RETURN TO PLAY AFTER ANTERIOR CRUCIATE LIGAMENT RECONSTRUCTION

The anterior cruciate ligament (ACL) tear represents more than half of all knee injuries in sports that involve body rotations and sudden changes of direction. Discharging the athlete for return to play (RTP) post-ACL reconstruction (ACLR) is a difficult task with multidisciplinary responsibility. For many years, a six-month period post-ACLR was adopted as the only criterion for RTP. However, it is now suggested that RTP should not be exclusively time-based, but to clinical data and systematic assessments. Despite the importance of post-ACLR factors for RTP, pre- and peri-ACLR factors must also be considered. Historically, ACLR is performed with the hamstring or autologous patellar tendons, although the choice of graft is still an open and constantly evolving theme. Anterolateral ligament reconstruction and repair of meniscal ramp tear associated with ACLR have recently been suggested as strategies for improving knee joint stability. Subjective questionnaires are easy to apply, and help identify physical or psychological factors that can hamper RTP. Functional tests, such as hop tests and strength assessment by means of isokinetic dynamometers, are fundamental tools for decision making when associated with clinical evaluation and magnetic resonance imaging. Recently, the capacity to generate force explosively has been incorporated into the muscle strength assessment. This is quantified through the rate of torque development (RTD). Due to characteristics inherent to the practice of sport, there is an extremely short time available for produce strength. Thus, RTD seems to better represent athletic demands than the maximum strength assessment alone. This review investigates the pre-, peri- and post-ACLR factors established in the literature, and shares our clinical practice, which we consider to be best practice for RTP. Descritores: Ligamento cruzado anterior; ao Articulação do


INTRODUCTION
The rupture of the anterior cruciate ligament (ACL) represents up to 64% of all knee injuries in sports that involve rotations and changes in direction. 1 With the evolution of surgical techniques and rehabilitation/ training protocols, there is currently an increase in the number of athletes who return to pre-injury sports after surgery for ACL reconstruction (ACLR). 2,3 The return to play (RTP) can be defined as the time necessary for the post-ACLR athlete to be able to compete at the same previous level. 4 For many years, the determining criterion for the RTP post-ACLR was the six-month period elapsed. 3 However, recent findings discourage using the temporal criterion in isolation. [5][6][7][8] The current trend is to characterize the RTP as a continuous process with three subphases: return to participation, return to playing and return to performance. These subphases should be guided not only by time frames, but by objective clinical criteria and physical testing. 6 Post-ACLR patients commonly demonstrate neuromuscular deficits in the lower limbs, 6,7,9 which can last up to two years. 10 Such deficits are considered a risk factor for re-rupture, ACL rupture in the contralateral limb and early development of osteoarthritis. 5,8 Thus, monitoring post-ACLR patients through physical tests with objective control parameters is mandatory to release the patient/athlete to the RTP without restrictions. 11 Although proposals have been made, 8,12 there is still no standardized battery of tests. 13,14 However, it is not uncommon for studies to report the use of assessments of the muscular strength of knee extensors (EXT) and flexors (FLEX) 6,15 and/or the different variations of hop tests. 6,16 In addition to these, it is common to use self-assessment questionnaires that are easy to apply and low cost.
Shelbourne et al. 17 followed more than 1400 athletes, for a minimum period of 5 years, submitted to ACLR to assess the rate of recurrence of ipsilateral and contralateral injuries. They observed that women have more than twice the chance of having a contralateral injury, as well as about 17% of athletes under 18 years old suffered graft rupture and that number progressively decreased with advancing age. In professional athletes, the main concern is neglected associated injuries and accelerated rehabilitation protocols. The anxiety of the athlete and the coaching staff can anticipate their return to competitions and be decisive for the graft failure. The graft rupture is potentially catastrophic and the RTP rate is much lower when dealing with a surgical revision. Among the causes of failure, four must be considered before scheduling a review: increased tilt of the plateau (tibial slope), varus alignment, anterolateral ligament injury (ALL) and meniscal injury. Currently, the correction of these factors is encouraged when identified. 18 In order to be successful post-ACLR, multidisciplinary rehabilitation/ training is indispensable together with the determination of objective criteria. Although the literature has advanced, there are still no established and widely accepted parameters for the athlete's release to the RTP, making this a challenging decision. Therefore, the objective of this review was to descrive and discuss factors in the pre-, per-and post-operative period established in the literature, as well as to share our clinical practice, which we consider to be the best practices for a safer RTP.

