In 1966, pure titanium was introduced as an alternative implant material to stainless steel for internal fixation.13 Stainless steel has several disadvantages for use in fracture care. Corrosion occurs with stainless steel implants, and allergic reactions to the nickel alloy have been documented.4 Stainless steel has a high modulus of elasticity that causes stress shielding and bone loss beneath applied plates. Titanium has a modulus of elasticity more similar to bone and is more resistant to corrosion.4 It is less reactive than stainless steel and considered more biocompatible.17 No allergic reactions have been reported with titanium, and titanium does not cause signal interference with magnetic resonance imaging (MRI).
Soft issue encapsulation occurs when a foreign implant is used. The thickness and adhesion of the soft tissue layer reflects the biocompatibility of the implant material and the surface conditions.11,18-20 Titanium implants with rough surfaces appear to lead to the most adherent, thinnest soft issue reactive layer.18,19 Electropolished stainless steel typically results in a nonadherent, substantially thicker soft tissue reactive layer with a liquid film at the interface. This dead space can allow bacteria to multiply with poor penetration of the body's immune defenses. The close soft tissue adherence to a titanium implant may have advantages with respect to infection.1,2
The use of titanium small implants for the peripheral skeleton has been reported clinically.13 There was more overgrowth of connective or bone tissue with the titanium implants compared with similar stainless steel implants. This tissue reaction could be related to the better tissue compatibility of titanium compared with stainless steel.13 However, in situations where tendons are in close approximation to implants, such as in the wrist and hand, this soft tissue adherence may not be advantageous. Direct tissue contact and adhesion to the implant may diminish tendon excursion. Furthermore, if titanium is problematic, it is unclear what role if any titanium debris plays. Titanium particles have been found adjacent to implanted miniplates.5,7-9,12 The role of a proliferative foreign body and/or immune response in eliciting these tendon changes cannot be ruled out.
Tenosynovitis and extensor tendon ruptures have been reported with the use of a low profile titanium distal radius plate.3,6,10,14 These clinical observations may be related to the plate design (including the boss edges), the titanium, the matte finish, or other factors. A modification of the original plate design involved the addition of a sloped, ramped surface to the distal plate edges to allow for a smoother surface for extensor tendon gliding. However, no consensus exists on the etiology of tendon disturbance. This raises two questions which we sought to answer in this study. Are tendon reactions observed clinically secondary to metallurgy, specifically the titanium alloy and its surface morphology? Secondly, are these tendon reactions caused by plate design, specifically the sharp edges at the distal surface of the original implant?
MATERIALS AND METHODS
In an effort to separate plate design versus metallurgy effects, the original titanium boss pi plate was compared with a similar version in stainless steel as well as a titanium plate with smoother, ramped distal edges. The tendon and bony anatomy of the distal radius in canines is very similar humans, allowing for comparisons using the dorsal pi plate. Fourteen skeletally mature male mixed-breed hound-type dogs (weight range, 33-38 kg) were used for the study. A commercially available distal radius dorsal pi plate (Synthes, Paoli, PA) was applied bilaterally in all cases. In seven dogs, Group I, the original titanium boss pi plate design was utilized on one side versus an identical stainless steel plate on the contralateral limb. In seven dogs, Group II, the original titanium plate was utilized on one side versus a modified titanium pi plate with ramped edges on the contralateral limb (Fig 1). (Supplemental materials are available via the Article Plus feature at www.corronline.com. You may locate this article, then click on the Article Plus link on the right.) Plate placement with respect to side was randomized. Range of motion of the wrist was measured preoperatively, following plate placement and at 3, 6 and 9 months, at which time the dogs were euthanized and the forearms with the plates in situ were sectioned and prepared for histologic examination. The principle outcome variables of the study were loss of wrist motion and the prevalence of histological changes (tendon degeneration, neovascularity, and inflammatory cell infiltrates).
