- Email: [email protected]
Radiofrequency-generated glow discharge treatment: potential benefits for polyester ligaments
J Orthop Sci (2003) 8:198–206
Radiofrequency-generated glow discharge treatment: potential benefits for polyester ligaments John Richard James Rowland1, Satoshi Tsukazaki2, Toshiyuki Kikuchi2, Kyosuke Fujikawa2, John Kearney3, Richard Lomas3, Edward Wood4, and Bahaa Botros Seedhom5 1
Xiros plc, 28-30 Blenheim Terrace, Leeds LS2 9HD, UK Department of Orthopaedic Surgery, National Defense Medical College, Saitama, Japan 3 Tissue Services R&D, Trent Blood Centre, Sheffield, UK 4 School of Biochemistry and Molecular Biology, University of Leeds, UK 5 Academic Unit of Musculo-Skeletal and Rehabilitation Medicine, University of Leeds, UK 2
Abstract This multicenter study has revealed that treating a woven polyethylene terephthalate (polyester) ligament with a radiofrequency (RF)-generated glow discharge (RFGD) produces marked benefits in terms of increased cell attachment and proliferation on the implant surface. In vitro tests of the same material revealed that the number of synovial fibroblasts attached to the treated samples after 14 days was four times that of the untreated material. Many of the cells were spread over the surface of a single filament, and some formed bridges between one filament and the next. The incorporation of [3H]-thymidine by synovial stromal cells (a measure of the amount of cell division) growing on the treated material was five times that on the untreated samples. The amount of DNA present on the treated material was also found to be almost an order of magnitude greater than that on untreated samples. This increase in cell attachment and proliferation is almost certainly related to a notable increase in wettability of the polyester surface induced by treatment. Mechanical tests revealed that, for ligaments with a nominal ultimate tensile strength of 2100 N, RF-generated glow treatment reduced the ligament’s strength by 12% but increased its stiffness by 15%. After a medium-term fatigue test (10.8 million cycles), however, there appeared to be recovery of the mechanical properties, with the strength and stiffness of untreated and treated samples being essentially the same. After exhaustive fatigue tests (more than 62 million cycles) the residual strength of the treated ligaments was only 9% lower than that of the unfatigued and untreated ligaments. Key words Radiofrequency-generated glow discharge · Leeds-Keio ligament · Cell attachment · Strength · Wettability
Introduction Because knees with a ruptured anterior cruciate ligament (ACL) frequently degenerate, there is a
Offprint requests to: J.R.J. Rowland Received: July 1, 2002 / Accepted: October 22, 2002
tendency to reconstruct this ligament. During the 1980s and early 1990s synthetic ligaments were commonly used for this procedure. Because of various problems encountered, however, most of the current reconstructive procedures use autogenous tissue. Bone–patellar tendon–bone grafts using the central third of the patellar tendon have been popular for some years, as these grafts have been reported to have an initial strength as high as 2977 N (10-mm strip) or 4389 N (15mm strip) and stiffness of around 1150 N mm1.5,15 More recently, many surgeons have preferred to use hamstring (semitendinosus) grafts because of postoperative problems such as parapatellar pain and difficulty when kneeling and in some cases subsequent fracture of the patella — problems related to harvesting of the graft. A review by Meins and Vierhout15 analyzed various complications cited in the literature. They included limited mobility and patellofemoral problems with both bone–patellar tendon–bone grafts and semitendinosus grafts, though the incidence of problems was lower with the latter. Furthermore, the main problem with autografts containing living cells is that because they no longer have their own blood supply they necrose and steadily lose strength until they have been repopulated with autologous fibroblasts and remodeled. This process takes around 2 years, and structures rarely attain their original strength.4 There is an awareness among orthopedic surgeons that the ideal (“gold standard”) method of ACL reconstruction has not yet been found. The harvesting of a tendon from the patient’s own knee can be seen as “robbing Peter to pay Paul,” with the consequent risk of further compromising the injured (or contralateral) knee. Alternatives that attempt to overcome problems encountered with autografts include allografts and xenografts. Both of these grafts must be carefully prepared to remove infective agents and immunogenic cells, leaving a collagenous scaffold on which the
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
patient’s cells may grow. In recent years, however, several concerns have led many surgeons to abandon these two graft sources. Allografts carry a small but worrying risk of viral infection by (among others) the human immunodeficiency virus (HIV) and the hepatitis viruses. Xenografts might lead to infective agents (e.g., porcine viruses, although currently not known to infect humans) gaining a foothold in the human population. One current research interest is seeding a threedimensional textile scaffold, made of a bioabsorbable material, with fibroblasts and incubating it in a culture medium in which the cells would proliferate and cover the scaffold with tissue. This is then implanted into the joint, giving the reconstruction a “running start.” No such ligament has yet progressed far enough to be used in a clinical trial. It is likely that major problems remain to be overcome before tissue engineering can be successfully applied to knee ligaments. They include (1) the identification and development of suitable resorbable scaffolds with an initial strength comparable to that of the natural ligament and able to retain adequate strength for 10–18 months until the neoligament is fully functional, and (2) the ex vivo cell culture techniques needed to prepare the implant. Disadvantages of this approach will probably include both cost (not known for ligaments but estimated at between £3000 and £7000 per treatment for cartilage repair in the United Kingdom11) and the greater difficulty of anchoring such an implant than is encountered with a “dryer,” less slippery structure. Furthermore, the postoperative fate of such a structure may be similar to that of autogenous grafts. Consideration of all of these difficulties has led some orthopedists to reconsider the role synthetic ligaments may play in the future.9 In this study we investigated the potential of an implant made of a synthetic material with bioenhanced surface properties produced by exposure of the woven polyester structure to radiofrequency glow discharge (RFGD) treatment. Such enhancement is designed to accelerate the recruitment and proliferation of cells, particularly synovial fibroblasts, once the device is implanted in the joint. The cyclic strains experienced by an implant within the joint have long been reported to encourage remodeling of the induced tissue and differentiation of the cells (originally synovial fibroblasts) into spindle-shaped ligament fibroblasts capable of synthesizing type I collagen.8,14 This process can take up to 18 months. Speeding it up would not only mean faster rehabilitation for the patient, it would limit potential damage to the device caused by abrasion against bony tunnel surfaces. Work by France et al. at the University of Sheffield7 has shown that introduction of a relatively small proportion (optimum 2.3%) of carboxylic acid groups at
199
the surface of certain polymers significantly increases the affinity of the material for cell attachment. Moreover, there is evidence that although RF glow treatment of polyethylene terephthalate (PET) in itself facilitates a stronger attachment of cells, this attachment is even more secure in the presence of proteins such as fibronectin.16 In this study the treatment was applied to a woven scaffold, rather than films or disks of the polymer generally used in cell attachment studies. The basic woven scaffold was identical to that used in the LeedsKeio ligament, one of the few commercial synthetic ligaments to survive in the marketplace. Since 1984 more than 50 000 of these ligments have been implanted in Japan and Europe. Clinical success using the LeedsKeio ligament (as in most ACL reconstruction methods) has been mixed. Whereas a fully functional neoligament has been formed in many cases,14,15 in others tissue did not regenerate satisfactorily on the polyester scaffold, or it regenerated too slowly13 — hence the reason for this study. It is often stated that such treatment has no effect on the bulk mechanical properties of the material, as it only modifies the surface molecules to a depth of a few nanometers. The treated ligaments were subjected to a series of tensile and fatigue tests to establish whether this is the case. The effects of the RFGD treatment on the wettability (hydrophilicity) of the material surface and the attachment and growth of various fibroblasts were also investigated. Experimental Materials The material used was the same as that used in the Leeds-Keio synthetic PET knee ligament, which has been marketed in Europe and Japan since 1984 and has a long history of use with low levels of adverse material reaction. It has a woven, open mesh-like structure and is manufactured in two versions: one with a minimum strength of 2100 N and the other with 3300 N. Prior to treatment the ligament was solvent cleaned to remove any traces of fiber-processing materials (e.g., lubricants). This type of solvent cleaning did not have any effect on the mechanical strength of the material. In tests carried out on 3300 N ligaments the loads to failure averaged 3570 and 3567 N for untreated and solventcleaned samples (n 4). Stiffness values were 333.8 and 348.3 N mm1, respectively. The glow discharge equipment used in this study was a proprietary unit, with the predominant reaction induced at the material’s surface being oxidation. The control for this study was the standard Leeds-Keio ligament without glow treatment. All test material was
200
