|Home | About | Journals | Submit | Contact Us | Français|
To systematically review the current evidence for conservative and surgical treatment of anterior cruciate ligament (ACL) tears in skeletally immature patients.
A systematic search of PubMed, CINAHL, EMBASE, CCTR, and CDSR was performed for surgical and/or conservative treatment of complete ACL tears in immature individuals. Studies with less than six months of follow-up were excluded. Study quality was assessed and data were collected on clinical outcome, growth disturbance, and secondary joint damage.
We identified 48 studies meeting the inclusion criteria. Conservative treatment was found to result in poor clinical outcomes and a high incidence of secondary defects, including meniscal and cartilage injury. Surgical treatment had only very weak evidence for growth disturbance, yet strong evidence of good postoperative stability and function. No specific surgical treatment showed clearly superior outcomes, yet the studies using physeal-sparing techniques had no reported growth disturbances at all.
The current best evidence suggests that surgical stabilization should be considered the preferred treatment in immature patients with complete ACL tears. While physeal-sparing techniques are not associated with a risk of growth disturbance, transphyseal reconstruction is an alternative with a beneficial safety profile and a minimal risk of growth disturbance. Conservative treatment commonly leads to meniscal damage and cartilage destruction and should be considered a last resort.
Level IV, systematic review of Level II, III, and IV studies.
The management of anterior cruciate ligament (ACL) injuries in adults attracts a considerable share of interest in ongoing research. However, the management of ACL tears in children is less well studied (1). A considerable surge in the incidence of such injuries, paired with the substantial spectrum and gravity of secondary damages, underlines the necessity of more, in-depth research in this field (2-6).
Historically, transphyseal ACL reconstruction has been avoided in skeletally immature patients because drilling across the growth plate carries a risk of future physeal malfunction and resultant growth disturbance and angular deformity (4,7). Thus, traditional care of the skeletally immature patient with an ACL tear has relied on bracing and activity modification until the young athlete is close enough to skeletal maturity to undergo transphyseal reconstruction (1, 8). Recently, surgeons have developed physeal sparing ACL reconstruction techniques, including transepiphyseal tunnel placement (9), and intra- and extra-articular stabilization without transosseous tunnels (10). Clinical results of each of these techniques have been reported individually; however, less is known about how these techniques compare with transphyseal reconstruction or conservative treatment in this patient population.
Our hypothesis was that there would be significant differences in patient outcomes with each different treatment method. A systematic review of the literature to address this hypothesis was performed.
This systematic review had three objectives. The first was to comprehensively and systematically review the current evidence for operative versus nonoperative treatment of immature patients with ACL tears. The second objective was to systematically assess the outcomes of different types of surgical treatment available to these patients. The third objective was to review the study quality and level of evidence of the current literature for management options of immature ACL injuries.
The systematic review was performed following the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) statement (11, 12). The PRISMA statement (www.prisma-statement.org), put forward by the CONSORT group (www.consort-statement.org), is a evidence-based guideline for conducting and reporting systematic reviews, and was formerly known as the QUOROM (Quality Of Reporting Of Meta-analysis) statement (13).
Studies were included if they reported on the clinical outcomes of surgical and/or conservative treatment of complete ACL tears in immature individuals. Immature individuals were defined either as patients with radiological proof of open physes, or those at appropriate Tanner staging (stage IV or below). Chronological age was not used as an inclusion criterion. Studies with less than 6 months of follow-up were excluded, as were studies of partial ACL tears and tibial spine avulsions.
The online databases PubMed, CINAHL, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL), and Cochrane Database of Systematic Reviews (CDSR) were searched for relevant publications. All dates and languages were included. The last search was performed on August 31, 2010.
The search algorithm was “((ACL) OR (anterior cruciate ligament)) AND ((young) OR (child) OR (pediatric) OR (paediatric) OR (immature)) AND (“humans”[MeSH] NOT “animals”[MeSH])” and was replicated using the keywords as MeSH terms as well (Figure 1). All searches were unlimited, i.e., considering publications in all languages and regardless of publication date. In addition to the online searches, the bibliographies of the included studies were reviewed by hand to identify further publications.
