- Email: [email protected]
Flipped patellar tendon autograft anterior cruciate ligament reconstruction
Flipped Patellar Tendon Autograft Anterior Cruciate Ligament Reconstruction F. Alan Barber, M.D.
Summary: To determine the efficacy of an anterior cruciate ligament (ACL) graft that customizes length and facilitates anatomic outlet fixation, a prospective study of the ‘‘flipped’’ patellar tendon autograft ACL reconstruction began in 1995. This technique shortens the tendon portion to match the intra-articular length by rotating 1 bone plug 180° proximally onto the tendon, thus flipping the bone plug over its ligamentous insertion. Bioscrews (poly L-lactic acid; Linvatec, Largo, FL) secured the grafts. All patients undergoing this procedure with a minimum 21 months follow-up were reviewed. Preoperative and postoperative Tegner, Lysholm, and IKDC activity scores, and Lachman and pivot shift tests were obtained. Postoperative KT testing and radiographs were obtained. Fifty patients were followed-up for an average of 28 months (range, 21 to 39 months). Average patient age was 34 years (range, 16 to 52 years). Tegner scores increased from 2.0 preoperatively to 6.0 postoperatively. Lysholm scores increased from 46 preoperatively to 93 at follow-up, with 86% excellent (66%) or good (20%). IKDC activity scores increased from 3.1 preoperatively to 1.7 postoperatively. KT manual-maximum difference at follow-up averaged 0.7 mm, with 74% less than 3-mm, 18% 3- to 5-mm, and 8% greater than 5-mm difference. Postoperative Lachman results were 0 in 45 patients and 1⫹ in 5 patients. Postoperative pivot shift was absent in all but 1 patient. Full extension was achieved in all cases and flexion averaged 136° with no patient having less than 120° flexion. No lytic bone changes or tunnel widening were seen. The flipped patellar tendon autograft reduces graft length to its intra-articular portion, increasing graft stability, isometry, and stiffness, and avoiding tunnel graft mismatch with clinically excellent results. Key Words: ACL reconstruction—Flip—Patellar tendon autograft—Interference fixation—Bioscrew—Isometric—Tunnel mismatch.
T
he success of an arthroscopically assisted patellar tendon autograft (PTA) anterior cruciate ligament (ACL) reconstruction is influenced by many factors including surgical technique, graft strength, and fixation.1 The inherent strength and the availability of the PTA plus its secure fixation to the attached bone plugs in the tibial and femoral tunnels make it a commonly used procedure.2-5 The ability to use interference screw fixation with this graft is attractive.6,7 Biodegradable interference fixation screws avoid revision and imag-
From the Plano Orthopedic and Sports Medicine Center, Plano, Texas, U.S.A. Address correspondence and reprint requests to F. Alan Barber, M.D., Plano Orthopedic and Sports Medicine Center, 5228 West Plano Pkwy, Plano, TX 75093, U.S.A. r 2000 by the Arthroscopy Association of North America 0749-8063/00/1605-2169$3.00/0 doi:10.1053/jars.2000.4384
ing problems and are less likely to cause graft damage than are metal implants.8 Morgan et al.9 recently described a technique of flipping one end of the PTA, which permits selection of a precise graft length, eliminates graft tunnel mismatch (both length and diameter), and allows for graft fixation at both femoral and tibial outlets. Other techniques have been described to accomplish this goal, but they require additional graft material.10 The common thread is graft fixation at the femoral and tibial outlets, which eliminates graft micromotion in the tunnels leading to tunnel widening or fixation failure. Reducing the graft length to its intra-articular portion increases graft stability and graft isometry, minimizes the potential for creep failure within the tunnel, and increases graft stiffness.11,12 The purpose of this study was to determine the clinical efficacy of this flip PTA technique for ACL reconstruction.
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 16, No 5 (July-August), 2000: pp 483–490
483
484
F. A. BARBER METHODS
A prospective study of flipped PTA ACL reconstructions secured by Bioscrew biodegradable interference screws (Linvatec, Largo, FL) started in June 1995. The inclusion criteria consisted of a complete ACL tear with knee instability demonstrated by both positive Lachman and positive pivot shift tests, adequate bone density, a PTA with a tendon segment measuring at least 4.5 cm, and a commitment to return for at least 2 years of follow-up. Exclusionary criteria included prior knee infection, concurrent multiple ligament reconstruction, prior ACL reconstruction, and a torn posterior cruciate ligament. Preoperative assessment included medical history, physical examination, radiographs, and preoperative Lysholm,13 Tegner,14 and IKDC activity scales. Baseline KT measurements and manual-maximum side-toside difference were obtained. The surgical findings and procedures performed were recorded. All patients underwent the same procedure: an arthroscopically assisted ACL reconstruction using a flipped PTA as described by Morgan et al.9 (Fig 1) fixed by Bioscrew interference screw fixation at both the proximal and distal graft sites. Postoperative evaluations were obtained at intervals of 3, 6, 12, and 24 months, and yearly thereafter. These assessments included the Lysholm, Tegner, and IKDC scales; meniscal tests; ligament tests; KT tests; and radiographs. The biodegradable interference screws made from purified poly L-lactide (containing approximately 45% crystalline polymer) were used for all reconstructions. The femoral biodegradable screw is placed through a separate portal in the patellar tendon chosen to match the angle of the femoral tunnel. The cannulated screw is advanced over the nitinol guide wire, which pre-
FIGURE 1. The flipped PTA is maintained under tension on the work station after preparation until insertion.
vents divergence, and is then inserted by a triangular fluted screwdriver. For these cases, screws measuring 8 ⫻ 20 mm were the most commonly selected devices. The 8-mm diameter screws were chosen because of concerns about breakage with 7-mm screws and the personal observation that 9-mm screws can push the bone plug into the adjacent cancellous bone, especially in the tibia. There were no cases of screw breakage in this series. This may be due to the larger core diameter of the 8-mm screw and customizing of the screw insertion portal to precisely parallel the angle of the femoral tunnel. Surgical Technique The distinctive feature of this technique is how the graft is prepared. The PTA is harvested through an anterior longitudinal incision removing a 10-mm wide strip of central patellar tendon with attached bone plugs from the tibia and patella that measure 10 ⫻ 25 mm and 10 mm deep. After determining that the tendinous portion of the graft is at least 4.5 cm long, the tibial bone segment is trimmed to allow it to be flipped back over the adjacent patellar tendon and reduced to 2 cm in length. Flipping the bone plug rotates it so that the cancellous portion faces away from the tendon. In this manner, interference screw fixation can be achieved between the cancellous bone of the plug and the cancellous bone of the tunnel wall. This also provides an adequate amount of bone to achieve a good interference fixation.15 If the bone plug were rotated the other way (to place the cancellous portion against the tendon), the screw fixation would be against the tendon-covered cortical side of the bone plug. While recent data suggest that the load-to-failure strength of this construct is no different from that of the tendon-tendon-bone arrangement, the method of failure is by pullout rather than by tendon stripping.16 Once properly trimmed, the flipped bone plug is fixed to the tendon by several encircling sutures of No. 1 PDS. Two sutures of No. 5 braided polyester are vertically placed through the flipped bone plug and tendon to achieve control for insertion and tensioning. Once prepared, the construct is kept tensioned on the workstation (Arthrex, Naples, FL) until inserted (Fig 1). Tensioning is applied by manually pushing the holding posts apart with the thumb and finger. Tibial and femoral tunnels are prepared with the same technique regardless of whether or not a flipped graft is used. The tibial tunnel is initially drilled to a 10-mm diameter and enlarged if necessary to be the same size as the flipped portion of the graft. Oversizing is not necessary for the biodegradable screw.