sterilized by gamma irradiation, receiving an absorbed dose of approximately 30 kGy. Wettability tests Wettability tests were performed at the School of Textile Industries, University of Leeds, England using a modified British Standard Test.2 The test pieces were conditioned in a standard atmosphere for at least 48 h at 20°C and 65% relative humidity (RH) before being placed on a glass plate at constant tension for testing. A drop of solution containing 50% (w/w) sucrose in distilled water was applied from a burette fitted with a high-performance liquid chromatography (HPLC)grade stainless steel tube with a tip of internal diameter 0.76 mm positioned 6 mm above the material. The sucrose solution was chosen because distilled water would be absorbed by the treated fabric too quickly to measure. (In the standard, 50% sucrose solution is stated to take approximately 12 times longer to be absorbed in this test than distilled water.) The time taken for the drop to be completely absorbed into the fabric was measured using a stopwatch. Measurements were made at intervals of approximately 50 mm along the length of the ligament (starting 25 mm from the end). Between each measurement the ligament was turned over so alternate measurements were made from either side of the device. The test samples were approximately 300 mm long, and results were recorded for six points on each sample. Samples were tested 5 weeks after the date of treatment (RFGD-treated material) and 3 weeks after irradiation (both materials). Mechanical testing Ligaments were tested on an Instron 8031 servohydraulic materials testing machine (High Wycombe, England) with a calibrated load cell. The test pieces were gripped in steel clamps lined with 1- to 2-mm thick rubber, and the clamp bolts were tightened in a specified sequence to a torque of 9 Nm. For tests to break, the load versus elongation curve was recorded on an X-Y plotter. The tests performed and the conditions of the test were as follows. Tensile test. Ligaments with nominal strengths of 2100 and 3300 N were tested to failure in air at room temperature. The stiffness was calculated by dividing the maximum load by the extension to failure. The gauge length was 40 mm, and the strain rate was 50% per second. Fatigue test — medium term. Ligaments with a nominal strength of 2100 N were subjected to cyclic sinusoidal loading between 50 and 500 N for 10.8 106 cycles (5 consecutive days) at 25 Hz in air at room tem-
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
perature. The gauge length for the cyclic loading was 150 mm. After the cyclic loading was completed the ligaments were subjected to a tensile loading test to failure to determine their residual strength as in the tensile test. Exhaustive fatigue test. This test is extremely timeconsuming; consequently, only two samples of the treated ligament were tested by this method. The nominal 3300 N variant of the ligament was used, in contrast to the 2100 N ligament used in the other fatigue tests. The ligaments were fatigued under conditions closer to those that pertain in vivo in physiological saline solution at 37°C for 62 106 cycles (uniaxial, 50–500 N, and at 25 Hz at night and on weekends and at 20 Hz during the day) using a gauge length of 40 mm. At the end of the test the samples were tested for residual strength as in the tensile test. The test took approximately 30 days to complete. Cell attachment and proliferation Cell attachment and proliferation tests were conducted at three collaborating centers. Each center received the samples identified only by a treatment code and was not informed of the type of treatment represented by the code until after the tests were completed. Skin fibroblast attachment to single fibers The preliminary tests of skin fibroblast attachment to single fibers were conducted at the School of Biochemistry and Molecular Biology, University of Leeds to assess in general terms the differences in cell attachment between treated and untreated single polyester fibers (approximately 22 µm diameter) removed from the woven meshes. Dermal fibroblasts were isolated from child foreskin by the explant method and maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 5% newborn calf serum, penicillin 100 units ml1, streptomycin 100 µg ml1, and 2 mM glutamine at 37°C in a humidified atmosphere of 5% CO2 in air. Two fibroblast strains from different individuals aged 1 and 3 years old were used in this work. The cultures were initially established on “tissue culture plastic” where they settled and proliferated and then were passaged after trypsinization. Fibroblasts were then transferred to bacteriological plastic dishes (100 000 cells/dish) containing culture medium and short fibers from the test material (e.g., about 5 mm); they were then observed as they attached to and spread on the fibers. For the short-term experiments the number of attached cells and fully spread cells were counted at 2, 4, and 6 h after seeding. After 24 h in the case of the treated sample, the fiber surface was almost completely
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
covered by adherent cells. Discrete lengths of fiber (10 pieces selected at random for either treatment) were measured, and the number of cells adhering were counted. The result was expressed as the number of cells per 5 mm length of fiber (up to 6 h) and as a percentage representing the extent of “coverage” of the fiber by cells (24–72 h). Initially (e.g., after 1–2 h), the cells were more or less spherical and had minimal contact with the fiber. However, after 4 h the cells had started to spread and wrap themselves around the fibers. The cells that had attached only at a single point of contact (and were spherical) were excluded from the counting; only cells that had clearly adhered and were starting to spread were counted. There was a slight difference in the type of cell attachment between the ends and the middle of the fibers, presumably resulting from a change in the physical surface properties of the fiber material due to the cutting. Cells at the ends of the fibers were excluded from the counting. For the extended observation study, the length of fiber covered with cells was measured after 24, 48, and 72 h; and the relative coverage of the fiber (as an index of proliferation) was expressed as a percentage of the total length. In a separate pilot study with dermal fibroblasts, observation was extended to 20 days; the cells were then stained with toluidine blue and photographs obtained under microscopy to visualize coverage of the mesh. Incorporation of synovial fibroblasts into meshes After the preliminary studies using dermal fibroblasts, tests were conducted at two centers using different methodologies to assess the attachment and proliferation of synovial fibroblasts to the woven meshes. The study at the Yorkshire Regional Tissue Bank, Wakefield, England used human synovial fibroblasts derived from synovial tissue obtained at remedial synovectomy or arthroplasty procedures (supplied by Dr. D. Woolley of the University of Manchester). The cells were obtained as either frozen cultures (requiring resuscitation) or growing cultures at passage 8 or 9. The cultures were maintained in DMEM supplemented with 10% fetal calf serum, 2 mM l-glutamine, penicillin (100 units ml1), and streptomycin (100 µg ml1), all obtained from Sigma Aldrich (St. Louis, MO, USA). The cultures were incubated at 37°C in 5% CO2 in air atmosphere. Strips of sterile woven mesh of approximately 100 mm length and 10 mm width from each treatment group were supplied for testing as two blinded groups. The meshes were aseptically cut into sections of approximately 1 cm2 area and placed in individual wells of non-tissue-culture-treated (NTC) polystyrene 24well dishes (Sarstedt). Subconfluent cultures of synovial fibroblasts were passaged using trypsin/EDTA [0.05% trypsin (Difco 1 : 250)/0.02% EDTA (Sigma) in Hank’s
201
balanced salt solution (HBSS)] to create a cell suspension. The suspension was counted and the concentration adjusted to 2.5 103 cells ml1. A 1-ml aliquot of the suspension was added to each well containing a mesh sample. The plates were then incubated for 24 h with 100% RH and gentle agitation, after which the meshes were rinsed briefly in HBSS and transferred to fresh NTC plates. At this point, 10 meshes were removed from each treatment group and incubated with the trypsin/EDTA solution to remove adherent cells, as above. The resulting cell suspension was mixed with an equal volume of standard culture medium to neutralize the trypsin; it was concentrated by centrifugation (800 g for 10 min) and counted with a hemocytometer (Improved Neubauer). Preliminary experiments (data not shown) had indicated this to be the most accurate method of counting low numbers of cells. The total number of cells removed from each mesh was recorded, and the meshes were dried for 3 days in a desiccator. They were then weighed and the number of cells removed calculated as cells per milligram dry weight of mesh. The remaining mesh samples were incubated in culture medium for 3, 6, 11, and 14 days, with the culture medium being changed every few days. At each time point, another 10 meshes from each group were selected, and the cells were removed and counted as above. The raw data were adjusted for the size (weight) of the mesh in each case and presented as cells counted per milligram of dry mesh. One-way and two-way ANOVA were carried out. Coefficients of variation (CVs) were calculated for each data set as follows: CV (s 100)/Y, where s is the sample standard deviation, and Y is the sample mean. In the tests conducted at the National Defense Medical College, Saitama, Japan, human synovial stromal cells (HSSCs) (Applied Cell Biology Research Institute) were cultured in CS-C medium (Standard Tools for Cell Biology) containing 10% serum and growth factor. They were seeded at a density of 1 105 cells ml1 per well on each mesh sample (approximately 10 10 mm) in a 24-well plate. The plate was changed 1 day after inoculation (day 1), moving the pieces of mesh to new wells to eliminate cells attached to the plate. The medium was changed on days 2 and 4. [3H]-Thymidine incorporation and DNA assays were carried out 1, 3, and 5 days after inoculation. For each sample [3H]-thymidine (2 µCi/well) was added, and the cells were incubated at 37°C for 3 h. After incubation, the medium was removed and the cells were washed for 15 min with phosphate-buffered saline (PBS). This washing was carried out three times for each sample. Then the cells on the mesh were solubilized in 1 M NaOH. After neutralization with 1 M HCl, the aliquots were analyzed for [3H] radioactivity
202
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
by scintillation counting. A preliminary study had demonstrated that negligible amounts of thymidine were retained by either untreated or treated mesh in the absence of cells. After digesting the cells attached to the samples for DNA assay using papain at 60°C overnight, the aliquots of the digest were mixed with Hoechst 33258 dye solution. Fluorescence was measured using a Titerek Fluoroskan II fluorometer with excitation and emission wavelengths set at 355 and 450 nm. Calf thymus DNA was used as a standard. Finally, scanning electron microscopy (SEM) was carried out on day 5 to visualize the cells on the materials.