Graft type
The graft determination for primary ACLR is still a controversial topic. 19 Many factors are studied in order to determine the best graft, as well as its ideal characteristics, such as diameter, failure rate and RTP. 20 Bjornsson et al. 20 compared the use of hamstring tendons to the patellar tendon and did not find different rates of failure or complications. Recently, the use of quadriceps tendon has received attention for showing equivalent functional results, patient satisfaction and low rates of failure and complications at the donor site. 19 The diameter of the graft seems to have an influence on the ACLR's risk of failure. Grafts of eight millimeters in diameter or more have lower failure rates than thinner grafts. 21 Legnani et al. 22 found no significant difference in the RTP rate for patients who used autologous contralateral hamstrings compared to the use of allografts. However, patients who used an autologous graft demonstrated an RTP in 7.7 months, while those who used a tissue bank graft took an average of 9.8 months. Despite all efforts in the search for the best graft, the autologous hamstring and patellar tendons continue to be the main choices.

Reconstruction of the anterolateral ligament and repair of the meniscal ramp lesion associated with the anterior cruciate ligament reconstruction
The importance of ALL in post-ACLR knee stability is a widely discussed topic. [23][24][25] DePhillipo et al., 26 observed only 1.1% graft rupture when ACLR was associated with ALL reconstruction (ALLR), against 16.3% of failures in isolated ACLR. Delaloye et al. 26 reported excellent clinical results with a two-year follow-up post-ACLR combined with ALLR, with a graft rupture rate of 10.8% (quadruple hamstring graft), 6.8% (patellar graft) and only 4.1% with the associated ALLR, regardless of the graft used in the ACLR. In addition, they observed a reduction in the reoperation rate to 8.7% when associated with ALLR compared to isolated ACLR.
Redler et al. 24 evaluated patients who underwent ACLR revision with ALL reinforcement performed with an iliotibial band with an average follow-up of 10 years. Their results showed that only 7.6% cases evolved with loosening (Lachman > 5mm contralateral) and there was no graft rupture in any case, suggesting that ALL reinforcement decreases the risk of failure in an ACLR review. Yoo et al. 25 found no clinical differences in knee stability after isolated ACLR in patients who had concomitant injury of ALL. However, when submitting the patient to post-ACLR arthroscopy, they found a lower graft tension in those patients who had concomitant ALL injury (87.5%) versus 20.5% in the group with intact ALL, suggesting that the neglected ALL injury can cause graft loosening and rupture.
Recent work suggests that many failures in ACLR may be due to the neglect of the meniscus-capsular lesion in the posterior horn of the medial meniscus, or zone IV, also called lesions in the meniscal ramp. 27,28 Di Vico et al. 28 evaluated patients with ACL rupture, making posteromedial portal in all those who had some degree of instability in the posterior horn of the medial meniscus. They found a 9.6% prevalence of injuries on the ramp among all patients with ACL injuries. It was also observed that the existence of this meniscal injury was greater in patients treated six months after the injury, suggesting that ACL insufficiency is a predisposing factor. Results that corroborate the observations of Liu et al., 29 who reported 16.6% of this association in patients with ACL rupture, especially men below 30 and lesions lasting more than six months (6 months = 12.7% and >6 months = 22.8%).
The ramp injury can reach 23% of cases in the ACL presence and rupture, 30 reinforcing the importance of its diagnosis. They signaled the need to conduct research through the posteromedial portal, at risk of diagnostic loss in up to 15% of cases. In addition, patients who underwent repair of the lesion on the ramp and had ALL reconstruction associated with ACL were twice as less likely to need revision of the meniscal suture than patients with isolated ACLR. 31 When there is an associated lesion of the menisci, there is a greater tibial acceleration in the rotation movements, thus, the integrity and repair of the menisci and the extracapsular anterolateral reinforcement (for rotational instability), should be performed whenever possible. 32 Recently, Heilpern et al. 33 simulated cadaveric ramp injuries and repaired with an all-inside suture technique. In their results, only 2.5% of the devices used failed, and all that had good fixation were efficient in maintaining the stability of the posterior horn of the medial meniscus. The efficacy of integrated approaches to treating injuries to the medial meniscus ramp and ALL injuries still lack further evidence. However, most results encourage the surgeon to perform ALL reconstruction whenever there is injury or rotational instability, and perform meniscal repair whenever possible to decrease the rate of graft failure and overload. In competitive athletes, indicating reconstruction of the ALL and suturing the menisci can represent a safer RTP and save them from a new surgery, which often shortens their career.