FLIPPED PATELLA TENDON AUTOGRAFT Selection of the graft as a suitable candidate for flipping is based on several criteria. Principal among these is the intra-articular distance from the femoral outlet to the tibial outlet. This distance usually ranges from 25 to 30 mm, but can be as little as 21 to 33 mm.17 Some femoral aiming devices feature a scale on the shaft that allows for a direct measurement associated with the femoral guide wire insertion. Another criterion is patellar tendon length. This graft can be constructed only if the tendon section is at least 4.5 cm long. Although shorter patients may require shorter intra-articular grafts, longer patellar tendons certainly facilitate this process. This technique is not advised for the graft that has a tendon shorter than 4.5 cm. With knowledge of the precise intraarticular length, the graft is placed on the workstation and the bone plugs are crafted. A squared off shape at both ends of the tibial bone plug is created with an ideal bone plug length of 2 cm.15 The square end allows for better rotation of the bone plug and a more intimate fit with of the bone adjacent to the tendon. Remember that the tendon must be long enough to provide for the measured intra-articular graft length and to accommodate the 2-cm bone plug. Postoperative Program In this series, the postoperative rehabilitation program was no different from that used for unflipped PTAs. Specifically, in the first days and weeks, achieving full extension was emphasized. Progressive weight bearing as tolerated was permitted and half of these patients were off crutches by 1 week. All were walking without crutches 2 weeks after surgery. A fullextension postoperative brace was used at night during the first 2 weeks, and continuous passive motion was used for 6 to 8 hours a day until at least 90° flexion was achieved. Progressive closed-chain exercises were started 1 to 2 weeks after surgery. These closed-chain exercises included stationary biking, stair climbing, and cross-country ski machines. Straight ahead, halfspeed running was started between 6 and 8 weeks after surgery. Pivoting, noncontact sports were initiated at 12 weeks, and a derotational brace was ordered. Once the brace was received, the patients were allowed to participate in full unlimited contact sports using the brace. Open-chain exercises against resistance in the terminal 30° of extension were avoided until 6 months after surgery. RESULTS During the study period from June 1995 through December 1996, a total of 55 knees underwent ACL
485
reconstruction using the flip technique. There were 50 knees in 50 patients (91%) available for long-term follow-up; these are the subjects of this report. Follow-up data are complete for all included patients with a minimum of 21 months follow-up (average, 28 months; range, 21 to 39 months). The average age was 34 years (range, 16 to 52 years). There were 35 male and 15 female patients. Reconstructions were performed in 21 right knees and 29 left knees. There is no indication that the 5 patients unavailable for the minimum 21-month follow-up had an outcome that was different from that of the group as a whole. Associated surgical findings (e.g., chondromalacia, meniscal tears) were reviewed. There were 18 fullthickness medial meniscus tears requiring treatment (2 repaired, 16 debrided). There were 20 lateral meniscus tears requiring treatment (none were repairs). Meniscal injury that did not require a specific treatment was not recorded. Femoral condyle damage was observed in 11 patients, patellar chondromalacia in 4, and medial tibial plateau chondral damage in 1 patient. Loose bodies were removed from 5 knees, 3 of which also had medial femoral chondromalacia. The time from injury to surgery was reviewed. There were 12 knees considered to be chronic (operated on more than 12 weeks after injury). No correlation between the length of time from injury and the incidence of meniscal tears was found; however, meniscal repairs were only performed in patients with recent injuries. Patients were asked for both preinjury and preoperative postinjury Tegner scores. The average preinjury score was 6.3 (range, 1-9). The average preoperative Tegner score was 2.0 (range, 0-6). The average postoperative Tegner score was 6.0 (range, 1-10) at an average follow-up of 28 months. The average preoperative Lysholm score was 45.8 (range, 13-93) and increased to 93.3 (range, 65-100) at an average follow-up of 28 months. The follow-up Lysholm scores showed excellent results (95-100) in 33 patients (66%) and good results (84-94) in 10 patients (20%). The IKDC form has an activity scale that classifies athletic activity into 4 groups, ranging from sedentary (housework, activities of daily living), to light activity (jogging, running), to moderate activity (skiing, tennis), to strenuous activity (jumping, pivoting, hard cutting). These 4 groupings are assigned a numerical score from the most strenuous (level 1) to the sedentary (level 4). The IKDC activity score was recorded preoperatively and averaged 3.1. This rank falls just below the light activity level. The IKDC activity level
486
F. A. BARBER
FIGURE 2. IKDC scores improved from the preoperative to postoperative measurements.
improved postoperatively to 2.3 at 14 months average follow-up and to 1.7 at 28 months follow-up (Fig 2). At follow-up, KT manual-maximum differences between both knees averaged 0.7 mm. The results for all knees at the 2-year follow-up were tabulated into 3 ranges: less than 3-mm difference (74%), 3- to 5-mm difference (18%), and greater than 5-mm difference (8%). Supine goniometer-measured motion showed no loss of extension in any patient at the 2-year follow-up. Every patient achieved at least 120° of flexion and the average flexion at the 2-year follow-up was 136° (range, 120° to 145°). The preoperative Lachman test results averaged 2.2 (range, 1⫹ to 4⫹). The 1-year postoperative Lachman test results were 0 in 48 patients and 1⫹ in 2 patients. At final follow-up, the Lachman test results were 0 in 45 patients and 1⫹ in 5. Preoperatively, the pivot shift
was 1⫹ in 10 patients and 2⫹ or greater in 40 patients. At follow-up, the pivot shift was 0 in all but 1 patient. There were no problems resulting from the use of the biodegradable interference screw. Specifically, no instances of screw breakage or graft laceration occurred. There were no joint effusions at follow-up. The immediate postoperative radiographs allowed confirmation of tunnel placement and bone plug location (Fig 3). Radiographs did not show any lytic changes or tunnel widening at the biodegradable interference screw insertion sites. Rapid bone plug incorporation and graft maturation were consistently observed (Fig 4). The use of the biodegradable interference screw permitted clear radiographic assessments of the healing PTA and an evaluation of the tunnel placement. Over time, the PTA and tunnels were difficult to view on radiographs (Fig 5) because of the rapid and progressive healing between the bone graft and the adjacent tunnel. DISCUSSION The flipped PTA offers the advantages of outlet graft fixation with the strength and availability of a patellar tendon. Outlet fixation eliminates concerns about micromotion of the tendinous portion of the graft in the tunnel. Concerns about tunnel widening and subsequent fixation failure are consequently lessened. A
FIGURE 3. (A) Anterior and (B) lateral radiographs: The Bioscrew permits a clear radiographic assessment of both graft and tunnel placement in the immediate postoperative (1 week) period, which can serve as a baseline for comparison with subsequent radiographs. The tibial bone plug (arrow) reaches the tunnel outlet in the tibial spine.