The titanium pi plates were fabricated of commercially pure titanium, and the titanium ramped plates were made of titaniummolybdenum alloy. Both types of titanium plates had an anodic oxide surface treatment. The stainless steel plates were made of iron-chromium-nickel-manganese alloy and had an electropolished surface. Screws of the corresponding alloys were used fix the plates to the radius. Scanning electron micrographs of the titanium and stainless steel plates were made to document the appearance of the implant surfaces (Fig 2). The elemental composition of the plates was confirmed using energy-dispersive xray analysis.
The protocol was approved by the Institutional Animal Care and Use Committee at Rush University Medical Center. The thoracic forearms and manus of each animal was prepared and draped. The animals were anesthetized using subcutaneous acepromazine (0.05 mg/kg) and morphine (0.5 mg/kg), intravenous thiobarbiturate (8-16 mg/kg), and isoflurane to maintain the surgical plane. A cephalosporin antibiotic was administered intravenously (1 gm). Preoperative radiographs were taken to ensure skeletal maturity and normal anatomy of the distal forelimbs. Lateral wrist radiographs were taken in maximum flexion and extension, and corresponding range of motion (ROM) measurements were obtained with the use of a handheld goniometer.
Through a dorsal approach to each distal radius, the extensor carpi radialis (ECR), common digital extensor (CDE), and abductor pollicus longus tendons were identified, elevated and retracted medially and laterally to expose the distal one-quarter of the radius. The extensor retinaculum and the tendon subsheaths of the ECR and CDE tendons were opened along the length of the plate to allow for exposure and correct plate positioning. Plates were contoured and placed in similar positions on the dorsal surface of the intact radius in all forelimbs. The horizontal bar of the plate was positioned just proximal to the dorsal margin of the articular surface per the implant design, conforming well to the bone and spanning from the radial styloid to the radioulnar joint laterally. The most proximal screw hole of the lateral arm and extreme medial screw hole of the horizontal arm were removed from each implant using the manufacturer's plate cutter. All of the plates were fixed in an identical manner using three 2.4-mm self-tapping screws and one locking buttress pin through the horizontal arm distally, three 2.7-mm screws through alternate screw holes in the medial arm, and two 2.7-mm screws through the lateral arm (Fig 1). The paw was placed through a range of motion to ensure gliding of the tendons over the plate arms. The retinaculum and soft tissues were closed in the usual manner and orthogonal radiographs were obtained. Compressive dressings and coaptation splints were then applied to the limbs.
Postoperative analgesia was provided using subcutaneous buprenorphine (10 to 30 μg/kg) at 12-hour intervals for 2 days. Acetaminophen was administered thereafter to any dogs exhibiting clinical signs of pain. A cephalosporin antibiotic was administered orally (22 mg/kg) three times daily for 5 postoperative days. Immobilization was discontinued after 14 days, after which no activity restrictions were enforced. The animals were allowed to bear weight as tolerated in their cages and were exercised daily. Range of motion (ROM) measurements using a handheld goniometer and flexion and extension radiographs were obtained at 3, 6, and 9 months under general anesthesia (Fig 3). (Supplemental materials are available via the Article Plus feature at www.corronline.com. You may locate this article, then click on the Article Plus link on the right.) Orthogonal frontal and lateral radiographs of all limbs were obtained at these time periods as well. At 9 months postoperatively the animals were euthanized using intravenous sodium pentobarbital.
After euthanasia, the forearms were dissected free from the upper limb and the skin was removed. The carpus was placed through a range of motion to assess any impairment of tendon excursion by adhesions to the implant. The carpus was placed in neutral position, the extensor carpi radialis (ECR) and common digital extensor (CDE) tendons were resected distally at the level of the carpus, and the carpus and manus were removed. The tissues were left intact over the remainder of the plate in order to maintain the relationship of the tendons to the implants. No attempt was made to remove any of the screws or the plate.
The specimens, consisting of the radius, ulna, fixation plate and the extensor tendons with the peritendonous soft tissues, were fixed in 10% neutral-buffered formalin and then cut transversely at 5-mm intervals from the horizontal bar to the proximal end of the plate. Contact radiographs were obtained of the cut sections to evaluate bone reaction around the implants. From one distal and one section in the middle of the plate, plastic embedded undecalcified ground slides with the implant in place were prepared and stained with basic fuschsin and toluidine blue. In five of the remaining transverse sections spaced evenly along the length of the plate, the tendons and surrounding soft tissues were dissected carefully from the bones and implant and processed to produce slides stained with hematoxylin and eosin.
