Objective To assess the factors that influence how particles might become fixed in tissues or migrate from them, by measuring the size of the injectable particles, their susceptibility to phagocytosis and their affinity for fibroblast attachment in culture.
Materials and methods The particle size of three types of particulate unphysiological bioinjectable material, i.e. Urocol (Genesis Medical, Ltd., London), Macroplastique™ (Uroplasty Ltd., Reading, UK) and Urethrin (Mentor Medical Systems, Wantage, UK) was analysed using phase-contrast light microscopy and confocal microscopy. Human monocytes from peripheral blood were incubated with the three materials in phagocytic studies, where ingestion was determined by confocal microscopy. A fibroblast cell line was used to ascertain the ability of the particles to act as a substrate for cell attachment in culture.
Results The mean ( sem) maximum particle diameters of Macroplastique, Urethrin and Urocol were 209 (5.10) µm, 49 (1.52) µm and 14 (0.39) µm, respectively. Rat peritoneal macrophages and human peripheral blood monocytes commonly ingested Urocol particles; the phagocytosis of Urethrin was rare and that of Macroplastique was not detected. Fibroblasts adhered to Urocol paste and Urethrin particles, but not to Macroplastique.
Conclusion Published reports of particle size and phagocytosis are confusing, but a relationship clearly exists. Macroplastique is the largest particle and is least likely to be phagocytosed by human mononuclear phagocytes. Urocol paste is the slowest to dissipate in culture conditions; the flat surfaces of Urethrin, but not Macroplastique, can serve as a substrate for fibroblast anchorage.
The use of injectable materials was established in the 1980s for various urological conditions, e.g. urinary incontinence [1,2] and VUR . More recent publications include some long-term follow-up assessments  and at least two reviews [5–7]. The four most commonly used nonautologous substances are bovine collagen (Contigen, Bard UK), PTFE (Urethrin, Mentor Medical Systems, Wantage, UK), particulate siloxane (Macroplastique, Uroplasty Ltd., Reading, UK) and a ceramic suspension of 60% hydroxylapatite and 40% tricalcium phosphate (Urocol, Genesis Medical Ltd., London, UK). Of these, only Contigen is a degradable organic substance and therefore not compromised by the risk of particle migration. However, it begins to degrade after 12 weeks and degradation is complete after 9–19 months; thus booster injections are required in most patients .
Ideally an injectable material should have anatomical integrity, bonding to local tissues with minimal inflammation and maintaining its injected volume. It must not be toxic and should not migrate to vital organs. Particle migration depends on particle size and the adherence characteristics of the particles to local tissues at the site of injection. Particles can migrate directly into the vascular compartment or by macrophage uptake. The former is thought to be a rare event, but may occur . Phagocytosis of particles by macrophages depends on the size, shape and surface characteristics of the particles.
We assessed three nonautologous inorganic injectable materials (Macroplastique, Urethrin and Urocol) to ascertain their particle size and susceptibility to phagocytosis by macrophages in culture.
Materials and methods
The three unphysiological bioinjectable materials were obtained from their respective manufacturers in their ‘for-use’ form; 0.4 mL of each was transferred to 500 mL of deionized water and vortexed gently until fully dispersed. An aliquot (10 mL) of this mixture was then filtered through cell-culture inserts (0.4 µm pore size, Falcon, Marathon Lab Supplies, London, UK) attached by a side-arm flask to a filter pump. The filters were allowed to dry overnight and transferred to slides for particle size analysis (National Institutes of Health image-analysis software package) using an inverted light microscope. The system was calibrated beforehand using glass beads (Sigma Chemical Co., Poole, Dorset, UK) of 150 µm diameter. Ten filters were prepared and 500 particles measured for each material. For imaging, samples of the particles were prepared for microscopy in aqueous mountant. Images were obtained of each material by reflection and/or differential interference contrast (DIC) microscopy.
