Restoring Knee Flexor Strength Symmetry Requires 2 Years After ACL Reconstruction, But Does It Matter for Second ACL Injuries? A Systematic Review and Meta-analysis

Abstract

Background: It is unknown whether knee flexor strength recovers after anterior cruciate ligament (ACL) reconstruction with a hamstring tendon (HT) autograft and whether persistent knee flexor strength asymmetry is associated to a second ACL injury.

Objective: We aimed to systematically review (1) whether knee flexor strength recovers after ACL reconstruction with HT autografts, and (2) whether it influences the association with a second ACL injury. A third aim was to summarize the methodology used to assess knee flexor strength.

Design: Systematic review and meta-analysis reported according to PRISMA.

Methods: A systematic search was performed using the Cochrane Library, Embase, Medline, PEDRo, and AMED databases from inception to December 2021 and until completion in January 2023. Human clinical trials written in English and conducted as randomized controlled trials, longitudinal cohort, cross-sectional, and case-control studies on patients with index ACL reconstructions with HT autografts harvested from the ipsilateral side were considered. Knee flexor strength was measured isokinetically in both the reconstructed and uninjured limb to enable the calculation of the limb symmetry index (LSI). The Risk of Bias Assessment Tool for Non-Randomized Studies was used to assess risk of bias for non-randomized studies and the revised Cochrane Risk of Bias tool was used for randomized controlled trials. For the meta-analysis, the LSI (mean ± standard error) for concentric knee flexor strength at angular velocities of 60°/second (s) and 180°/s preoperatively and at 3 months, 6 months, 12 months, and 24 months were pooled as weighted means with standard errors.

Results: The search yielded 64 studies with a total of 8378 patients, which were included for the assessment of recovery of knee flexor strength LSI, and a total of 610 patients from four studies that investigated the association between knee flexor strength and second ACL injuries. At 1 year after ACL reconstruction, the knee flexor strength LSI had recovered to 89.0% (95% CI 87.3; 90.7%) and 88.3% (95% CI 85.5; 91.1%) for the velocities of 60°/s and 180°/s, respectively. At 2 years, the LSI was 91.7% (95% CI 90.8; 92.6%) and 91.2% (95% CI 88.1; 94.2%), for velocities of 60°/s and 180°/s, respectively. For the association between knee flexor strength and second ACL injuries, there was insufficient and contradictory data.

Conclusions: There was low to very low certainty of evidence indicating that the recovery of knee flexor strength LSI, defined as ≥ 90% of the uninjured side, takes up to 2 years after ACL reconstruction with HT autografts. Whether knee flexor strength deficits influence the association of second ACL injuries is still uncertain. There was considerable heterogeneity in the methodology used for knee flexor strength assessment, which together with the low to very low certainty of evidence, warrants further caution in the interpretation of our results.

Registration number: CRD42022286773.

Keywords: Anterior cruciate ligament; Hamstring tendon autograft; Knee flexor strength; Knee flexor strength methodology; Limb symmetry index; Second ACL injuries.

Fig. 1

Flowchart of inclusion and exclusion of articles for review. ACL, anterior cruciate ligament; HT, hamstring tendon; LSI, limb symmetry index; n, number

Fig. 2

The recovery of knee flexor strength symmetry presented as the limb symmetry index (LSI) preoperatively. The LSI mean value is presented with the standard error and a 95% confidence interval using a random effect model. A limb symmetry index of 90% is represented by the black line. Knee flexor strength at 60°/second was 80.8% ± 2.5% (95% CI 75.9; 85.7%, I² = 89.5%) and at 180°/second 75.6% ± 9.8% (95% CI 56.4; 94.8%, I² = 97.6%). Baba et al. 2019 A [25] = Surgery < 1 months; Baba et al. 2019 B [25] = Surgery 1–3 months; Baba et al. 2019 C [25] = Surgery > 3 months; Harilainen et al. 2005 A [44] = Transfix; Harilainen et al. 2005 B [44] = Interference screw; Suh et al. 2021 A [80] = Dominant leg; Suh et al. 2021 B [80] = Non-dominant leg; Tsuda et al. 2009 A [85] = Women; Tsuda et al. 2009 B [85] = Men

Fig. 3

The recovery of knee flexor strength symmetry presented as the limb symmetry index (LSI) at 3 months. The LSI mean value is presented with the standard error and a 95% confidence interval using a random effect model. A limb symmetry index of 90% is represented by the black line. Knee flexor strength at 60°/second was 81.7% ± 2.8% (95% CI 76.2; 87.2%, I² = 95.4%) and at 180°/second 88.8% ± 1.4% (95% CI 86.2; 91.5%, I² = 58.4%). Iriuchishima et al. 2010 A [51] = Standard rehabilitation; Iriuchishima et al. 2010 B [51] = Accelerated rehabilitation; Koutras et al. 2013 A [60] = Anteromedial portal; Koutras et al. 2013 B [60] = Transtibial portal; Sengoku et al. 2022 A [77] = Semitendinosus alone; Sengoku et al. 2022 B [77] = Semitendinosus + gracilis; Tsuda et al. 2009 A [85] = Women; Tsuda et al. 2009 B [85] = Men