Rehabilitation protocols
It has been suggested that the RTP depends more on a rigorous rehabilitation program than on the surgical technique or the choice of graft used. 1 Like the final RTP criteria, most post-ACLR rehabilitation protocols are solely time-based. In fact, it is important to respect the healing and "ligamentization" period of the grafts so that they are not subjected to stresses that can lead to loosening, incompetence and/ or risk of new ruptures, especially at times when the sensorimotor and neuromuscular functions are still not able to react properly. But recently, it has been suggested to design the rehabilitation protocols in phases, and to use objective criteria for progression between them. [34][35][36] This strategy improve communication between the members of the rehabilitation team (physicians, physiotherapists, physical educators, nutritionists, psychologists, etc.) providing a better understanding on wich phase of the rehabilitation process the athlete is. Treatment and training become more effective since they are carried out at the necessary intensity, respecting the principle of overload, but with a low risk of injury.

Image criteria
Li et al. 37 initially did not find clinical importance in graft signal changes in magnetic resonance imaging (MRI) performed post-ACLR and did not associate the findings with graft strength/resistance, but suggest that the decrease in graft signal intensity is directly related to better scores on the International Knee Documentation Committee (IKDC), especially after 12 months of surgery. Hofbauer et al. 38 compared MRI of post-ACLR knees with those of healthy knees. After six months of ACLR, they observed a difference in the intensity of the graft signal in relation to the original ACL in its substance and in the femoral insertion, and without difference in the tibial insertion. They suggested that this period is precocious for RTP, even if the functional and muscle strength criteria are met, as there is no biological maturity of the graft ("ligamentization").
The combination of graft signal intensity and volume should be considered for the assessment of graft incorporation in ACLR surgery. Wider grafts with low signal intensity were associated with better performances after five years. These patients had better results in the evaluation of functional tests, pain, quality of life questionnaires and symptoms. 3 In a recent review, the MRI graft signal with six months of ACLR was evaluated and repeated with 12 months in parallel to the IKDC questionnaire. They observed that in the evolution from six to 12 months there was a decrease in the signal on MRI, and an improvement in the IKDC score. The results suggest that the greater the graft maturity, the better the surgery functional results. 37 These results corroborate those found by Biercevicz et al., 39 who correlated the graft maturity, measured by intensity and diameter at MRI, and the clinical and functional results (residual laxity and single hop-test).