FLIPPED PATELLA TENDON AUTOGRAFT
FIGURE 4. At 2 years, the radiographs show progressive incorporation of the PTA without tunnel widening or lytic change. Rapid bone plug incorporation and graft maturation were consistently observed.
graft with a tendon portion precisely matching the intra-articular length, and with a bone plug–interference screw–tendon mass that eliminates tunnel diameter mismatch by completely filling the tibial and femoral tunnel outlets, should avoid graft movement problems (a possible cause of the windshield wiper effect).9 Plugging the tibial outlet will also prevent synovial fluid leaking into the tibial tunnel, which may be another cause of tunnel widening and lytic change. Fixation of this shorter intra-articular graft at the anatomic ACL insertion outlet for the femur prevents graft abrasion by the anterior edge of the femoral tunnel18 and provides optimal strain patterns that mimic normal ACL isometry.12 Outlet fixation at the tibia minimizes creep failure that can occur within the tunnel because the points of graft fixation are moved away from the anatomic origin and insertion.9 A
487
fixation position more distal in the tibial tunnel can increase the elongation forces on the graft as the knee goes through a range of motion.9 Ishibashi et al.11 have shown that proximal tibial tunnel attachment provides the most stable knee reconstruction, with less tibial displacement and axial rotation than these distal placements in the tunnel. Presuming that the elastic modulus of the graft is constant along its length, increasing the effective graft length by more distal tibial tunnel fixation decreases graft stiffness.11 Consequently, proximal attachment at the tibial ACL outlet minimizes the functional length and maximizes the graft stiffness. PTA mismatch can occur not only between the tendon portion of the graft and tunnel diameter, but also between the length of the graft and the length of the tunnel.17 The immediate advantage of the flipped PTA is the technical facility to deal with a patellar tendon that is too long or with a tibial tunnel inadvertently placed too high on the tibia. Either of these circumstances may result in a tendon mismatched to the tunnel. The hang-out situation has been reported to occur in 25%17 or more19 of cases. The flip eliminates this with the additional bonus of also eliminating any windshield wiper effect because the graft now fully fills the tunnels where it resides and provides the biomechanical advantages of a stiffer graft with less possibility of creep. Minor variations in the alignment of an endoscopic drill guide can affect the isometricity of the graft location. The ideal isometric site for the tibial fibers is the anteromedial area of the tibial tunnel.20 Unfortunately, the standard PTA places the fibers at the posterior portion of this area. When the graft is twisted, the fibers may be better oriented. The flip technique not only achieves outlet fixation, but the tendon fibers are reversed and located in the desired anteromedial position without twisting. An additional advantage of the flipped graft is its mechanism of failure. In general, graft failure can occur by bone plug fracture, bone plug pullout from the interference fixation screw, or tendon rupture. The mode of failure for interference screw fixation with the standard PTA is influenced by the screw placement. When the screw is placed on the cancellous side of the bone plug, the most common mode of failure is bone plug pullout.21 In contrast, when the screw is placed on the cortical side of the bone plug, the most common mode of failure is by tendon rupture or bone plug fracture.22 Both of these situations are all-or-nothing events. As excessive load is placed on the flipped graft, failure occurs by the stripping of the tendon from the
488
F. A. BARBER
FIGURE 5. (A) Anterior and (B) lateral radiographs 3 years postoperatively show no evidence of any localized response radiologically to the Bioscrew, and cancellous material completely fills the tunnel sites.
rotated bone plug.16 This may actually provide a safety valve for overconstrained fixation or for tunnels placed in less than ideal sites, because the graft will not fail all at once. Instead, it will stretch out incrementally and reach a steady state in which the elongation force is reduced by the stretching graft to a level below that able to damage the ligament further. The length of the tendon portion of these grafts was selected to match the intra-articular distance between the femoral and tibial tunnel outlets measured by the transtibial guide. The intra-articular length of the ACL is reported to be about 24 mm in fresh-frozen cadaver specimens19,23 and 26 mm in patients undergoing ACL reconstruction (range, 21 to 33 mm).17 The length of the patellar tendon measures from 33 to 63 mm long with an average length variously reported as 43 to 48 mm.17,19 When the tendon portion of the harvested PTA is at least 45 mm long, the graft can be flipped. With a 2-cm bone plug, this construct produces an
intra-articular graft length of 25 mm. The femoral bone can be whatever length the surgeon prefers; I usually use one that is at least 20 mm long. The tibial plug for the flipped PTA should be 20 mm as well. This length is sufficient for secure graft fixation.24 The interference fixation screw is made of poly L-lactide and requires 3 to 4 years to completely biodegrade. Biodegradable screws have been shown clinically8 and in vitro21 to provide fixation equivalent to metal interference fixation screws for unflipped PTA ACL reconstructions. The biodegradable screw also works well with the flipped PTA, avoiding problems with graft laceration found with metal screws and allowing excellent postsurgical imaging, which facilitates assessment of graft location. The screw at the tunnel outlet provides the compression that allows rapid healing of the graft into the tunnel.24 Because of its poly L-lactide composition, it does not cause the intense inflammatory reactions that can lead to lytic
FLIPPED PATELLA TENDON AUTOGRAFT areas in the bone or draining wounds reported with polyglycolic acid or polyglycolic acid copolymers as early as 8 weeks after implantation.25-27 Serial radiographic examinations showed progressive autograft incorporation into the bone tunnels without adverse occurrences (Figs 3-5). There was no evidence of tunnel widening in any of the follow-up radiographs. No evidence of osteolysis or other osseous reaction was found by magnetic resonance imaging (Fig 6). The poly L-lactide polymer is less rigid and more malleable than metal. It will conform to and follow the tunnel rather than cut a new direction and will not diverge during insertion. Biodegradable materials reduce problems with postsurgical imaging and minimize difficulties that may be encountered during revision procedures or with subsequent internal fixation around the knee.28 This prospective evaluation of the flipped PTA ACL reconstruction shows improved Tegner and Lysholm
489
scores with 86% excellent or good results by Lysholm scoring; an average KT manual-maximum difference of 0.7 mm 28 months after surgery, and with 74% of the patients having a KT manual-maximum difference of less than 3 mm. Recent data show that this technique provides fixation strengths that are comparable to conventional means, whether secured by metal or biodegradable screws. In addition, in vitro graft failure of this technique is by tendon stripping from its ligamentous attachments, which may provide a self-adjusting effect if the initial fixation is overconstrained. In conclusion, the flipped PTA permits selection of a precise graft length, eliminates the problem of too long a graft hanging out the anterior tibia, and allows for fixation of the graft at both femoral and tibial outlets. Graft fixation at the femoral and tibial outlets eliminates graft micromotion in the tunnels, graft tunnel size mismatch, and tunnel widening or fixation failure.
