Bionate Biocompatibility: In Vivo Study in Rabbits.

Response to foreign materials includes local tissue reaction, osteolysis, implant loosening, and migration to lymph nodes and organs. Bionate 80A human explants show minor wear and slight local tissue reaction, but we do not know the response at the spinal cord, nerve roots, lymph nodes, or distant organs. This study aims to figure out reactions against Bionate 80A when implanted at the spinal epidural space of 24 20-week-old New Zealand white rabbits. In one group of 12 rabbits, we implanted Bionate 80A on the spinal epidural space, and another group of 12 rabbits was used as the control group. We studied tissues, organs, and tissue damage markers on blood biochemistry, urine tests, and necropsy. The animals' clinical parameters and weight showed no statistically significant differences. At 3 months, the basophils increased slightly in the implant group, platelets decreased in all, and at 6 months, implanted animals showed slight eosinophilia, but none of these changes was statistically significant. External, organ, and spinal tissue examination showed neither toxic reaction, inflammatory changes, or noticeable differences between groups or survival periods. Under microscopic examination, the Bionate 80A particles induced a chronic granulomatous response always outside the dura mater, with giant multinucleated cells holding phagocytized particles and no particle migration to lymph nodes or organs. Thus, it was concluded that Bionate particles, when implanted in the rabbit lumbar epidural space, do not generate a significant reaction limited to the surrounding soft tissues with giant multinucleated cells. In addition, the particles did not cross the dura mater or migrate to lymph nodes or organs.

Introduction

Body response to implanted foreign materials and their wear and tear particles[1] is known particularly in devices with movable pieces, such as in major joint arthroplasty.[2] The particles and metallic ions can induce local tissue reactions with osteolysis[3] and implant loosening,[4] migrate through the regional lymph nodes,[5] and reach distant organs (liver and spleen).[6] The number and size of those particles correlate with the probability of these unwanted events.[7] This response is particularly severe with ultrahigh molecular weight polyethylene (UHMWPE),[8] less with metallic alloys[9,10] and minimal with polymeric materials,[1] particularly polycarbonate urethanes (PCU).[11] These latter materials have a negligible wear rate, generating fewer but larger particles than other polymeric materials.[12,13] That correlates with a more limited macrophage response[14] and minimal osteolytic capacity.[11,15]

The PCU Bionate 80A (a polycarbonate urethane from Polymer Technology Group, Inc., Sunnyvale, CA, USA) has shown excellent mechanical resistance and biocompatibility both in vitro(16−19) and in vivo.(20) However, hip acetabular articular cartilage substitution with this material in sheep has a slight long-term wear and tear rate[20,21] with minor superficial cracks and limited particle liberation.[22] The mechanically resistant and biocompatible properties are present in its bulk form, but we are unsure about its effects in particulate form. Additionally, we know that sterilization affects its chemical composition[12,19,23] and mechanical resistance[23] as it induces chain scission and cross-linking[24−26] with the liberation of other chemical products like N,N′-methylene dianiline.[27]

Human data come from explanted implants containing Bionate 80A (Bryan cervical disc and Dynesis lumbar dynamic stabilization). In both cases, wear particles and local tissue reactions were negligible, but nobody studied lymph node or distant organ migration as these implants came from live patients and not postmortem studies. Additionally, and for the same reason, no data is available on the possible reaction to nearby nervous structures (spinal cord or nerve roots).

After literature research, we found no publications about body reactions against Bionate 80A reaction when implanted in particulate form at the lumbar spine epidural space, a gap this study intended to cover. The data obtained will be a step before the clinical application of a Bionate 80A discal nucleus replacement previously published.[28]

Materials and Methods

We followed UNE-EN ISO 10993-6:2007,[29] UNE-ISO 10993-11:2006,[30] and ASTM F 763-04[31] for experimental design, surgical procedure, and data analysis. The Ethical Committee from the Polytechnic University of Valencia, Spain, approved this study and certified that we performed the experiments according to national and local guidelines and regulations.