Title and abstracts from all search results were screened for eligibility. Studies were excluded if title and/or abstract clearly refuted eligibility. Full texts were obtained for all studies matching the inclusion criteria and all with unclear eligibility. The obtained full texts were reviewed to confirm eligibility. All study selections were done independently in duplicate and cross-referenced. Disagreement was resolved by consensus.
All identified studies were categorized by type of treatment (conservative, surgical/intra-articular, surgical/extra-articular) and level of evidence (I to V) using the ranking system published by the Journal of Bone and Joint Surgery, American edition (http://www2.ejbjs.org/misc/instrux.dtl) (14, 15). Data were extracted independently and in duplicate. Duplicate data extractions were compared for difference and disagreement was resolved by consensus.
Data were extracted for the endpoints limb-length or angular deformity, clinical outcome (scores), secondary problems, and anteroposterior (AP) laxity to allow for gross comparison between techniques. Levels of Evidence were assessed for all included studies.
The risk of bias was assessed through categorization by level of evidence. We decided against using composite scores of study quality because these scores have been shown to be unreliable in some of the included study types, and because there are no scores that allow a valid assessment across different study designs (16). Studies with particularly high risk of bias are pointed out explicitly in “Study Description” subsection of the “Results” section.
Given the substantial clinical heterogeneity, the poor quality of the evidence from the overall literature, and limited number of studies reporting the same outcome measures, we did not perform a quantitative data synthesis, but report all data descriptively as a systematic review. To provide a more comprehensive overview of the literature, we included studies from all levels of evidence. As subanalyses, we also analyzed the data for the youngest 15th of patients, and for the highest level of evidence studies (Level II and III), individually. Results are given as mean ± SD.
Our search produced 247 results in tota;. 92 publications were obtained and reviewed based on the criteria described above; 2 additional papers were identified by bibliographic cross-reference. Finally, 48 papers reporting on a total of 1,217 patients who were followed-up for an average 44.7 ± 18.7 months were included in the analysis (2, 4, 7-10, 17-58) (Figure 1). These articles were published between 1986 and 2010 in English, German, and French.
The average age across all studies was 13.3 ± 1.2 years. All but 2 studies reported at least radiologically confirmed open physes as the criterion for immaturity; 16 studies included a Tanner score for description of maturity. Subgrouping by method of determination of skeletal age showed an average age of 13.2 ± 1.1 years for those studies reporting open physes and 12.9 ± 1.5 years for those reporting Tanner stages (P = .510). A median of 19 patients (range, 1 to 129; interquartile range, 10 to 43) are given per study. Thirty-nine studies reported on intra-articular stabilization, and 5 on extra-articular stabilization, although 3 of these included procedures with both an intra- and an extra-articular component. Twelve studies reported on natural history or had a conservative treatment group in their populations. Table 1 summarizes the characteristics of these studies (Table 1).
The level of evidence for the included papers ranged from Level II to Level IV. There was one Level II study (33), 10 Level III studies (8, 17, 28, 30, 33, 35, 42, 43, 48, 55, 57), and 37 Level IV studies (Table 1). Most studies were longitudinal analyses of single cohorts without controls and without randomization; this situation is representative of the studied field (59). We categorized all studies by level of evidence to underscore the differences in the likelihood of bias of their respective results. However, it should be noted that the objective of this systematic review was to give as comprehensive an analysis as possible of all current evidence and that longitudinal designs are an adequate design to study feasibility and long-term outcomes of (surgical) procedures.