FIGURE 6. (A) Anterior and (B) lateral magnetic resonance imaging scans 3 years after surgery show no localized inflammatory response, no osteolysis nor other osseous reaction, and good graft incorporation.
490
F. A. BARBER
Reducing the graft length to its intra-articular portion increases graft stability and graft isometry, minimizes the potential for creep failure within the tunnel, and increases graft stiffness. The clinical results of the flipped PTA are excellent and this technique offers many advantages. Acknowledgment: The author appreciates the assistance of Nancy Alexander with data collection and the surgical assistance of James N. Click, P.A.-C., and Brian Howard, C.S.T., in this study.
REFERENCES 1. Paulos L, Noyes FR, Grood E, Butler DL. Knee rehabilitation after anterior cruciate ligament reconstruction and repair. Am J Sports Med 1981;9:140-149. 2. Brostrom L, Gillquist J, Liljedahl S-O, Lindvall N. Behandling av invetererad ruptur av framre korsbandet. Lakrtidn 1968;64: 44-77. 3. Gillquist J, Liljedahl S-O, Lindvall N. Reconstruction for old rupture of the anterior cruciate ligament. Injury 1971;4:271278. 4. Alm A, Gillquist J. Reconstruction of the anterior cruciate ligament by using the medial third of the patellar ligament. Acta Chir Scand 1974;140:289-296. 5. Clancy WG, Nelson DA, Reider B, Narechania RG. Anterior cruciate ligament reconstruction using one-third of the patellar ligament, augmented by extra-articular tendon transfers. J Bone Joint Surg Am 1982;64:352-359. 6. Lambert KL. Vascularized patellar tendon graft with rigid internal fixation for anterior cruciate ligament insufficiency. Clin Orthop 1983;172:85-89. 7. Kurosaka M, Yoshiya S, Andrish JT. A biomechanical comparison of different surgical techniques of graft fixation in anterior cruciate ligament reconstruction. Am J Sports Med 1987;15:225229. 8. Barber FA, Elrod BF, McGuire DA, Paulos LE. Preliminary results of an absorbable interference screw. Arthroscopy 1995; 11:537-548. 9. Morgan CD, Kalman VR, Grawl DM. Isometry testing for anterior cruciate ligament reconstruction revisited. Arthroscopy 1995;11:647-659. 10. Fowler BL, DiStefano VJ. Tibial tunnel bone grafting: A new technique for dealing with graft-tunnel mismatch in endoscopic anterior cruciate ligament reconstruction. Arthroscopy 1998;14: 224-228. 11. Ishibashi Y, Rudy TW, Livesay GA, Stone JD, Fu FH, Woo SLY. The effect of anterior cruciate ligament graft fixation site at the tibia on knee stability: Evaluation using a robotic testing system. Arthroscopy 1997;13:177-182.
12. Shaffer B. The effect of graft recession on isometry in ACL reconstruction. Arthroscopy 1998;14:424 (abstr). 13. Lysholm J, Gillquist J. Evaluation of knee ligament surgery results with special emphasis on use of a scoring scale. Am J Sports Med 1982;10:150-154. 14. Tegner Y, Lysholm J. Rating systems in the evaluation of knee ligament injuries. Clin Orthop 1985;198:43-49. 15. Pomeroy G, Baltz M, Pierz K, Nowak M, Post W, Fulkerson JP. The effects of bone plug length and screw diameter on the holding strength of bone-tendon-bone grafts. Arthroscopy 1998;14:148-152. 16. Hoffmann R, Peine R, Bail H, Su¨dkamp N, Weiler A. Initial fixation strength of modified patellar tendon grafts for anatomic fixation in anterior cruciate ligament reconstruction. Arthroscopy 1999;15:392-399. 17. Shaffer B, Gow W, Tibone JE. Graft-tunnel mismatch in endoscopic anterior cruciate ligament reconstruction: A new technique of intraarticular measurement and modified graft harvesting. Arthroscopy 1993;9:633-646. 18. Graf BK, Henry J, Rothenberg M, Vanderby R. Anterior cruciate ligament reconstruction with patellar tendon. An ex vivo study of wear related damage and failure at the femoral tunnel. Am J Sports Med 1994;22:131-135. 19. Olszewski AD, Miller MD, Ritchie JR. Ideal tibial tunnel length for endoscopic anterior cruciate ligament reconstruction. Arthroscopy 1998;14:9-14. 20. Cooper DE, Urrea L, Small J. Factors affecting isometry of endoscopic anterior cruciate ligament reconstruction: The effect of guide offset and rotation. Arthroscopy 1998;14:164170. 21. Adate JA, Fadale PD, Hulstyn MJ, Walsh WR. Initial fixation strength of polylactic acid interference screws in anterior cruciate ligament reconstruction. Arthroscopy 1998;14:278284. 22. Rupp S, Seil R, Krauss PW, Kohn DM. Cortical versus cancellous interference fixation for bone–patellar tendon–bone grafts. Arthroscopy 1998;14:484-488. 23. Miller MD, Olszewski AD. Cruciate ligament graft intraarticular distances. Arthroscopy 1997;13:291-295. 24. Rodeo SA, Arnoczky SP, Torzilli PA, Hidaka C, Warren RF. Tendon healing in a bone tunnel. A biomechanical and histological study in the dog. J Bone Joint Surg Am 1993;75:1795-1803. 25. Bo¨stman OM. Intense granulomatous inflammatory lesions associated with absorbable internal fixation devices made of polyglycolide in ankle fractures. Clin Orthop 1992;278:193199. 26. Edwards DJ, Hoy G, Saies AD, Hayes MG. Adverse reactions to an absorbable shoulder fixation device. J Shoulder Elbow Surg 1994;3:230-233. 27. Bo¨stman OM. Osteolytic changes accompanying degradation of absorbable fracture fixation implants. J Bone Joint Surg Br 1991;73:679-682. 28. Roberts C, John C, Seligson D. Prior anterior cruciate ligament reconstruction complicating intramedullary nailing of a tibia fracture. Arthroscopy 1998;14:779-783.