Using ultrapure water, we obtained the Bionate 80A particles through the ″pin-on-cylinder″ wear technique[21] using a polyurethane piece pressed under a constant 50 N force against a rotating metallic cylinder (Figure ), an equipment developed at the Institute of Biomechanics of Valencia (IBV), Spain, using ultrapure water as the fluid test medium. We extracted the particle samples from the test fluid and filtered and analyzed it by SEM (scanning electron microscopy) for their size characterization. Most of the particles obtained are smaller than 1 μm (75%). Ninety-two percent of the particles are smaller than 10 μm, which is the size limit for a particle to be phagocytable by the cells,[7] thus being susceptible to producing biological reactions (Figure ).

Figure 1

Pin-on-cylinder wear technique. F = force (50 N).

Figure 2

Particle size distribution according to the equivalent circle diameter (ECD) with the pin-on-cylinder wear technique and analyzed by SEM (shown in the image above the chart).

We undertook the study in the BV with methods and facilities approved by the Dirección General de Producción Agraria de la Consellería de Agricultura, Pesca, y Alimentación under registry number ES 4625000010006. Throughout the study, we followed Spanish animal research laws.[32]

We selected 20-week-old nullipara female, parasite-free New Zealand white rabbits for the study, weighing each animal and measuring its food consumption daily.

We submitted 24 animals to the same surgical procedure. In 12, we implanted a Bionate 80A particle sample at the L6 epidural space, and in another 12 (control group), we inserted no foreign material. Then, we split both groups into two subgroups of six animals each, with 3 and 6 month survival, respectively.

For pre-anesthesia and after weighing each animal, we administered xylazine 2% (2.5 mg/kg) and ketamine hydrochloride (17.5 mg/kg) as well as an antibiotic, enrofloxacin (3.5 mg/kg), all intramuscularly. Next, we administered intravenous propofol 1% in a marginal ear vein with an initial dose of 3 mg/kg, followed by continuous infusion (21 mg/kg/h).

We placed the rabbit in the prone position, shaved the lumbar area with an electrical clipper, scrubbed it with povidone iodine, and sprayed it with a mixture of povidone iodine and 80% ethanol.

Under aseptic conditions, we made a midline lumbar 4 cm incision, exposed the L5 and L6 spinous processes, excised the latter, and removed the ligamentum flavum at the L5–L6 interspace, accessing the epidural space. We then implanted a gamma radiation sterilized Bionate 80A 0.7 mm3 (=0.8 mg) powder particle sample in the epidural space at the junction between the nerve root and thecal sac. The gamma sterilization used 2.5 Mrad in a gamma cell 220 Co-irradiating Unit (IONISOS Ibérica, Tarancón, Spain). The control group’s surgical technique was identical but without particle implantation. Finally, we closed the incision in the muscle, subcutaneous tissue, and skin layers.

For 3 days post-operation, we administered meglumine (anti-inflammatory, 1 mg/kg/day, subcutaneously), butorfanol (analgesic, 0.4 mg/kg/day, intramuscularly), and enrofloxacin (antibiotic, 0.4 mL/kg/day, intramuscularly).

Post-operation rabbits ate ad libitum and moved freely inside their cages. We weighed them and controlled their food intake daily, following ASTM F750.[33]

We inspected the rabbits’ skin, fur, eyes, mucous membranes, secretions, excretions, autonomic activity, gait, posture, response to handling, clonic or tonic movements, stereotypes, or any bizarre behavior daily (ISO-10993-11[30]) (Figure ).

Figure 3

Toxicity study clinical inspection parameters: (a) behavior and posture, (b) leg extension reflex, (c) pupillary light reflex, and (d) ears pricking up.

Table details the parameters controlled during these inspections.

Table 1

Toxicity Clinical Parameters Evaluated

We requested blood biochemistry tests pre- and post-operation at monthly intervals and before euthanasia, looking for toxic effects in tissues, kidneys, liver, and tissue damage markers.