The stained slides were studied using light microscopy by two of the authors (TMT, RMU) who were blinded to the study group at the time of evaluation. The plastic embedded slides were examined to determine the nature of the membrane surrounding the plate, the spatial relationship of the fixation plate and membranes to the tendons, and the presence of metal alloy particles within the peri-implant tissues. The hematoxylin and eosin stained slides were evaluated for possible reactive or degenerative changes in the tendons due to the surgical procedure and the proximity of the metal implant. A minimum of five slides of the ECR and CDE tendons from each limb were used to assess the prevalence of specific histologic changes to the tendons. After an initial review of the slides, three primary variables were identified and their presence or absence was recorded for the tendons of each limb. The variables were tendon degeneration (chondroid metaplasia with associated hypocellularity and hyalinization of blood vessels), neovascularization, and inflammatory cell infiltrates (including changes secondary to metallic particulate debris).
We used SPSS for Windows (Version 13, SPSS Inc, Chicago, IL) for data management and statistical analysis. Nonparametric statistical methods were used to analyze the data. The McNemar test for paired proportions was used to analyze the primary histologic variables of the presence or absence of tendon degeneration, neovascularity, and inflammatory cell infiltrate. Percentages for plates from different groups were compared using Fisher's exact test. The Friedman test was used to compare the bilateral ROM measurements within each of the groups. The Friedman test also was used to compare the initial, 3-month, 6-month, and 9-month data separately for each plate type with respect to preoperative ROM values. When a result was significant, additional Friedman tests were done to compare the time periods two at a time to determine differences. In addition, the Mann-Whitney test was used to compare each type of plate for Group I dogs to each type of plate for Group II dogs. A 0.05 significance level was used for all statistical tests. No one-sided statistical tests were done.
Inclusion of seven dogs per group insured an 80% power for detecting a difference between a plate type for Group I dogs and another plate type for Group II dogs if the probability that an observation in one group was less than an observation in the other group (0.932, based on a two-sided Mann-Whitney test with a 0.05 significance level).
All surgical wounds healed without complications. Dogs in both groups exhibited clinical weight bearing in the coaptation splints. Following splint removal at 2 weeks, all animals resumed and maintained normal clinical function for the remainder of the study. No migration of the screws and plate, fracture of the implants, or radiolucencies at the screw-bone or plate-bone interfaces was apparent on the serial clinical radiographs (Fig 4).
There was no significant difference (p > 0.05) in loss of wrist flexion between the different plates in either Group I or Group II at any of the postoperative intervals (Figs 5, 6). In Group I, the average loss of flexion at 9 months was 11 ± 23° for the titanium pi plates and 12 ± 25° for the stainless steel plates. In Group II, the average loss of flexion after 9 months was 19 ± 21° for titanium pi plates and 26 ± 24° for the titanium ramped plates. For each type of plate in Group II, the loss of flexion was significant at 3, 6, and 9 months (p < 0.03). The mean loss of extension at 9 months was less than 2° for all of the different types of plates.
A substantial loss of flexion of 35 to 65° after 9 months was recorded in three dogs in Group I and four dogs in Group II. However, no association between this substantial motion loss and plate material or design was identified. In Group I, marked limitation of flexion occurred with both the titanium boss and the stainless steel plates in one dog, with only a titanium boss plate in one dog, and with only a stainless steel plate in another dog. In Group II, substantial losses in flexion were present with both the titanium boss and a titanium ramped plates in one dog, with only a titanium boss plate in one dog and with only a titanium ramped plate in two dogs.