Summary: To determine the efficacy of an anterior cruciate ligament (ACL) graft that customizes length and facilitates anatomic outlet fixation, a prospective study of the ‘‘flipped’’ patellar tendon autograft ACL reconstruction began in 1995. This technique shortens the tendon portion to match the intra-articular length by rotating 1 bone plug 180° proximally onto the tendon, thus flipping the bone plug over its ligamentous insertion. Bioscrews (poly L-lactic acid; Linvatec, Largo, FL) secured the grafts. All patients undergoing this procedure with a minimum 21 months follow-up were reviewed. Preoperative and postoperative Tegner, Lysholm, and IKDC activity scores, and Lachman and pivot shift tests were obtained. Postoperative KT testing and radiographs were obtained. Fifty patients were followed-up for an average of 28 months (range, 21 to 39 months). Average patient age was 34 years (range, 16 to 52 years). Tegner scores increased from 2.0 preoperatively to 6.0 postoperatively. Lysholm scores increased from 46 preoperatively to 93 at follow-up, with 86% excellent (66%) or good (20%). IKDC activity scores increased from 3.1 preoperatively to 1.7 postoperatively. KT manual-maximum difference at follow-up averaged 0.7 mm, with 74% less than 3-mm, 18% 3- to 5-mm, and 8% greater than 5-mm difference. Postoperative Lachman results were 0 in 45 patients and 1⫹ in 5 patients. Postoperative pivot shift was absent in all but 1 patient. Full extension was achieved in all cases and flexion averaged 136° with no patient having less than 120° flexion. No lytic bone changes or tunnel widening were seen. The flipped patellar tendon autograft reduces graft length to its intra-articular portion, increasing graft stability, isometry, and stiffness, and avoiding tunnel graft mismatch with clinically excellent results. Key Words: ACL reconstruction—Flip—Patellar tendon autograft—Interference fixation—Bioscrew—Isometric—Tunnel mismatch.
T
he success of an arthroscopically assisted patellar tendon autograft (PTA) anterior cruciate ligament (ACL) reconstruction is influenced by many factors including surgical technique, graft strength, and fixation.1 The inherent strength and the availability of the PTA plus its secure fixation to the attached bone plugs in the tibial and femoral tunnels make it a commonly used procedure.2-5 The ability to use interference screw fixation with this graft is attractive.6,7 Biodegradable interference fixation screws avoid revision and imag-
From the Plano Orthopedic and Sports Medicine Center, Plano, Texas, U.S.A. Address correspondence and reprint requests to F. Alan Barber, M.D., Plano Orthopedic and Sports Medicine Center, 5228 West Plano Pkwy, Plano, TX 75093, U.S.A. r 2000 by the Arthroscopy Association of North America 0749-8063/00/1605-2169$3.00/0 doi:10.1053/jars.2000.4384
ing problems and are less likely to cause graft damage than are metal implants.8 Morgan et al.9 recently described a technique of flipping one end of the PTA, which permits selection of a precise graft length, eliminates graft tunnel mismatch (both length and diameter), and allows for graft fixation at both femoral and tibial outlets. Other techniques have been described to accomplish this goal, but they require additional graft material.10 The common thread is graft fixation at the femoral and tibial outlets, which eliminates graft micromotion in the tunnels leading to tunnel widening or fixation failure. Reducing the graft length to its intra-articular portion increases graft stability and graft isometry, minimizes the potential for creep failure within the tunnel, and increases graft stiffness.11,12 The purpose of this study was to determine the clinical efficacy of this flip PTA technique for ACL reconstruction.
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 16, No 5 (July-August), 2000: pp 483–490
483
484
F. A. BARBER METHODS
A prospective study of flipped PTA ACL reconstructions secured by Bioscrew biodegradable interference screws (Linvatec, Largo, FL) started in June 1995. The inclusion criteria consisted of a complete ACL tear with knee instability demonstrated by both positive Lachman and positive pivot shift tests, adequate bone density, a PTA with a tendon segment measuring at least 4.5 cm, and a commitment to return for at least 2 years of follow-up. Exclusionary criteria included prior knee infection, concurrent multiple ligament reconstruction, prior ACL reconstruction, and a torn posterior cruciate ligament. Preoperative assessment included medical history, physical examination, radiographs, and preoperative Lysholm,13 Tegner,14 and IKDC activity scales. Baseline KT measurements and manual-maximum side-toside difference were obtained. The surgical findings and procedures performed were recorded. All patients underwent the same procedure: an arthroscopically assisted ACL reconstruction using a flipped PTA as described by Morgan et al.9 (Fig 1) fixed by Bioscrew interference screw fixation at both the proximal and distal graft sites. Postoperative evaluations were obtained at intervals of 3, 6, 12, and 24 months, and yearly thereafter. These assessments included the Lysholm, Tegner, and IKDC scales; meniscal tests; ligament tests; KT tests; and radiographs. The biodegradable interference screws made from purified poly L-lactide (containing approximately 45% crystalline polymer) were used for all reconstructions. The femoral biodegradable screw is placed through a separate portal in the patellar tendon chosen to match the angle of the femoral tunnel. The cannulated screw is advanced over the nitinol guide wire, which pre-
FIGURE 1. The flipped PTA is maintained under tension on the work station after preparation until insertion.
vents divergence, and is then inserted by a triangular fluted screwdriver. For these cases, screws measuring 8 ⫻ 20 mm were the most commonly selected devices. The 8-mm diameter screws were chosen because of concerns about breakage with 7-mm screws and the personal observation that 9-mm screws can push the bone plug into the adjacent cancellous bone, especially in the tibia. There were no cases of screw breakage in this series. This may be due to the larger core diameter of the 8-mm screw and customizing of the screw insertion portal to precisely parallel the angle of the femoral tunnel. Surgical Technique The distinctive feature of this technique is how the graft is prepared. The PTA is harvested through an anterior longitudinal incision removing a 10-mm wide strip of central patellar tendon with attached bone plugs from the tibia and patella that measure 10 ⫻ 25 mm and 10 mm deep. After determining that the tendinous portion of the graft is at least 4.5 cm long, the tibial bone segment is trimmed to allow it to be flipped back over the adjacent patellar tendon and reduced to 2 cm in length. Flipping the bone plug rotates it so that the cancellous portion faces away from the tendon. In this manner, interference screw fixation can be achieved between the cancellous bone of the plug and the cancellous bone of the tunnel wall. This also provides an adequate amount of bone to achieve a good interference fixation.15 If the bone plug were rotated the other way (to place the cancellous portion against the tendon), the screw fixation would be against the tendon-covered cortical side of the bone plug. While recent data suggest that the load-to-failure strength of this construct is no different from that of the tendon-tendon-bone arrangement, the method of failure is by pullout rather than by tendon stripping.16 Once properly trimmed, the flipped bone plug is fixed to the tendon by several encircling sutures of No. 1 PDS. Two sutures of No. 5 braided polyester are vertically placed through the flipped bone plug and tendon to achieve control for insertion and tensioning. Once prepared, the construct is kept tensioned on the workstation (Arthrex, Naples, FL) until inserted (Fig 1). Tensioning is applied by manually pushing the holding posts apart with the thumb and finger. Tibial and femoral tunnels are prepared with the same technique regardless of whether or not a flipped graft is used. The tibial tunnel is initially drilled to a 10-mm diameter and enlarged if necessary to be the same size as the flipped portion of the graft. Oversizing is not necessary for the biodegradable screw.