Subjective questionnaires
The use of questionnaires is a valuable and complementary tool in decision-making about the RTP. They are easy to apply instruments and help to identify factors, whether physical or psychological, that may hinder or prevent RTP. Barber-Westin and Noyes 9 observed that only one of 264 studies used validated questionnaires as a criterion for the RTP. More recently, Grassi et al. 40 observed that 90% of the studies used one or more clinical scores for evaluation, demonstrating that an increasing importance has been attributed to these forms of evaluation. Specific tools that report on the patient's personal experience such as the IKDC, the Tampa Scale for Kinesiophobia (TSK) and the Anterior Cruciate Ligament -Return to Sport after Injury Scale (ACL-RSI) have been documented as important measures in the post-ACLR assessment. 9,[41][42][43][44][45] The IKDC allows global assessment for ACL injuries, and has been used as a criterion for post-ACLR RTP. 46 Low IKDC scores seem to be associated with a greater likelihood of unsatisfactory results in muscle strength tests, hop tests and the degree of patient satisfaction. 43,46 The Tampa Scale for Kinesiophobia (TSK) assesses fear related to pain caused by movement. 43 Zwolski et al. 46 found that in the TSK short version, TSK-11, patients who did not return to sport had slightly higher scores.
The ACL-RSI assesses the psychological impact on the post-ACLR RTP. Patients who returned to sport had higher scores (76.8 ± 15.0 points) when compared to patients who did not return (48.7 ± 27.2 points) 43,47 suggesting that psychological factors may influence the outcome and that strategies to increase athletes' confidence are beneficial. 43,47 The ACL-RSI is composed of 12 items subdivided into three themes: emotional, performance and risk assessment. 48 Likewise, Nawasreh et al. 14

Functional tests
Noyes et al. 16 proposed a set of four jump tests (single leg, triple and crossover hop for distance; and 6m timed hop) with the proposal to provide an objective and functional measure related to sports demands that involve jumping, pivoting and constant changes of direction. Hop tests can be performed easily, requiring little specialized equipment and training for their application. In addition, it has already proved to be a reliable measure to monitor patients in post-ACLR. 49 Studies report a gradual improvement in jumping performance, reaching a limb symmetry index (LSI) ≥ 90% in the period between 6 to 9 months after ACLR. 9,11 Recently, Herrington et al. 50 evaluated the correlation between EXT knee strength and performance in functional tests post-ACLR in 15 professional soccer players. At the time of RTP, they observed that 80% of the athletes failed the LSI criterion (≥ 90%) in the strength test, although 75% of them passed the same criterion for the hop test. Barfod et al. 51 concluded that the single leg hop test cannot be used to replace the knee EXT strength measure, since satisfactory values in the jump test can be achieved without the strength levels being achieved as well. In addition, there is evidence that the LSI measured for distance should not be used as an isolated RTP method. 52 The Knee Santy Athletic Return to Sport Test (K-STARTS) is a score that provides data on post-ACLR conditions. 53

