Anterior cruciate ligament (ACL) reconstruction surgery has been continuously evolving. There are multiple options for femoral-sided graft fixation in ACL reconstruction, one of which uses a femoral cortical button (FCB). Cortical buttons have been used with multiple techniques and have demonstrated good biomechanical properties for fixation.1 One such technique uses a device where the femoral cortical button is attached to an adjustable loop around which the femoral side of the graft is secured. The cortical button sits on the outer cortex of the femur, allowing the graft to be suspended through a retrograde-drilled tunnel. This device allows for minimally invasive stab incisions and the ability to tension the graft after cycling through the adjustable loop.2
Traditionally, the FCB is placed without direct intraoperative visualization by simply applying tension to the graft once the button has slid through the tunnel. However, misplacement of the FCB has been cited as a source of complications of this technique, and problems with this type of graft fixation have been reported in the literature, such as migration of the button through the posterior cortex, migration into the knee joint, loss of fixation, and the button becoming lodged into the tunnel.1,3–6 These misplacements are often not noted until postoperative radiographs are performed.4,5
Descriptions of several techniques have been published attempting to address misplacement of the FCB.2,7–9 Sonnery-Cottet et al9 described a technique where an arthroscopic camera was placed into the lateral gutter and visualized button placement through an arthrotomy created over the lateral femoral condyle. To the current authors' knowledge, no studies have compared placement positions using this technique vs the standard technique.
This study aimed to retrospectively compare postoperative radiographs of patients in whom the FCB was placed via direct vs indirect visualization. The authors' hypothesis was that, on postoperative radiographs, FCBs placed with direct visualization would demonstrate improved position compared with those placed using indirect methods.
Materials and Methods
Electronic medical records from January 2013 to December 2016 were searched using the Current Procedural Terminology code 29888 (arthroscopically aided anterior cruciate ligament repair/augmentation or reconstruction). Eighty-four patients were identified who underwent ACL reconstruction with autologous semitendinosus tendon graft. All patients underwent ACL reconstruction using an adjustable loop cortical fixation button device. The device used was the ACL TightRope (Arthrex, Naples, Florida). Anteroposterior and lateral radiographs were obtained at 1 week postoperatively. For patients with questionable button placement, additional internal rotation radiographs were performed. A single surgeon (C.R.) performed all procedures. All charts were retrospectively reviewed via the electronic medical record after appropriate institutional review board approval was obtained.
Inclusion criteria consisted of patients who underwent all-inside ACL reconstruction using the adjustable loop cortical fixation device using either the technique for direct button visualization or previously used indirect techniques. In addition, patients included were required to have postoperative radiographs performed at 1 week. Exclusion criteria consisted of those patients who did not have postoperative radiographs for evaluation of button position, patients who underwent revision ACL reconstruction, and cases where intraoperative fluoroscopy was used.
Technique for Direct Visualization
To visualize placement of the cortical button, several steps are performed during the procedure to visualize the extra-articular location of button placement on the lateral femoral condyle. After the passing strands for the cortical button are passed through the tunnel and out the lateral cortex of the femur, the arthroscopic shaver is inserted through the same stab incision through which the passing sutures were drawn (Figure 1). With the arthroscopic camera positioned in the inferior lateral portal, the camera is directed to the lateral gutter. The shaver is pressed against the lateral capsule from its extra-articular position. When this is located, the shaver is used to perform a capsulotomy, and the sutures are localized (Figure 2). The shaver is then used to clear soft tissue in the area of the femur where the button would be positioned. During this step, the sutures are held tight to prevent iatrogenic tearing by the shaver. The sutures are also kept short on the femoral side so as to allow advancement if the suture is frayed by the shaver. The FCB is then advanced using the passing suture and flipped onto the lateral cortex of the femur under direct visualization (Figure 3). An arthroscopic probe may be used through the stab incision to position the FCB if it does not sit directly on the lateral cortex of the femur. The authors' technique is a modification of a technique previously described by Sonnery-Cottet et al.9
Intraoperative photographs during right knee anterior cruciate ligament reconstruction. Graft-passing sutures shown exiting though stab incision on lateral aspect of knee after being passed through the femoral tunnel (A). Arthroscopic shaver inserted through same stab incision. Notice tension constantly held on passing sutures (B).