FLIPPED PATELLA TENDON AUTOGRAFT Selection of the graft as a suitable candidate for flipping is based on several criteria. Principal among these is the intra-articular distance from the femoral outlet to the tibial outlet. This distance usually ranges from 25 to 30 mm, but can be as little as 21 to 33 mm.17 Some femoral aiming devices feature a scale on the shaft that allows for a direct measurement associated with the femoral guide wire insertion. Another criterion is patellar tendon length. This graft can be constructed only if the tendon section is at least 4.5 cm long. Although shorter patients may require shorter intra-articular grafts, longer patellar tendons certainly facilitate this process. This technique is not advised for the graft that has a tendon shorter than 4.5 cm. With knowledge of the precise intraarticular length, the graft is placed on the workstation and the bone plugs are crafted. A squared off shape at both ends of the tibial bone plug is created with an ideal bone plug length of 2 cm.15 The square end allows for better rotation of the bone plug and a more intimate fit with of the bone adjacent to the tendon. Remember that the tendon must be long enough to provide for the measured intra-articular graft length and to accommodate the 2-cm bone plug. Postoperative Program In this series, the postoperative rehabilitation program was no different from that used for unflipped PTAs. Specifically, in the first days and weeks, achieving full extension was emphasized. Progressive weight bearing as tolerated was permitted and half of these patients were off crutches by 1 week. All were walking without crutches 2 weeks after surgery. A fullextension postoperative brace was used at night during the first 2 weeks, and continuous passive motion was used for 6 to 8 hours a day until at least 90° flexion was achieved. Progressive closed-chain exercises were started 1 to 2 weeks after surgery. These closed-chain exercises included stationary biking, stair climbing, and cross-country ski machines. Straight ahead, halfspeed running was started between 6 and 8 weeks after surgery. Pivoting, noncontact sports were initiated at 12 weeks, and a derotational brace was ordered. Once the brace was received, the patients were allowed to participate in full unlimited contact sports using the brace. Open-chain exercises against resistance in the terminal 30° of extension were avoided until 6 months after surgery. RESULTS During the study period from June 1995 through December 1996, a total of 55 knees underwent ACL
485
reconstruction using the flip technique. There were 50 knees in 50 patients (91%) available for long-term follow-up; these are the subjects of this report. Follow-up data are complete for all included patients with a minimum of 21 months follow-up (average, 28 months; range, 21 to 39 months). The average age was 34 years (range, 16 to 52 years). There were 35 male and 15 female patients. Reconstructions were performed in 21 right knees and 29 left knees. There is no indication that the 5 patients unavailable for the minimum 21-month follow-up had an outcome that was different from that of the group as a whole. Associated surgical findings (e.g., chondromalacia, meniscal tears) were reviewed. There were 18 fullthickness medial meniscus tears requiring treatment (2 repaired, 16 debrided). There were 20 lateral meniscus tears requiring treatment (none were repairs). Meniscal injury that did not require a specific treatment was not recorded. Femoral condyle damage was observed in 11 patients, patellar chondromalacia in 4, and medial tibial plateau chondral damage in 1 patient. Loose bodies were removed from 5 knees, 3 of which also had medial femoral chondromalacia. The time from injury to surgery was reviewed. There were 12 knees considered to be chronic (operated on more than 12 weeks after injury). No correlation between the length of time from injury and the incidence of meniscal tears was found; however, meniscal repairs were only performed in patients with recent injuries. Patients were asked for both preinjury and preoperative postinjury Tegner scores. The average preinjury score was 6.3 (range, 1-9). The average preoperative Tegner score was 2.0 (range, 0-6). The average postoperative Tegner score was 6.0 (range, 1-10) at an average follow-up of 28 months. The average preoperative Lysholm score was 45.8 (range, 13-93) and increased to 93.3 (range, 65-100) at an average follow-up of 28 months. The follow-up Lysholm scores showed excellent results (95-100) in 33 patients (66%) and good results (84-94) in 10 patients (20%). The IKDC form has an activity scale that classifies athletic activity into 4 groups, ranging from sedentary (housework, activities of daily living), to light activity (jogging, running), to moderate activity (skiing, tennis), to strenuous activity (jumping, pivoting, hard cutting). These 4 groupings are assigned a numerical score from the most strenuous (level 1) to the sedentary (level 4). The IKDC activity score was recorded preoperatively and averaged 3.1. This rank falls just below the light activity level. The IKDC activity level
486
F. A. BARBER
FIGURE 2. IKDC scores improved from the preoperative to postoperative measurements.
improved postoperatively to 2.3 at 14 months average follow-up and to 1.7 at 28 months follow-up (Fig 2). At follow-up, KT manual-maximum differences between both knees averaged 0.7 mm. The results for all knees at the 2-year follow-up were tabulated into 3 ranges: less than 3-mm difference (74%), 3- to 5-mm difference (18%), and greater than 5-mm difference (8%). Supine goniometer-measured motion showed no loss of extension in any patient at the 2-year follow-up. Every patient achieved at least 120° of flexion and the average flexion at the 2-year follow-up was 136° (range, 120° to 145°). The preoperative Lachman test results averaged 2.2 (range, 1⫹ to 4⫹). The 1-year postoperative Lachman test results were 0 in 48 patients and 1⫹ in 2 patients. At final follow-up, the Lachman test results were 0 in 45 patients and 1⫹ in 5. Preoperatively, the pivot shift
was 1⫹ in 10 patients and 2⫹ or greater in 40 patients. At follow-up, the pivot shift was 0 in all but 1 patient. There were no problems resulting from the use of the biodegradable interference screw. Specifically, no instances of screw breakage or graft laceration occurred. There were no joint effusions at follow-up. The immediate postoperative radiographs allowed confirmation of tunnel placement and bone plug location (Fig 3). Radiographs did not show any lytic changes or tunnel widening at the biodegradable interference screw insertion sites. Rapid bone plug incorporation and graft maturation were consistently observed (Fig 4). The use of the biodegradable interference screw permitted clear radiographic assessments of the healing PTA and an evaluation of the tunnel placement. Over time, the PTA and tunnels were difficult to view on radiographs (Fig 5) because of the rapid and progressive healing between the bone graft and the adjacent tunnel. DISCUSSION The flipped PTA offers the advantages of outlet graft fixation with the strength and availability of a patellar tendon. Outlet fixation eliminates concerns about micromotion of the tendinous portion of the graft in the tunnel. Concerns about tunnel widening and subsequent fixation failure are consequently lessened. A
FIGURE 3. (A) Anterior and (B) lateral radiographs: The Bioscrew permits a clear radiographic assessment of both graft and tunnel placement in the immediate postoperative (1 week) period, which can serve as a baseline for comparison with subsequent radiographs. The tibial bone plug (arrow) reaches the tunnel outlet in the tibial spine.
