Abstract
Background:
Primary anterior cruciate ligament reconstructions (ACLRs) with extensor mechanism autografts (bone–patellar tendon–bone [BTB] or quadriceps tendon [QT]) have been reported to have lower quadriceps strength symmetry outcomes 6 months postoperatively as compared with those using a hamstring tendon (HT) autograft. However, the influence of a recurrent ipsilateral extensor mechanism graft harvest for a revision ACLR on quadriceps strength symmetry recovery is not well known.
Purpose:
To investigate the effect of ipsilateral revision ACLR autograft selection on the recovery of quadriceps limb symmetry index (Q-LSI) 6 months after a revision ACLR.
Study Design:
Cohort study; Level of evidence, 3.
Methods:
After revision ACLR, 53 patients were divided into 3 groups based on the ipsilateral autograft utilized for the primary and revision ACLR: hamstring to extensor (HT/Ext; ie, BTB or QT), extensor to hamstring (Ext/HT), and extensor to extensor (Ext/Ext). Six months after revision ACLR, patients completed an isometric quadriceps strength testing protocol on an isokinetic dynamometer; Q-LSI and quadriceps peak torque relative to body weight were then calculated and used for analysis. Separate analyses of covariance were conducted for each outcome using a multivariable linear regression framework, with group as the primary independent variable and age, sex, body mass index, and time since surgery included as covariates.
Results:
The Ext/HT group exhibited significantly greater quadriceps strength symmetry (Q-LSI = 88%) than the Ext/Ext group (Q-LSI = 69%; P = .028) and the HT/Ext group (Q-LSI = 70%; P = .005) 6 months after a revision ACLR. At the 6-month testing timepoint after a revision ACLR, no significant differences in Q-LSI were observed between the Ext/Ext group (Q-LSI = 69%) and HT/Ext group (Q-LSI = 70%; P = .99).
Conclusion:
As compared with HT autografts, an ipsilateral BTB or QT autograft for a revision ACLR is associated with significantly lower Q-LSI values at the 6-month postoperative timepoint, regardless of the autograft type used for the primary ACLR.
Keywords
After an anterior cruciate ligament (ACL) tear, an ACL reconstruction (ACLR) is generally recommended for athletes returning to pivoting, landing, and change-of-direction sports. 42 Overall revision ACLR rates after a primary ACLR have been reported to be as high as 7.5%. 30 Reported reasons for recurrent ACL reinjury and graft failure include the inherent risk of returning to sport,2,32 early return despite not meeting criteria for return to sport,16,23 nonanatomic reconstruction/surgical techniques,18,31,35 anatomic risk factors for ACL injury,21,35,48 unaddressed concomitant injury or injuries at the time of primary ACLR,37,45 and nonmodifiable patient factors.36,46,48 After a recurrent ACL graft injury or failure, a revision ACLR is usually recommended with an autograft to optimally restore knee stability. 14
Bone–patellar tendon–bone (BTB) and hamstring tendon (HT) autografts are the most common grafts utilized for a revision ACLR; however, in the last decade, the utilization of a quadriceps tendon (QT) autograft has increased.27,28,47 The utilization of an ipsilateral autograft from the extensor mechanism for a primary ACLR (ie, BTB or QT grafts) has been reported to result in lower quadriceps strength symmetry outcomes (ie, quadriceps limb symmetry index [Q-LSI]) when compared with HT autografts at 3, 6, and 9 months after an ACLR. 39 Moreover, it is plausible that recurrent ipsilateral harvesting of an extensor mechanism autograft (ie, BTB or QT autograft) for a revision ACLR may produce a greater impairment in Q-LSI outcomes than for a primary ACLR. 1
Previous literature has reported Q-LSI outcomes after a revision ACLR6,9,13,22,43; however, less is known about the influence that a repeat ipsilateral extensor mechanism autograft harvest has on quadriceps strength recovery in revision ACLR cohorts.9,43 An improved understanding of quadriceps strength recovery after a revision ACLR may facilitate improved awareness for areas of focus for postoperative rehabilitation, a more successful return to sport, and most critically, a reduced risk of recurrent ACL injury as seen in primary ACLR cohorts. 16 Therefore, the purpose of this study was to investigate the effect of ipsilateral revision ACLR autograft selection on the recovery of the Q-LSI 6 months after a revision ACLR. We hypothesized that individuals who underwent both procedures (ie, primary and revision ACLR) with 2 consecutive ipsilateral extensor mechanism autograft harvests would demonstrate inferior strength symmetry (Q-LSI) outcomes 6 months after a revision ACLR, as compared with those who had only a single autograft harvested from the extensor mechanism during either procedure.