We euthanized the animals with a sodium pentobarbital overdose (100 mg/kg) administered through an ear vein.

After sacrifice, we submitted each animal to a complete necropsy (Figure ), including external and internal exams. We looked at their fur and natural orifices as well as their cranial, thoracic, and abdominal cavities and its contents.

Figure 4

Ventral and dorsal rabbit necropsy.

We opened the rib cage and abdominal cavity, observed the general aspect, clamped the esophagus and rectum, and removed the abdominal contents.

We visually examined their thoracic and abdominal cavities, looking for injury signs or organic liquids. Next, we compared the size, color, and appearance of the implanted rabbits’ spinal cord and organs to those of control animals. Finally, we incised the whole rabbit’s back, observed the aspect of the dorsal musculoskeletal structures, isolated the anatomical area where we had implanted the Bionate 80A particles, and removed it as a piece including the spine and spinal cord from L5 to L7.

Finally, we performed a histopathological examination in search of lesions in the brain, spinal cord, pituitary, eyes, thyroid, parathyroid, thymus, esophagus, salivary glands, stomach, small and large intestines, liver, gall bladder, pancreas, kidneys, adrenals, spleen, heart, trachea, lungs, aorta, ovaries, uterus, mammary gland, urinary bladder, lymph nodes, peripheral nerve, bone marrow, and skin.

We performed microscopic assessment in tissues from the implantation zone (spinal cord and surrounding tissues and musculoskeletal tissues from caudal to adjacent cranial zones from L5 to L7) and organs (kidneys, liver, and spleen).

We collected the kidneys, liver, spleen, and tissues surrounding the implanted Bionate 80A particles, including the muscles, epidural space, dura mater, nerve root, and L5 and L7 spinous processes. Then, we removed the Bionate 80A with sufficient unaffected surrounding tissue (ASTM F619[34]) to evaluate the local response.

At the particle implantation site, we evaluated with a 0 to 5 scale (0 for no response, 5 for severe) the number and cell type and whether the reaction was focal or diffuse (ASTM F1904[35]). We also analyzed the presence of particles in regional lymph nodes and distant organs and histological responses compared to the control animals.

We decalcified the L5 and L7 spinous processes (so we could slice them easily), preserved the organs and tissue samples in formaldehyde, included them in paraffin, sectioned them into 5 and 10 μm slices, and stained them with hematoxylin–eosin. Then, we looked at signs of cellular abnormalities, inflammatory reactions, hemorrhage, hemosiderin, adipocytic necrosis, macrophage, granulocyte, lymphocyte or plasmatic cells, or migration of implanted Bionate particles outside the implant zone, mainly if they had crossed the dura mater. We searched for migrated Bionate 80A particles in the organs and signs of cellular abnormality or inflammatory response. We studied the epidural space, dura mater, subdural space, and spinal cord, looking for signs of inflammatory reaction, hemorrhage, adipocytic necrosis, and the presence of macrophages, granulocytes, lymphocytes, and plasma cells as well as possible migration of released particles that could cross the dura mater and escape from the implant zone.

We also did a microscopical analysis of the Bionate particles that we could isolate from the implant side to see if they had undergone any material degradation.

Statistical Analysis

Once we obtained the data from all blood samples, we grouped them in an MS Excel (Microsoft Corporation, Redmond, WA, USA) sheet, which we also used to conduct statistical tests. We used the free statistical analysis software R (R Development Core Team) with Student’s t tests to find statistically significant differences between the implanted and control rabbits. First, we used Student’s t tests to find whether statistically significant differences existed between the implanted and control rabbits. Next, we compared these values with reported hemogram reference values for rabbits.[36] We considered a p value of <0.05 as statistically significant.

Results

Animal Behavior Observations

The parameters most altered post-operation were the locomotion pattern, hind limb weakness and trembling, and decreased crawl reflex, which recovered gradually until it returned to preoperative status. The absence of statistically significant differences between groups indicates their relationship with the surgical procedure and not the implanted particles.

