Background: Anterior cruciate ligament (ACL) graft tear and contralateral ACL injury have devastating effects on patient outcomes following ACL reconstruction. Long-term results after ACL reconstruction are being reported with greater frequency, allowing a detailed analysis of the long-term risk of failure.
Methods: A systematic review of prospective studies that recorded the risk of ACL graft rupture and contralateral ACL injury following primary ACL reconstruction was performed. All studies included intra-articular ACL reconstruction with modern techniques and patellar tendon or hamstring autograft with a minimum follow-up of ten years.
Results: Nine studies met the inclusion and exclusion criteria. The overall ACL graft rupture risk was 7.9% (211 ruptures in 2682 reconstructions) and ranged from 3.2% to 11.1% in the individual studies. The overall risk of ACL injury in the contralateral limb was 12.5% (335 injuries in 2682 reconstructions) and ranged from 0.6% to 22.7% in the individual studies. Among the seven homogeneous studies that were pooled, the overall risk of a contralateral ACL tear was 1.63 (95% confidence interval [CI]: 1.30 to 2.04) times that of a graft tear during the follow-up period.
Conclusions: At a minimum follow-up time of ten years, the risk of contralateral ACL tear significantly exceeds the risk of ACL graft rupture. Further work is necessary to understand and potentially reduce this risk of graft and contralateral ACL tears.
Level of Evidence: Therapeutic Level II. See Instructions for Authors for a complete description of levels of evidence.
The anterior cruciate ligament (ACL) is commonly injured and is the most frequently reconstructed ligament of the knee. Current reconstructive techniques generally result in a stable knee, which allows patients to return to desired activities in the short term; however, long-term outcomes are not universally good. A tear of the ACL graft or an ACL injury in the contralateral knee is a devastating event for the patient.
Risk of graft failure following ACL reconstruction has been reported to be between 2% and 6% at short-term (less than five-year) follow-up1-3. As expected, longer follow-up (five to ten years) has demonstrated a higher failure risk, ranging from 3% to 10% in various studies4-6. Several recent studies have reported that the failure risk at least ten years following ACL reconstruction is similar to the risk at five to ten years7,8. Data regarding the risk of injury to the contralateral ligament are sparser, but have shown that the risk of contralateral ACL tear within two years after surgery is similar to the risk of graft rupture3. Interestingly, a systematic review by Wright et al. noted that, in studies with at least a five-year follow-up, the injury risk for the contralateral limb is nearly double that of graft rupture in the surgically treated limb9.
The long-term risk of ACL graft rupture or contralateral ACL rupture is crucial information to share with patients who undergo ACL reconstruction. The primary aim of this systematic review is to compare the risk of ACL graft rupture with the risk of contralateral ACL rupture through the use of studies with prospective data collection and ten-year minimum follow-up. We hypothesize that (1) the contralateral knee is at higher risk than the reconstructed knee, and (2) the failure risk of both knees is higher than that noted at a shorter-term follow-up.
Materials and Methods
A search of MEDLINE and Scopus (Elsevier) was performed to identify all publications from January 1, 1966 through July 15, 2013 in which long-term outcomes of ACL reconstruction were reported. Searches including the terms “ACL” or “anterior cruciate,” “reconstruction,” and “long-term” yielded 480 articles in MEDLINE and 803 articles in Scopus. After elimination of 441 duplicate citations, the titles and abstracts of 842 articles were reviewed. The 799 studies that failed to meet inclusion and exclusion criteria (Table I) were eliminated. Review of the references of the remaining forty-three studies yielded three other potential studies for inclusion. A full-text review of the forty-six studies was undertaken, and sixteen studies were removed because of the following: they were duplicate publications (three studies), they reported a shorter follow-up of patients who were reported on elsewhere with longer follow-up (five studies), they included ACL repair and augmentation rather than reconstruction (one study), or they reported on open procedures (seven studies). Seventeen of the remaining thirty studies were excluded due to retrospective data collection and four were excluded because they did not report failure risk. The remaining nine studies made up the data set for this review7,8,10-16. The literature search is summarized in Figure 1.
Extracted data included study characteristics, patient demographics, associated injuries, surgical technique, graft choice, details of rehabilitation, length of follow-up, and graft and contralateral ACL failure risk. Failures were determined on the basis of the reports of the authors, with use of criteria established by each individual author. Data were extracted by two authors independently, and discrepancies were resolved by consensus17,18.