Rat peritoneal macrophages were obtained from Wistar rats which had been injected intraperitoneally with water 48 h before harvesting. A midline laparotomy was made after the animals were killed by carbon dioxide anaesthesia, followed by cervical dislocation. Through the laparotomy, 10 mL of PBS was instilled into the peritoneal cavity and re-aspirated. The suspension was centrifuged at 200 g for 5 min. The supernatant was discarded and the cell pellet transferred to 25 cm2 culture-grade flasks containing 5 mL of RPMI 1640 medium, 10% fetal calf serum and 1%l-glutamine/penicillin-streptomycin cocktail. The cells were allowed to adhere for 48 h before transferring 1 mL of the previously prepared particle mixture. Incubation was continued for 8 h before fixation with 70% ethanol in water. The fixed unstained culture dishes were then transferred for DIC microscopy. Alternatively, cells were viewed ‘live’ by fluorescence confocal microscopy immediately after exposure to acridine orange (3 µg/mL final concentration). This supravital dye stains nuclei green, indicating viability and helping to locate cells on difficult substrates, e.g. opaque Urocol or on the convoluted surfaces of the transparent particulates.
Monocytes were separated from normal human peripheral venous blood samples by step-gradient sedimentation. Blood (10 mL) was collected into Universal tubes containing 1 mL of a batch solution of 2.7% (w/v) EDTA and 6% dextran 500 (Pharmacia & Upjohn, Uppsala, Sweden), dissolved in distilled water at pH 7.4. After sedimentation the leukocyte-rich plasma was carefully layered over 3 mL of Nycodenz Monocyte solution (Nycomed, Sigma-Aldrich, Poole, UK) and differentially centrifuged at 500 g and 4 °C for 15 min. The cell-rich layer was aspirated and the cells washed twice with ice-cold Ca2+- and Mg2+-free PBS by centrifugation at 50 g for 5 min. Finally, the pellets of cells were resuspended in 100 µL of PBS and transferred to Petri dishes, where they were exposed to particles in the same way as the rat peritoneal macrophages.
Particle aggregation/carrier stickiness
All three materials are provided as opaque pastes with water-miscible bases. However, the paste in which Urocol is suspended is more resistant to dissolution than the carriers of Urethrin and Macroplastique, such that when applied to cell-culture dishes and incubated (see cell adhesion experiments, below), the Urocol alone remains as a discrete layer stuck to the plastic. In aqueous suspension both Urethrin and Macroplastique particles were relatively discrete, Macroplastique more so than Urethrin, in which relatively small fragments were found associated with larger particles. Urocol appeared to contain very small basic units but with a strong tendency to aggregate. There were always abundant free particles in preparations and the aggregates were susceptible to dispersion by mechanical agitation.
On low-power microscopy, Urocol particles appeared black, gritty and solid; Macroplastique and Urethrin particles gave the appearance of crumpled translucent sheets with smooth surfaces ( Fig. 1). The density of (basic) particles per unit volume of paste was much higher in Urocol than in the other materials.
For each of the 500 particles measured from each material, the longest dimension was recorded. The derived statistics are shown in Table 1 and the distribution of sizes in Fig. 2. The upper quartile value for the basic particles of Urocol was about six times smaller than the lower quartile for Macroplastique. Even taking small aggregates into account, there was virtually no size overlap amongst the materials. Urethrin was of intermediate size, with a mean size 3.5 times larger than Urocol and four times smaller than Macroplastique.
Table 1. The mean sizes of 500 particles from preparations of each material
IQR, interquartile range.
The murine fibroblast cell line C3H10T1/2 did not adhere to Macroplastique in normal culture medium (containing serum). Cells were seen attached to Urethrin particles and to a layer of Urocol paste, with sufficient strength to withstand irrigation using a Pasteur pipette. Figure 3 shows adherent cells on Urethrin and Urocol. A fluorescence marker viewed by epi-illumunation was required to detect the presence of cells on the opaque Urocol layer; acridine orange supravital staining fulfilled this need and confirmed viability.
Peritoneal macrophages from rats and human peripheral blood monocytes phagocytosed Urocol ( Fig. 4). There was occasional evidence for phagocytosis of smaller Urethrin particles. No phagocytosis of Macroplastique was detected; none of this material was of the same order of size as mononuclear cells. The size comparison is highlighted in Fig. 1 by including a peritoneal macrophage in one panel. Figure 5 shows a cytospun preparation of zymosan-stimulated human monocytes in which multinucleate giant cells have formed, illustrating the probable maximum size of naturally occurring human phagocytes.
It is apparent that the particles examined here have very different physical and biological properties. Using in vitro systems, three properties were assessed which might be expected to affect the processing of injected material by the tissues into which they are placed. There was no intention to show the relative effectiveness of each material. Each of the properties tested has a theoretical risk/benefit profile, but how these compare between products and which factors are more significant is subjective, and compounded by cost implications. The correct choice may be different for each application. The present study also concentrated on the particulate content of the preparations; although this is considered to be the active fraction, how the body handles and dissipates the carrier may also be important.