Soft tissue thickening over the dorsal radius (clinically) and proliferation of the peritendonous tissues overlying and surrounding the plates (histologically) was identified in all dogs, regardless of plate design or material. This was palpable clinically by the 3-month evaluation and remained for the duration of the study. Histologic examination confirmed proliferation of peritendonous tissue overlying the plates on gross examination and in the transverse sections of the radius, ulna, fixation plate and dorsal soft tissues (Fig 7). (Supplemental materials are available via the Article Plus feature at www.corronline.com. You may locate this article, then click on the Article Plus link on the right.) The ECR tendon, in particular, appeared deformed by its close proximity to the medial arm of the plate.
Fibrous membranes encapsulating the plates and infiltrating the screw holes were contiguous with the side of the tenosynovium facing the plate. Peritendinous adhesions were also apparent on gross dissection at the distal level of the plates, extending from the adjacent tissues to the sheaths of the ECR and CDE tendons (Fig 8). (Supplemental materials are available via the Article Plus feature at www.corronline.com. You may locate this article, then click on the Article Plus link on the right.) The adhesions consisted of thin filamentous or wider bands of fibrous tissue, which appeared to emanate from the edges of the plates.
The fibrous tissue membranes surrounding the plates had fibers oriented parallel to the implant surfaces in a similar manner regardless of the implant material or the plate design. Deposits of hemosiderin and associated macrophages were seen within the membranes adjacent to the majority of the specimens. In addition, small focal accumulations of metal alloy particles, also associated with macrophages, were present in the membranes between the plates and bone and adjacent to the plate and screw junctions of both titanium and stainless steel implants.
The prevalence of degenerative changes in the architecture, vascularity and cellularity on the side of the tendon facing the fixation plate was not significantly different (p > 0.05) whether the titanium plate was compared to a stainless steel plate in Group I or to a titanium ramped plate in Group II. The undersides of the tendons, especially the ECR, were flattened or concave in the area adjacent to the plate. The degenerative changes included fibrillation of the tendon surface and chondroid metaplasia of the paratenon and tendon parenchyma accompanied by areas of hypocellularity and hyalinization of small and mediumsized vessels (Fig 9). (Supplemental materials are available via the Article Plus feature at www.corronline.com. You may locate this article, then click on the Article Plus link on the right.) Specifically, in Group 1, the degenerative changes occurred with titanium and stainless steel plates in four dogs and only with a titanium plate in one dog. In Group 2, these changes were present with titanium and titanium ramped plates in three dogs and only with titanium plate in 2 dogs. Degeneration of the body of the tendon was particularly severe with the titanium and the stainless steel plates in one dog and with a titanium ramped plate in another dog. In these tendons, chondroid metaplasia, hypocellularity, and hyalinization of blood vessels extended into the ECR tendon for up to ⅓ of its thickness (Fig 10).
The prevalence of neovascularity, consisting of budding and branching small blood vessels in the parenchyma of the tendon opposite the plates, was not different (p > 0.05) whether the titanium boss plate was compared with a stainless steel plate in Group I or to a titanium ramped plate in Group II. Neovascularity was present in the tendons of Group I dogs adjacent to six of seven titanium and four of seven stainless steel plates and in the tendons of Group II dogs adjacent to seven of seven titanium and six of seven titanium ramped plates.
A minimal inflammatory cell infiltrate limited to the tenosynovium or the paratenon consisting of small, focal aggregates of lymphocytes, plasma cells, and macrophages containing hemosiderin was observed with similar prevalence (p > 0.05) whether the titanium plate was compared to a stainless steel plate in Group I or to a titanium ramped plate in Group II. In Group I, focal inflammatory cell infiltrates were observed with the titanium and the stainless steel plates in three dogs, with only the titanium plate in one dog, and with only the stainless steel plate in another dog. In one of the dogs in Group I, the inflammatory cell infiltrate in the tenosynovium and paratenon was more extensive than any of the other dogs and occurred to a similar extent with the titanium plate and the stainless steel plate in this animal. In addition, a minimal perivascular inflammatory cell infiltrate was noted in the body of the tendon of this dog on the side hosting the stainless steel plate. In Group II, infiltration of inflammatory cells was present in association with a titanium ramped plate in two dogs and a titanium boss plate in one dog. Polymorphonuclear leukocytes were not observed in any of the specimens. No metal alloy particles were noted within the tendons, paratenon, or the tenosynovium in any specimens.