FLIPPED PATELLA TENDON AUTOGRAFT
FIGURE 4. At 2 years, the radiographs show progressive incorporation of the PTA without tunnel widening or lytic change. Rapid bone plug incorporation and graft maturation were consistently observed.
graft with a tendon portion precisely matching the intra-articular length, and with a bone plug–interference screw–tendon mass that eliminates tunnel diameter mismatch by completely filling the tibial and femoral tunnel outlets, should avoid graft movement problems (a possible cause of the windshield wiper effect).9 Plugging the tibial outlet will also prevent synovial fluid leaking into the tibial tunnel, which may be another cause of tunnel widening and lytic change. Fixation of this shorter intra-articular graft at the anatomic ACL insertion outlet for the femur prevents graft abrasion by the anterior edge of the femoral tunnel18 and provides optimal strain patterns that mimic normal ACL isometry.12 Outlet fixation at the tibia minimizes creep failure that can occur within the tunnel because the points of graft fixation are moved away from the anatomic origin and insertion.9 A
487
fixation position more distal in the tibial tunnel can increase the elongation forces on the graft as the knee goes through a range of motion.9 Ishibashi et al.11 have shown that proximal tibial tunnel attachment provides the most stable knee reconstruction, with less tibial displacement and axial rotation than these distal placements in the tunnel. Presuming that the elastic modulus of the graft is constant along its length, increasing the effective graft length by more distal tibial tunnel fixation decreases graft stiffness.11 Consequently, proximal attachment at the tibial ACL outlet minimizes the functional length and maximizes the graft stiffness. PTA mismatch can occur not only between the tendon portion of the graft and tunnel diameter, but also between the length of the graft and the length of the tunnel.17 The immediate advantage of the flipped PTA is the technical facility to deal with a patellar tendon that is too long or with a tibial tunnel inadvertently placed too high on the tibia. Either of these circumstances may result in a tendon mismatched to the tunnel. The hang-out situation has been reported to occur in 25%17 or more19 of cases. The flip eliminates this with the additional bonus of also eliminating any windshield wiper effect because the graft now fully fills the tunnels where it resides and provides the biomechanical advantages of a stiffer graft with less possibility of creep. Minor variations in the alignment of an endoscopic drill guide can affect the isometricity of the graft location. The ideal isometric site for the tibial fibers is the anteromedial area of the tibial tunnel.20 Unfortunately, the standard PTA places the fibers at the posterior portion of this area. When the graft is twisted, the fibers may be better oriented. The flip technique not only achieves outlet fixation, but the tendon fibers are reversed and located in the desired anteromedial position without twisting. An additional advantage of the flipped graft is its mechanism of failure. In general, graft failure can occur by bone plug fracture, bone plug pullout from the interference fixation screw, or tendon rupture. The mode of failure for interference screw fixation with the standard PTA is influenced by the screw placement. When the screw is placed on the cancellous side of the bone plug, the most common mode of failure is bone plug pullout.21 In contrast, when the screw is placed on the cortical side of the bone plug, the most common mode of failure is by tendon rupture or bone plug fracture.22 Both of these situations are all-or-nothing events. As excessive load is placed on the flipped graft, failure occurs by the stripping of the tendon from the
488
F. A. BARBER
FIGURE 5. (A) Anterior and (B) lateral radiographs 3 years postoperatively show no evidence of any localized response radiologically to the Bioscrew, and cancellous material completely fills the tunnel sites.
rotated bone plug.16 This may actually provide a safety valve for overconstrained fixation or for tunnels placed in less than ideal sites, because the graft will not fail all at once. Instead, it will stretch out incrementally and reach a steady state in which the elongation force is reduced by the stretching graft to a level below that able to damage the ligament further. The length of the tendon portion of these grafts was selected to match the intra-articular distance between the femoral and tibial tunnel outlets measured by the transtibial guide. The intra-articular length of the ACL is reported to be about 24 mm in fresh-frozen cadaver specimens19,23 and 26 mm in patients undergoing ACL reconstruction (range, 21 to 33 mm).17 The length of the patellar tendon measures from 33 to 63 mm long with an average length variously reported as 43 to 48 mm.17,19 When the tendon portion of the harvested PTA is at least 45 mm long, the graft can be flipped. With a 2-cm bone plug, this construct produces an
intra-articular graft length of 25 mm. The femoral bone can be whatever length the surgeon prefers; I usually use one that is at least 20 mm long. The tibial plug for the flipped PTA should be 20 mm as well. This length is sufficient for secure graft fixation.24 The interference fixation screw is made of poly L-lactide and requires 3 to 4 years to completely biodegrade. Biodegradable screws have been shown clinically8 and in vitro21 to provide fixation equivalent to metal interference fixation screws for unflipped PTA ACL reconstructions. The biodegradable screw also works well with the flipped PTA, avoiding problems with graft laceration found with metal screws and allowing excellent postsurgical imaging, which facilitates assessment of graft location. The screw at the tunnel outlet provides the compression that allows rapid healing of the graft into the tunnel.24 Because of its poly L-lactide composition, it does not cause the intense inflammatory reactions that can lead to lytic
FLIPPED PATELLA TENDON AUTOGRAFT areas in the bone or draining wounds reported with polyglycolic acid or polyglycolic acid copolymers as early as 8 weeks after implantation.25-27 Serial radiographic examinations showed progressive autograft incorporation into the bone tunnels without adverse occurrences (Figs 3-5). There was no evidence of tunnel widening in any of the follow-up radiographs. No evidence of osteolysis or other osseous reaction was found by magnetic resonance imaging (Fig 6). The poly L-lactide polymer is less rigid and more malleable than metal. It will conform to and follow the tunnel rather than cut a new direction and will not diverge during insertion. Biodegradable materials reduce problems with postsurgical imaging and minimize difficulties that may be encountered during revision procedures or with subsequent internal fixation around the knee.28 This prospective evaluation of the flipped PTA ACL reconstruction shows improved Tegner and Lysholm
489
scores with 86% excellent or good results by Lysholm scoring; an average KT manual-maximum difference of 0.7 mm 28 months after surgery, and with 74% of the patients having a KT manual-maximum difference of less than 3 mm. Recent data show that this technique provides fixation strengths that are comparable to conventional means, whether secured by metal or biodegradable screws. In addition, in vitro graft failure of this technique is by tendon stripping from its ligamentous attachments, which may provide a self-adjusting effect if the initial fixation is overconstrained. In conclusion, the flipped PTA permits selection of a precise graft length, eliminates the problem of too long a graft hanging out the anterior tibia, and allows for fixation of the graft at both femoral and tibial outlets. Graft fixation at the femoral and tibial outlets eliminates graft micromotion in the tunnels, graft tunnel size mismatch, and tunnel widening or fixation failure.