Muscular strength
Different methods have been used to assess muscle strength preand post-ACLR. 8,14 Most studies used isokinetic dynamometers, but the protocols have a lot of variation as to the type of muscle action, angular speed, number of repetitions and range of motion. 8,51 However, it must be said that the vast majority of studies used concentric actions with a speed of 60°/s for EXT and FLEX of the knee. 15 The weakness of the EXT and FLEX of the knee are commonly reported post-ACLR at different time frames. 15,54 However, data suggest that recovery/strength gain depends on adherence to the rehabilitation/training program. 54 In the absence of normative data for comparison, analyses are performed through LSI. The LSI is used to establish a proportionality value between the limb involved with the non-operated contralateral, 90% is considered as an absence of deficit. 6, 7 A recent study suggests that obtaining an appropriate LSI together with the nine-month post-ACLR time criterion reduces the chance of graft rupture by 84%. 5 However, it is noteworthy that the use of LSI can mask the existence of bilateral deficits. 55,56 In addition to the LSI among homologous muscle groups, maintaining the FLEX/EXT ratio of the knee is an indirect indicator of the joint functional stability. Thus, deficit in the FLEX/EXT ratio has been considered a risk factor for ACL injuries. 57 This assumption emerges from the fact that the knee FLEXs act as ACL synergists, acting in the prevention of the anterior translation of the tibia in relation to the femur. 8 In fact, Kyritsis et al. 8 evaluated a total of 158 athletes and observed that the subjects who presented lower values for the FLEX/EXT ratio at a speed of 60°/s had a greater chance of graft rupture (Figure 1).
In addition to the EXT and FLEX groups of the knee, the abductor and hip external rotator muscles are also important in reducing the risk of ACL rupture, and should also be monitored. 58 The weakness in these muscles can cause the femur internal rotation, together with the knee dynamic valgus, movement patterns that were previously associated with ACL rupture. 59 However, few references are found that assessed hip strength in post-ACLR. 11,60 Conflicting results are reported regarding the effect of the type of graft used for ACLR on the EXT and FLEX knee strength. 11,61 A meta-analysis demonstrated that after 12 months of ACLR, patients who received a graft from the patellar tendon presented reduced strength values for EXT, while patients who received the flexor tendon demonstrated a strength deficit for FLEX, suggesting that such findings are related to the graft donor area, a recently reinforced view. 15,62 Recently, it has been suggested that the ability to produce explosive strength in short periods of time (<300ms) is more representative of sporting capacity than maximum strength (eg, peak isokinetic or isometric torque). 63 During sports activities, the time available to produce muscle strength/torque is infinitely less than the time needed to reach maximum muscle capacity. Previous studies have observed that the time between contact with the ground and ACL rupture lasted on average 50-60ms, 59 suggesting that this is the window of time available for any attempt at dynamic joint stabilization provided by muscle contraction. The rate of torque development (RTD) assesses this condition by measuring how much strength/torque an individual is capable of producing in fractions of time between 50-250ms from the contraction onset. 63 Therefore, it has been proposed to include it in batteries for the evaluation and control of the athlete for 64 RTP. (Figure 2)   Anterior cruciate ligament reconstruction (ACLR), six months after ACLR (6M), nine months after ACLR (9M), Limb symmetry index (LSI), maximum isometric voluntary contraction (MIVC), rate of torque development (RTD), RTD measured in 50ms (RTD50, also called early RTD), RTD measured in 200ms (RTD200, also called late RTD). Blue lines and bars represent the contralateral member, while red lines and bars represent the involved member (ACL). On the left side, explosive torque curves measured for knee extension at different times can be observed. On the right side, the calculated values of MIVC and RTDs are shown through a bar graph. Note that the LSI is referenced to the right y-axis and the dashed lines mark the clinical criterion of the LSI ≥ 90%. Note that the behavior of the RTD50 and RTD200 are different from each other and compared to MIVC.    Data suggest that maximum strength can be restored prior to RTD. 65,66 Kuenze et al. 65 observed the existence of a RTD deficit after one year of ACLR in athletes, even though they passed the 90% LSI criterion for maximum strength. Angelozzi et al. 66 evaluated 45 professional soccer athletes who underwent the same post-ACLR rehabilitation/training protocol. They observed that in 6 months post-ACLR the maximum isometric strength of EXT reached the criterion of 90% of the LSI and with values close to those measured before the injury, but the RTD reached the release rates only after 12 months of the ACLR, with athletes performing a rehabilitation/training protocol focusing on the development of muscle power. Although few studies have evaluated post-ACLR RTD, the current findings underscore its importance and open space for further investigation.

Aerobic criteria
Despite the maximum VO 2 (VO 2max ) expressing cardiovascular conditioning, little information is associated with RTP post-ACLR. Almeida et al. 67 evaluated the VO 2max of 20 athletes post-ACLR in the pre-and after six months of ACLR. While the control group had an average VO 2max of 56.9 mL/kg/min, the group with ACLR had 45.2 mL/kg/min. Although the index improved after 6 months to 48.9 mL/kg/min, it was still lower than the control group. The authors hypothesized that athletes with low VO 2max are more likely to suffer a new ACL injury and that 6 months were insufficient to reach an acceptable VO 2max index suggesting postponement of the RTP. 67