Collected data were tabulated and summarized. In order to compare the risk of graft failure with the risk of contralateral ACL rupture, a random effects model was developed in Review Manager 5, which used the method of DerSimonian and Laird to calculate an overall pooled estimate of effect19. Study heterogeneity was assessed qualitatively by comparing the populations and designs of individual studies, as well as quantitatively using a chi-square test. A sensitivity analysis was subsequently performed by adding any excluded studies back into the final meta-analysis, both individually and as a group, to determine if their exclusion had any effect on the pooled estimate of effect. A funnel plot was subsequently created to evaluate for the presence of publication bias.
The nine studies identified in the literature review include three randomized controlled trials (two comparing patellar tendon with hamstring grafts12,16, and one evaluating the effect of bracing during rehabilitation14), three prospective cohort studies8,11,15, and three case series with prospective data collection7,10,13. Study design was reported as was initially assigned by the authors and/or journal at the time of publication; however, all nine studies feature prospective data collection.
There were 3079 patients included in the nine studies, with a mean age at the time of ACL reconstruction of 26.0 years. Mean patient age for the individual studies ranged from twenty-three to thirty-eight years. Overall, 67.1% of patients were male, with the percentage in the individual studies ranging from 48% to 72%. The chronicity of the ACL injuries at the time of reconstruction varied among the studies. Exclusion criteria among the studies varied, but all studies excluded knees with multiple complete ligament injuries. Study demographic information is detailed in Table II.
Autograft tissue (patellar tendon or hamstring) was used in all studies via either an all-arthroscopic, arthroscopic-assisted, or mini-arthrotomy technique. Femoral tunnel drilling technique and fixation methods varied among the studies. Data regarding surgical technique are summarized in Table III. Seven of the nine studies reported details of their rehabilitation protocol. Most allowed immediate full weight-bearing, and brace use was limited to two studies for time periods ranging from three to eight weeks. Table IV details the rehabilitation protocols for each study.
Follow-up was available for 2682 (87.1%) of the 3079 patients initially enrolled in the studies, with individual study follow-up ranging from 65.6% to 96.9%. Mean follow-up in all studies was at least ten years and ranged from 10.3 to 16.9 years in the individual studies. As shown in Table V, the overall ACL graft rupture risk was 7.9% (211 of 2682) and ranged from 3.9% to 11.1% in the individual studies. The overall contralateral ACL injury risk was 12.5% (335 of 2682) and ranged from 0.6% to 22.7% in the individual studies. Two studies utilized patient populations that were qualitatively different than those in the other seven studies. The patient population in the Felmet study was much older (mean age, thirty-eight years) than the patient populations in the remaining studies10, and the study by Möller et al. was the only one that included more females than males14. Statistical heterogeneity was also noted when including these studies in a pooled analysis; therefore, data from these studies were not pooled. A forest plot (Fig. 2) demonstrates the differences in graft rupture risk and contralateral ACL rupture in each pooled study and demonstrates that the overall risk of a contralateral ACL tear was 1.63 (95% confidence interval [CI] = 1.30 to 2.04) times that of a graft tear during the follow-up period in these studies. Sensitivity analysis demonstrated no difference in the overall pooled risk ratio with the addition of the Felmet10 study (relative risk = 1.57; 95% CI = 1.18 to 2.08), the study by Möller et al.14 (relative risk = 1.57; 95% CI = 1.18 to 2.08), or both studies (relative risk = 1.49; 95% CI = 1.08 to 2.04). The funnel plot revealed a symmetrical shape, suggesting that publication bias was not a concern (Fig. 3).
The primary finding of this study was that the risk of ACL graft tear at a minimum follow-up time of ten years was 7.9%, while the risk of contralateral ACL injury during the same period was 12.5%. Upon pooling of the seven homogeneous studies, the overall risk of a contralateral ACL tear was 1.63 times that of a graft tear. These findings mirror those demonstrated in a prior systematic review of prospective studies9. The prior review demonstrated a graft failure risk of 5.8% and a contralateral ACL tear risk of 12.5% in the five studies with five to ten years of follow-up. The similarity between the failure risks in the two reviews with significantly different follow-up suggests that a relatively low proportion of grafts or contralateral ACL tears are torn beyond ten years after operation, as has been demonstrated by the authors of one longitudinal follow-up study7. It should be noted that the study by Shelbourne and Gray was included in both reviews8. However, exclusion of this study did not alter the graft or contralateral failure risk described by Wright et al.9.
