Traditional fixation for anterior cruciate ligament (ACL) reconstruction has been often performed with metal interference screws1. This fixation technique provides high initial fixation strength while promoting early osseous integration2,3. Despite favorable reports on metal interference screws, there are concerns regarding the distortion on postoperative magnetic resonance imaging (MRI) evaluation as well as the requirement for implant removal during revision surgery4. To avoid these potential problems, bioabsorbable screws have been proposed for graft fixation. However, several adverse complications, such as fractured screws, transcutaneous migration of a screw, cyst formation with osteolysis, abscess formation, and inflammatory reaction, have been reported with bioabsorbable screws5-8.
Tibial tunnel bone-grafting to address graft-tunnel mismatch9 or for revision ACL reconstruction10 has been well demonstrated. Recently, Jagodzinski et al.11 reported that tibial press-fit fixation decreased the amount of proximal bone-tunnel enlargement. However, they used a xenogenic spongiosa cylinder instead of autogenous bone. To our knowledge, no previous studies have compared the clinical outcomes of bioabsorbable screws with those of autogenous bone plugs for tibial fixation of tendon grafts.
The purpose of this study was to compare two different methods of tibial tunnel fixation with respect to clinical scores, results of stability testing, complication rates, and tunnel widening as assessed with imaging. Our hypothesis was that use of an autogenous bone plug for tibial tunnel fixation would reduce the complication rate and tibial tunnel widening without inducing instability compared with what is seen after use of a bioabsorbable interference screw.
Study Population
For this prospective study, we identified 102 patients (102 knees) who had undergone ACL reconstruction only with fresh-frozen Achilles tendon allograft between January 2000 and January 2006. Twenty-one patients were not included in the study because four had been lost to follow-up, fifteen had inadequate follow-up MRI and postoperative functional scores, and two had died. Inclusion criteria were (1) an ACL reconstruction and (2) a patient age of twenty to fifty years. Exclusion criteria were (1) an associated posterior cruciate ligament (PCL) injury, (2) a full-thickness articular cartilage lesion, and (3) previous knee surgery due to meniscal or collateral ligament injury.
The final study group consisted of eighty-one patients (eighty-one knees). Patients were randomized with use of a computer-generated random list to ensure equal distribution of patients into each group. In forty-one of the eighty-one patients, a bioabsorbable interference screw was used for tibial tunnel fixation (group I). These patients were compared with forty patients (group II) in whom an autogenous bone plug obtained from the tibial tunnel was used for fixation (see Appendix). The patients and physiotherapists were blinded to the method used. Our institutional review board approved the study protocol. An informed consent form concerning the operative technique to be performed was signed by all patients. The average age of the patients at the time of surgery was 32.0 years (range, twenty to fifty years). There were fifty-four men and twenty-seven women. The minimum duration of follow-up was 5.5 years (average, 7.5 years; range, 5.5 to 10.9 years).
Surgical Technique
All procedures were performed at one institution by a single surgeon (H.-C.L.). A complete ACL tear was reconstructed in all of the patients. Following diagnostic arthroscopy, an Achilles tendon-bone graft measuring 10 mm in diameter was prepared from a thawed allograft. In patients with severe impingement of the graft with the lateral wall and roof of the notch, an adequate intercondylar notchplasty was done to allow for easier graft passage and visualization. The tibial tunnel was drilled at a 45° angle and positioned according to technique described by Howell and Barad12. All tunnels were a uniform size in which to place a 10-mm-diameter tendon. Once the guide pin was in the proper position, the outer cortex of the tibia was reamed with a standard cannulated reamer, which was the same size as the coring reamer to be used. The guide pin was then exchanged for a collared guide pin over which the coring reamer was passed. With use of the coring reamer, we created the tibial tunnel accurately while simultaneously harvesting cylindrical bone blocks. The coring reamer was passed into the joint and was seen to spin freely to ensure that all soft-tissue attachments had been severed. The reamer was removed from the joint, and a cored cancellous bone plug of the length of the tibial tunnel was obtained from within the reamer. The cone-shaped bone plug of 8 mm in diameter (sized with a ruler) required to fill the tibial tunnel was prepared by trimming (in group II) (Fig. 1 and Appendix). For the femoral tunnel, with the knee flexed 90°, a long Beath pin was passed up through the tibial tunnel, into the previously marked location for the femoral tunnel, and out the anterolateral aspect of the thigh. Over this guide pin, the femoral tunnel was reamed to the desired depth. The graft was pulled up into the joint and the Achilles tendon bone plug of 30 mm in length within the femoral tunnel was secured with absorbable RIGIDFIX (DePuy Mitek, Edinburgh, United Kingdom). The knee was moved through a full range of motion with traction on the free tibial end to verify that the position of the graft was satisfactory. The graft was tensioned between 15 and 20 N before it was secured with the knee in 10° of flexion. On the tibial side, we used a cancellous screw and ligament washer as the primary tibial fixation method, either with an additional 8-mm bioabsorbable screw (Arthrex, Naples, Florida) or an 8-mm bone plug, in all cases. Then a final check of knee motion was made and a Lachman test was performed.