Results Wettability Figure 1 illustrates the marked difference in wettability after RF glow treatment. The droplet has clearly spread out and penetrated the treated material but has
remained more or less on the surface of the untreated material. The mean time taken for the droplet to disperse on the control material was well over 200 s (n 24), whereas in the case of the treated material it took only 5.5 s (SD 9.6 s; n 18). Mechanical properties In tensile tests of unfatigued ligaments (Table 1), for both 2100 and 3300 N ligaments, the load to failure of the treated material was found to be lower than that for the untreated material (12% in the case of the 2100 N ligament). This was accompanied by an approximately 11%–15% increase in stiffness. After having undergone a medium-term fatigue test (10.8 106 cycles) (Table 2), both treated and untreated 2100 N ligaments (n 4 each) had virtually the same load and extension to failure (2300 N and 11.2 mm, respectively). The exhaustive fatigue test (62 106 cycles) was conducted on only two samples of treated ligaments owing to the length of the test program. A comparison can be made, however, with control samples from the identical batch but obtained before treatment and not subjected to any cyclic loading before the mechanical test (Table 3). The untreated, unfatigued ligaments had a UTS of 3689.0 N and stiffness of 338.4 N mm1, 91% and 115%, respectively, of the values obtained with the controls. Although the numbers are small, this gives some assurance of the durability of the RFGD-treated product under these conditions. Cell recruitment and proliferation
Fig. 1. Droplet test for wettability. Note the difference between the radiofrequency (RF) glow treated (left) and untreated (right) material. For the purposes of photography, methylene blue dye was added to the sucrose solution to make the droplet easier to see
Short-term observations (6 h) In tests with small sections of single fibers, there was little attachment of dermal fibroblasts to the control material, whereas the treated material exhibits markedly (⬃35 times) more attachment after 6 h, with most of the attached cells being fully spread (Fig. 2). Longer-term observations (24–72 h and up to 14 days) In tests with single-fiber pieces, both materials supported steady spreading of dermal fibroblasts over
Table 1. Tensile test results from control and RF glow-treated ligaments: strength and stiffness (n 4) 2100 N Ligament Parameter (UTS) SD (N) Extension to failure SD (mm) Stiffness SD (N/mm)
3300 N Ligament
Control
Treated
Percent of control
Control
Treated
Percent of control
2288.0 70.0 12.7 0.45
2013.0 45.7 9.7 0.46
88.0 76.4
3282 126 12.3 1.17
3177 298 10.6 0.60
96.8 86.2
179.9 5.99
207.2 5.26
115.2
267.6 21.1
299.5 19.4
111.9
These tests were conducted on nonfatigued ligaments with a nominal strength of 2100 N RF, radio frequency; UTS, ultimate tensile strength
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
203
Table 2. Medium-term (10 800 000 cycles) fatigue test results: residual strength and stiffness (n 4) Control
RF glow treated
Percent of control
2322.0 34.8 11.20 0.34
2319.00 15.17 11.20 0.38
99.9 100
207.4 4.9
206.80 7.76
99.7
Parameter UTS SD (N) Extension to failure SD (mm) Stiffness SD (N/mm)
These tests were conducted on ligaments with a nominal strength of 2100 N
Table 3. Effect of exhaustive fatigue testing (62 126 000 cycles) of Leeds-Keio ligament-II treated with glow: residual strength and stiffness (n 2) Condition Not fatigued Fatigued Change
UTS (N)
Overall stiffness (N/mm)
3689.0 3347.5 9.3%
338.40 389.75 15.2%
Fig. 3. Coverage of RF glow treated and untreated (control) fibers with epithelial fibroblasts 24, 48, and 72 h after inoculation (mean SD)
Fig. 2. Total number of spreading and fully spread epithelial fibroblasts on 5-mm lengths of RF glow treated and untreated (control) fibers 2, 4, and 6 h after inoculation
the 72-h period; about twice the coverage was obtained with the treated material compared with the control (Fig. 3). Figure 4 presents a view of individually treated fibers obtained 72 h after inoculation, illustrating the abundant coverage by fibroblasts and bridging of cells between the fibers. The results from the 14-day study on meshes are presented in Fig. 5. Although the initial number of synovial fibroblasts recovered from the pieces of mesh at day 0 was greater in the control samples than in the treated samples, by day 6 the situation had reversed. At the end of the experiment (day 14) there were approximately four times the number of cells on the
Fig. 4. Photograph taken with an optical microscope showing epithelial fibroblasts growing on and bridging RF glow treated fibers. Hematoxylin stain was applied 72 h after inoculation. 200
treated samples as on the control. The increase in cell number, as a percentage of the initial count, was 83.5% in the case of the control, and 1461% in the case of the treated material. Initial two-way ANOVA demonstrated that there were significant variations over time and between each group, although the error margin was
204
too large to distinguish between the individual data sets. Using one-way ANOVA, the differences between treatments could be shown to be significant at the 95% confidence level at day 0 and from day 6 onward. The incorporation of [3H]-thymidine by synovial stromal cells growing on the RF glow-treated samples was significantly greater than by those growing on the control, even at day 1. The difference increased with time, from approximately threefold at day 1 to approximately fivefold at day 5 (Fig. 6). This incorporation is a measure of cell division, and the results indicate a faster rate of cell multiplication on the treated material. The total DNA content of the cells growing on the treated samples (1.484 µg ml1) was significantly greater than that of those growing on the control samples (0.156 µg ml1) (P 0.001).
Fig. 5. Average number of synovial fibroblasts recovered from RF glow treated and control (untreated) mesh at each time point per milligram dry weight of mesh (n 10). Significantly fewer cells (P 0.05) were recovered from the treated material at day 0. Significantly more cells (P 0.05) were recovered from the treated material at days 6, 11, and 14
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
Figure 7 shows scanning electron micrographs of cells on samples of control and treated mesh obtained on day 5 after inoculation. In the case of the control material, little cell spreading can be seen. By contrast, cells have almost completely covered the treated fibers, and the cells are generally aligned along the length of the fibers. These images are consistent with the conclusions from the DNA and [3H]-thymidine measurements and clearly illustrate more rapid colonization of the treated mesh by cells.