There was a tendency to lose weight on the second and third day post-surgery due to temporary anorexia, weakness, and discomfort. However, one week after surgery, all animals recovered their preoperative weight with no statistically significant differences between groups or survival periods (Table ).

Table 2

Weight Changes in Kilograms during the Study (Means ± SD)

Blood Tests

We found no statistically significant differences between the control and implanted rabbits at 3 and 6 months post-operation (Table ), and the mean values for the different parameters fell within the published ranges.[36] At 3 months, the percentage of basophils (although not its count) increased slightly, while platelets decreased in both implanted and control rabbits. At 6 months, implanted rabbits showed slight eosinophilia but no statistically significant differences from the control group.

Table 3

Hemogram Results for the Implanted and Control Rabbits (Means ± SD)

Postmortem Macroscopic Assessment

External exams during the necropsy study did not reveal any abnormality signs in any animal. No noticeable differences were observed between the implanted groups and blank controls either. Internal exams of organs and spinal tissues did not reveal any abnormality or inflammatory signs in any animal. Again, we observed no noticeable differences between the implanted groups and blank controls.

External animal exams and internal organ and spinal tissue examinations showed no abnormalities, inflammatory changes, or noticeable differences between treatments (particle implantation and control groups) or survival periods (3 and 6 months). In addition, we found no signs of toxicity reactions to Bionate 80A particles implanted in the New Zealand rabbit L6 epidural space.

Histological Analysis of Organs and Tissues

We found no inflammatory reaction in the control rabbits 3 or 6 months after surgery.

Analysis of the drained lymph nodes and organs from the rabbits revealed no microscopic changes or Bionate particles in any rabbit, neither at 3 nor 6 months. In addition, the Bionate particles did not migrate from their original location, next to the spinal cord, and we found no inflammatory reaction. The slices we analyzed were two for the kidneys (one from the right and one from the left), two for the liver (including the gallbladder), and one for the spleen. In all cases, we used midsections of the organs. Concerning the analysis of the spinal cord and its adjacent tissues, we identified macroscopically the zone where we implanted the particles based on the apparent lack of vertebral lamina due to the hemilaminectomy we performed in the surgery.

We located the Bionate particles and identified an associated inflammatory reaction in all implanted rabbits, finding a chronic granulomatous inflammatory response with giant multinucleated cells. This response is the expected reaction to foreign bodies, such as the Bionate particles. At 400× per field, we saw an average of 17 ± 5.22 SD giant cells containing the particles at 3 months and 25 ± 7.63 SD cells at 6 months, thus showing their phagocytosis. We identified no acute inflammatory reaction in any of the cases. The chronic inflammatory response affected only the soft tissues (mainly muscle) surrounding the spinal cord. Furthermore, the reaction always happened outside the dura mater (Figure ).

Figure 5

These two micrographs depict the inflammatory reaction caused by the Bionate particles (birefringent structures). The arrow points to this reaction on the left image, which is entirely extradural. On the right image, the indicator shows a giant multinucleated cell.

Under microscopic examination, the Bionate 80A particles recovered from the implantation side had not changed and looked like those that were not implanted.

Discussion

Bionate 80A shows excellent in vitro resistance to oxidative challenge,[19,37,38] hydrolytic tests,[18] cholesterol esterase enzymes,[39,40] and wear and tear,[11] better than other polymeric[14,38,41] or metallic materials.[9] Also, in vitro, it performs well under good hydration, and long-term cracks grow very slowly.[22] Moreover, it has no L929 fibroblast cell in vitro cytotoxicity, with good cell cytocompatibility and adherence.[16]

In vitro as an acetabular replacement in partial hip joint arthroplasty shows an excellent performance with minimal wear and tear[42] but moderate femoral head cartilage volume loss.[43] Thus, its properties, especially when hydrated, are the closest to the articular cartilage but not quite the same.[22]

In vivo, Bionate 80A showed excellent results in the acetabular sheep replacement with no signs of material degradation, local wear particle release, lymph node, or distant organ migration.[20] The results of our study corroborate it.