Rehabilitation
The postoperative protocol, which was uniform for all patients, consisted of accelerated rehabilitation with early motion and physiotherapy. Continuous passive motion was utilized twice a day during the first seven days. Full extension was easily achieved, and full weight-bearing was allowed within the first two weeks. At four weeks, use of the knee brace and crutches was discontinued. At the third month after the operation, patients began isotonic and isokinetic exercises. They were allowed to return to sport activities four to six months after the operation. The patients were seen for clinical and radiographic follow-up at two, four, eight, and twelve weeks and then every three months thereafter.
Postoperative Evaluation
All patients were seen by the senior author (H.-C.L.) at an outpatient clinic and examined for any signs of complications after surgery. The physical examination assessed joint effusion, signs of infection at the tibial fixation site, limitation of knee motion, and ligament laxity.
All radiographs were measured with use of StarPACS PiView STAR 5.0.6.0 software (Infinitt Healthcare, Seoul, Korea). All patients had preoperative MRI scans of the injured knee and second MRI studies performed five years postoperatively, which were reviewed by three independent experienced radiologists. Images were obtained with 1.5-T magnets providing assessment in three planes. High-resolution sequences (3 to 4-mm sections with at least a 256 × 256 matrix) were used to assess ligament and meniscal tears. T1 and T2-weighted axial sequences were obtained perpendicular to the long axis of the tibial tunnel 1 cm below the articular surface, 1 cm above the distal tibial tunnel exit, and at the midpoint of the tunnel. The cross-sectional area of the tunnel was measured digitally with use of a computer-generated best-fit circle13 (Fig. 2).
Clinical scores were assessed with use of the International Knee Documentation Committee (IKDC), Lysholm, and Tegner activity scores. Knee motion was measured at an outpatient clinic by an experienced physiotherapist using a long-arm goniometer. The Lachman test was graded as negative (−) when there was a hard end point and a side-to-side difference of <3 mm, as slightly positive (+) when there was a side-to-side difference of 3 to 5 mm, and as clearly positive (++) when there was a side-to-side difference of >5 mm. Correspondingly, the pivot-shift test was graded as negative (−), glide (+), or clearly positive (++). Laxity of both knees was measured with a KT2000 knee ligament arthrometer (MEDmetric, San Diego, California) at 30 lb. Restoration of anteroposterior stability was considered good to excellent if the manual maximum side-to-side difference was <3 mm. The knee scores were completed by the patients, and the KT2000 arthrometer measurements were performed by a physiotherapist. These examinations were repeated at one, two, three, four, and five years postoperatively. The patients and the persons who assessed the outcomes were blinded to group assignment.
Statistical Analysis
Data were recorded using Microsoft Excel 2007 version (Microsoft, Redmond, Washington) and analyzed with use of SPSS software (SPSS, Chicago, Illinois). The chi-square test was used to evaluate differences in the IKDC score and complication rates between groups. Clinical scores and various indices were described by a mean and standard deviation. The Student t test was used to compare the clinical parameters of group I with those of group II. Significance was reported at the 95% confidence level (p < 0.05). A power analysis was performed with use of the results of the Lysholm and Tegner scores as the primary variables and data from a surgical population at our institution that were representative of the study population. This analysis indicated that a sample size of at least twenty-seven patients per group was necessary to detect a between-group difference in the Lysholm scores with an alpha of 0.05 and a power of 80%. For the Tegner scores, an alpha of 0.05 and a power of 86% were indicated. Radiographic parameters were tested for reproducibility with interobserver studies with use of Pearson correlation coefficients; the correlation coefficients ranged from 0.871 to 0.961.
Source of Funding
There was no external funding source for this investigation.
There are numerous ways to fix the graft into the bone tunnels for ACL reconstruction; interference fixation with a screw of metal or bioabsorbable material has been used primarily for this purpose14. Concerns associated with the use of bioabsorbable interference screws include intraoperative screw breakage; the possibility of an inflammatory response to the progressive absorption, or even the incomplete absorption, of the bioabsorbable compound; the increased cost compared with that of metallic screws; and the potential for bone tunnel widening15,16. The purpose of this study was to evaluate bone tunnel widening and complication rates after the use of an autogenous bone plug or a bioabsorbable interference screw for secondary fixation of Achilles tendon allografts in ACL reconstruction.
Bone plugs seem to be advantageous for ACL reconstruction as the complication rates were significantly lower than those associated with bioabsorbable interference screws. Bioabsorbable interference screws are known to be associated with delayed intra-articular inflammatory reaction17. In contrast, autogenous bone plugs are biocompatible. Bone graft provides a foundation or scaffold for the growth of new bone. Impacting autogenous bone into the distal tibial tunnel exit enhanced the healing of the graft to the tunnel in a circumferential way.