Size is closely but not inextricably related to adherence and phagocytosis in vitro. In general, a susceptibility to phagocytosis is not a desirable characteristic. It is a cause for concern because it is a mechanism for particle migration and relocation. An insult to remote reticulo-endothelial tissue might result in an inflammatory response or some degree of blockade.
It is sometimes stated (and incorrectly attributed to Travis et al.) that 80 µm represents a threshold particle diameter for ingestion by macrophages. Certainly Smith et al., experimenting with dogs, used radiolabelled polystyrene beads of 80 µm as a tracer mixed with polydimethylsiloxane particles, 93% of which were > 50 µm. However, ultimately this assumption is attributable to Malizia et al., working with dogs and primates. These authors also directly studied the migration of Polytef particles; 90% of these were < 40 µm in diameter. Particles migrated to the draining nodes, lungs and brain in the dogs. The largest particles found were 4–80 µm and this would appear to be where the concept of an ‘80 µm maximum’ originated. Precise location is given only for the lungs, where Polytef was found free in alveoli, blood vessels and bronchi, with no sizes given.
That Malizia et al. found free particles in the lungs is interesting, as these are the organs first encountered by blood-borne substances from much of the body. Radiolabelled microspheres of 16 µm diameter are used in haemodynamic studies , where they are assumed to become trapped in the first capillary bed encountered. Malizia et al. also used strontium beads of 8.7 µm diameter, but these are of little relevance in the current context, being of a different material and very small (cited by Malizia et al. in relation to haemodynamics).
The maximum size of particle which can be ingested by a macrophage has not been reported in any published study. The question of a discrete maximum size may not be meaningful. A solitary cell can only accommodate a phagosome with which its cytoplasm can cope. This is probably appreciably less than 80 µm diameter (note the peritoneal macrophage in Fig. 1). Giant cells ( Fig. 5) can amass much cytoplasm, but are unlikely to be capable of migration.
No Macroplastique particles measured (or seen) in the present study were small enough to be suspected of ingestion by phagocytes. Urethrin particles cross any plausible size limit; the modal size (21 µm maximum diameter) is arguably small enough for ingestion by single phagocytes. Experimentally, apparent intracellular material was detected, but only as a rare event; perhaps the surface properties of Urethrin are unfavourable for ingestion. The basic particles and smaller aggregates of Urocol are readily phagocytosed. Migration patterns of this material would be affected by the cohesion of aggregates in vivo and the speed at which small particles become enmeshed in fibrin or newly deposited collagen.
The carrier fraction of the material is also potentially important. Of the physiological materials sometimes used in this context, heterologous collagen is essentially water-miscible, and autologous fat effectively an emulsion, both of which are ultimately absorbed in vivo. The inorganic particles studied here are suspended in aqueous pastes or gels. Urocol and Macroplastique are prepared in polyvinyl pyrrolidone (PVP), and Urethrin in glycerol. Urocol matrix is, in culture, more resistant to dissolution than either Macroplastique (or Urethrin). This may result from differences in the size of PVP polymer used; the Urocol carrier also contains chlorhexidine.
Surfaces of large particles may offer platforms for stromal cell anchorage. The promotion of adherence can be construed as a positive attribute. The extent to which a substance is a ready substrate for cell adhesion would influence tissue remodelling. Its incorporation into a fibrotic response might also be a factor in the immobilization of particles. Cells attach readily to Urethrin particles, as they do to Urocol paste. The fundamental Urocol solids presumably have an affinity for cells, or the basic particles would not be so readily phagocytosed. Macroplastique is not a substrate to which murine fibroblasts readily adhere. PVPs readily coat some particulates; it is this property which transforms toxic colloidal silica  into the harmless density-gradient medium which is Percoll®. Cells fractionated on this material readily lose any adhering material on washing and re-suspension. If Macroplastique particles become similarly coated with PVP, their lack of affinity for cells is perhaps not surprising.
It seems likely that the mechanism of scar formation will be different with the three agents. Their efficacy must be judged clinically, but specific attributes or sequelae may be predicted from the properties described here and assessments of any future developments similarly influenced by this information.
L.Z. Solomon, FRCS(Urol), Specialist Registrar in Urology.