While the original Bionate 80A is colorless, the gamma sterilized implants have a yellowish color, attesting that although minimal, this sterilization method induces some deleterious effects[12] (i.e., 9% reduction in its mechanical resistance[23]). For our study, this has no meaning, but it might play a role in long-term discal replacements or dynamic spinal fusions. It is an area that needs further evaluation.

In vivo, long-term reaction data are limited and come from live patient device explants. The only study on explanted Bryan cervical disc prosthesis[15] describes a minimal periprosthetic inflammatory reaction with polymeric debris globular in shape and 3.89 μm in diameter. However, the response elicited could not be separated from the one induced by the metallic wear and tear titanium particles. Meanwhile, several studies deal with explanted lumbar Dynesis dynamic stabilization devices. All have shown minimal wear,[39] rubbing surface burnishing,[12,39,44] and small cracks[12,13,39] in the exposed surfaces[13,39] not deeper than 10 μm,[12] so the effect was negligible. The particle size was >10 μm and in no case under 0.46 μm[44] (the most inflammatory inductive particles[45]). Local tissue reaction was limited to macrophage proliferation and fibroblastic capsular formation but no chronic inflammatory response or wear debris.[44] These changes persist unchanged even after 7 years.[13] Unfortunately, as already happened with the Bryan cervical disc, it is not easy to separate the reaction induced by the implant’s Bionate 80A, polyethylene-terephthalate cord, and titanium alloy.[12,44]

In vivo, there are published studies on epidural foreign particle implantation in rabbits with PEEK,[46,47] stainless steel,[48] cobalt–chromium,[48,49] nitinol,[9] titanium,[9,48] UHMWP,[48] ceramic,[48] polytetrafluorethylene,[48] silicone,[48] polyethene terephthalate,[48] polyester,[48] polyether ketone,[48] and a non-specified polycarbonate urethane.[48] Animals were euthanized at 3 and 6 months after follow-up. There was epidural fibrosis in all of them but intradural particle diffusion only in the metallic groups, with additional intradural fibrosis. Epidural cobalt oxide implantation in the rat induced chronic deafferentation pain,[50] an event not described in any of the studies mentioned above, perhaps because it is a different chemical agent or research animal. There was no intrathecal particle penetration in polymeric implants, and the spinal cord, subdural space, and nerve roots did not develop any inflammatory response. No group showed a reticuloendothelial or systemic reaction. Our study confirms that the Bionate 80A does not penetrate the dura mater or migrate to regional lymph nodes or distant organs. Our hematological and urine periodical tests data rule out acute or toxic Bionate 80A responses.

A future line of investigation is the analysis of the possible degradation products generated as a consequence of the long-term use and deterioration of this material (“Extractables and Leachables” according to ISO 10993) as reported by other research groups,[27] as well as the evaluation of the possible biological effects of these resulting products of degradation.

Limitations

The number of animals is small, and the postoperative follow-up is short. A more considerable number of animals would supply more reliable statistical values, and a more prolonged rabbit survival would show its long-term effects. Furthermore, in the clinical setting, the wear particles are liberated continuously and not in a highly concentrated dose as used in the present study.

Strengths

We have done behavioral, hematological, urinary, macroscopic, and microscopic studies for each animal, including draining lymph nodes and distant organs. We have also studied the local peri-implant tissues and, most importantly, the intradural responses induced by Bionate 80A particles.

Conclusions

According to the results obtained with clinical behavioral parameters, blood and urine tests, necropsy, and histopathological studies, the Bionate particles, when implanted in the rabbit lumbar epidural space, do not generate a significant immunological reaction limited to the soft tissues surrounding the implanted particles, with giant multinucleated cells typical of responses against a foreign body. In addition, the particles did not migrate from the original implantation site to draining lymph nodes or inner organs or cross the dura mater. Thus, we consider that the Bionate wear particles effect is not severe enough to recommend against using this material in spinal implants.