New bone formation along the plate was observed on the contact specimen radiographs in seven of seven titanium plates and five of seven stainless steel plates of Group I and in four of seven titanium plates and three of seven titanium ramped plates in Group II (p > 0.05). In the stained undecalcified sections and in the radiographs of these sections, the reactive bone when present had developed along the medial aspect of the lateral arm of the plate, in the area between the two plate arms, lateral to the medial arm of the plate, and through screw holes. The bone extended to the height of the plate and occasionally above the plate.
Titanium has several theoretical advantages compared with stainless steel. It has a lower degree of stiffness, is more corrosion resistant, is more biocompatible, and the alloy does not cause any known allergic reactions.4 It has been suggested, however, that titanium plates may lead to greater soft tissue and tendon irritation than stainless steel implants.8,16 Specifically, tenosynovitis and extensor tendon ruptures have been reported with the use of a low profile, precontoured titanium plate of mesh design for the distal radius.3,6,10,14 The etiology of these problems remains unclear. Titanium or its matte finish has been implicated. The design of the plate has been considered, leading to the development of a plate with ramped distal edges to minimize extensor tendon irritation.
We used a canine distal radius model to replicate the anatomical positioning and tendon-plate relationships with this implant in order to separate out potential effects of metallurgy and plate edge design. The canine distal forelimb anatomy is very analogous to that of man, specifically with respect to bony anatomy and the pathways of the wrist and digital extensor tendons. Canine carpal range of motion is also similar; but the canine has less extension (10-15°) and greater flexion (approximately 150°) than a human. This does represent a potential study limitation. However, the greater flexion of the canine wrist is not necessarily a disadvantage when studying extensor tendon interactions with a dorsally applied radial plate. An additional limitation is that we only evaluated the animals at 9 months, and thus we may not have appreciated the evolution of histological changes earlier after plate implantation.
We found striking changes in the extensor tendons and soft tissues adjacent to the plates in all animals, irrespective of implant material (titanium versus stainless steel) or plate design (boss or ramped edges). These included deformations of the extensor tendons with tendon degeneration (fibrillation, chondroid metaplasia, hypocellularity and hyalinization of blood vessels) on the tendon undersurface adjacent to the plate, peritendonous adhesions, and neovascularity in the parenchyma opposite the plate with minimal inflammatory cell infiltrates. Of note, we have observed similar histologic changes in the extensor tendons and peritendonous soft tissues in several of our patients who developed tendon dysfunction and rupture following pi plate implantation (Fig 11). In our study, metal alloy particles were only identified adjacent to the plate and screw junctions, but not within the tendons, paratenon or the tenosynovium in all dogs. Although dogs in all groups lost wrist motion with time, no differences were again observed between groups.
Our findings showing no effect of implant composition on tendon changes are consistent with the clinical series of Rozental et al,15 who found extensor tendon complications occurring similarly in patients treated with either stainless steel or titanium pi plates. These findings, however, differ from those of Sinicropi et al,16 who reported moderate to severe intratendon inflammation associated with titanium in a canine distal radius fracture model. These authors implicated titanium specifically as the “inflammatory trigger.” Differences between this study and ours include first the timing of sacrifice (3 months versus 9 months), which might affect the presence or absence of inflammation. In addition, the plate design (straight plate versus mesh design of the pi plate), and the surface characteristics of the titanium were dissimilar. Under electron microscopy, we identified vastly different microtopography of the titanium pi plate and the 2.7-mm titanium compression plates they studied (Fig 12). Therefore, these two models are not directly comparable.
Placement of metal plates dorsally beneath the extensor tendons of the wrist and the digits can alter the mechanocellular interactions and cause strain within the extensor tendons, leading to a soft tissue response. Tendon problems observed with the dorsal pi plate do not seem to be specific to a particular metal or edge design (boss or sloped). This implant is applied to the entire distal radius width and spans from the radial styloid to the sigmoid notch within millimeters of the radiocarpal joint. It is over this joint margin that the extensor tendons sharply angulate during flexion of the wrist. This aspect of the plate design may be partially responsible for some of the tendon changes observed. When dorsal fixation is used, other options such as smaller fragment specific implants that do not envelope the entire dorsal surface of the distal radius may be beneficial. Further study is warranted. When problems do occur with dorsal plates beneath extensor tendons, it would appear that titanium or stainless steel cannot specifically be implicated.