Discussion Wettability Polyethylene terephthalate (PET), a material with a long history of safe implant use, usually elicits little adverse tissue reaction. Being hydrophobic, however, cells do not attach and spread on it as readily as if it were a hydrophilic material. Radiofrequency-generated glow discharge treatment is an established method for
Fig. 6. [3H]-Thymidine incorporation by synovial stromal cells growing on RF glow treated and control samples
Fig. 7. Scanning electron microscopy (SEM) showing portions of control (left) and RF glow treated (right) mesh on day 5. Note the greater coverage of the treated materials by synovial stromal cells. 500
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
modifying a material’s surface to increase its chemical reactivity and improve its wettability. This treatment does not involve coating the surface of the PET filaments with any substance, nor does it require high temperatures. It exposes the material to an energized atmosphere generated by RF discharge excitation in the presence of a gas at low pressure without significantly raising the temperature of the material. Excited ions and atoms react with the material surface to form new, ionizable surface radicals. In biomedical applications RF glow treatment has been suggested as a method for improving the attachment of cells to implant surfaces.3,10 Depending on the nature of the gas medium, various chemical changes can be produced in the polymer surface. Dekker et al.6 reported that the thickness of the modified layer in the case of polytetrafluoroethylene (PTFE) was only about 1 nm, and that not only was hydrophilicity increased but proteins including fibronectin were absorbed more readily from a serumcontaining medium onto the modified material. This treatment can also be used for more specific surface modification by coupling particular biochemical entities to the surface molecules of the material, such as glycine or fibronectin1 and heparin.12 The increased wettability brought about by the RFGD treatment was not unexpected, as Canonico et al.3 reported that the critical surface tension for wetting RF glow-treated PET fabrics was significantly greater than for conventional PET fabrics. They showed by electron spectroscopy for chemical analysis (ESCA) that subjecting PET to an oxidative glow discharge results mostly in the introduction of carboxyl and carbonyl functionalities onto the polymer surface, and this increases its wettability. Mechanical properties It has often been reported or claimed that RF glow treatment affects only the outermost layer of materials. Glow discharges, however, emit ultraviolet light, which can cause crosslinking and possibly some chain realignment. The occurrence of either of these phenomena could account for the changes in the initial mechanical properties observed in this study. The apparent “recovery” of strength by the material after cyclic loading is not easy to explain. The strength and stiffness of the ACL in the young have been estimated by Woo et al. at around 2500 N and 292 N mm1, respectively.17 Although the RF glow treatment reduces the tensile strength of the artificial ligament, in the case of the 3000 N ligament used in the exhaustive fatigue test it is still greater than this figure even after 62 106 loading cycles. The mean stiffness of the latter (389.75 N mm1) is 33% greater than that of the natural ACL ligament. However, this is still
205
comparable with the stiffness of quadrupled semitendinosus grafts and less than that of central-third patellar tendon grafts. Cell attachment and proliferation The three studies of cell attachment and proliferation on RFGD-treated woven polyester meshes all demonstrated that the test cells (dermal, synovial, and stromal cells) attach, proliferate, or both more quickly over the first few days of culture than did the untreated controls. Consistently, there are more cells on the treated material than on the controls at the end of each study. In the study using synovial fibroblasts, it is not clear why significantly fewer cells adhered initially to the treated mesh than to the control, but this effect was also found in a preliminary study with these cells (using this method but not otherwise reported here). It could be that the cells are not actually attached to the mesh, but that the hydrophobic nature of the untreated control reduced the efficiency of the rinsing procedure to remove unattached cells in this study.
Conclusions The RF glow treatment of the PET ligament reported here greatly enhances its wettability. This results in marked improvements in terms of increased cell attachment and proliferation on the surface of the implant, a consistent finding of studies undertaken in three research establishments using different in vitro approaches and methodologies. Adverse effects on the strength of the woven ligament have been found to be relatively small and can be readily overcome by small design changes in the ligament’s load-bearing structure. The changes induced by the treatment are not adversely affected by subsequent, terminal sterilization of the implant by gamma radiation (25–50 kGy) (confirmed by other studies not reported here); hence we believe that RF-generated glow discharge treatment has significantly enhanced the potential of polyester for use in ligament implants. The RF glow discharge process investigated in the studies reported in this paper is now used to provide increased functionality on the new range of synthetic ligaments known as the Leeds-Keio Ligament-II (also known as the ESP3000 ligament in Europe). Acknowledgments. The authors thank Xiros plc for providing ligament samples. We also thank the following individuals for their contribution to the work described in this study: Mr. Les Johnson, School of Textile Industries, University of Leeds, for carrying out wettability tests; Dr. David Woolley, Department of Medicine, University of Manchester for supplying cells for the study at Yorkshire Regional Tissue Bank; Mr. John
206 Tresnan, Xiros plc, for carrying out the mechanical tests. We especially acknowledge the contribution of Sangjin Kang to the work carried out at the School of Biochemistry and Molecular Biology, University of Leeds.
References 1. Anderheiden D, Klee D, Höcher H, et al. Surface modification of a biocompatible polymer based on polyurethane for artificial blood vessels. J Mater Sci Mater Med 1992;3:1–4. 2. BS 4554 method of test for wettability of textile fabrics. London: British Standards Institution, 1970. 3. Canonico P, Cassinelli C, Morra M. Surface-modified polyester fabrics. Med Device Technol 1997;8:16–21(April). 4. Clancy WG, Narechania RG, Rosenberg TD, et al. Anterior and posterior cruciate ligament reconstruction in rhesus monkeys. J Bone Joint Surg Am 1981;63:270–84. 5. Cooper DE, Deng XH, Burstein AL, et al. The strength of the central third patellar tendon graft. Am J Sports Med 1993;21:818– 24. 6. Dekker A, Reitsma K, Beugeling T, et al. Adhesion of endothelial-cells and adsorption of serum-proteins on gas plasmatreated polytetrafluoroethylene. Biomaterials 1991;12:130–8. 7. France RM, Short R, Dawson RA, et al. Attachment of human keratinocytes to plasma co-polymers of acrylic acid/octa-1,7-diene and allyl amine/octa-1,7-diene. J Mater Chem 1998;8:1–6. 8. Fujikawa K, Seedhom BB, Matsumoto H, et al. The Leeds-Keio ligament. In: Strover A, editor. Intra-articular reconstruction of the anterior cruciate ligament. London: Butterworth Heinemann; 1993. p. 173–207.
J.R.J. Rowland et al.: RFGD-treated polyester ligaments 9. Groner C. Synthetic ACL repair may get another chance. Biomechanics 1998;32:79–84(May). 10. Jansen JA, van der Waerden JPCM, de Groot K. Fibroblast and epithelial cell interactions with surface-treated implant materials. Biomaterials 1991;12:25–31. 11. Jobanputra P, Parry D, Fry-Smith A, et al. Effectiveness of autologous chondrocyte transplantation for hyaline cartilage defects in knees. Report commissioned by NHS HTA Programme. Birmingham, UK: West Midlands Development and Evaluation Service, The University of Birmingham. July 2000. 12. Kang IK, Kwon OH, Lee YM, et al. Preparation and surface characterization of functional group-grafted and heparin-immobilized polyurethanes by plasma glow discharge. Biomaterials 1996;17:841–7. 13. Macnicol MF, Penny ID. Leeds-Keio ligament anterior cruciate replacement: a clinical, arthroscopic and histological review. J Bone Joint Surg Br 1990;72(suppl. 1):1. 14. Marcacci M, Gubellini P, Buda R, et al. Histological and ultrastructural findings of tissue ingrowth: the Leeds-Keio prosthetic anterior cruciate ligament. Clin Orthop 1991;267:115– 21. 15. Meins G, Vierhout PAM. Arthroscopic reconstruction after rupture of anterior cruciate ligament with the Leeds-Keio ligament. Enschede, The Netherlands: Medische Spectrum Twente, Department of Surgery, 1995. 16. Pratt KJ, Williams SK, Jarrell BE. Enhanced adherence of human adult endothelial cells to plasma discharge modified polyethylene terephthalate. J Biomed Mater Res 1989;23:1131–47. 17. Woo SL-Y, Young EP, Kwan MK. Fundamental studies in knee ligament mechanics. In: Daniel D, Akeson W, O’Connor J, editors. Knee ligaments: structure, function, injury and repair. New York: Raven; 1990. p. 115–34.