FIGURE 6. (A) Anterior and (B) lateral magnetic resonance imaging scans 3 years after surgery show no localized inflammatory response, no osteolysis nor other osseous reaction, and good graft incorporation.
490
F. A. BARBER
Reducing the graft length to its intra-articular portion increases graft stability and graft isometry, minimizes the potential for creep failure within the tunnel, and increases graft stiffness. The clinical results of the flipped PTA are excellent and this technique offers many advantages. Acknowledgment: The author appreciates the assistance of Nancy Alexander with data collection and the surgical assistance of James N. Click, P.A.-C., and Brian Howard, C.S.T., in this study.
REFERENCES 1. Paulos L, Noyes FR, Grood E, Butler DL. Knee rehabilitation after anterior cruciate ligament reconstruction and repair. Am J Sports Med 1981;9:140-149. 2. Brostrom L, Gillquist J, Liljedahl S-O, Lindvall N. Behandling av invetererad ruptur av framre korsbandet. Lakrtidn 1968;64: 44-77. 3. Gillquist J, Liljedahl S-O, Lindvall N. Reconstruction for old rupture of the anterior cruciate ligament. Injury 1971;4:271278. 4. Alm A, Gillquist J. Reconstruction of the anterior cruciate ligament by using the medial third of the patellar ligament. Acta Chir Scand 1974;140:289-296. 5. Clancy WG, Nelson DA, Reider B, Narechania RG. Anterior cruciate ligament reconstruction using one-third of the patellar ligament, augmented by extra-articular tendon transfers. J Bone Joint Surg Am 1982;64:352-359. 6. Lambert KL. Vascularized patellar tendon graft with rigid internal fixation for anterior cruciate ligament insufficiency. Clin Orthop 1983;172:85-89. 7. Kurosaka M, Yoshiya S, Andrish JT. A biomechanical comparison of different surgical techniques of graft fixation in anterior cruciate ligament reconstruction. Am J Sports Med 1987;15:225229. 8. Barber FA, Elrod BF, McGuire DA, Paulos LE. Preliminary results of an absorbable interference screw. Arthroscopy 1995; 11:537-548. 9. Morgan CD, Kalman VR, Grawl DM. Isometry testing for anterior cruciate ligament reconstruction revisited. Arthroscopy 1995;11:647-659. 10. Fowler BL, DiStefano VJ. Tibial tunnel bone grafting: A new technique for dealing with graft-tunnel mismatch in endoscopic anterior cruciate ligament reconstruction. Arthroscopy 1998;14: 224-228. 11. Ishibashi Y, Rudy TW, Livesay GA, Stone JD, Fu FH, Woo SLY. The effect of anterior cruciate ligament graft fixation site at the tibia on knee stability: Evaluation using a robotic testing system. Arthroscopy 1997;13:177-182.
12. Shaffer B. The effect of graft recession on isometry in ACL reconstruction. Arthroscopy 1998;14:424 (abstr). 13. Lysholm J, Gillquist J. Evaluation of knee ligament surgery results with special emphasis on use of a scoring scale. Am J Sports Med 1982;10:150-154. 14. Tegner Y, Lysholm J. Rating systems in the evaluation of knee ligament injuries. Clin Orthop 1985;198:43-49. 15. Pomeroy G, Baltz M, Pierz K, Nowak M, Post W, Fulkerson JP. The effects of bone plug length and screw diameter on the holding strength of bone-tendon-bone grafts. Arthroscopy 1998;14:148-152. 16. Hoffmann R, Peine R, Bail H, Su¨dkamp N, Weiler A. Initial fixation strength of modified patellar tendon grafts for anatomic fixation in anterior cruciate ligament reconstruction. Arthroscopy 1999;15:392-399. 17. Shaffer B, Gow W, Tibone JE. Graft-tunnel mismatch in endoscopic anterior cruciate ligament reconstruction: A new technique of intraarticular measurement and modified graft harvesting. Arthroscopy 1993;9:633-646. 18. Graf BK, Henry J, Rothenberg M, Vanderby R. Anterior cruciate ligament reconstruction with patellar tendon. An ex vivo study of wear related damage and failure at the femoral tunnel. Am J Sports Med 1994;22:131-135. 19. Olszewski AD, Miller MD, Ritchie JR. Ideal tibial tunnel length for endoscopic anterior cruciate ligament reconstruction. Arthroscopy 1998;14:9-14. 20. Cooper DE, Urrea L, Small J. Factors affecting isometry of endoscopic anterior cruciate ligament reconstruction: The effect of guide offset and rotation. Arthroscopy 1998;14:164170. 21. Adate JA, Fadale PD, Hulstyn MJ, Walsh WR. Initial fixation strength of polylactic acid interference screws in anterior cruciate ligament reconstruction. Arthroscopy 1998;14:278284. 22. Rupp S, Seil R, Krauss PW, Kohn DM. Cortical versus cancellous interference fixation for bone–patellar tendon–bone grafts. Arthroscopy 1998;14:484-488. 23. Miller MD, Olszewski AD. Cruciate ligament graft intraarticular distances. Arthroscopy 1997;13:291-295. 24. Rodeo SA, Arnoczky SP, Torzilli PA, Hidaka C, Warren RF. Tendon healing in a bone tunnel. A biomechanical and histological study in the dog. J Bone Joint Surg Am 1993;75:1795-1803. 25. Bo¨stman OM. Intense granulomatous inflammatory lesions associated with absorbable internal fixation devices made of polyglycolide in ankle fractures. Clin Orthop 1992;278:193199. 26. Edwards DJ, Hoy G, Saies AD, Hayes MG. Adverse reactions to an absorbable shoulder fixation device. J Shoulder Elbow Surg 1994;3:230-233. 27. Bo¨stman OM. Osteolytic changes accompanying degradation of absorbable fracture fixation implants. J Bone Joint Surg Br 1991;73:679-682. 28. Roberts C, John C, Seligson D. Prior anterior cruciate ligament reconstruction complicating intramedullary nailing of a tibia fracture. Arthroscopy 1998;14:779-783.