Fig. 4

The recovery of knee flexor strength symmetry presented as the limb symmetry index (LSI) at 6 months. The LSI mean value is presented with the standard error and a 95% confidence interval using a random effect model. A limb symmetry index of 90% is represented by the black line. Knee flexor strength at 60°/second was 89.0% ± 0.9% (95% CI 87.3; 90.7%, I² = 66.2%) and at 180°/second 91.3% ± 1.9% (95% CI 87.5; 95.1%, I² = 94.7%). Carter and Edinger 1999 A [29] = Semitendinosus alone; Carter and Edinger 1999 B [29] = Semitendinosus + gracilis; Iriuchishima et al. 2010 A [51] = Standard rehabilitation; Iriuchishima et al. 2010 B [51] = Accelerated rehabilitation; Królikowska et al. 2019 A [62] = Supervised < 6 months; Królikowska et al. 2019 B [62] = Supervised > 6 months; Koutras et al. 2013 A [60] = Anteromedial portal; Koutras et al. 2013 B [60] = Transtibial portal; Sengoku et al. 2022 A [77] = Semitendinosus alone; Sengoku et al. 2022 B [77] = Semitendinosus + gracilis; Severyns et al. 2022 A [78] = No graft failure; Severyns et al. 2022 B [78] = Graft failure; Suh et al. 2021 A [80] = Dominant leg; Suh et al. 2021 B [80] = Non-dominant leg; Tsuda et al. 2009 A [85] = Women; Tsuda et al. 2009 B [85] = Men

Fig. 5

The recovery of knee flexor strength symmetry presented as the limb symmetry index (LSI) at 12 months. The LSI mean value is presented with the standard error and a 95% confidence interval using a random effect model. A limb symmetry index of 90% is represented by the black line. Knee flexor strength at 60°/second was 89.0% ± 0.9% (95% CI 87.3; 90.7%, I² = 66.2%) and at 180°/second 88.3% ± 1.4% (95% CI 85.5; 91.1%, I² = 80.3%). Araki et al. 2011 A = Single-bundle; Araki et al. 2011 B; Double-bundle; Harilainen et al. 2005 A [44] = Transfix; Harilainen et al. 2005 B [44] = Interference crew; Suh et al. 2021 A [80] = Dominant leg; Suh et al. 2021 B [80] = Non-dominant leg; Tsuda et al. 2009 A [85] = Women; Tsuda et al. 2009 B [85] = Men

Fig. 6

The recovery of knee flexor strength symmetry presented as the limb symmetry index (LSI) at 24 months. The LSI mean value is presented with the standard error and a 95% confidence interval using a random effect model. A limb symmetry index of 90% is represented by the black line. Knee flexor strength at 60°/second was 91.7% ± 0.5% (95% CI 90.8; 92.6%, I² = 5.4%) and at 180°/second 91.2% ± 1.5% (95% CI 88.1; 94.2%, I² = 82.6%). Baba et al. 2019 [25] A = Surgery < 1 month, Baba et al. 2019 B [25] = Surgery 1–3 months; Baba et al. 2019 C [25] = Surgery > 3 months; Gifstad et al. 2013 A [40] = Ezloc; Gifstad et al. 2013 B [40] = Bone mulch; Hamada et al. 2001 A [42] = Single-socket; Hamada et al. 2001 B [42] = Bi-socket; Harilainen et al. 2005 A [44] = Transfix; Harilainen et al. 2005 B [44] = Interference screw; Inagaki et al. 2013 A [50] = Semitendinosus alone; Inagaki et al. 2013 B [50] = Semitendinosus + gracilis; Koga et al. 2015 A [55] = 0°; Koga et al. 2015 B [55] = 20°; Koga et al. 2015 A [55] = 45°; Kondo et al. 2012 A [57] = Endobutton-CL; Kondo et al. 2012 B [57] = Endobuttion-CL-BTB; Kouloumentas et al. 2019 A [59] = Semitendinosus alone; Kouloumentas et al. 2019 B [59] = Semitendinosus + gracilis; Nakamura et al. 2002 A [70] = Semitendinosus alone; Nakamura et al. 2002 B [70] = Semitendinosus + gracilis; Nishio et al. 2018 A [71] =  > 40 years; Nishio et al. 2018 B [71] =  < 40 years