Methods
Participants
This study was approved by an institutional review board (08.19.2019 NS_Haus; IntegReview IRB). All participants provided their informed consent. Parental and youth consents were obtained for all participants aged <18 years. Between August 2018 and December 2023, 53 competitive and recreational athletes who underwent a revision ACLR procedure were included in this analysis. Participants were included if they had undergone a revision ACLR with a concomitant meniscal repair, meniscectomy, and/or a lateral extra-articular procedure (eg, lateral extra-articular tenodesis or anterolateral ligament reconstruction). These concomitant procedures were included because of the increased utilization of lateral extra-articular procedures and the high reported prevalence of meniscal pathology at the time of a revision ACLR.5,15,20,37
Participants were divided into 3 groups by the autograft utilized for the primary-to-revision ACLR: hamstring to extensor (HT/Ext), extensor to hamstring (Ext/HT), and extensor to extensor (Ext/Ext). The extensor autograft consisted of individuals receiving either an ipsilateral BTB or QT autograft at the time of primary and/or revision ACLR.
Exclusion criteria for the analysis included the following: (1) those who underwent >1 revision ACLR, (2) a revision ACLR performed with a contralateral autograft or an allograft, (3) a revision surgery for a multiligament knee injury, (4) an index procedure consisting of an ACL repair, and (5) a staged procedure owing to bone grafting of ACL femoral and tibial tunnels and/or an osteotomy. All participants in this study were aged 15 to 35 years at the time of rehabilitation and performance testing.
Clinical Testing
Anthropometric measurements were collected, including height and weight. All participants underwent quadriceps strength testing 6 months after a revision ACLR. A standardized dynamic warm-up was completed before testing, as led by a physical therapist, athletic trainer, or sports performance coach. 39 A 90° knee flexion isometric quadriceps strength protocol on an isokinetic dynamometer was utilized (Biodex Medical Systems, Inc). Participants were seated with the lateral femoral epicondyle of the knee aligned with the dynamometer's axis of rotation. Thigh, waist, and 2 chest straps were used to lock the participant to the chair, while the dynamometer's force arm was secured superior to the lateral malleolus of the ankle. Participants completed 4 isokinetic knee extension trials through a self-selected range of motion, followed by 1 maximal-effort 90° quadriceps isometric contraction on the dynamometer before testing as part of a warm-up protocol and to familiarize themselves with the task.
Before testing, participants were given verbal instructions to apply as much force as possible against the fixed arm of the dynamometer throughout the duration of the test. 39 Loud verbal encouragement was provided to facilitate maximal torque output. Participants completed three 5-second maximal-effort trials, followed by a 30-second rest interval between trials. The average of the 3 peak torque trials was analyzed. Testing was completed on the nonsurgical limb first, followed by the surgical limb.
Quadriceps Strength Assessment
Strength data were divided into average quadriceps peak torque output (Avg Q-PT), average Q-LSI, and average quadriceps peak torque relative to body weight (Q-PT/BW). 39 Quadriceps strength symmetry (Q-LSI) was calculated using the following formula:
Q-PT/BW was calculated using the following formula:
Q-LSI and Q-PT/BW were the variables used in our analysis.
Statistical Analysis
The primary dependent variables were average Q-LSI and average Q-PT/BW for the surgical and nonsurgical limbs. Separate analyses of covariance (ANCOVAs) were conducted for each outcome via a multivariable linear regression framework, with group as the primary independent variable and age, sex, body mass index, and time since surgery included as covariates.
Model assumptions were evaluated by visual inspection of residuals, the Shapiro-Wilk test, and Levene test and were deemed acceptable. No extreme outliers were identified; outliers were defined as observations exceeding 3 times the interquartile range above the third quartile or below the first quartile. 44
The primary group effect was tested by analysis of variance on the fitted model. When a significant group effect was observed, adjusted pairwise comparisons were conducted through estimated marginal means with Tukey correction for multiple comparisons. 44 Descriptive data are presented as mean and standard deviation, while adjusted effects are reported with corresponding effect sizes and confidence intervals. The alpha level was set a priori at P≤ .05.
