- Email: [email protected]
Magnetic Resonance Imaging Evaluation of Biodegradable Transfemoral Fixation Used in Anterior Cruciate Ligament Reconstruction
Magnetic Resonance Imaging Evaluation of Biodegradable Transfemoral Fixation Used in Anterior Cruciate Ligament Reconstruction Andrew J. Cossey, F.R.C.S.(Tr&Orth), Yegappan Kalairajah, M.A., M.Phil., F.R.C.S.(Tr&Orth), Russell Morcom, F.R.A.N.Z.C.R., and Anthony J. Spriggins, F.R.A.C.S.
Purpose: The study was designed to evaluate bioabsorbable transfemoral fixation in anterior cruciate ligament (ACL) reconstruction using hamstring tendon as graft. Magnetic resonance imaging (MRI) was used to assess the continuity of the bioabsorbable implant at different stages of the patients’ rehabilitation. Type of Study: Retrospective case series. Methods: Forty-nine patients underwent ACL reconstruction performed by a single surgeon. The graft, a tensioned quadrupled semitendinosus tendon, was fixed proximally using a bioabsorbable TransFix implant (Arthrex, Naples, FL). The patients underwent an accelerated rehabilitation program and were assessed clinically at regular intervals postoperatively using MRI, with specific attention focused on the implant. Results: The average time from surgery to MRI was 28 weeks (range, 4 to 54 weeks). All implants were fully visible with no evidence of resorption. Five implants were fractured at an average of 20 weeks postoperatively (range, 9 to 47 weeks). Three implants showed deformation but no definite fracture at an average of 14 weeks (range, 12 to 17 weeks). This amounts to 16% of implants with fractures or deformation, many close to the period of theoretical graft incorporation. All patients were clinically stable with no symptoms or signs or instability on clinical review and all had returned to preinjury sporting activities. Conclusions: Transfemoral biodegradable implants have the potential to fracture or deform during their postoperative course in tensioned hamstring tendon ACL reconstruction. Although no apparent detrimental effect was found in our series, further research is needed on this device before it can be recommended for ACL reconstruction. We also question the idea that rigid fixation for the ACL graft for the entire healing process is required. Level of Evidence: Level IV, retrospective case series. Key Words: ACL reconstruction—Bioabsorbable implant—Transfemoral fixation—Knee—Magnetic resonance imaging.
H
amstring tendons are being used increasingly in anterior cruciate ligament (ACL) reconstruction with this technique claiming lower harvest-site morbidity when compared with patellar tendon grafts.1 With the continued advances in reconstructive techniques allowing for the early return of neuromuscular
From the Orthopaedic Division, SPORTSMED-SA, Adelaide, South Australia, Australia. Address correspondence and reprint requests to Yegappan Kalairajah, M.A., M.Phil., F.R.C.S.(Tr&Orth), Orthopaedic Division, SPORTSMED-SA, 32 Payneham Road, Adelaide, South Australia 5069, Australia. E-mail: [email protected] © 2006 by the Arthroscopy Association of North America 0749-8063/06/2202-05-1$32.00/0 doi:10.1016/j.arthro.2005.08.044
function, graft fixation has to withstand not only these early physiologic forces but also facilitate biological incorporation of the graft construct. With hamstring tendons, the histologic tendon-to-bone anchorage is not fully understood despite animal experimental2 and human studies.3 This constitutes a potential weak point of the technique2 with tendon-to-bone fixation being of the utmost importance for postoperative rehabilitation and resumption of athletic activities.3 ACL fixation methods have improved significantly over the past decade. Currently, many methods of fixation are available, including suture-post constructs, staples, soft-tissue washers, buttons, crosspins, and interference screws. Autograft hamstring ACL reconstruction with femoral cross-pin fixation is
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 22, No 2 (February), 2006: pp 199-204
199
200
A. J. COSSEY ET AL.
a technique popularized by the original work of Clark et al. in the 1990s.4 This study provided biomechanical evidence that femoral cross-pin fixation of the graft mimics the stress-strain relationship of the native ACL. Transfemoral cross-pin fixation using metal cross pins has been shown to have strength and stiffness superior to any other form of femoral fixation.5 The use of interference screw fixation can be associated with possible complications including screw divergence, intraoperative graft damage, bone regrowth within the tunnel, decreased space for biological fixation, and difficult hardware removal.6 Transfemoral fixation offers the theoretical advantage of an increased potential for biological fixation with a 360° bone-to-graft contact area and a decrease in graft damage (when compared with interference screw) during fixation, thus enabling a more complete regrowth and ligamentization of the graft.7 The use of bioabsorbable fixation to attach soft tissue to bones is increasing. Bioabsorbable implants limit stress shielding of bone, gradually apply load as they degrade, obviate removal of hardware procedures, and facilitate postoperative radiological imaging. Graft fixation strength in ACL reconstruction is critical in the period from initial fixation to osseous integration of the graft.8 This period ranges from 6 weeks for patella tendon grafts, to approximately 16 weeks for hamstring fixation. This means that the biodegradable implant must maintain its structural integrity during the entire interval. Fixation devices were developed in biodegradable forms to address the problems associated with metal implants. These problems include artifact production on magnetic resonance imaging (MRI), cold intolerance, and difficulty with revision procedures, especially with hardware removal. However, concerns have been raised about the orthopaedic use of bioabsorbable implants, with specific complications including foreign-body reaction, cyst formation, fluid collection, sterile drainage, lack of complete osseous in growth into the defect left on degradation, early loss of pullout strength resulting from hydrolysis, and breakage of the implant.9,10 MRI is an excellent noninvasive, nonionizing imaging modality that affords excellent visualization of the tunnels, grafts, and fixation devices following ACL reconstruction.11-13 Our study used this investigative procedure to assess the TransFix (Arthrex, Naples, FL) transfemoral biodegradable implant used as a fixation device in ACL reconstructive surgery. The manufacturer states that these absorb between 3 and 5 years. The study group underwent opportunistic MRI
FIGURE 1.
