellowish-red fibres were mainly mature collagen (MC), whilst fine netlike greenish-yellow fibres were mainly immature collagen (IC) . The images were evaluated by an image analyzer (Image Pro Plus 6.0, Media Cybernetics Inc., Silver Spring, MD, USA), which can automatically calculate the area of the defined regions (yellow-red or green) in each section. Results were expressed as area of mature (SM) and immature (SI) collagen fibres. The ratios of SM/SI were calculated to represent the content of mature and immature collagen fibres. In the preliminary experiment, the authors found that reproducible values, not affected by further counts, were achieved by observing eight non-inflammation and non-overlapping regions of capsule wall. Therefore, in each section, eight regions at 200× magnification were analysed.

For the image obtained by polarized microscopy, segmentation was done and the desired pseudo colour image was obtained by substitution with pure black colour. For MC, the orange-red part was obtained by segmentation and the rest of the image was substituted with pure black, which had a grey level of zero. Thus any pixel with a grey-scale level greater than zero represented MC. The image was converted to an eight-bit grey image, and the result of the conversion was an image consisting of grey collagen fibres on a black background. A histogram of the brightness of each pixel in the image was plotted. Thus, collagen content could be expressed as the mean fraction of pixels with a grey-scale level greater than zero. For the IC, a similar procedure was followed. After segmentation, substitution, and subtraction, the green part of the image was converted to an eight-bit grey level for quantification. The ratio of MC to IC was calculated . The authors also observed eight non-inflammation and non-overlapping regions of each section (at 200× magnification) for analysis.

Statistical analysis

To eliminate subjective bias, images were evaluated independently by two observers. Data were expressed as mean ± standard error of the mean. Statistical differences between groups were analysed using SPSS 13.0 software (SPSS Inc, Chicago, Illinois, USA). One-way Analysis of Variance (AVONA) was also carried out. Differences were considered to be significant when P < 0.05.