Given the fixed sample of patients undergoing revision ACLR, a sensitivity analysis was used to characterize the magnitude of group effects detectable by the ANCOVA models. For the primary outcome (Q-LSI), the adjusted group effect demonstrated a partial η2 of approximately 0.20 (Cohen f2≈ 0.25), corresponding to a large effect size, indicating that the available sample size was sufficient to detect moderate to large group differences in limb symmetry. All analyses were performed in RStudio (Version 2023.9.1.494; Posit Software, PBC).
Results
A summary of the participants’ characteristics is outlined in Table 1. There was a significant difference present in age and body mass index. Specifically, the Ext/Ext group was younger than the Ext/HT group (P = .002) and HT/Ext group (P = .04). Additionally, the Ext/Ext group had a lower body mass index than the HT/Ext group (P = .041). The means for Q-LSI and Q-PT/BW are reported in Table 2. Figure 1 shows all participants’ Q-LSI values stratified by graft type.
Descriptive Data for First-Time Ipsilateral Autograft Revision Anterior Cruciate Ligament Reconstruction by Group (53 Patients) a
Values are expressed as mean ± SD unless otherwise noted. Groups are expressed by autograft utilized for primary-to-revision surgical procedure. Ext, extensor mechanism autograft (bone–patellar tendon–bone or quadriceps tendon); HT, hamstring tendon autograft; ns, nonsignificant.
Statistically significant difference between Ext/Ext and HT/Ext groups.
Statistically significant difference between Ext/Ext and Ext/HT groups.
Normative Strength Outcomes by Autograft Harvest Group at 6 Months After Revision Anterior Cruciate Ligament Reconstruction a
Values are expressed as mean ± SD unless otherwise noted. Groups are expressed by autograft utilized for primary-to-revision surgical procedure. Ext, extensor mechanism autograft (bone–patellar tendon–bone or quadriceps tendon); HT, hamstring tendon autograft; ns, nonsignificant; Q-LSI, quadriceps limb symmetry index; Q-PT/BW, average quadriceps peak torque relative to body weight.
Statistically significant difference between Ext/Ext and Ext/HT groups.
Statistically significant difference between Ext/HT and HT/Ext groups.

Quadriceps limb symmetry index outcomes at 6 months after revision ACLR by sex: (A) male and (B) female. Groups are presented by autograft type utilized at the time of revision ACLR. Values are expressed as means. ACLR, anterior cruciate ligament reconstruction; Ext, extensor mechanism autograft (bone–patellar tendon–bone or quadriceps tendon); HT, hamstring tendon autograft; LSI, limb symmetry index.
After adjustment for age, sex, body mass index, and time since surgery, a significant group effect was observed for Q-LSI (F2,46 = 2.529; P = 0.006; partial η2≈ 0.20). Adjusted pairwise comparisons demonstrated that the Ext/HT group exhibited significantly greater Q-LSI as compared with the Ext/Ext group (P = .028) and the HT/Ext group (P = .005), while no difference was observed between the Ext/Ext and HT/Ext groups (P = .99). No significant differences in Q-PT/BW among groups were observed between the nonsurgical limb (F1,46 = 1.466; P = .241) and surgical limb (F1,46 = 0.686; P = .508) when accounting for covariates (Table 2). Overall, there were no between-group differences in average Q-PT/BW; however, between the nonsurgical limb (sex, F1,46 = 13.678; P = .0006) and surgical limb (sex, F1,46 = 14.912; P = .0004), there was a statistically significant difference in strength of both limbs based on sex (Appendix Tables A1 and A2).
Discussion
The most significant finding of this study was that utilizing either an ipsilateral BTB or QT autograft for a revision ACLR was associated with significantly impaired quadriceps strength symmetry at the 6-month postoperative timepoint relative to a revision ACLR with an HT autograft, regardless of the autograft utilized for the primary ACLR. In addition, the hypothesis was not supported that individuals undergoing 2 consecutive ipsilateral extensor mechanism autograft harvests (primary and revision ACLR) would demonstrate inferior strength symmetry (Q-LSI) outcomes 6 months after a revision ACLR, as compared with those with a single ipsilateral extensor autograft harvest for the revision ACLR.