Twelve articles reported on conservative treatment and natural history (8, 17, 19, 20, 30, 33, 35, 42, 43, 45, 48, 55, 58), 8 of which were Level III studies (8, 17, 30, 33, 35, 42, 43, 48, 55). Six of these studies compared conservative treatment with a surgical treatment group (17, 30, 42, 43, 48, 55). These reports provide data for 476 patients followed-up for 52.7 ± 11.9 months on average, and they consistently show high proportions of unstable, symptomatic patients with early, severe meniscal degeneration and cartilage defects requiring surgical stabilization (average of 50.2%, range 17.4% to 87.6%) during the period of observation. Interestingly, in contrast to the others, one study found no increase in secondary injury rates in immature patients with conservative treatment and delayed surgical repair of the ACL deficiency after the physes had closed (8).
Three types of surgical procedures are presented in the current literature: (1) intra-articular, transphyseal, transosseous reconstruction, (2) intra-articular, physeal-sparing, transosseous reconstruction, and (3) combined intra- and extra-articular, physeal-sparing, extraosseous stabilization.
Thirty-eight studies presented results of intra-articular, transosseous stabilization. The average age for the patients in this group was 13.2 ± 1.2 years. Nine reports describe physeal-sparing techniques (2, 9, 10, 31, 32, 35, 47, 49, 53), and two describe physeal-sparing tunnel placement on the femoral, but not the tibial side. The remainder (n=27) reported on transphyseal reconstruction. Six studies offered comparison between surgical and conservative treatment (Table 2) (17, 30, 42, 43, 48, 55), and 3 between immediate and delayed ACL reconstruction (Table 3) (8, 33, 57). These studies reported better Lysholm scores (83 v 7), better subjective outcomes, and fewer secondary pathologies after immediate surgical reconstruction. Thirty-one articles report on ACL stabilization with at least one transphyseal tunnel in 479 patients of 13.6 ± 0.9 years of age followed for 42.32 ± 18.7 months on average (7, 9, 17, 18, 20-23, 25-30, 33, 34, 36-44, 48, 50-52, 54, 55). In this group of almost 500 subject, 3 angular deformities and 2 limb-length discrepancies (1.3 cm and 2 cm) were observed. Another 10 patients had MRI results consistent with physeal narrowing, but without angular or limb-length deformities. Across these studies, the Lysholm scores for the surgically treated patients ranged from 83 (at 63 months) to 98 (at 78 months). There was no significant difference in results with the use of one transphyseal tunnel (tibia only) versus two tunnels (tibia and femur). No other secondary problems attributable to the reported type of procedures were reported.
Five articles include at least one group of patients undergoing physeal-sparing, intra-articular, transosseous stabilization, which is usually done by placing tunnels proximally to the tibial physis and distal from the femoral physis (31, 32, 47, 49, 53). The average age of this group was 12.7 ± 1.8 years. No limb-length or angular deformities were seen in this group. Unfortunately, these authors did not use the Lysholm score, but data on the OAK (Orthopädische Arbeitsgruppe Knie) score (98 patients) and IKDC (International Knee Documentation Committee) score (96 patients) are available. The average difference in AP laxity compared with healthy, contra-lateral knees was 1.5 mm.
The results of extra-physeal stabilization techniques in 106 patients, with an average age of 12.1 ± 1.2 years of age, were presented in 6 reports (2, 10, 24, 28, 30, 42). Strictly speaking, these were all combined intra- and extra-articular, physeal-sparing, extraosseous reconstructions, i.e., modifications of the technique designed by McIntosh and Darby (60). Briefly, the iliotibial band was incised, tubularized, and brought to the over-the-top position by wrapping it around, and suturing it to, the lateral femoral condyle. At this position, it was inserted into the knee through the posterior knee capsule. From there the iliotibial band was brought to the front of the tibial ACL footprint, led through a groove made underneath the intermeniscal ligament, and attached to the tibial cortex with staples or sutured to the periosteum. This configuration created extra-articular, anteriorposterior stabilization between Gerdy’s tubercle and the lateral femoral condyle as well as an intra-articular stabilizer against AP translation and rotation. No growth deformities were seen in these patients at an average follow-up of 47.3 ± 20.7 months. Lysholm scores at the latest follow-up were in the range of 94.3 to 97.4, with no instabilities. Brief used a somewhat different approach in his study with a semitendinosus and gracilis autograft left in situ at its tibial insertion and passed underneath the anterior horn of the medial meniscus and attached to the femur with staples (24). All of these patients reported satisfying results, but none returned to sports without a brace. One study included both extraphyseal stabilization and transphyseal reconstruction and reported no difference in functional outcomes at 32 months of follow-up (28).
