Introduction
In the last few years, non-operative management after anterior cruciate ligament (ACL) injury has gained a lot of popularity and surgery may not always be recommended. Specific criteria such as age, desired sport and work activities, intensity of sporting activities and associated lesions might be able to inform patients and practitioners on the possibility of non-operative management. A battery of functional tests and Patient Reported Outcomes Measures (PROM) have been shown to identify patients who might respond well to conservative treatment (copers) versus those who might not (non-copers).
In parallel to studies focusing on outcomes without ACL reconstruction, there is emerging evidence that the ACL may spontaneously heal offering similar perceived quality of life and sport participation than patients with reconstructed ACL in the long term.1 Within this context, a review of this new literature is necessary to best inform patients of the options available to them.
What determines whether an ACL can heal?
It is well documented that the ACL attempts to heal, and a decrease and reduction in tibial translation may be observed in some patients.2 However, the process of healing is not considered optimal; the ACL may not heal onto the anatomical femoral attachment, and different patterns of scar formation may influence ligament length and subsequent knee stability. In a retrospective study, Costa Paz et al. demonstrated the possibility of ACL healing using magnetic resonance imaging (MRI).3 It should be noted that two patients over fourteen had a re-injury (and needed a reconstruction) and 30% had knee-related deficiencies on clinical examination (Lachman and pivot shift tests).3These results are similar to Fujimoto et al. who showed that 26% of patients with a hypothetical ACL healing (MRI and KT-2000 arthrometer were performed) had to subsequently undergo reconstruction.3,4 Recently, a secondary analysis of the KANON trial showed that within the non-reconstructed group, those who were considered to have healed ACL’s based on MRI findings, had better functional scores than the rest of the non-reconstructed group as well as the ACL reconstructed group at 2 years follow-up.1 These results may challenge current clinical practices and management of ACL injury. However, the term “healed” should not only refers to ACL’s fibers continuity observed on MRI but also to functional recovery. The “healed” ligament may have different mechanical properties than the original. This could explain why within the healed ACL group 25% had an abnormal pivot shift test, 56% had an abnormal Lachman’s test and an increased tibial translation was observed in laximetry testing.
As highlighted in a recent systematic review, studies on this topic are of low quality and there results could not lead to a reliable and generalizable conclusion.5 This review also pointed out that healing capacity of the ACL may be dependent on the rupture’s location and the conservation of the sheath of the ACL. Proximal ruptures having more chance to heal than distal ruptures. Another important factor observed by authors was that in many included studies rehabilitation protocols were not detailed. Rehabilitation protocols are different and have the potential to alter the outcomes between groups in studies.5 For example, Razi et al. recommended from the first week of injury to perform active range of motion out of brace, isometric quadriceps exercise, close kinetic chain exercise based on each individual’s pain tolerance, and stationary biking from the third week. In patient with valgus knees, they delay weight bearing until 6 weeks.6 During the acute phase Fujimoto et al. allowed early range of motion and quadriceps muscle strengthening exercises with a brace that had a 20° initial range of motion restriction. Weight-bearing was also allowed, as tolerated, with crutches initially. Full weight-bearing without the use of crutches was generally achieved within 4 weeks after the trauma.4 According to Jacobi et al., full weight bearing was allowed from the start of the treatment. Range of movement to the extent possible in the brace was allowed, giving patients a range of flexion of about 0° to 100°. Removal of the brace was allowed in 90° of knee flexion (sitting position) without quadriceps contraction. With the knee in flexion it was also the recommended position to take a shower. After four months, the brace was removed and exercises and physiotherapy were started to aid the recovery of muscle strength and full mobility. Sporting activity, including cutting and pivoting, was allowed after six months.7
According to Blanke et al. ACL healing happened in low-demand patients with femoral single bundle lesions without increased posterior tibial slope.8 In general, evidence is lacking regarding which criteria may indicate whether ACL healing is possible after an injury.5 The healing process of the ACL is still poorly understood, and it is difficult to know which patients are likely to have a healed ACL. It is also questionable whether some ACLs healing are not confused with ACLs scarring to the posterior cruciate ligament. This may happen in studies using MRI outcomes instead of an arthroscopic assessment which is the gold standard for ACL rupture.