This study is consistent with previously published work from several groups, indicating that the risk of contralateral ACL tear is higher than the risk of graft tear following ACL reconstruction7-9. In contrast, Hettrich et al., in reporting MOON (Multicenter Orthopaedic Outcomes Network) group data at six years after ACL reconstruction, demonstrated no significant differences in the risk of graft rupture as compared with contralateral ACL tear20. This difference may be due to the longer follow-up in the current review. Hettrich et al. did note a trend toward earlier graft rupture as compared with contralateral ACL tear, with 63% of their graft ruptures and only 40% of the contralateral ACL tears occurring within the first two years postoperatively20. Another potential factor known to influence the risk of both graft and contralateral ACL tear is postoperative activity level21. Only six of the included studies quantified postoperative activity level, with Tegner scores between four and six generally reported. Paterno et al. confirmed that the ACL reconstructed population is at a much higher risk of ipsilateral and especially contralateral ACL injury than a control population is22. While the reasons for this finding are not definitively known, the most commonly cited reason is the tendency for the uninjured contralateral knee to protect the injured knee and thus bear more load22,23. An alternative theory is that injury to one ACL can cause loss of proprioception and subsequent higher risk of further injury to either knee. Rehabilitation that is focused on the reconstructed knee may leave the contralateral limb at a higher risk of injury24. Paterno et al. demonstrated significantly increased ground reaction forces in the contralateral limb two years following ACL reconstruction25, and others investigators have demonstrated significantly altered kinematics of both involved and uninvolved sides following ACL reconstruction26.
The strengths of this study are the inclusion of only prospective studies with long-term follow-up. The overall follow-up rate is near 90%, which is a testament to the work of the authors of the included studies. All studies included intra-articular ACL reconstructions in which modern techniques (e.g., all-arthroscopic, arthroscopic-assisted, or mini-arthrotomy methods) were used. The sensitivity analysis demonstrated our findings to be robust, and the funnel plot showed no evidence of publication bias.
Limitations of the study include the definition of graft failure. For the purposes of this review, the failure risk quoted by the study authors was utilized. Failure criteria were generally poorly defined and inconsistent among the individual studies. As recently demonstrated by Crawford et al., these reported failure numbers likely underestimate the number of failed reconstructions relative to the number of failures demonstrated through objective clinical measures27. Unfortunately, only five of the studies provided objective laxity measures in a manner that could identify additional failures. In these five studies, a total of six additional potential graft failures could be identified, with no significant effect noted on the overall results. No studies provide sufficient detail with Knee injury and Osteoarthritis Outcome Scores (KOOS) to potentially identify failures on the basis of a KOOS quality-of-life (QOL) score of <44, as has recently been suggested28. These limitations may result in an under-detection of failed ACL grafts relative to contralateral injuries. Future long-term follow-up studies should include multiple assessments of failure, including laxity assessments and clinical outcome scores in addition to clinically diagnosed failures to aid future analyses of the true failure risk following ACL reconstruction. A final limitation of this study is that the heterogeneity in the reported data among the studies precludes a detailed assessment of the ways in which factors such as patient age, the chronicity of the ACL tear, sex, graft type, and activity level may have an impact on the risk of graft failure and ipsilateral ACL tear.
At a minimum follow-up time of ten years, the risk of contralateral ACL tear significantly exceeds the risk of ACL graft rupture. Further work is necessary to understand and potentially reduce this risk.
The authors acknowledge that the project was partially funded by grant numbers 5R01 AR053684 (K.P.S.) and K23 AR063767 (R.A.M.) from the National Institutes of Health/National Institute of Arthritis and Musculoskeletal and Skin Diseases.
Source of Funding: This study was supported by a grant from the National Institute of Health - National Institute of Arthritis and Musculoskeletal and Skin Diseases.
Investigation performed at The Ohio State University Medical Center, Columbus, Ohio
Disclosure: One or more of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of an aspect of this work. In addition, one or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.
- Copyright © 2015 by The Journal of Bone and Joint Surgery, Incorporated