Six studies present data on the youngest 15% of patients, ranging from 10.3 to 12.1 years of age at Tanner stage I and II (2, 31, 37, 61-63). Four studies used either intra-articular, physeal-sparing, transosseous stabilization (31, 62) or the modified McIntosh technique (intra- and extra-articular, physeal-sparing, extraosseous reconstruction) (2, 61), and 2 studies used intra-articular transphyseal reconstruction (37, 63). Liddle et al. (37) followed-up on 17 prepubescent (Tanner I and II) patients aged 12.1 years (range 9.5 to 14.0) for 44 months (range 25 to 100) after transphyseal reconstruction with a quadrupled hamstring graft, which produced 15 excellent and 1 good result. There were 2 complications, 2 graft re-rupture during a playground accident, and 1 superficial wound infection, but no leg-length discrepancies. One patient developed a 5° valgus deformity without functional disturbance according to these authors. Streich et al. (63) treated 12 patients nonoperatively and 16 surgically with semitendinosus and gracilis grafts (median age, 11 years; range, 9 to 12 years) and followed them up for 70 months. At the final follow-up, the patients had grown an average of 20.3 ± 6.9 cm, but no angular deformities or leg-length discrepancies (defined by Streich and coworkers as ≥ 15 mm side-to-side difference) were observed. Unsurprisingly, the surgical group had significantly better results for laxity and functional scores. Seven (58%) of the 12 patients receiving nonoperative treatment proceeded to undergo surgical stabilization within 2 years after the initial injury.
Ten studies ranked as Level II and Level III evidence (8, 17, 28, 30, 33, 43, 48, 57, 64). These studies compared surgical with conservative, nonsurgical treatments (n=6), immediate with delayed surgical treatment (n=2), and surgical treatment in mature with immature patients (n=1) or two different surgical treatments (n=1). Table 4 summarizes their outcomes in detail. Briefly, in alignment with the overall findings, and the findings for the youngest 15% of patients, the highest level of evidence studies unanimously report significantly better results in clinical scores and knee laxity after surgical ACL reconstruction when compared with conservative treatment. At the same time, there was no difference in the risk of growth disturbances. The studies that looked specifically at the timing of surgical repair, support immediate treatment over delays.
This systematic review of conservative versus surgical treatment provides evidence that surgical treatment of the immature, torn ACL produces superior clinical outcomes in stability and in the prevention of secondary injury. Few risks are associated with surgical stabilization, while many patients initially selected for conservative treatment suffer from secondary damages and cross over to surgical stabilization, thus potentially combining the risk profiles of both types of treatments. The specific procedure chosen for surgical stabilization appears to have less clinical impact than the selection of surgical treatment.
Currently, many consider nonsurgical treatment to be the most appropriate initial approach to the torn ACL in immature patients until they reach skeletal maturity (1, 8). The rationale of this approach is to allow the physes to close before a surgical intervention, primarily because it is feared that transphyseal tunnel placement would cause sufficient growth plate damage to result in limb-length differences or angular deformities due to the formation of bony bridges along the tunnel across the growth plate (4, 7). The exact mechanisms and risk factors for such deformities have been the subject of a number of animal studies suggesting that the risks of growth disturbance can be minimized by adherence to several basic principles. Factors associated with increased risk of physeal malfunction in animals include posterior tunnel placement (65, 66), a high ratio of tunnel diameter to physeal surface area (31, 32, 67), excessive graft tensioning (68), incomplete tunnel filling by the graft (69, 70), and graft fixation across the physis (71). If these factors are considered, transphyseal reconstruction can be performed in immature ovine knees without subsequent growth disturbance (72). In human patients, the vast majority of growth disturbances and angular deformities have been associated with graft fixation devices or bone-plugs leading to bony bars across the lateral distal femoral physis (54% of angular deformities) or epiphysiodetic effects of fixation devices crossing the tibial physis (27% of angular deformities) (4). Other frequentistically noteworthy reasons for deformities included tunnel placement and tunnel diameter (4).