Radiofrequency-generated glow discharge treatment: potential benefits for polyester ligaments John Richard James Rowland1, Satoshi Tsukazaki2, Toshiyuki Kikuchi2, Kyosuke Fujikawa2, John Kearney3, Richard Lomas3, Edward Wood4, and Bahaa Botros Seedhom5 1
Xiros plc, 28-30 Blenheim Terrace, Leeds LS2 9HD, UK Department of Orthopaedic Surgery, National Defense Medical College, Saitama, Japan 3 Tissue Services R&D, Trent Blood Centre, Sheffield, UK 4 School of Biochemistry and Molecular Biology, University of Leeds, UK 5 Academic Unit of Musculo-Skeletal and Rehabilitation Medicine, University of Leeds, UK 2
Abstract This multicenter study has revealed that treating a woven polyethylene terephthalate (polyester) ligament with a radiofrequency (RF)-generated glow discharge (RFGD) produces marked benefits in terms of increased cell attachment and proliferation on the implant surface. In vitro tests of the same material revealed that the number of synovial fibroblasts attached to the treated samples after 14 days was four times that of the untreated material. Many of the cells were spread over the surface of a single filament, and some formed bridges between one filament and the next. The incorporation of [3H]-thymidine by synovial stromal cells (a measure of the amount of cell division) growing on the treated material was five times that on the untreated samples. The amount of DNA present on the treated material was also found to be almost an order of magnitude greater than that on untreated samples. This increase in cell attachment and proliferation is almost certainly related to a notable increase in wettability of the polyester surface induced by treatment. Mechanical tests revealed that, for ligaments with a nominal ultimate tensile strength of 2100 N, RF-generated glow treatment reduced the ligament’s strength by 12% but increased its stiffness by 15%. After a medium-term fatigue test (10.8 million cycles), however, there appeared to be recovery of the mechanical properties, with the strength and stiffness of untreated and treated samples being essentially the same. After exhaustive fatigue tests (more than 62 million cycles) the residual strength of the treated ligaments was only 9% lower than that of the unfatigued and untreated ligaments. Key words Radiofrequency-generated glow discharge · Leeds-Keio ligament · Cell attachment · Strength · Wettability
Introduction Because knees with a ruptured anterior cruciate ligament (ACL) frequently degenerate, there is a
Offprint requests to: J.R.J. Rowland Received: July 1, 2002 / Accepted: October 22, 2002
tendency to reconstruct this ligament. During the 1980s and early 1990s synthetic ligaments were commonly used for this procedure. Because of various problems encountered, however, most of the current reconstructive procedures use autogenous tissue. Bone–patellar tendon–bone grafts using the central third of the patellar tendon have been popular for some years, as these grafts have been reported to have an initial strength as high as 2977 N (10-mm strip) or 4389 N (15mm strip) and stiffness of around 1150 N mm1.5,15 More recently, many surgeons have preferred to use hamstring (semitendinosus) grafts because of postoperative problems such as parapatellar pain and difficulty when kneeling and in some cases subsequent fracture of the patella — problems related to harvesting of the graft. A review by Meins and Vierhout15 analyzed various complications cited in the literature. They included limited mobility and patellofemoral problems with both bone–patellar tendon–bone grafts and semitendinosus grafts, though the incidence of problems was lower with the latter. Furthermore, the main problem with autografts containing living cells is that because they no longer have their own blood supply they necrose and steadily lose strength until they have been repopulated with autologous fibroblasts and remodeled. This process takes around 2 years, and structures rarely attain their original strength.4 There is an awareness among orthopedic surgeons that the ideal (“gold standard”) method of ACL reconstruction has not yet been found. The harvesting of a tendon from the patient’s own knee can be seen as “robbing Peter to pay Paul,” with the consequent risk of further compromising the injured (or contralateral) knee. Alternatives that attempt to overcome problems encountered with autografts include allografts and xenografts. Both of these grafts must be carefully prepared to remove infective agents and immunogenic cells, leaving a collagenous scaffold on which the
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
patient’s cells may grow. In recent years, however, several concerns have led many surgeons to abandon these two graft sources. Allografts carry a small but worrying risk of viral infection by (among others) the human immunodeficiency virus (HIV) and the hepatitis viruses. Xenografts might lead to infective agents (e.g., porcine viruses, although currently not known to infect humans) gaining a foothold in the human population. One current research interest is seeding a threedimensional textile scaffold, made of a bioabsorbable material, with fibroblasts and incubating it in a culture medium in which the cells would proliferate and cover the scaffold with tissue. This is then implanted into the joint, giving the reconstruction a “running start.” No such ligament has yet progressed far enough to be used in a clinical trial. It is likely that major problems remain to be overcome before tissue engineering can be successfully applied to knee ligaments. They include (1) the identification and development of suitable resorbable scaffolds with an initial strength comparable to that of the natural ligament and able to retain adequate strength for 10–18 months until the neoligament is fully functional, and (2) the ex vivo cell culture techniques needed to prepare the implant. Disadvantages of this approach will probably include both cost (not known for ligaments but estimated at between £3000 and £7000 per treatment for cartilage repair in the United Kingdom11) and the greater difficulty of anchoring such an implant than is encountered with a “dryer,” less slippery structure. Furthermore, the postoperative fate of such a structure may be similar to that of autogenous grafts. Consideration of all of these difficulties has led some orthopedists to reconsider the role synthetic ligaments may play in the future.9 In this study we investigated the potential of an implant made of a synthetic material with bioenhanced surface properties produced by exposure of the woven polyester structure to radiofrequency glow discharge (RFGD) treatment. Such enhancement is designed to accelerate the recruitment and proliferation of cells, particularly synovial fibroblasts, once the device is implanted in the joint. The cyclic strains experienced by an implant within the joint have long been reported to encourage remodeling of the induced tissue and differentiation of the cells (originally synovial fibroblasts) into spindle-shaped ligament fibroblasts capable of synthesizing type I collagen.8,14 This process can take up to 18 months. Speeding it up would not only mean faster rehabilitation for the patient, it would limit potential damage to the device caused by abrasion against bony tunnel surfaces. Work by France et al. at the University of Sheffield7 has shown that introduction of a relatively small proportion (optimum 2.3%) of carboxylic acid groups at
199
the surface of certain polymers significantly increases the affinity of the material for cell attachment. Moreover, there is evidence that although RF glow treatment of polyethylene terephthalate (PET) in itself facilitates a stronger attachment of cells, this attachment is even more secure in the presence of proteins such as fibronectin.16 In this study the treatment was applied to a woven scaffold, rather than films or disks of the polymer generally used in cell attachment studies. The basic woven scaffold was identical to that used in the LeedsKeio ligament, one of the few commercial synthetic ligaments to survive in the marketplace. Since 1984 more than 50 000 of these ligments have been implanted in Japan and Europe. Clinical success using the LeedsKeio ligament (as in most ACL reconstruction methods) has been mixed. Whereas a fully functional neoligament has been formed in many cases,14,15 in others tissue did not regenerate satisfactorily on the polyester scaffold, or it regenerated too slowly13 — hence the reason for this study. It is often stated that such treatment has no effect on the bulk mechanical properties of the material, as it only modifies the surface molecules to a depth of a few nanometers. The treated ligaments were subjected to a series of tensile and fatigue tests to establish whether this is the case. The effects of the RFGD treatment on the wettability (hydrophilicity) of the material surface and the attachment and growth of various fibroblasts were also investigated. Experimental Materials The material used was the same as that used in the Leeds-Keio synthetic PET knee ligament, which has been marketed in Europe and Japan since 1984 and has a long history of use with low levels of adverse material reaction. It has a woven, open mesh-like structure and is manufactured in two versions: one with a minimum strength of 2100 N and the other with 3300 N. Prior to treatment the ligament was solvent cleaned to remove any traces of fiber-processing materials (e.g., lubricants). This type of solvent cleaning did not have any effect on the mechanical strength of the material. In tests carried out on 3300 N ligaments the loads to failure averaged 3570 and 3567 N for untreated and solventcleaned samples (n 4). Stiffness values were 333.8 and 348.3 N mm1, respectively. The glow discharge equipment used in this study was a proprietary unit, with the predominant reaction induced at the material’s surface being oxidation. The control for this study was the standard Leeds-Keio ligament without glow treatment. All test material was
200