Our main findings should influence rehabilitation specialists’ exercise selection and quadriceps loading strategies used during the postoperative rehabilitation recovery of a revision ACLR to focus on restoration of quadriceps strength during this early time frame. Restoration of Q-LSI outcomes appears to be positively associated with critical neuromuscular performance metrics, patient-reported outcome measures, and the decreased likelihood of symptomatic osteoarthritis.4,7,10,16,38,49 It is important that clinicians be able to differentiate the cause of persistent Q-LSI deficits, as it may be a result of muscle atrophy, subtherapeutic exercises selection, insufficient loading stimulus, arthrogenic muscle inhibition (AMI), or a combination of these factors.17,24,25,33,41 Solie et al 41 described an evidence-based framework suggesting that integration of open and closed kinetic chain loading exercises performed at various angles, contraction modes, and external loads executed at appropriate intensities may best stimulate hypertrophy and force production of the quadriceps complex. Specifically, utilization of open kinetic chain exercises between 0° and 130° of knee flexion may facilitate regional hypertrophy of the vastus intermedius, vastus medialis, and vastus lateralis secondary to the quadriceps length-tension relationship. Additionally, inclusion of neuromuscular electrical stimulation has been proposed to address AMI and promote motor unit recruitment and rate coding.3,26 Therefore, we theorize that including neuromuscular electrical stimulation may help address the cortical excitability deficits that result in lower descending action potential arriving at the quadriceps complex.3,33 Future work should analyze the influence that Q-LSI restoration has on objective performance metrics and patient-reported outcome measures in revision ACLR cohorts to better quantify performance and physical impairments, as well as define strategies to address them.
Our findings appear to suggest faster initial quadriceps strength symmetry recovery at the 6-month timepoint after a revision ACLR with an HT autograft relative to the revision ACLR with an ipsilateral extensor mechanism autograft. We observed that the Ext/HT group had significantly greater mean Q-LSI outcomes (18%) as compared with the HT/Ext group at 6 months after a revision ACLR. Our findings are consistent with those of previous groups who observed a 10%-15% higher mean Q-LSI value at the 6-month postoperative time point in individuals who underwent a revision ACLR with an HT autograft versus those who received an ipsilateral extensor mechanism autograft.8,9 As such, it appears that utilizing an HT autograft for a revision ACLR is associated with higher Q-LSI outcomes at the 6-month postoperative timepoint as compared with a revision ACLR with an ipsilateral extensor mechanism autograft.
Previous studies8,9 have analyzed strength outcomes in revision ACLR cohorts; however, our data include patients with ipsilateral QT autografts, whereas prior groups defined their cases of recurrent extensor mechanism autografts as individuals receiving only a contralateral BTB autograft at the time of revision ACLR. Our results illustrate a significant decrease in the Q-LSI for the Ext/Ext group, but we are limited in making a direct comparison with previous work8,9 given the cohort differences. Setliff et al 40 aimed to investigate how recurrent ipsilateral extensor mechanism autografts for a revision ACLR would delay the time to achieve an 80% Q-LSI outcome or a return to jogging status. They observed that on average, their Ext/Ext group and HT/Ext group achieved an 80% Q-LSI outcome or returned to jogging status 132 and 150 days postoperatively, respectively. We are unable to perform a direct comparison of the present study's results with those of Setliff et al, secondary to their primary outcome being defined as criterion and task based. Additionally, they were able to obtain isokinetic Q-LSI outcomes for only 20% of their sample, and the other 80% was obtained via electronic medical record documentation. It is possible that the Q-LSI deficits noted at the midway timepoint of postoperative recovery (ie, 6 months) observed within this study may be explained by graft harvest–related AMI, subtherapeutic exercise selection, and insufficient loading stimulus after a revision ACLR with a recurrent ipsilateral extensor mechanism autograft.1,33
We observed a significant difference in age among groups: the Ext/Ext group was younger than the Ext/HT group, but the clinical implications of this finding may be minimal since all groups were still relatively young. Although our data highlight that ipsilateral extensor mechanism autografts were associated with lower Q-LSI outcomes at the approximate midway timepoint in postoperative rehabilitation (ie, 6 months) after a revision ACLR, lower graft failure rates have been observed with BTB and QT autografts relative to HT autografts in primary and revision ACLR cohorts.29,34 Taking into consideration the higher HT autograft failure rates in younger patients, this may have influenced graft selection at time of revision ACLR in more active and younger cohorts. Therefore, this potentially explains the observed differences in age among groups in the present study. Nonetheless, careful consideration of graft availability, advantages, and disadvantages must be weighed at the time of revision ACLR.