TransFix implant.
scanning at different stages during their rehabilitation program as well clinical assessments to determine any specific problems associated with this form of fixation. Our aim was thus to determine if there were any problems associated with the use of a bioabsorbable transfemoral fixation device in tensioned hamstring reconstruction and, in particular, to assess its integrity throughout the period of graft incorporation. The hypothesis was that this type of graft fixation is as good as that with the metal cross pins from which they have evolved.14-18 METHODS Between December 2002 and December 2003, 49 patients under went ACL reconstruction performed by the senior author (A.J.S.) for knee instability. Twentynine patients were male and 20 were female. The average age of the patient at the time of surgery was 27 years (range 18 to 45 years). The average time from injury to surgery was 18 weeks (range, 1 to 43 weeks). Thirty-three patients were either elite or club standard athletes, with the remainder being recreational athletes or manual workers; 26 were left knees and 23 were right knees. The graft used in the reconstruction was a quadrupled semitendinosus tendon harvested through a 3-cm incision on the medial border of the tibial tuberosity. The graft was fixed proximally using the biodegradable TransFix implant, which is composed of poly-Llactic acid (Fig 1). This technique makes use of a Nitinol wire to pull the graft through the femoral tunnel, the graft being secured using a transfemoral approach utilizing a biodegradable intraosseous im-
EVALUATION OF BIODEGRADABLE TRANSFEMORAL FIXATION
201
Study approval was obtained from our hospital board and patients consented fully. All patients had regular clinical reviews at 2 weeks, 6 weeks, 3 months, 6 months, and 1 year postoperatively, assessment being conducted by the senior author using Lachman and pivot-shift tests as well as return to original sporting activity. RESULTS
FIGURE 2. Three-part fracture of TransFix implant. MRI revealed 5 fractured TransFix implants (out of 49) identified an average of 20 weeks postoperatively (range, 9 to 47 weeks). Two types of fracture patterns occurred to the implant. This image along with Figs 3 and 4 shows a 3-part fracture with the central third being disassociated from the proximal and distal fragment with no evidence of implant resorption. Only minor migration of fragments had occurred, with the fractured implant fragments remaining in close continuity with the graft.
plant. Distally, the graft was fixed using a suture-post construct (secure strand and Arthrex post), the graft was tensioned to 150 N using a pretensioning device (Smith & Nephew, Andover, MA) hooked to the distal secure strand suture intraoperatively. Postoperatively, all patients underwent an accelerated braceless rehabilitation program under the guidance of both a physiotherapist and the senior author. This is a functionally based program with an early aim of independent motion and weight bearing exercise, bike riding between 2 and 4 weeks, running performed at 6 to 12 weeks, and a return to sport at 6 months if predetermined goals are achieved by the patient. Open-chain quadriceps exercises (leg extension) were only carried out after 10 weeks to avoid anterior shear of the tibia and possible stretching of the graft. All patients were radiologically assessed using a single opportunistic MRI scan at different stages of their rehabilitation program with specific attention being focused on the transfemoral biodegradable implant and graft continuity. MRI was performed on a 1.5-T Echospeed with EXCITE system (General Electric, Milwaukee, WI). A quadrature knee coil was used with a field of view of 15 ⫻ 15 cm. Matrix was 384 ⫻ 224. Proton density (TR/TE 2700/30) and fat-suppressed T2 FSE (TR/TE 3600/47) sequences were performed (2 NEX). Slice thickness was 3.5 mm, and the interslice gap was 0.5 mm. All MRIs were independently assessed by a senior consultant radiologist.
The average time from surgery until MRI scan was 28 weeks (range, 4 to 54 weeks). MRI evaluation showed all 49 implants fully visible with no evidence of implant resorption. MRI also revealed 5 fractured TransFix implants identified at 9, 13, 16, 19, and 47 weeks (average, 20 weeks postoperatively; range, 9 to 47 weeks). Two types of fracture patterns occurred to the implant, a 2-part fracture with the fracture through the tip of the implant or a 3-part fracture with the central third being disassociated from the proximal and distal fragments (Figs 2-5). Only minor migration of fragments had occurred, with the fractured implant fragments remaining in close continuity with the graft. Three patients showed implant deformation but no fracture on MRI (Fig 6). These were identified at 12, 14, and 17 weeks postoperatively (average, 14 weeks postoperatively; range, 12 to 17 weeks). None of the knees was lost to clinical follow-up. The knees were all clinically stable at follow-up at each of their visits as assessed by Lachman and pivotshift testing (2 weeks, 6 weeks, 3 months, 6 months, and 1 year). No patient exceeded a Lachman ⫹⫹ and all pivot-shift tests were negative. All individuals had returned to their original sporting activity with no
FIGURE 3.
Three-part fracture of TransFix implant.
202
A. J. COSSEY ET AL.
FIGURE 4.
Three-part fracture of TransFix implant.
reported problems, including the patients with fractured/deformed implants. DISCUSSION Bioabsorbable poly-L-lactic acid was first used in orthopaedic surgery by Nakamura et al.19 in 1988 and Rozema et al.20 In 1995, Barber et al.21 published work relating to bioabsorbable implants in ACL reconstructive surgery. Since then, the development and commercial release of bioabsorbable implants in ACL reconstructive surgery have multiplied at the expense of metallic implants.21-25 However, bioabsorbable implants are not without complications.9,10 Over recent years, bioabsorbable implant fracture in ACL reconstructive surgery has been repeatedly reported in the
FIGURE 5. Two-part fracture of TransFix implant. This MRI shows the second type of fracture pattern observed, a 2-part fracture with the fracture through the tip of the implant.