sterilized by gamma irradiation, receiving an absorbed dose of approximately 30 kGy. Wettability tests Wettability tests were performed at the School of Textile Industries, University of Leeds, England using a modified British Standard Test.2 The test pieces were conditioned in a standard atmosphere for at least 48 h at 20°C and 65% relative humidity (RH) before being placed on a glass plate at constant tension for testing. A drop of solution containing 50% (w/w) sucrose in distilled water was applied from a burette fitted with a high-performance liquid chromatography (HPLC)grade stainless steel tube with a tip of internal diameter 0.76 mm positioned 6 mm above the material. The sucrose solution was chosen because distilled water would be absorbed by the treated fabric too quickly to measure. (In the standard, 50% sucrose solution is stated to take approximately 12 times longer to be absorbed in this test than distilled water.) The time taken for the drop to be completely absorbed into the fabric was measured using a stopwatch. Measurements were made at intervals of approximately 50 mm along the length of the ligament (starting 25 mm from the end). Between each measurement the ligament was turned over so alternate measurements were made from either side of the device. The test samples were approximately 300 mm long, and results were recorded for six points on each sample. Samples were tested 5 weeks after the date of treatment (RFGD-treated material) and 3 weeks after irradiation (both materials). Mechanical testing Ligaments were tested on an Instron 8031 servohydraulic materials testing machine (High Wycombe, England) with a calibrated load cell. The test pieces were gripped in steel clamps lined with 1- to 2-mm thick rubber, and the clamp bolts were tightened in a specified sequence to a torque of 9 Nm. For tests to break, the load versus elongation curve was recorded on an X-Y plotter. The tests performed and the conditions of the test were as follows. Tensile test. Ligaments with nominal strengths of 2100 and 3300 N were tested to failure in air at room temperature. The stiffness was calculated by dividing the maximum load by the extension to failure. The gauge length was 40 mm, and the strain rate was 50% per second. Fatigue test — medium term. Ligaments with a nominal strength of 2100 N were subjected to cyclic sinusoidal loading between 50 and 500 N for 10.8 106 cycles (5 consecutive days) at 25 Hz in air at room tem-
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
perature. The gauge length for the cyclic loading was 150 mm. After the cyclic loading was completed the ligaments were subjected to a tensile loading test to failure to determine their residual strength as in the tensile test. Exhaustive fatigue test. This test is extremely timeconsuming; consequently, only two samples of the treated ligament were tested by this method. The nominal 3300 N variant of the ligament was used, in contrast to the 2100 N ligament used in the other fatigue tests. The ligaments were fatigued under conditions closer to those that pertain in vivo in physiological saline solution at 37°C for 62 106 cycles (uniaxial, 50–500 N, and at 25 Hz at night and on weekends and at 20 Hz during the day) using a gauge length of 40 mm. At the end of the test the samples were tested for residual strength as in the tensile test. The test took approximately 30 days to complete. Cell attachment and proliferation Cell attachment and proliferation tests were conducted at three collaborating centers. Each center received the samples identified only by a treatment code and was not informed of the type of treatment represented by the code until after the tests were completed. Skin fibroblast attachment to single fibers The preliminary tests of skin fibroblast attachment to single fibers were conducted at the School of Biochemistry and Molecular Biology, University of Leeds to assess in general terms the differences in cell attachment between treated and untreated single polyester fibers (approximately 22 µm diameter) removed from the woven meshes. Dermal fibroblasts were isolated from child foreskin by the explant method and maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 5% newborn calf serum, penicillin 100 units ml1, streptomycin 100 µg ml1, and 2 mM glutamine at 37°C in a humidified atmosphere of 5% CO2 in air. Two fibroblast strains from different individuals aged 1 and 3 years old were used in this work. The cultures were initially established on “tissue culture plastic” where they settled and proliferated and then were passaged after trypsinization. Fibroblasts were then transferred to bacteriological plastic dishes (100 000 cells/dish) containing culture medium and short fibers from the test material (e.g., about 5 mm); they were then observed as they attached to and spread on the fibers. For the short-term experiments the number of attached cells and fully spread cells were counted at 2, 4, and 6 h after seeding. After 24 h in the case of the treated sample, the fiber surface was almost completely
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
covered by adherent cells. Discrete lengths of fiber (10 pieces selected at random for either treatment) were measured, and the number of cells adhering were counted. The result was expressed as the number of cells per 5 mm length of fiber (up to 6 h) and as a percentage representing the extent of “coverage” of the fiber by cells (24–72 h). Initially (e.g., after 1–2 h), the cells were more or less spherical and had minimal contact with the fiber. However, after 4 h the cells had started to spread and wrap themselves around the fibers. The cells that had attached only at a single point of contact (and were spherical) were excluded from the counting; only cells that had clearly adhered and were starting to spread were counted. There was a slight difference in the type of cell attachment between the ends and the middle of the fibers, presumably resulting from a change in the physical surface properties of the fiber material due to the cutting. Cells at the ends of the fibers were excluded from the counting. For the extended observation study, the length of fiber covered with cells was measured after 24, 48, and 72 h; and the relative coverage of the fiber (as an index of proliferation) was expressed as a percentage of the total length. In a separate pilot study with dermal fibroblasts, observation was extended to 20 days; the cells were then stained with toluidine blue and photographs obtained under microscopy to visualize coverage of the mesh. Incorporation of synovial fibroblasts into meshes After the preliminary studies using dermal fibroblasts, tests were conducted at two centers using different methodologies to assess the attachment and proliferation of synovial fibroblasts to the woven meshes. The study at the Yorkshire Regional Tissue Bank, Wakefield, England used human synovial fibroblasts derived from synovial tissue obtained at remedial synovectomy or arthroplasty procedures (supplied by Dr. D. Woolley of the University of Manchester). The cells were obtained as either frozen cultures (requiring resuscitation) or growing cultures at passage 8 or 9. The cultures were maintained in DMEM supplemented with 10% fetal calf serum, 2 mM l-glutamine, penicillin (100 units ml1), and streptomycin (100 µg ml1), all obtained from Sigma Aldrich (St. Louis, MO, USA). The cultures were incubated at 37°C in 5% CO2 in air atmosphere. Strips of sterile woven mesh of approximately 100 mm length and 10 mm width from each treatment group were supplied for testing as two blinded groups. The meshes were aseptically cut into sections of approximately 1 cm2 area and placed in individual wells of non-tissue-culture-treated (NTC) polystyrene 24well dishes (Sarstedt). Subconfluent cultures of synovial fibroblasts were passaged using trypsin/EDTA [0.05% trypsin (Difco 1 : 250)/0.02% EDTA (Sigma) in Hank’s
201
balanced salt solution (HBSS)] to create a cell suspension. The suspension was counted and the concentration adjusted to 2.5 103 cells ml1. A 1-ml aliquot of the suspension was added to each well containing a mesh sample. The plates were then incubated for 24 h with 100% RH and gentle agitation, after which the meshes were rinsed briefly in HBSS and transferred to fresh NTC plates. At this point, 10 meshes were removed from each treatment group and incubated with the trypsin/EDTA solution to remove adherent cells, as above. The resulting cell suspension was mixed with an equal volume of standard culture medium to neutralize the trypsin; it was concentrated by centrifugation (800 g for 10 min) and counted with a hemocytometer (Improved Neubauer). Preliminary experiments (data not shown) had indicated this to be the most accurate method of counting low numbers of cells. The total number of cells removed from each mesh was recorded, and the meshes were dried for 3 days in a desiccator. They were then weighed and the number of cells removed calculated as cells per milligram dry weight of mesh. The remaining mesh samples were incubated in culture medium for 3, 6, 11, and 14 days, with the culture medium being changed every few days. At each time point, another 10 meshes from each group were selected, and the cells were removed and counted as above. The raw data were adjusted for the size (weight) of the mesh in each case and presented as cells counted per milligram of dry mesh. One-way and two-way ANOVA were carried out. Coefficients of variation (CVs) were calculated for each data set as follows: CV (s 100)/Y, where s is the sample standard deviation, and Y is the sample mean. In the tests conducted at the National Defense Medical College, Saitama, Japan, human synovial stromal cells (HSSCs) (Applied Cell Biology Research Institute) were cultured in CS-C medium (Standard Tools for Cell Biology) containing 10% serum and growth factor. They were seeded at a density of 1 105 cells ml1 per well on each mesh sample (approximately 10 10 mm) in a 24-well plate. The plate was changed 1 day after inoculation (day 1), moving the pieces of mesh to new wells to eliminate cells attached to the plate. The medium was changed on days 2 and 4. [3H]-Thymidine incorporation and DNA assays were carried out 1, 3, and 5 days after inoculation. For each sample [3H]-thymidine (2 µCi/well) was added, and the cells were incubated at 37°C for 3 h. After incubation, the medium was removed and the cells were washed for 15 min with phosphate-buffered saline (PBS). This washing was carried out three times for each sample. Then the cells on the mesh were solubilized in 1 M NaOH. After neutralization with 1 M HCl, the aliquots were analyzed for [3H] radioactivity
202
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
by scintillation counting. A preliminary study had demonstrated that negligible amounts of thymidine were retained by either untreated or treated mesh in the absence of cells. After digesting the cells attached to the samples for DNA assay using papain at 60°C overnight, the aliquots of the digest were mixed with Hoechst 33258 dye solution. Fluorescence was measured using a Titerek Fluoroskan II fluorometer with excitation and emission wavelengths set at 355 and 450 nm. Calf thymus DNA was used as a standard. Finally, scanning electron microscopy (SEM) was carried out on day 5 to visualize the cells on the materials.