Functional Knee Brace Management
To support ACL healing, some current protocols propose functional bracing management for periods of up to 12 weeks.4,7,9,10 Few studies have assessed the effect of functional bracing management on ACL healing.4 Fujimoto et al. showed that functional bracing management can help ACL healing but in a population with low intensity of physical activity.4 Functional bracing management seems to reduce ACL strain, which may improve the healing process. More recently, a case series in which they immobilized patients at 90° ok knee flexion for four weeks after ACL injury was published.10 The purpose was to diminish the distance between the origin and the insertion of the ACL to favor the healing process. They obtained interesting result with 90% of patients showing signs of ACL healing at 3 months post injury (MRI and Lachmann test). However, 50% were classified as grade 2-3 on the ACL Osteoarthritis Score which may be indicative of a non-functional ACL explaining significantly lower scores of this group on the ACL-Quality Of Life score.
It is to remember that current guidelines do not recommend functional bracing management after ACL injuries or ACL reconstruction. Strict bracing, as used in the case series study, can lead to knee joint disuse, which may alter muscle activation, spinal excitability and intracortical inhibition.10,11 The role of functional bracing on postural control is not clear either. Birmingham et al. concluded that functional bracing may improve performance during simple tasks but not during more functional or daily living tasks.12 From a broader perspective, a systematic review looked at the effect of functional bracing on patient-reported outcome measures and functional outcomes. No difference was observed between patients who wore a knee brace after ACLR and those who did not.13
Overall, there is very limited evidence that functional bracing influences ACL healing or is of any benefit for the patient and wearing a brace for a prolonged period may have undesirable effects.
Return to sport and spontaneous ACL healing
Filbay et al. presents the most advanced study in terms of objective results on spontaneous ACL healing with their secondary analysis of the KANON trial. Two-year outcomes were better in the healed ACL group (n=16) compared with the non-healed group (n=14) (mean difference (95% CI) KOOS-Sport/Rec: 25.1 (8.6-41.5); KOOS-QOL: 27.5 (13.2-41.8)).1 It is to notice that the battery of tests used in this study (KOOS-Sport/Rec, KOOS-QOL, KOOS pain, KOOS symptoms, Tegner Activity Score, KT-1000, Lachman and pivot shift tests, radiography) does not correspond to validated return to sports evaluations found in the current international literature,1 and KOOS subscales use as outcome measure in this study has been shown to be of poor responsiveness for patients with ACL injury.14 Objective data on muscle strength, functional tests and the psychological aspects of return to sport are lacking. Regarding the level of activity, authors recorded a mean pre-injury Tegner score of 8, which corresponds to high-intensity activity level. Unfortunately, they didn’t compare Tegner scores from the 2 and 5-year follow-up between healed and non-healed group. Similarly, the addition and comparison of a pre- and post-Marx activity scale would have provided information on the homogeneity of the groups in terms of their involvement in pivoting sports, which are responsible for most ACL injuries. It should also be noted that professional athletes were excluded from this study. It is therefore impossible to extrapolate the results to this specific population.
Patients’ perceived function is very well represented in their study and it is a fundamental aspect of treatment’s success. However, the authors may not have chosen the most appropriate questionnaire to assess this aspect. Future research also needs to include more objective data on strength, function, and psychological aspects. This is especially true for populations involved in pivoting activities with a high level of intensity. To date, there is a lack of evidence regarding return to sport, especially at the elite level, after ACL ‘healing’ and too much uncertainty to recommend this type of protocol.
Conclusion
Different options to treat patient with an ACL rupture exist. Non-surgical options may not be suitable to every patient and should be taken with extreme caution and truthfully discuss with the patient and within the medical team. There is evidence that conservative treatment can be successful in the general population, with some people healing their ACL. This is not the case for elite athletes. The emerging evidence regarding the ability for the ACL to heal is intriguing and may change clinical practice in the future, but we urge clinicians to take these results with extreme caution as this may only be suitable for a very small percentage of the population. Much more research is needed before recommendations for this option can be generalized.