FIGURE 6. Deformation of TransFix implant. Three of 49 implants showed implant deformation without a fracture on MRI. These were identified at 12, 14, and 17 weeks postoperatively (average, 14 weeks). These images show the potential for bioabsorbable transfemoral fixations to fracture or deform at a relatively early stage postoperatively (earliest observed was at 9 weeks) and may occur well in advance of the projected time of osseous integration of the graft of 16 weeks.
orthopaedic literature21,26 with the potential of fracture fragment migration and collateral damage to the articular surfaces within the knee.27,28 The orthopaedic literature relating to transfemoral biodegradable fixation is sparse.14-16 Numerous studies have warned of potential problems that may arise when early aggressive rehabilitation is conducted with transfemoral fixation devices. Concerns of fatigue fracture at the point where the graft goes over the implant were raised in Clark’s original work,4 which led to changes in the design of the implant. Biomechanical studies using bioabsorbable RigidFix23 (Mitek, Ethicon, Westwood, MA) and clinical studies using bioabsorbable TransFix17 showed increased residual displacement of the implant, with both studies recommending caution in the rehabilitation of patients with these implants. All our patients underwent early, aggressive rehabilitation programs with this regimen having the theoretical potential of causing the implant fractures discovered on MRI. The degradation rates of biodegradable implants vary depending on the implant design, composition, processing, enzyme concentrations, and local stresses at the operation site. Numerous studies using MRI as an imaging modality have quoted different rates of degradation ranging from 6 months to 5 years, the majority of these studies being conducted on bioab-
EVALUATION OF BIODEGRADABLE TRANSFEMORAL FIXATION sorbable interference screw implants.11-13 Our study showed no mass degradation of the TransFix implant on MRI, although one potential cause for the implant to fracture may have been the weakening of the implant during the degradation process. Theoretically, the implants may have become structurally weakened after a certain period in situ and, with continued stresses being applied, the potential to fracture is possible. This theory is supported by the deformations seen in 3 of the implants in our study. These deformations may be a prefracture phenomenon. ACL graft tensioning is a significant variable in ACL reconstruction.29,30 Factors that affect graft tension include graft type,31 tunnel placement,31,32 fixation method,32-34 and the angle of the knee at the time of fixation.35 The optimum graft tension required to produce functional stability after ACL reconstruction has not yet been determined despite numerous studies in cadaveric and animal models, although a general consensus does prevail that an undefined “wide safe window” exists between too loose and too tight.36 Our surgical technique used an intraoperative tensioning force of 150 N at 30° of flexion, which has the theoretical potential to cause intraoperative fracture of the implant. However, we believe this tension provides the graft with the best environment for the proper orientation of the newly formed collagen during the remodeling phase of the graft,36,37 especially when taking into account that initial tension decreases considerably (more than 60%) postoperatively due to the viscoelastic properties of the graft.38 Furthermore, during early rehabilitation, the graft fixation is subject to repetitive, submaximal loading as well as activities of daily living. It is estimated that the graft is loaded to approximately 150 to 500 N during these activities.39,40 Biomechanical theories concerning the different fracture patterns we encountered in this study may relate to eccentric loading of the implant. Nonconcentric loading of the device has the potential to propagate microfractures at the distal tunnel-bone interface, this coinciding with the implant taper. These forces are increased if the implant does not cross the bone tunnel and penetrate into the bone stock on the far side of the tunnel by a considerable extent. We hypothesize that once the distal fracture occurs, a significant increase in force is focused on the proximal implantbone interface, resulting in fracture propagation and the formation of a 3-part fracture. This study has shown the potential for bioabsorbable transfemoral fixations to fracture or deform at a relatively early stage postoperatively (the earliest ob-
203
served was 9 weeks postoperatively) and may occur well in advance of the projected time of osseous integration of the graft of 16 weeks4,5 in at least 10% of cases. With an early failure of the femoral fixation, the potential for graft failure is high. Fortunately, no problems have arisen within our study group in terms of stability of the reconstruction or associated collateral damage to the chondral surfaces of the knee at this stage postoperatively. However, the potential for migration of the fractured implants is still present, as has been shown in case reports such as ours prior to the instigation of this study.28 These patients will thus be kept under regular review. We have suggested 3 possible theories to account for the fractured/deformed implants: fracture at the time of intraoperative graft tensioning, fracture at the time of early accelerated physiotherapy, or the implants may have become structurally weakened after only a relatively short period in situ. Any of these theories would suggest that this form of fixation is inadequate in comparison with the metal cross pins, which have no reports of structural inadequacy. Unfortunately our study cannot support any one particular theory entirely; hence a prospective study has now been instigated with other types of bioabsorbable fixation devices in anticipation of providing answers to the cause for fractures of transfemoral bioabsorbable implants. CONCLUSIONS Transfemoral biodegradable implants have the potential to fracture or deform during their postoperative course in tensioned hamstring tendon ACL reconstruction. Although no apparent detrimental effect has been shown in our series, further research is needed on this device before it can be recommended for ACL reconstruction. We also question the idea that rigid fixation for the ACL graft for the entire healing period is required. REFERENCES 1. Otero A, Hutcheson L. A comparison of double semitendinosus/gracilis and central third of the patellar tendon autograft in arthroscopic ACL reconstruction. Arthroscopy 1993;9:143148. 2. Rodeo S, Izawa K. Tendon to bone healing. Tech Orthop 1999;14:22-33. 3. Robert H, Es-Sayeh J, Heyman D, Passuti N Eliot S, Vaneenoge E. Hamstring insertion site healing after ACL reconstruction in patients with symptomatic hardware or repeat rupture: An histological review in 12 patients. Arthroscopy 2003;19:948-954.