The bioabsorbable material used in this study might have affected graft healing in a way that differed from what occurred with the bone plugs. This may be one of the factors that explains the differences in postoperative complication rates between the groups. Only intraoperative breakage of the material used for fixation (the bone plug in group II and the screw in group I) occurred more frequently in group II. Delivering and impacting the graft into the narrow tibial tunnel with arthroscopic instruments is technically difficult. However, the technique for tibial tunnel grafting is relatively easier than that for femoral tunnel grafting because the tibial tunnel allows direct access under the skin10. To avoid intraoperative bone breakage while obtaining the bone plug, the surgeon should see the coring reamer spin freely in the joint to ensure that all soft-tissue attachments have been severed. Also, the bone plug must be trimmed to a conical shape before impaction for easier bone passage.
There have been several recent reports on bone tunnel widening following ACL reconstruction18-20. Jagodzinski et al.11 demonstrated that tibial press-fit fixation decreased the amount of proximal bone tunnel widening. However, they used computed tomography scans and radiographs for evaluation. Also, a xenogenic spongiosa cylinder was used instead of autogenous bone, and their study population included only ten per group. Other studies of both tibial and femoral tunnel enlargement following autograft ACL reconstruction implicate mechanical instability as a main contributory factor21,22. Most studies are prone to inaccuracies as the authors obtained the analytical data from two-dimensional measurements derived from radiographs, which often have poor tunnel visualization with associated image magnification, resulting in errors of calculation18,21. Therefore, we used a previously reported technique for evaluating tunnel enlargement13, involving accurate digital MRI analysis of cross-sectional area, and the changes were correlated with clinical outcome. We are not aware of any previous MRI study of tunnel widening comparing bioabsorbable screws with bone plugs in ACL reconstruction. Our study showed a significant difference in the tibial tunnel widening according to whether a bioabsorbable screw had been used to fix an Achilles tendon allograft or an autogenous bone plug had been used. We found no correlation between tunnel enlargement and clinical outcome at this early stage. This finding might be due to the fact that an amorphous loose connective-tissue envelope surrounds the graft, providing either a temporary or a permanent anchorage. However, this vascular connective tissue might lead to the late failure of graft anchorage with subsequent knee instability23. Furthermore, enlarged tunnels have the potential to cause problems with graft positioning and fixation in revision ACL surgery.
We found satisfactory subjective and objective clinical results for both types of fixation. There were no significant differences between bioabsorbable interference screws and bone plugs with regard to clinical outcomes such as IKDC, Lysholm, and Tegner activity scores. The overall mean Tegner activity score was very slightly higher in group II (p = 0.167). Also, the anterior-posterior tibial translation measured with the KT2000 arthrometer showed no significant difference between groups. A slightly (0.2-mm) greater side-to-side difference was noted in group II at the time of final follow-up, but this is probably not clinically detectable. Long-term follow-up studies of these clinical results are needed.
In our study, the bone plug and bioabsorbable screw were used for secondary fixation, not for primary fixation. They were supplemental and did not provide substantial fixation strength. On the tibial side, the Achilles tendon is often fixed without an additional interference screw or bone plug24,25. The primary fixation of the graft on the tibia is obtained with the use of a cancellous screw and ligament washer. This also helps to explain why there was no difference in laxity between the two groups.
Our study has some limitations. First, we were unable to perform studies comparing biomechanical load-to-failure levels at functionally relevant strain rates. While there have been several studies comparing the strength between bioabsorbable screws and metal screws26, the fixation strength that is obtained with a bone plug needs to be investigated. Although biomechanical studies are required to determine the exact pullout strength of this construct, it appears that filling the tibial tunnel with bone has the potential advantage of reducing motion between the tendon and the tunnel, improving the healing of the graft to the tunnel, and allowing additional anatomical fixation of the tibial end of the graft. A second limitation of our study is that, although we attempted to make the two groups as equivalent as possible in terms of age and sex and average pretreatment score, there may be certain dissimilarities between the two groups. Another limitation is that there was no group with only extra tunnel fixation without a bone plug or bioabsorbable screw in the tibial tunnel with which to compare tunnel widening and fixation strength.
Our clinical results associated with bioabsorbable screws and bone plugs were not significantly different. Laxity evaluation demonstrated no significant differences between bioabsorbable screws and bone plugs. Compared with bioabsorbable interference screws, autogenous bone plugs for tibial tunnel fixation were associated with a lower complication rate and less tibial tunnel widening without inducing instability. Considering clinical scores, function, cost, and overall failure rates over time, we conclude that autogenous bone plugs are a reasonable option for ACL reconstruction with Achilles tendon allograft.