Results Wettability Figure 1 illustrates the marked difference in wettability after RF glow treatment. The droplet has clearly spread out and penetrated the treated material but has
remained more or less on the surface of the untreated material. The mean time taken for the droplet to disperse on the control material was well over 200 s (n 24), whereas in the case of the treated material it took only 5.5 s (SD 9.6 s; n 18). Mechanical properties In tensile tests of unfatigued ligaments (Table 1), for both 2100 and 3300 N ligaments, the load to failure of the treated material was found to be lower than that for the untreated material (12% in the case of the 2100 N ligament). This was accompanied by an approximately 11%–15% increase in stiffness. After having undergone a medium-term fatigue test (10.8 106 cycles) (Table 2), both treated and untreated 2100 N ligaments (n 4 each) had virtually the same load and extension to failure (2300 N and 11.2 mm, respectively). The exhaustive fatigue test (62 106 cycles) was conducted on only two samples of treated ligaments owing to the length of the test program. A comparison can be made, however, with control samples from the identical batch but obtained before treatment and not subjected to any cyclic loading before the mechanical test (Table 3). The untreated, unfatigued ligaments had a UTS of 3689.0 N and stiffness of 338.4 N mm1, 91% and 115%, respectively, of the values obtained with the controls. Although the numbers are small, this gives some assurance of the durability of the RFGD-treated product under these conditions. Cell recruitment and proliferation
Fig. 1. Droplet test for wettability. Note the difference between the radiofrequency (RF) glow treated (left) and untreated (right) material. For the purposes of photography, methylene blue dye was added to the sucrose solution to make the droplet easier to see
Short-term observations (6 h) In tests with small sections of single fibers, there was little attachment of dermal fibroblasts to the control material, whereas the treated material exhibits markedly (⬃35 times) more attachment after 6 h, with most of the attached cells being fully spread (Fig. 2). Longer-term observations (24–72 h and up to 14 days) In tests with single-fiber pieces, both materials supported steady spreading of dermal fibroblasts over
Table 1. Tensile test results from control and RF glow-treated ligaments: strength and stiffness (n 4) 2100 N Ligament Parameter (UTS) SD (N) Extension to failure SD (mm) Stiffness SD (N/mm)
3300 N Ligament
Control
Treated
Percent of control
Control
Treated
Percent of control
2288.0 70.0 12.7 0.45
2013.0 45.7 9.7 0.46
88.0 76.4
3282 126 12.3 1.17
3177 298 10.6 0.60
96.8 86.2
179.9 5.99
207.2 5.26
115.2
267.6 21.1
299.5 19.4
111.9
These tests were conducted on nonfatigued ligaments with a nominal strength of 2100 N RF, radio frequency; UTS, ultimate tensile strength
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
203
Table 2. Medium-term (10 800 000 cycles) fatigue test results: residual strength and stiffness (n 4) Control
RF glow treated
Percent of control
2322.0 34.8 11.20 0.34
2319.00 15.17 11.20 0.38
99.9 100
207.4 4.9
206.80 7.76
99.7
Parameter UTS SD (N) Extension to failure SD (mm) Stiffness SD (N/mm)
These tests were conducted on ligaments with a nominal strength of 2100 N
Table 3. Effect of exhaustive fatigue testing (62 126 000 cycles) of Leeds-Keio ligament-II treated with glow: residual strength and stiffness (n 2) Condition Not fatigued Fatigued Change
UTS (N)
Overall stiffness (N/mm)
3689.0 3347.5 9.3%
338.40 389.75 15.2%
Fig. 3. Coverage of RF glow treated and untreated (control) fibers with epithelial fibroblasts 24, 48, and 72 h after inoculation (mean SD)
Fig. 2. Total number of spreading and fully spread epithelial fibroblasts on 5-mm lengths of RF glow treated and untreated (control) fibers 2, 4, and 6 h after inoculation
the 72-h period; about twice the coverage was obtained with the treated material compared with the control (Fig. 3). Figure 4 presents a view of individually treated fibers obtained 72 h after inoculation, illustrating the abundant coverage by fibroblasts and bridging of cells between the fibers. The results from the 14-day study on meshes are presented in Fig. 5. Although the initial number of synovial fibroblasts recovered from the pieces of mesh at day 0 was greater in the control samples than in the treated samples, by day 6 the situation had reversed. At the end of the experiment (day 14) there were approximately four times the number of cells on the
Fig. 4. Photograph taken with an optical microscope showing epithelial fibroblasts growing on and bridging RF glow treated fibers. Hematoxylin stain was applied 72 h after inoculation. 200
treated samples as on the control. The increase in cell number, as a percentage of the initial count, was 83.5% in the case of the control, and 1461% in the case of the treated material. Initial two-way ANOVA demonstrated that there were significant variations over time and between each group, although the error margin was
204
too large to distinguish between the individual data sets. Using one-way ANOVA, the differences between treatments could be shown to be significant at the 95% confidence level at day 0 and from day 6 onward. The incorporation of [3H]-thymidine by synovial stromal cells growing on the RF glow-treated samples was significantly greater than by those growing on the control, even at day 1. The difference increased with time, from approximately threefold at day 1 to approximately fivefold at day 5 (Fig. 6). This incorporation is a measure of cell division, and the results indicate a faster rate of cell multiplication on the treated material. The total DNA content of the cells growing on the treated samples (1.484 µg ml1) was significantly greater than that of those growing on the control samples (0.156 µg ml1) (P 0.001).
Fig. 5. Average number of synovial fibroblasts recovered from RF glow treated and control (untreated) mesh at each time point per milligram dry weight of mesh (n 10). Significantly fewer cells (P 0.05) were recovered from the treated material at day 0. Significantly more cells (P 0.05) were recovered from the treated material at days 6, 11, and 14
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
Figure 7 shows scanning electron micrographs of cells on samples of control and treated mesh obtained on day 5 after inoculation. In the case of the control material, little cell spreading can be seen. By contrast, cells have almost completely covered the treated fibers, and the cells are generally aligned along the length of the fibers. These images are consistent with the conclusions from the DNA and [3H]-thymidine measurements and clearly illustrate more rapid colonization of the treated mesh by cells.