204
A. J. COSSEY ET AL.
4. Clark R, Olsen R, Larson B, Gobble E, Farrer R. Cross-pin femoral fixation: A new technique for hamstring ACL reconstruction of the knee. Arthroscopy 1998;14:258-267. 5. Kline J, Lintner D, Downs D, Vavrenka K. The incidence and significance of femoral tunnel widening after quadrupled hamstring ACL reconstruction using femoral cross-pin fixation. Arthroscopy 2003;19:470-476. 6. Mariani P, Camillieri G, Margheritini F. Transcondylar screw fixation in ACL reconstruction. Arthroscopy 2001;17:717-723. 7. Brown G, Pena F, Grontvedt T, Labadie D, Engebresten L. Fixation strength of interference screw fixation in bovine, young human and elderly human cadaver knees: Influence of insertion torque, tunnel bone gap and interference. Knee Surg Traumatol Arthrosc 1996;3:238-244. 8. Ciccone W, Motz C, Bentley C, Tasto J. Bioabsorbable implants in orthopaedics: New developments and clinical applications. J Am Academy Orthop Surg 2001;9:280-287. 9. McGuire D, Barber A, Milchgrub S, Wolchock J. A post mortem examination of poly-L-lactic acid interference screws four months after implantation during ACL reconstruction. Arthroscopy 2001;17:988-992. 10. Bostman O, Hirvensalo E, Makinen J, Rokkanen P. Foreign body reactions to fracture fixation implants of biodegradable synthetic polymers. J Bone Joint Surg Br 1990;72:592-596. 11. Bach F, Carlier R, Elis J, et al. ACL reconstruction with bioabsorbable polyglycolic acid interference screws: MRI follow-up. Radiology 2002;225:541-550. 12. Lajtai G, Schmiedhuber G, Unger F, et al. Bone tunnel remodeling at the site of biodegradable interference screws used for ACL reconstruction: Five year follow-up. Arthroscopy 2001; 17:597-602. 13. Warden W, Friedman R, Teresi L, Jackson W. MRI of bioabsorbable polylactic acid interference screws during the first two years after ACL reconstruction. Arthroscopy 1999;15:474480. 14. Milano G, Mulas PD, Ziranu F, Severini G. Anterior cruciate ligament reconstruction with doubled semitendinosus and gracilis tendon graft in rugby players. Knee Surg Sports Traumatol Arthrosc 2005;13:2-7. 15. Fabbriaciani C, Mulas PD, Ziranu F, Deriu L, Zarelli D, Milano G. Mechanical analysis of fixation methods for anterior cruciate ligament reconstruction with hamstring tendon graft. An experimental study in sheep knees. Knee 2005;12:135-138. 16. Milano G, Mulas PD, Sanna-Passino E, Careddu GM, Ziranu F, Fabbriciani C. Evaluation of bone plug and soft tissue anterior cruciate ligament graft over time using transverse femoral fixation in a sheep model. Arthroscopy 2005;21:532539. 17. Brand J, Weiler A, Carbon D. Graft fixation in cruciate ligament reconstruction. Am J Sports Med 2000;28:769-774. 18. Mariani P, Camillieri G, Margheritini F. Transcondylar screw fixation in ACL reconstruction. Arthroscopy 2001;17:717-723. 19. Nakamura T, Shimamoto T, Watanabe S. Uber die anwendung von in vivo zersetzbaren polymeren in der thoraxchirurgie. Markromol Chem Marcromol 1988;19:201-208. 20. Rozema F, Bos R, Boering G. Absorbable boneplates and screws for the fixation of malar fractures. Implant materials in biofunction. In: Putter C et al., eds. Advances in biomaterials. Amsterdam: Elsevier, 1998;251-255. 21. Barber F, Elrod B, McGuire D. Preliminary results of an absorbable interference screw. Arthroscopy 1995;11:537-548.
22. Kousa P, Jarvinen T, Kannus P, Jarvinen M. Initial fixation strength of bioabsorbable and titanium interference screws in ACL reconstruction. Am J Sports Med 2001;29:420-425. 23. Kousa P, Jarvinen T, Vihavainen M, Kannus P, Jarvinen M. Biomechanical evaluation of six different fixation devices for ACL hamstring grafts—Tibial site. RigidFix Surgical Manual. Westwood, MA: Mitek, Ethicon; 2003. 24. Rokkanen P, Bostman O, Hairvensalo E, et al. Bioabsorbable fixation in orthopaedic surgery and traumatology. Biomaterials 2000;21:2607-2613. 25. Fink C, Benedetto K, Hackl W, Hoser C, Freund M, Rieger M. Bioabsorbable polyglyconate interference screw fixation in an ACL reconstruction: A prospective computed tomography controlled study. Arthroscopy 2000;16:491-498. 26. Kotani A, Ishii Y. Reconstruction of the ACL using poly-Llactide interference screws or titanium screws: A comparative study. Knee 2001;8:311-331. 27. Macdonald P, Arneja S. Biodegradable screw presents as a loose intra-articular body after anterior cruciate ligament reconstruction. Arthroscopy 2003;19:E22-E24. 28. Cossey AJ, Paterson RS. Loose intra-articular body following ACL reconstruction. Arthroscopy 2005;21:348-350. 29. Boylan D, Greis P, West J, Kent B, Bachus N, Burks R. Effects of initial graft tension on knee stability after ACL reconstruction using hamstring tendons. Arthroscopy 2003;19:700-705. 30. Burks R, Lelan R. Determination of graft tension before fixation in ACL reconstruction. Arthroscopy 1988;4:260-266. 31. Markolf K, Burchfield D, Shapiro M. Biomechanical consequences of replacement of the ACL. J Bone Joint Surg Am 1996;78:1728-1734. 32. Howell S, Wallace M, Hull M. Evaluation of the single incision arthroscopic technique for ACL replacement: A study of tibial tunnel replacement, intraoperative graft tension and stability. Am J Sports Med 1999;27:284-293. 33. Giurea M, Zorilla P, Amis A. Comparative pull-out strength tests of anchorage of hamstring tendon in ACL reconstruction. Am J Sports Med 1999;27:621-625. 34. Hamner D, Brown C, Steiner M. Hamstring tendon grafts for reconstruction of the ACL: Biomechanical evaluation of the use of multiple strands of tensioning techniques. J Bone Joint Surg Am 1999;81:549-557. 35. Bylski-Austow D, Grood E, Hefzy M. ACL ligament replacements: A mechanical study of femoral attachment, flexion angle at tensioning and initial tension. J Orthop Res 1990;8: 522-531. 36. Van-Kampen A, Wymenga A, van der Heide H. The effect of different graft tensioning in ACL reconstruction: A prospective randomized study. Arthroscopy 1998;14:845-850. 37. Frank C, Amiel D, Akeson W. Healing of the medial collateral ligament of the knee: A morphological and biomechanical assessment in rabbits. Acta Orthop Scand 1983;54:917-923. 38. Nurmi J, Kannus P, Sievanen H, Jarvela T, Jarvinen M, Jarvinen T. Interference screw fixation of soft tissue grafts in ACL reconstruction. Am J Sports Med 2004;32:418-424. 39. Holden J, Grood E, Korvick D. In vivo forces in ACL reconstruction. Direct measurements during walking and trotting in a quadruped. J Biomech 1994;27:517-526. 40. Noyes F, Butler D, Grood E. Biomechanical analysis of human ligament grafts used in knee ligament repairs and reconstructions. J Bone Joint Surg Am 1984;66:344-352.