In the present study, sex had an influence in average Q-PT/BW. Specifically, the female cohort demonstrated a lower Q-PT/BW outcome in the surgical limb as compared with the nonsurgical limb. Schwery et al 39 noted similar findings, as they observed that males had a significantly greater Q-PT/BW as compared with females 6 months after a primary ACLR. Additionally, Ebert et al 11 found that females demonstrated a lower Q-LSI after a primary ACLR with an extensor mechanism autograft. Our analysis did not observe any differences in Q-LSI outcomes when accounting for sex. These observations highlight how biological differences in females, such as hormonal and neuromuscular performance, may explain the differences in quadriceps strength recovery after an ACLR, although further work is needed to understand the exact mechanism. 12
Limitations
There were a few limitations in our investigation. We did not conduct an a priori power analysis; however, we did perform a sensitivity analysis to characterize the magnitude of group effects detectable by our ANCOVA model, which demonstrated that our sample size was sufficient to detect a moderate to large group difference for our primary outcome. Additionally, given the retrospective design of our study, we are unable to account for baseline quadriceps strength and Q-LSI for our sample. It is plausible that these individuals had a lower Q-LSI, which may have influenced subsequent ACL reinjury, as prior work has observed an association between higher Q-LSI, as well as postponed reentry to sport (>9 months), and lower knee reinjury rate after a primary ACLR. 16 Furthermore, we did not account for limb dominance in our cohort. Recent work has also found that patients who underwent a primary ACLR of the dominant limb were able to achieve higher Q-LSI 7 months postoperatively. 19 In addition, we were not able to assess the influence that postoperative rehabilitation may have had on quadriceps strength recovery, owing to the retrospective design of this analysis. Nonetheless, it is important to recognize that expression of quadriceps strength in this complex population may be influenced by various factors, such as graft harvest–related AMI, the presence of concomitant injury and procedure, subtherapeutic exercise selection, and insufficient loading stimulus.1,5,14,20,25,33,37,41 This investigation was also limited to solely capturing quadriceps strength production and did not analyze any other neuromuscular performance measures, such as power production (ie, countermovement jump). Furthermore, we observed quadriceps strength recovery at only the 6-month timepoint, which is often considered the midway point for recovery after revision surgery. This did not allow us to examine if differences among groups normalized in the later stages of recovery. Finally, this study did not assess associations between differences of concomitant injuries (eg, meniscal pathology) and surgical techniques.
Conclusion
Harvesting an ipsilateral BTB or QT autograft for a revision ACLR, as opposed to an HT autograft, is associated with significantly lower Q-LSI values at the 6-month postoperative timepoint, regardless of the autograft type used for the primary ACLR.
Footnotes
Appendix
Normative Quadriceps Strength Outcomes by Sex at 6 Months After Revision ACL Reconstruction a
| Q-PT/BW, N·m/kg | |||
|---|---|---|---|
| Group: Sex | Surgical Limb | Nonsurgical Limb | Q-LSI, % |
| Ext/Ext | |||
| Female | 2.04 ± 0.83 b | 3.15 ± 0.83 b | 63.4 ± 18.4 |
| Male | 2.60 ± 0.52 | 3.50 ± 0.55 | 74.6 ± 12.9 |
| Ext/HT | |||
| Female | 1.97 ± 0.35 b | 2.23 ± 0.39 b | 89.8 ± 16.7 |
| Male | 2.69 ± 0.60 | 3.15 ± 0.49 | 85.6 ± 14.4 |
| HT/Ext | |||
| Female | 1.66 ± 0.35 b | 2.49 ± 0.79 b | 68.6 ± 14.4 |
| Male | 2.48 ± 0.54 | 3.55 ± 0.60 | 70.0 ± 11.7 |
Values are expressed as mean ± SD unless otherwise noted. Groups are expressed by autograft utilized for primary-to-revision surgical procedure. ACL, anterior cruciate ligament; Ext, extensor mechanism autograft (bone–patellar tendon–bone or quadriceps tendon); HT, hamstring tendon autograft; ns, nonsignificant; Q-LSI, quadriceps limb symmetry index; Q-PT/BW, average quadriceps peak torque relative to body weight.
Statistically significant difference in Q-PT/BW (male vs female; P≤ .05).
Final revision submitted April 2, 2026; accepted April 28, 2026.
The authors declared that they have no conflicts of interest in the authorship and publication of this contribution.
Ethical approval was obtained from IntegReview IRB (08.19.2019 NS_Haus).
