For this cohort study, we performed a database search for all surgically treated patients with ACL tears with clinical instability. A total of 396 patients were operated between 2012 and 2016, of which 351 patients fulfilled the inclusion criteria (primary ACL reconstruction with quadrupled semitendinosus tendon with suspensory fixation on the femoral and tibial side). There has already been a publication out of this patient collective, covering a different research question with different subsamples.(19) All patients with bilateral ACL reconstruction (n=12), complex meniscal lesions (subtotal/total meniscectomy, complete meniscal root avulsions and complete radial tears – n=15), as well as multiligamentous injuries (n=10) and high-grade chondral lesions requiring surgical treatment (n=9) were excluded from the study. A total of 305 patients were invited to participate in the study, and 75 patients gave their informed consent. Patients who underwent revision ACL surgery or were newly diagnosed with a re-rupture of the graft (n=8) were also excluded, leaving 67 eligible patients.
The follow-up examination included a standardized clinical assessment with measurement of the active range of motion (ROM) of both knees with goniometry, the thigh circumference, residual hyposensitivity in the innervation area of the infrapatellar branch of the saphenous nerve as well as all the other parameters of the International Knee Documentation Committee (IKDC) Objective Score. Patient-reported outcomes were evaluated using written questionnaires, including the IKDC Subjective Score, the Lysholm Knee Score, the Knee Injury and Osteoarthritis Outcome Score (KOOS) and the current Tegner Activity Scale. In addition, the medical records of all patients were assessed to determine demographics and details of the surgical procedure. Subsequently, knee laxity was evaluated with Telos stress x-rays and each patient had a magnetic resonance imaging (MRI) scan.
All operations were performed by one out of six surgeons specialized in knee surgery at our hospital, either with a fixed-loop (FL) system (Position ACL, B.Braun-Aesculap, Tuttlingen, Germany) or with an adjustable-loop (AL) technique (All-Inside ACL, Arthrex, Naples, FL, USA) according to the recommendations of the manufacturer. (Figure 1) Prior to the actual ACL reconstruction, meniscus and cartilage pathologies were treated arthroscopically, if necessary. A combination of anatomical landmarks (namely the remnants of the ACL, the lateral intercondylar ridge, the apex of the deep cartilage and the anterior horn of the lateral meniscus) were used to guide tunnel placement during surgery.
1. Fixed-Loop Fixation
The semitendinosus (ST) tendon alone was harvested through an incision over the pes anserinus superficialis. The two free ends were whip-stitched using non-resorbable high-strength suture material (Orthocord #2, DePuy Synthes Mitek Sports Medicine, Raynham, MA, USA), then the tendon was folded into an M-formation and inserted into the closed loop (Suture Plate, B.Braun-Aesculap, Tuttlingen, Germany) to obtain a four-stranded graft. The femoral socket was drilled according to anatomical landmarks through the anteromedial portal at 110-120° of flexion. The tibial tunnel was created in the middle of the remaining distal stump of the ACL. The graft was pulled transtibially into the femoral socket. Tibial fixation was performed using a suture button (Suture Disk, B.Braun-Aesculap, Tuttlingen, Germany). After passive cycling of the knee joint the transplant was tensioned in 30° of knee flexion by twisting of the tibial button.
2. Adjustable-Loop Fixation
The ST tendon was harvested through a minimally invasive popliteal approach.(20) (Figure 2) The tendon was symmetrically folded over two adjustable-length loops (ACL TightRope RT, Arthrex, Naples, FL, USA) in order to obtain a four-stranded graft. The graft was secured with two sutures at the tibial end and two sutures at the femoral end of the graft (FiberWire # 2; Arthrex, Naples, FL, USA).(21) A femoral and tibial socket was created at the anatomic ACL insertion site with an outside-in technique using a retrograde drill (FlipCutter, Arthrex, Naples, FL, USA). The graft was inserted through the AM portal and shuttled into the bone sockets via pull-through sutures. Cortical buttons were flipped, and the graft tensioned with the knee in extension. The knee was then passively cycled and the graft re-tensioned as required. Finally, the tensioning strands of the adjustable-loop were knotted with an arthroscopic knot pusher to reduce the risk of loop slippage.(9)
Post-operatively, no brace was used in the group of patients with isolated ACL tears (n=34) and full weight bearing was allowed after 2-4 weeks (after quadriceps control had been regained). The patients with meniscal repair (n=33) were restricted to partial weight-bearing for 4 weeks and flexion was limited to 60° for 4 weeks and to 90° for another 2 weeks with a knee brace.
Physiotherapy was started on the first post-operative day while still on admission. Return to sports was recommended not earlier than 9 to 12 months post-operatively, depending on thigh circumference, rehabilitation milestones and confidence to return to sports.
The same standardized rehabilitation protocol was used for both treatment groups.
Dynamic passive Lachman x-rays of both knees (operated knee and contralateral healthy knee) were taken by the department’s radiology technicians on a Telos GA-Ⅲ/E device (Telos GmbH Laubscher, Holstein, Switzerland) at 150 N. Differential (Dif AD) measurements (operated knee versus healthy contralateral knee) for the anterior drawer of the medial compartment were estimated, as described before.(22) (Figure 3)
All participants underwent MRI (1,5T, Avanto, Siemens AG Healthcare, Erlangen, Germany) to evaluate the integrity of the anterior cruciate ligament graft with a 15-channel phased-array transmit/receive knee coil (Siemens AG Healthcare, Erlangen, Germany) with the following sequences: sagittal (sag.) T1-weighted Turbo Spin Echo (TSE); sag., coronal (cor.), axial (ax.) T2-weighted Blade with Fat Saturation (FS); sag. Proton Density (PD) with Spatial Phase Coding (SPC).
Sample size analysis resulted in a need to recruit 24 patients per group to detect a clinically significant group difference (2.5 mm, SD 3.0 mm) in side-to-side difference of the anterior translation, given a significance level of 0.05 and a power of 0.80.(23) Statistical analyses were done using the free software environment R version 3.6.3 on a PC running Linux Ubuntu version 16.04.6 LTS.(24) Descriptive statistics are given as counts, percentages, or means, and standard deviations (SD) or ranges as appropriate. For inferential statistics, t-test and χ2-test were used, and p-values for tests performed are given rounded to three decimal places.