Purpose: The study was designed to evaluate bioabsorbable transfemoral fixation in anterior cruciate ligament (ACL) reconstruction using hamstring tendon as graft. Magnetic resonance imaging (MRI) was used to assess the continuity of the bioabsorbable implant at different stages of the patients’ rehabilitation. Type of Study: Retrospective case series. Methods: Forty-nine patients underwent ACL reconstruction performed by a single surgeon. The graft, a tensioned quadrupled semitendinosus tendon, was fixed proximally using a bioabsorbable TransFix implant (Arthrex, Naples, FL). The patients underwent an accelerated rehabilitation program and were assessed clinically at regular intervals postoperatively using MRI, with specific attention focused on the implant. Results: The average time from surgery to MRI was 28 weeks (range, 4 to 54 weeks). All implants were fully visible with no evidence of resorption. Five implants were fractured at an average of 20 weeks postoperatively (range, 9 to 47 weeks). Three implants showed deformation but no definite fracture at an average of 14 weeks (range, 12 to 17 weeks). This amounts to 16% of implants with fractures or deformation, many close to the period of theoretical graft incorporation. All patients were clinically stable with no symptoms or signs or instability on clinical review and all had returned to preinjury sporting activities. Conclusions: Transfemoral biodegradable implants have the potential to fracture or deform during their postoperative course in tensioned hamstring tendon ACL reconstruction. Although no apparent detrimental effect was found in our series, further research is needed on this device before it can be recommended for ACL reconstruction. We also question the idea that rigid fixation for the ACL graft for the entire healing process is required. Level of Evidence: Level IV, retrospective case series. Key Words: ACL reconstruction—Bioabsorbable implant—Transfemoral fixation—Knee—Magnetic resonance imaging.
H
amstring tendons are being used increasingly in anterior cruciate ligament (ACL) reconstruction with this technique claiming lower harvest-site morbidity when compared with patellar tendon grafts.1 With the continued advances in reconstructive techniques allowing for the early return of neuromuscular
From the Orthopaedic Division, SPORTSMED-SA, Adelaide, South Australia, Australia. Address correspondence and reprint requests to Yegappan Kalairajah, M.A., M.Phil., F.R.C.S.(Tr&Orth), Orthopaedic Division, SPORTSMED-SA, 32 Payneham Road, Adelaide, South Australia 5069, Australia. E-mail: [email protected] © 2006 by the Arthroscopy Association of North America 0749-8063/06/2202-05-1$32.00/0 doi:10.1016/j.arthro.2005.08.044
function, graft fixation has to withstand not only these early physiologic forces but also facilitate biological incorporation of the graft construct. With hamstring tendons, the histologic tendon-to-bone anchorage is not fully understood despite animal experimental2 and human studies.3 This constitutes a potential weak point of the technique2 with tendon-to-bone fixation being of the utmost importance for postoperative rehabilitation and resumption of athletic activities.3 ACL fixation methods have improved significantly over the past decade. Currently, many methods of fixation are available, including suture-post constructs, staples, soft-tissue washers, buttons, crosspins, and interference screws. Autograft hamstring ACL reconstruction with femoral cross-pin fixation is
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 22, No 2 (February), 2006: pp 199-204
199
200
A. J. COSSEY ET AL.
a technique popularized by the original work of Clark et al. in the 1990s.4 This study provided biomechanical evidence that femoral cross-pin fixation of the graft mimics the stress-strain relationship of the native ACL. Transfemoral cross-pin fixation using metal cross pins has been shown to have strength and stiffness superior to any other form of femoral fixation.5 The use of interference screw fixation can be associated with possible complications including screw divergence, intraoperative graft damage, bone regrowth within the tunnel, decreased space for biological fixation, and difficult hardware removal.6 Transfemoral fixation offers the theoretical advantage of an increased potential for biological fixation with a 360° bone-to-graft contact area and a decrease in graft damage (when compared with interference screw) during fixation, thus enabling a more complete regrowth and ligamentization of the graft.7 The use of bioabsorbable fixation to attach soft tissue to bones is increasing. Bioabsorbable implants limit stress shielding of bone, gradually apply load as they degrade, obviate removal of hardware procedures, and facilitate postoperative radiological imaging. Graft fixation strength in ACL reconstruction is critical in the period from initial fixation to osseous integration of the graft.8 This period ranges from 6 weeks for patella tendon grafts, to approximately 16 weeks for hamstring fixation. This means that the biodegradable implant must maintain its structural integrity during the entire interval. Fixation devices were developed in biodegradable forms to address the problems associated with metal implants. These problems include artifact production on magnetic resonance imaging (MRI), cold intolerance, and difficulty with revision procedures, especially with hardware removal. However, concerns have been raised about the orthopaedic use of bioabsorbable implants, with specific complications including foreign-body reaction, cyst formation, fluid collection, sterile drainage, lack of complete osseous in growth into the defect left on degradation, early loss of pullout strength resulting from hydrolysis, and breakage of the implant.9,10 MRI is an excellent noninvasive, nonionizing imaging modality that affords excellent visualization of the tunnels, grafts, and fixation devices following ACL reconstruction.11-13 Our study used this investigative procedure to assess the TransFix (Arthrex, Naples, FL) transfemoral biodegradable implant used as a fixation device in ACL reconstructive surgery. The manufacturer states that these absorb between 3 and 5 years. The study group underwent opportunistic MRI
FIGURE 1.
