One group implanted polymer membranes of varying pore size into the rat subcutis, and reported an 80-100 fold increase in vascularization at the membrane for those implants with the larger pore sizes (8µm versus 0.8µm) (Brauker 1995). A significantly higher level of vascularization was maintained for one year when the animals were then sacrificed. The authors of this study hypothesized that geometric factors caused the difference in tissue response.

These results were supported in a later study, which looked not only at the vascularity of tissue but also the cellular density and diffusion coefficient (Sharkawy 1997). Polymer implants of three different geometries (smooth, 60µm pores, and 350µm pores) were implanted into the rat subcutis for 4 weeks. Quantitative histology showed significant differences in the cell density immediately surrounding the implant, where the smooth implants had the highest percentage of cell bodies and fibrous tissue. Moreover, the diffusion coefficient was significantly smaller for tissue surrounding the smooth implant relative to normal subcutis. This was not true for the porous implants. The lower diffusion coefficient indicates that tortuosity has increased and volume fraction decreased, which implies that encapsulation density is inversely proportional to resistivity. Sharkawy’s histology also descriptively supported Brauker’s observations regarding vascularization. Again, implant geometry is hypothesized to alter the tissue response and diffusion properties.

Another study looked at fiber diameters in a polymer matrix ranging from 2 to 27µm (Sanders 2000). The results showed a decrease in capsular thickness proportional to the decrease in fiber diameter. While this study did not investigate fiber spacing, it again implicates geometry of the implant.