The authors would like to thank Susan Shott, PhD, for her expertise and assistance with the statistical analysis of the data.
1. Arens S, Schlechel U, Pritzen G. Influence of materials for fixation implants on local infection: an experimental study on steel versus titanium DCP in rabbits. J Bone Joint Surg
2. Gristina AG. Biomaterial-centered infection: microbial adhesion versus tissue integration. Science. 1987;l237:1588-1595.
3. Hahnloser D, Platz A, Amgwerd M, Trentz O. Internal fixation of distal radius fractures with dorsal dislocation: pi-plate or two ¼ tube plates? A prospective randomized study. J Trauma
4. Hallab NJ, Merritt K, Jacobs JJ. Metal sensitivity in patients with orthopaedic implants. J Bone Joint Surg
5. Jorgenson DS, Mayer MH, Ellenbogen RG. Detection of titanium in human tissues after craniofacial surgery. Plast Reconstr Surg
6. Kambouroglou GK, Axelrod TS. Complications of the AO/ASIF titanium distal radius plate system (π plate) in internal fixation of the distal radius: A brief report. J Hand Surg
7. Kane KR, Mochel DM, DeHeer DH. Influence of titanium particle size on the in-vitro activation on macrophages. Contemp Orthop
8. Katou F, Andoh N, Motegi K, Nagura H. Immuno-inflammatory responses in the tissue adjacent to titanium miniplates used in the treatment of mandibular fractures. J Craniomaxillofac Surg
9. Kim YK, Yeo HH, Lim SC. Tissue response to titanium plates: A transmitted electron microscopic study. J Oral Maxillofac Surg
10. Lowry KJ, Gainor BJ, Hoskins JS. Extensor tendon rupture secondary to the AO/ASIF titanium distal radius plate without associated plate failure: A case report. Am J Orthop
11. Meachim G, William DF. Changes in nonosseous tissue adjacent to titanium implants. J Biomed Mater Res
12. Moberg LE, Nordenram A, Kjellman O. Metal release from plates used in jaw fracture treatment. Int J Oral Maxillof Surg
. 1989;18: 311-314.
13. Pfeiffer KM, Brennwald J, Buchler U, Hanel D, Jupiter J, Lowka K, Mark J, Staehlin P. Implants of pure titanium for internal fixation of the peripheral skeleton. Injury
14. Ring D, Jupiter JB, Brennwald J, Buchler U, Hastings H. Prospective multicenter trial of a plate for dorsal fixation of distal radius fractures. J Hand Surg
15. Rozental TD, Beredjiklian PK, Bozentka DJ. Functional outcome and complications following two types of dorsal plating for unstable fractures of the distal part of the radius. J Bone Joint Surg
16. Sinicropi SM, Su BW, Raia FJ, Parisien M, Strauch RJ, Rosenwasser MP. The effects of implant composition on extensor tenosynovitis in a canine distal radius fracture model. J Hand Surg
17. Suzuki R, Frangos JA. Inhibition of inflammatory species by titanium surfaces. Clin Orthop Relat Res
18. Ungersbock A, Pohler OEM, Perren SM. Evaluation of the soft tissue interface at titanium implants with different surface treatments: experimental study on rabbits. Biomed Mater Eng
. 1994;4: 317-325.
19. Ungersbock A, Pohler OEM, Perren SM. Evaluation of soft tissue reactions at the interface of titanium limited contact dynamic compression plate implants with different surface treatments: An experimental sheep study. Biomaterials
© 2006 Lippincott Williams & Wilkins, Inc.
20. Woodward SC, Salthouse TM. The tissue response to implants and its evaluation by light microscopy. In: Von Recum AF, ed. Handbook of Biomaterials Evaluation
. London, England: Collier Macmillan Publishers; 1986:364-378.