TransFix implant.
scanning at different stages during their rehabilitation program as well clinical assessments to determine any specific problems associated with this form of fixation. Our aim was thus to determine if there were any problems associated with the use of a bioabsorbable transfemoral fixation device in tensioned hamstring reconstruction and, in particular, to assess its integrity throughout the period of graft incorporation. The hypothesis was that this type of graft fixation is as good as that with the metal cross pins from which they have evolved.14-18 METHODS Between December 2002 and December 2003, 49 patients under went ACL reconstruction performed by the senior author (A.J.S.) for knee instability. Twentynine patients were male and 20 were female. The average age of the patient at the time of surgery was 27 years (range 18 to 45 years). The average time from injury to surgery was 18 weeks (range, 1 to 43 weeks). Thirty-three patients were either elite or club standard athletes, with the remainder being recreational athletes or manual workers; 26 were left knees and 23 were right knees. The graft used in the reconstruction was a quadrupled semitendinosus tendon harvested through a 3-cm incision on the medial border of the tibial tuberosity. The graft was fixed proximally using the biodegradable TransFix implant, which is composed of poly-Llactic acid (Fig 1). This technique makes use of a Nitinol wire to pull the graft through the femoral tunnel, the graft being secured using a transfemoral approach utilizing a biodegradable intraosseous im-
EVALUATION OF BIODEGRADABLE TRANSFEMORAL FIXATION
201
Study approval was obtained from our hospital board and patients consented fully. All patients had regular clinical reviews at 2 weeks, 6 weeks, 3 months, 6 months, and 1 year postoperatively, assessment being conducted by the senior author using Lachman and pivot-shift tests as well as return to original sporting activity. RESULTS
FIGURE 2. Three-part fracture of TransFix implant. MRI revealed 5 fractured TransFix implants (out of 49) identified an average of 20 weeks postoperatively (range, 9 to 47 weeks). Two types of fracture patterns occurred to the implant. This image along with Figs 3 and 4 shows a 3-part fracture with the central third being disassociated from the proximal and distal fragment with no evidence of implant resorption. Only minor migration of fragments had occurred, with the fractured implant fragments remaining in close continuity with the graft.
plant. Distally, the graft was fixed using a suture-post construct (secure strand and Arthrex post), the graft was tensioned to 150 N using a pretensioning device (Smith & Nephew, Andover, MA) hooked to the distal secure strand suture intraoperatively. Postoperatively, all patients underwent an accelerated braceless rehabilitation program under the guidance of both a physiotherapist and the senior author. This is a functionally based program with an early aim of independent motion and weight bearing exercise, bike riding between 2 and 4 weeks, running performed at 6 to 12 weeks, and a return to sport at 6 months if predetermined goals are achieved by the patient. Open-chain quadriceps exercises (leg extension) were only carried out after 10 weeks to avoid anterior shear of the tibia and possible stretching of the graft. All patients were radiologically assessed using a single opportunistic MRI scan at different stages of their rehabilitation program with specific attention being focused on the transfemoral biodegradable implant and graft continuity. MRI was performed on a 1.5-T Echospeed with EXCITE system (General Electric, Milwaukee, WI). A quadrature knee coil was used with a field of view of 15 ⫻ 15 cm. Matrix was 384 ⫻ 224. Proton density (TR/TE 2700/30) and fat-suppressed T2 FSE (TR/TE 3600/47) sequences were performed (2 NEX). Slice thickness was 3.5 mm, and the interslice gap was 0.5 mm. All MRIs were independently assessed by a senior consultant radiologist.
The average time from surgery until MRI scan was 28 weeks (range, 4 to 54 weeks). MRI evaluation showed all 49 implants fully visible with no evidence of implant resorption. MRI also revealed 5 fractured TransFix implants identified at 9, 13, 16, 19, and 47 weeks (average, 20 weeks postoperatively; range, 9 to 47 weeks). Two types of fracture patterns occurred to the implant, a 2-part fracture with the fracture through the tip of the implant or a 3-part fracture with the central third being disassociated from the proximal and distal fragments (Figs 2-5). Only minor migration of fragments had occurred, with the fractured implant fragments remaining in close continuity with the graft. Three patients showed implant deformation but no fracture on MRI (Fig 6). These were identified at 12, 14, and 17 weeks postoperatively (average, 14 weeks postoperatively; range, 12 to 17 weeks). None of the knees was lost to clinical follow-up. The knees were all clinically stable at follow-up at each of their visits as assessed by Lachman and pivotshift testing (2 weeks, 6 weeks, 3 months, 6 months, and 1 year). No patient exceeded a Lachman ⫹⫹ and all pivot-shift tests were negative. All individuals had returned to their original sporting activity with no
FIGURE 3.
Three-part fracture of TransFix implant.
202
A. J. COSSEY ET AL.
FIGURE 4.
Three-part fracture of TransFix implant.
reported problems, including the patients with fractured/deformed implants. DISCUSSION Bioabsorbable poly-L-lactic acid was first used in orthopaedic surgery by Nakamura et al.19 in 1988 and Rozema et al.20 In 1995, Barber et al.21 published work relating to bioabsorbable implants in ACL reconstructive surgery. Since then, the development and commercial release of bioabsorbable implants in ACL reconstructive surgery have multiplied at the expense of metallic implants.21-25 However, bioabsorbable implants are not without complications.9,10 Over recent years, bioabsorbable implant fracture in ACL reconstructive surgery has been repeatedly reported in the
FIGURE 5. Two-part fracture of TransFix implant. This MRI shows the second type of fracture pattern observed, a 2-part fracture with the fracture through the tip of the implant.
FIGURE 6. Deformation of TransFix implant. Three of 49 implants showed implant deformation without a fracture on MRI. These were identified at 12, 14, and 17 weeks postoperatively (average, 14 weeks). These images show the potential for bioabsorbable transfemoral fixations to fracture or deform at a relatively early stage postoperatively (earliest observed was at 9 weeks) and may occur well in advance of the projected time of osseous integration of the graft of 16 weeks.