On the other hand, it has been reported repeatedly and consistently that conservative treatment leads to recurrent instability and results in increased intra-articular damage, specifically meniscal damage and cartilage degeneration (30, 35, 45). Hence it is not surprising that most patients treated conservatively eventually press for ACL reconstruction (average 50.2%, range 17.4% to 87.6%), some even when still at a young age (30, 35, 45). In light of these facts, conservative treatment might be an option for a few, very carefully selected, highly compliant patients with low demands and no other pathologies (8), but the notion that nonsurgical treatment is the most suitable approach for all immature cases, especially in an active patient, deserves critical re-evaluation. A number of studies have followed-up on immature patients after ACL reconstruction, using various techniques. What stands out from these studies is that surgical treatment of the immature, torn ACL produces convincing, beneficial results, at least in the short and intermediate term. Streich et al., in the most recent of the included studies, allocated only those patients with concomitant injury to surgical treatment and compared them with nonoperatively treated patients with isolated ACL ruptures without evidence of other injuries (55). Yet, interestingly, even this hand-selected group of conservatively treated patients, with unequivocally better initial conditions, performed significantly worse than their surgically treated counterparts with complex and extensive injuries, suggesting that any ACL rupture, even if isolated and without concomitant injuries, would benefit from surgical treatment.
Physeal-sparing procedures, both intra- and/or extra-articular, have evolved into valuable alternatives (2, 10, 61). Recent studies by Kocher et al. have shown that postoperative results of extra-physeal iliotibial band reconstruction are equivalent to transphyseal ACL reconstruction. Although this treatment was initially planned to be a temporizing procedure, it has functioned as a definitive reconstruction for a number of patients (10, 61). Similarly, a comparative study of physeal sparing ACL reconstruction with autologous fascia lata (n = 12) and transphyseal ACL reconstruction (hamstring, bone–patellar tendon–bone, and quadriceps tendon, n = 12 each) showed no differences in terms of functional outcome or the occurrence of growth disturbances (28). Lastly, transepiphyseal graft placement with tunnels placed in the tibial and femoral epiphyses has shown good outcomes in 8 patients at 4 ± 2 years postoperatively (9). However, to date there is no evidence on the effects of drilling close and parallel to the growth plate, which not only has a risk of directly injuring the physes but also of thermal damage from friction heat that cannot be seen at the time of the procedure, but which might manifest later (54, 73).
In addition to the overall systematic review, which had a liberal inclusion policy, we also separately analyzed the youngest patients and studies with the highest-level evidence. The findings from these subgroups were similar to that of the overall cohort of studies. Even for the youngest patients, there was no significant increased risk of growth deformities with surgical treatment, but there were significantly better outcomes in knee stability and function 70 months after surgical treatment compared with conservative treatment. Equivalent results were seen for Level II and III studies, which constitute the highest level of evidence available. In summary, the findings in these subgroups suggest that our overall interpretation of surgical treatment being more effective and no more complication prone than conservative treatment is accurate and valid, even for the youngest patients in this collective, and under the most stringent criteria used in the current literature.