Discussion Wettability Polyethylene terephthalate (PET), a material with a long history of safe implant use, usually elicits little adverse tissue reaction. Being hydrophobic, however, cells do not attach and spread on it as readily as if it were a hydrophilic material. Radiofrequency-generated glow discharge treatment is an established method for
Fig. 6. [3H]-Thymidine incorporation by synovial stromal cells growing on RF glow treated and control samples
Fig. 7. Scanning electron microscopy (SEM) showing portions of control (left) and RF glow treated (right) mesh on day 5. Note the greater coverage of the treated materials by synovial stromal cells. 500
J.R.J. Rowland et al.: RFGD-treated polyester ligaments
modifying a material’s surface to increase its chemical reactivity and improve its wettability. This treatment does not involve coating the surface of the PET filaments with any substance, nor does it require high temperatures. It exposes the material to an energized atmosphere generated by RF discharge excitation in the presence of a gas at low pressure without significantly raising the temperature of the material. Excited ions and atoms react with the material surface to form new, ionizable surface radicals. In biomedical applications RF glow treatment has been suggested as a method for improving the attachment of cells to implant surfaces.3,10 Depending on the nature of the gas medium, various chemical changes can be produced in the polymer surface. Dekker et al.6 reported that the thickness of the modified layer in the case of polytetrafluoroethylene (PTFE) was only about 1 nm, and that not only was hydrophilicity increased but proteins including fibronectin were absorbed more readily from a serumcontaining medium onto the modified material. This treatment can also be used for more specific surface modification by coupling particular biochemical entities to the surface molecules of the material, such as glycine or fibronectin1 and heparin.12 The increased wettability brought about by the RFGD treatment was not unexpected, as Canonico et al.3 reported that the critical surface tension for wetting RF glow-treated PET fabrics was significantly greater than for conventional PET fabrics. They showed by electron spectroscopy for chemical analysis (ESCA) that subjecting PET to an oxidative glow discharge results mostly in the introduction of carboxyl and carbonyl functionalities onto the polymer surface, and this increases its wettability. Mechanical properties It has often been reported or claimed that RF glow treatment affects only the outermost layer of materials. Glow discharges, however, emit ultraviolet light, which can cause crosslinking and possibly some chain realignment. The occurrence of either of these phenomena could account for the changes in the initial mechanical properties observed in this study. The apparent “recovery” of strength by the material after cyclic loading is not easy to explain. The strength and stiffness of the ACL in the young have been estimated by Woo et al. at around 2500 N and 292 N mm1, respectively.17 Although the RF glow treatment reduces the tensile strength of the artificial ligament, in the case of the 3000 N ligament used in the exhaustive fatigue test it is still greater than this figure even after 62 106 loading cycles. The mean stiffness of the latter (389.75 N mm1) is 33% greater than that of the natural ACL ligament. However, this is still
205
comparable with the stiffness of quadrupled semitendinosus grafts and less than that of central-third patellar tendon grafts. Cell attachment and proliferation The three studies of cell attachment and proliferation on RFGD-treated woven polyester meshes all demonstrated that the test cells (dermal, synovial, and stromal cells) attach, proliferate, or both more quickly over the first few days of culture than did the untreated controls. Consistently, there are more cells on the treated material than on the controls at the end of each study. In the study using synovial fibroblasts, it is not clear why significantly fewer cells adhered initially to the treated mesh than to the control, but this effect was also found in a preliminary study with these cells (using this method but not otherwise reported here). It could be that the cells are not actually attached to the mesh, but that the hydrophobic nature of the untreated control reduced the efficiency of the rinsing procedure to remove unattached cells in this study.
Conclusions The RF glow treatment of the PET ligament reported here greatly enhances its wettability. This results in marked improvements in terms of increased cell attachment and proliferation on the surface of the implant, a consistent finding of studies undertaken in three research establishments using different in vitro approaches and methodologies. Adverse effects on the strength of the woven ligament have been found to be relatively small and can be readily overcome by small design changes in the ligament’s load-bearing structure. The changes induced by the treatment are not adversely affected by subsequent, terminal sterilization of the implant by gamma radiation (25–50 kGy) (confirmed by other studies not reported here); hence we believe that RF-generated glow discharge treatment has significantly enhanced the potential of polyester for use in ligament implants. The RF glow discharge process investigated in the studies reported in this paper is now used to provide increased functionality on the new range of synthetic ligaments known as the Leeds-Keio Ligament-II (also known as the ESP3000 ligament in Europe). Acknowledgments. The authors thank Xiros plc for providing ligament samples. We also thank the following individuals for their contribution to the work described in this study: Mr. Les Johnson, School of Textile Industries, University of Leeds, for carrying out wettability tests; Dr. David Woolley, Department of Medicine, University of Manchester for supplying cells for the study at Yorkshire Regional Tissue Bank; Mr. John
206 Tresnan, Xiros plc, for carrying out the mechanical tests. We especially acknowledge the contribution of Sangjin Kang to the work carried out at the School of Biochemistry and Molecular Biology, University of Leeds.
References 1. Anderheiden D, Klee D, Höcher H, et al. Surface modification of a biocompatible polymer based on polyurethane for artificial blood vessels. J Mater Sci Mater Med 1992;3:1–4. 2. BS 4554 method of test for wettability of textile fabrics. London: British Standards Institution, 1970. 3. Canonico P, Cassinelli C, Morra M. Surface-modified polyester fabrics. Med Device Technol 1997;8:16–21(April). 4. Clancy WG, Narechania RG, Rosenberg TD, et al. Anterior and posterior cruciate ligament reconstruction in rhesus monkeys. J Bone Joint Surg Am 1981;63:270–84. 5. Cooper DE, Deng XH, Burstein AL, et al. The strength of the central third patellar tendon graft. Am J Sports Med 1993;21:818– 24. 6. Dekker A, Reitsma K, Beugeling T, et al. Adhesion of endothelial-cells and adsorption of serum-proteins on gas plasmatreated polytetrafluoroethylene. Biomaterials 1991;12:130–8. 7. France RM, Short R, Dawson RA, et al. Attachment of human keratinocytes to plasma co-polymers of acrylic acid/octa-1,7-diene and allyl amine/octa-1,7-diene. J Mater Chem 1998;8:1–6. 8. Fujikawa K, Seedhom BB, Matsumoto H, et al. The Leeds-Keio ligament. In: Strover A, editor. Intra-articular reconstruction of the anterior cruciate ligament. London: Butterworth Heinemann; 1993. p. 173–207.
J.R.J. Rowland et al.: RFGD-treated polyester ligaments 9. Groner C. Synthetic ACL repair may get another chance. Biomechanics 1998;32:79–84(May). 10. Jansen JA, van der Waerden JPCM, de Groot K. Fibroblast and epithelial cell interactions with surface-treated implant materials. Biomaterials 1991;12:25–31. 11. Jobanputra P, Parry D, Fry-Smith A, et al. Effectiveness of autologous chondrocyte transplantation for hyaline cartilage defects in knees. Report commissioned by NHS HTA Programme. Birmingham, UK: West Midlands Development and Evaluation Service, The University of Birmingham. July 2000. 12. Kang IK, Kwon OH, Lee YM, et al. Preparation and surface characterization of functional group-grafted and heparin-immobilized polyurethanes by plasma glow discharge. Biomaterials 1996;17:841–7. 13. Macnicol MF, Penny ID. Leeds-Keio ligament anterior cruciate replacement: a clinical, arthroscopic and histological review. J Bone Joint Surg Br 1990;72(suppl. 1):1. 14. Marcacci M, Gubellini P, Buda R, et al. Histological and ultrastructural findings of tissue ingrowth: the Leeds-Keio prosthetic anterior cruciate ligament. Clin Orthop 1991;267:115– 21. 15. Meins G, Vierhout PAM. Arthroscopic reconstruction after rupture of anterior cruciate ligament with the Leeds-Keio ligament. Enschede, The Netherlands: Medische Spectrum Twente, Department of Surgery, 1995. 16. Pratt KJ, Williams SK, Jarrell BE. Enhanced adherence of human adult endothelial cells to plasma discharge modified polyethylene terephthalate. J Biomed Mater Res 1989;23:1131–47. 17. Woo SL-Y, Young EP, Kwan MK. Fundamental studies in knee ligament mechanics. In: Daniel D, Akeson W, O’Connor J, editors. Knee ligaments: structure, function, injury and repair. New York: Raven; 1990. p. 115–34.