orthopaedic literature21,26 with the potential of fracture fragment migration and collateral damage to the articular surfaces within the knee.27,28 The orthopaedic literature relating to transfemoral biodegradable fixation is sparse.14-16 Numerous studies have warned of potential problems that may arise when early aggressive rehabilitation is conducted with transfemoral fixation devices. Concerns of fatigue fracture at the point where the graft goes over the implant were raised in Clark’s original work,4 which led to changes in the design of the implant. Biomechanical studies using bioabsorbable RigidFix23 (Mitek, Ethicon, Westwood, MA) and clinical studies using bioabsorbable TransFix17 showed increased residual displacement of the implant, with both studies recommending caution in the rehabilitation of patients with these implants. All our patients underwent early, aggressive rehabilitation programs with this regimen having the theoretical potential of causing the implant fractures discovered on MRI. The degradation rates of biodegradable implants vary depending on the implant design, composition, processing, enzyme concentrations, and local stresses at the operation site. Numerous studies using MRI as an imaging modality have quoted different rates of degradation ranging from 6 months to 5 years, the majority of these studies being conducted on bioab-
EVALUATION OF BIODEGRADABLE TRANSFEMORAL FIXATION sorbable interference screw implants.11-13 Our study showed no mass degradation of the TransFix implant on MRI, although one potential cause for the implant to fracture may have been the weakening of the implant during the degradation process. Theoretically, the implants may have become structurally weakened after a certain period in situ and, with continued stresses being applied, the potential to fracture is possible. This theory is supported by the deformations seen in 3 of the implants in our study. These deformations may be a prefracture phenomenon. ACL graft tensioning is a significant variable in ACL reconstruction.29,30 Factors that affect graft tension include graft type,31 tunnel placement,31,32 fixation method,32-34 and the angle of the knee at the time of fixation.35 The optimum graft tension required to produce functional stability after ACL reconstruction has not yet been determined despite numerous studies in cadaveric and animal models, although a general consensus does prevail that an undefined “wide safe window” exists between too loose and too tight.36 Our surgical technique used an intraoperative tensioning force of 150 N at 30° of flexion, which has the theoretical potential to cause intraoperative fracture of the implant. However, we believe this tension provides the graft with the best environment for the proper orientation of the newly formed collagen during the remodeling phase of the graft,36,37 especially when taking into account that initial tension decreases considerably (more than 60%) postoperatively due to the viscoelastic properties of the graft.38 Furthermore, during early rehabilitation, the graft fixation is subject to repetitive, submaximal loading as well as activities of daily living. It is estimated that the graft is loaded to approximately 150 to 500 N during these activities.39,40 Biomechanical theories concerning the different fracture patterns we encountered in this study may relate to eccentric loading of the implant. Nonconcentric loading of the device has the potential to propagate microfractures at the distal tunnel-bone interface, this coinciding with the implant taper. These forces are increased if the implant does not cross the bone tunnel and penetrate into the bone stock on the far side of the tunnel by a considerable extent. We hypothesize that once the distal fracture occurs, a significant increase in force is focused on the proximal implantbone interface, resulting in fracture propagation and the formation of a 3-part fracture. This study has shown the potential for bioabsorbable transfemoral fixations to fracture or deform at a relatively early stage postoperatively (the earliest ob-
203
served was 9 weeks postoperatively) and may occur well in advance of the projected time of osseous integration of the graft of 16 weeks4,5 in at least 10% of cases. With an early failure of the femoral fixation, the potential for graft failure is high. Fortunately, no problems have arisen within our study group in terms of stability of the reconstruction or associated collateral damage to the chondral surfaces of the knee at this stage postoperatively. However, the potential for migration of the fractured implants is still present, as has been shown in case reports such as ours prior to the instigation of this study.28 These patients will thus be kept under regular review. We have suggested 3 possible theories to account for the fractured/deformed implants: fracture at the time of intraoperative graft tensioning, fracture at the time of early accelerated physiotherapy, or the implants may have become structurally weakened after only a relatively short period in situ. Any of these theories would suggest that this form of fixation is inadequate in comparison with the metal cross pins, which have no reports of structural inadequacy. Unfortunately our study cannot support any one particular theory entirely; hence a prospective study has now been instigated with other types of bioabsorbable fixation devices in anticipation of providing answers to the cause for fractures of transfemoral bioabsorbable implants. CONCLUSIONS Transfemoral biodegradable implants have the potential to fracture or deform during their postoperative course in tensioned hamstring tendon ACL reconstruction. Although no apparent detrimental effect has been shown in our series, further research is needed on this device before it can be recommended for ACL reconstruction. We also question the idea that rigid fixation for the ACL graft for the entire healing period is required. REFERENCES 1. Otero A, Hutcheson L. A comparison of double semitendinosus/gracilis and central third of the patellar tendon autograft in arthroscopic ACL reconstruction. Arthroscopy 1993;9:143148. 2. Rodeo S, Izawa K. Tendon to bone healing. Tech Orthop 1999;14:22-33. 3. Robert H, Es-Sayeh J, Heyman D, Passuti N Eliot S, Vaneenoge E. Hamstring insertion site healing after ACL reconstruction in patients with symptomatic hardware or repeat rupture: An histological review in 12 patients. Arthroscopy 2003;19:948-954.