Psychological aspects
Psychological aspects are often overlooked in the RTP decision. 61,68 However, the literature mentions that 50% of athletes do not return to the previous competition level, mainly because of fear of a new injury, although they know their knees are able and prepared to resume physical activity post-ACLR. [69][70][71] Therefore, it has been suggested that the psychological factor is one of the factors that most hamper the RTP post-ACLR. 69,70 Tjong et al. 69 report that fear of a new injury/pain, change of priority in life and the patient's personality are the three main psychological factors involved in RTP. 69 Ardern et al. 72 observed that the main deleterious causes for RTP were lack of confidence in the knee (28%), fear of a new injury (24%) and impaired knee function (22%).
Personality and motivation play an important role in the post-ACLR rehabilitation process. 68,73 Individuals with strong personal traits and highly competitive personalities or self-motivated are associated with a higher rate of RTP. 68,73 On the other hand, the development of post-injury anxiety and depression is linked to a delay in post-ACLR rehabilitation. 68,73 Therefore, motivational strategies throughout the rehabilitation/training process to combat fear, promote self-confidence and define personal goals have shown promising results in raising awareness of the injury and reducing athletes' anxiety. 68,73

Our experience
We take into account the sport practiced for graft choice. For sports that require repeated jumping, such as basketball and volleyball, we avoid violating the extensor mechanism, giving preference to the flexor tendons. The same occurs for fighting athletes who make kneeling movements (eg. MMA, jiu-jitsu), given the common report of anterior knee pain post-ACLR with patellar graft. In athletes who need to do pivoting and sudden changes in direction (eg. soccer and rugby), we prefer to use the patellar or quadriceps graft to preserve the posterior thigh musculature and its synergistic function with the ACL.
Currently, for patients with explosive pivot (3+/3+), Segond fracture, age <20 years, generalized ligamentous hyper-laxity and in revision surgeries; we performed the tenodesis of the ilio-tibial band (ITB) as an associated lateral reinforcement (ALL). We always test the mobility of the meniscus, even if it does not present evident lesions. In case of any sign of hypermobility, we opted for the repair with an all-inside and/or inside-out and/or outside-in technique, according to the lesion pattern, aiming at stabilizing the meniscus-capsular insertion.
Undoubtedly, the subject "turn around time" is the most discussed in our service and always discussed with each athlete submitted to ACLR. After the ACLR, athletes are instructed to follow our Rehabilitation Protocol (Table 2). Prior to the release to the RTP, all clinical, functional, imaging (MRI) and isokinetic information of the athlete is discussed with the multidisciplinary group. It is worth mentioning that functional (questionnaires and jumps) and strength assessments are conducted pre-and post-ACLR. Post-ACLR assessments are carried out first at 6 months and successively every month until the target indexes are reached. We believe this to be the best practice and a determining factor for accumulating high RTP rates.

CONCLUSION
Guiding post-ACLR clinical practice during the RTP process is not an easy task and is constantly evolving. After reviewing the literature, together with the experience of our group, we concluded: the choice of graft must take into account the sport practiced; whenever possible, reconstruction of the ALL and suturing of the meniscus should be encouraged; the rehabilitation program must be structured in phases with objective progression criteria so that the athlete is treated/trained at the appropriate intensity, respecting the principle of overload. Subjective questionnaires should be incorporated, not just those established for functionality (eg, Lysholm and IKDC), as well as those for assessing psychological and trust aspects. Hop test and muscle strength measures are mandatory for understanding muscle function and dynamic joint stability. We believe these are the safest practices for the RTP.  started only if the progression criteria of the previous phase are met. # realized in the post-ACLR at 24 weeks and repeats every subsequent month until the total release for Return to Sport, release needs to reach LSI ≥ 90%, specifically for FLEX/EXT value between 60-80%.

AUTHORS' CONTRIBUTIONS:
Each author made significant individual contributions to this manuscript. RG and VC: outlined the themes and subthemes. RG, VC, BF and JN: general review and final approval of the manuscript. VC: writing the functional evaluation topics, preparation of the images. BF: writing topics related to physiotherapy and functional evaluation, elaboration of the tables. RG, AC, GS, RB and JN were responsible for writing the medical and surgical topics.