Two recent, noteworthy publications in Arthroscopy deal with the management of immature ACL ruptures (73, 74). Keading et al. published a systematic review of 13 studies (192 patients; median age, 13 years; follow-up, 45.6 months) of varied surgical treatments for ACL injuries in preadolescent patients (boys <15 years, girls <14 years of age, Tanner stage I to III) (74). They reported no differences in patient-reported outcomes, AP laxity, or leg-length discrepancy or angular deformities between physeal-sparing and transphyseal reconstruction for any of the surgical treatments, which is in alignment with our findings. These authors point out that they could not accrue sufficient data on Tanner I patients to reach a valid conclusion. However, in our study, by inclusion of non-English-language publications, we were able to produce some data on Tanner I and II patients that support the use of surgical reconstruction in the management of ACL tears even in those younger patients. In a second article, Frosch et al. presented results from a meta-analysis of 55 original studies (935 patients; median age, 13 years, median follow-up, 40 months) on surgical treatment options for immature ACL tears (73). This study showed that the risk of leg-length discrepancy or angular deformity after surgical treatment of an ACL tear in a skeletally immature individual was 1.8% (95% CI, 0% to 3.9%). The risk of rerupture in the same population was 3.8% (95% CI, 2.6% to 5.2%). However, this study included no comparison of surgical treatment with conservative treatment (Table 2). Interestingly, Frosch and colleagues found evidence for a significantly higher risk of angular deformity after physeal-sparing, transosseous reconstruction compared with transphyseal, transosseous reconstruction, with a risk ratio of 0.34 (95% CI, 0.14 to 0.81) in favor of transphyseal reconstruction. The authors argue that this difference in risk might stem from detrimental effects of drilling parallel to the growth plate or from a pressure/obstacle effect of the implant on the expanding growth plate (73).
Our study has potential shortcomings. First of all, the bulk of the literature in this field is situated at the base of the pyramid of levels of evidence and is most likely subject to some level of confounding and/or bias. We used levels of evidence to categorize the included studies, but decided against the use of composite quality scores because of the variations in study designs in this group of studies (16).
Lack of statistical power was a feature of several of the studies. This was likely due in part to the relatively small number of patients as well as the known biologic variability inherent in clinical outcome studies. Thus, relative effectiveness of several surgical techniques may be difficult to assess in each individual study. The systematic review was helpful in comparing techniques because cohorts and case series are appropriate tools to investigate long-term outcomes. It is also possible that some studies were not published in this controversial area, thus causing publication bias.
Finally, the biggest limitation of this study is the definition of skeletal immaturity. Tanner stages, physeal closure, and other parameters of skeletal age have been used in addition to chronologic age, but there is no universal method across the current literature, which complicates direct comparison of patient populations. However, our findings show convincing consistency for outcome differences across different age groups, suggesting that our collective was homogenous enough to insure valid conclusions.
The results of our systematic review of the current evidence for management of immature ACL tears suggest that early surgical treatment results in more favorable outcomes than conservative management. Thus, surgical stabilization should be considered as the first line of treatment for immature patients with ACL tears. The existing literature suggests that transphyseal reconstruction can be safely done in this population if a few rules are considered, and there are physeal-sparing procedures that provide excellent results with less theoretical risk to the growth plate. Conservative or delayed surgical treatment, which carries an increased risk of secondary joint injury, should be reserved for very compliant patients with both low demands and no other pathologies.
Disclosures: Item 1. Board member/owner/officer/committee appointments:___M.M.M. is a founder of Connective Orthopedics _______
Item 2. Royalties:___n/a_______
Item 3. Speakers bureau/paid presentations:___ n/a _______
Item 4A. Paid consultant or employee:___P.V. is a consultant for Connective Orthopedics_______
Item 4B. Unpaid consultant:___ n/a _______
Item 5. Research or institutional support from publishers:___ n/a _______
Item 6. Research or institutional support from companies or suppliers (data generated from such studies must be unrestricted):__This study was supported by NIH NIAMS Grant No. AR 054099. _________
Item 7. Stock or stock options:__M.M.M. is a stockholder in Connective Orthopedics________
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.