204
A. J. COSSEY ET AL.
4. Clark R, Olsen R, Larson B, Gobble E, Farrer R. Cross-pin femoral fixation: A new technique for hamstring ACL reconstruction of the knee. Arthroscopy 1998;14:258-267. 5. Kline J, Lintner D, Downs D, Vavrenka K. The incidence and significance of femoral tunnel widening after quadrupled hamstring ACL reconstruction using femoral cross-pin fixation. Arthroscopy 2003;19:470-476. 6. Mariani P, Camillieri G, Margheritini F. Transcondylar screw fixation in ACL reconstruction. Arthroscopy 2001;17:717-723. 7. Brown G, Pena F, Grontvedt T, Labadie D, Engebresten L. Fixation strength of interference screw fixation in bovine, young human and elderly human cadaver knees: Influence of insertion torque, tunnel bone gap and interference. Knee Surg Traumatol Arthrosc 1996;3:238-244. 8. Ciccone W, Motz C, Bentley C, Tasto J. Bioabsorbable implants in orthopaedics: New developments and clinical applications. J Am Academy Orthop Surg 2001;9:280-287. 9. McGuire D, Barber A, Milchgrub S, Wolchock J. A post mortem examination of poly-L-lactic acid interference screws four months after implantation during ACL reconstruction. Arthroscopy 2001;17:988-992. 10. Bostman O, Hirvensalo E, Makinen J, Rokkanen P. Foreign body reactions to fracture fixation implants of biodegradable synthetic polymers. J Bone Joint Surg Br 1990;72:592-596. 11. Bach F, Carlier R, Elis J, et al. ACL reconstruction with bioabsorbable polyglycolic acid interference screws: MRI follow-up. Radiology 2002;225:541-550. 12. Lajtai G, Schmiedhuber G, Unger F, et al. Bone tunnel remodeling at the site of biodegradable interference screws used for ACL reconstruction: Five year follow-up. Arthroscopy 2001; 17:597-602. 13. Warden W, Friedman R, Teresi L, Jackson W. MRI of bioabsorbable polylactic acid interference screws during the first two years after ACL reconstruction. Arthroscopy 1999;15:474480. 14. Milano G, Mulas PD, Ziranu F, Severini G. Anterior cruciate ligament reconstruction with doubled semitendinosus and gracilis tendon graft in rugby players. Knee Surg Sports Traumatol Arthrosc 2005;13:2-7. 15. Fabbriaciani C, Mulas PD, Ziranu F, Deriu L, Zarelli D, Milano G. Mechanical analysis of fixation methods for anterior cruciate ligament reconstruction with hamstring tendon graft. An experimental study in sheep knees. Knee 2005;12:135-138. 16. Milano G, Mulas PD, Sanna-Passino E, Careddu GM, Ziranu F, Fabbriciani C. Evaluation of bone plug and soft tissue anterior cruciate ligament graft over time using transverse femoral fixation in a sheep model. Arthroscopy 2005;21:532539. 17. Brand J, Weiler A, Carbon D. Graft fixation in cruciate ligament reconstruction. Am J Sports Med 2000;28:769-774. 18. Mariani P, Camillieri G, Margheritini F. Transcondylar screw fixation in ACL reconstruction. Arthroscopy 2001;17:717-723. 19. Nakamura T, Shimamoto T, Watanabe S. Uber die anwendung von in vivo zersetzbaren polymeren in der thoraxchirurgie. Markromol Chem Marcromol 1988;19:201-208. 20. Rozema F, Bos R, Boering G. Absorbable boneplates and screws for the fixation of malar fractures. Implant materials in biofunction. In: Putter C et al., eds. Advances in biomaterials. Amsterdam: Elsevier, 1998;251-255. 21. Barber F, Elrod B, McGuire D. Preliminary results of an absorbable interference screw. Arthroscopy 1995;11:537-548.
22. Kousa P, Jarvinen T, Kannus P, Jarvinen M. Initial fixation strength of bioabsorbable and titanium interference screws in ACL reconstruction. Am J Sports Med 2001;29:420-425. 23. Kousa P, Jarvinen T, Vihavainen M, Kannus P, Jarvinen M. Biomechanical evaluation of six different fixation devices for ACL hamstring grafts—Tibial site. RigidFix Surgical Manual. Westwood, MA: Mitek, Ethicon; 2003. 24. Rokkanen P, Bostman O, Hairvensalo E, et al. Bioabsorbable fixation in orthopaedic surgery and traumatology. Biomaterials 2000;21:2607-2613. 25. Fink C, Benedetto K, Hackl W, Hoser C, Freund M, Rieger M. Bioabsorbable polyglyconate interference screw fixation in an ACL reconstruction: A prospective computed tomography controlled study. Arthroscopy 2000;16:491-498. 26. Kotani A, Ishii Y. Reconstruction of the ACL using poly-Llactide interference screws or titanium screws: A comparative study. Knee 2001;8:311-331. 27. Macdonald P, Arneja S. Biodegradable screw presents as a loose intra-articular body after anterior cruciate ligament reconstruction. Arthroscopy 2003;19:E22-E24. 28. Cossey AJ, Paterson RS. Loose intra-articular body following ACL reconstruction. Arthroscopy 2005;21:348-350. 29. Boylan D, Greis P, West J, Kent B, Bachus N, Burks R. Effects of initial graft tension on knee stability after ACL reconstruction using hamstring tendons. Arthroscopy 2003;19:700-705. 30. Burks R, Lelan R. Determination of graft tension before fixation in ACL reconstruction. Arthroscopy 1988;4:260-266. 31. Markolf K, Burchfield D, Shapiro M. Biomechanical consequences of replacement of the ACL. J Bone Joint Surg Am 1996;78:1728-1734. 32. Howell S, Wallace M, Hull M. Evaluation of the single incision arthroscopic technique for ACL replacement: A study of tibial tunnel replacement, intraoperative graft tension and stability. Am J Sports Med 1999;27:284-293. 33. Giurea M, Zorilla P, Amis A. Comparative pull-out strength tests of anchorage of hamstring tendon in ACL reconstruction. Am J Sports Med 1999;27:621-625. 34. Hamner D, Brown C, Steiner M. Hamstring tendon grafts for reconstruction of the ACL: Biomechanical evaluation of the use of multiple strands of tensioning techniques. J Bone Joint Surg Am 1999;81:549-557. 35. Bylski-Austow D, Grood E, Hefzy M. ACL ligament replacements: A mechanical study of femoral attachment, flexion angle at tensioning and initial tension. J Orthop Res 1990;8: 522-531. 36. Van-Kampen A, Wymenga A, van der Heide H. The effect of different graft tensioning in ACL reconstruction: A prospective randomized study. Arthroscopy 1998;14:845-850. 37. Frank C, Amiel D, Akeson W. Healing of the medial collateral ligament of the knee: A morphological and biomechanical assessment in rabbits. Acta Orthop Scand 1983;54:917-923. 38. Nurmi J, Kannus P, Sievanen H, Jarvela T, Jarvinen M, Jarvinen T. Interference screw fixation of soft tissue grafts in ACL reconstruction. Am J Sports Med 2004;32:418-424. 39. Holden J, Grood E, Korvick D. In vivo forces in ACL reconstruction. Direct measurements during walking and trotting in a quadruped. J Biomech 1994;27:517-526. 40. Noyes F, Butler D, Grood E. Biomechanical analysis of human ligament grafts used in knee ligament repairs and reconstructions. J Bone Joint Surg Am 1984;66:344-352.