Arthroscopic view of lateral gutter of knee joint with arthroscopic shaver pressed against capsule positioned to create capsulotomy.
Arthroscopic view of lateral femur demonstrating direct visualization of femoral cortical button positioned flat on lateral cortex of femur during anterior cruciate ligament reconstruction.
Technique for Indirect Visualization
The passing strands for the femoral side are slowly pulled through the femur with adequate counter-resistance applied to the opposite end of the graft. The arthroscope is placed in the inferior medial portal and directed up the femoral canal to observe the FCB pass through the tunnel. The passing sutures and FCB are then pulled into the femoral-sided tunnel with slow advancement until the surgeon can feel the button flip on the lateral cortex of the femur. The passing sutures are then dropped, and tension is applied to the tibial side of the graft to seat the button on the lateral femoral cortex. The adjustable loop of the femoral side is lengthened prior to passing the button for both techniques.
A grading scheme was designed with the senior author (C.R.) to create standards for which button placement could be graded. This scale was created based on the senior author's experiences in reviewing FCB placement on postoperative radiographs. Button placements were allocated into 4 categories designated as type 1 through type 4 (Figure 4). It was deemed by the senior author that types 1 and 2 would be considered “optimal” and types 3 and 4 would be “suboptimal.” The senior author considered that a button that appeared with at least 1 end in contact with cortical bone (type 1 and type 2) would create a solid endpoint for graft tensioning.
Doss and Ruotolo grading scale for femoral cortical button placement.
Radiographs of each patient who met the inclusion criteria were collected. The operative reports for each case were then reviewed to determine the technique used during surgery (ie, direct vs indirect visualization). The compiled radiographs were then presented independently to 2 sports fellowship–trained orthopedic surgeons (R.M., C.R.) for grading. The graders were then educated on the grading system and provided cartoon examples of each type of placement (Figure 4). The graders were blinded to whether the button was placed under direct or indirect visualization.
The scores for each patient were recorded electronically as each radiograph was displayed to the grader on an electronic slide show. The chi-square test was used to analyze the difference between results determined to be optimal or suboptimal. Significance was determined when P<.05. Statistical analysis was performed using Excel for Mac version 14.0.0 (Microsoft, Redmond, Washington).
Ninety-nine patient charts were initially included for review. Of those, 15 patients did not have follow-up postoperative radiographs obtained, so these patients were excluded. Eight additional patients failed to meet inclusion criteria (4 of these patients had undergone revision ACL reconstruction and the other 4 were noted to have intraoperative fluoroscopy used during button placement). The index procedures for the 4 excluded patients who had undergone revision ACL reconstruction were performed at an outside institution by surgeon(s) unrelated to this study and were not resultant from failure using the authors' technique. Overall, 76 patients met the inclusion criteria and were included in the analysis. Forty-two cases were performed using direct visualization, whereas 34 cases used indirect methods.
Scores for each of the 76 cases analyzed were collected independently from the graders and used for analysis. Of the 42 patients in the direct visualization group, 0 were found to have suboptimal FCB placement. Of the 34 patients in the indirect visualization group, 3 were found to have suboptimal FCB placement. It was determined that using direct visualization provided statistically significant results for optimal placement of the FCB (chi-square test, P=.046). Comparing the 2 graders' individual scores for each radiograph, the authors determined moderate strength of interobserver reliability (kappa coefficient=0.62).10
Previous studies have shown that ACL graft fixation with cortical buttons provides secure biomechanical fixation.1 A current device used consists of an FCB attached to an adjustable femoral loop. This construct allows for minimally invasive incisions and tensioning of the graft after it is cycled and secured.2 Several techniques have been described to ensure proper FCB placement. The current study compared results using 2 different techniques for visualizing FCB placement. By comparing the 2 techniques, the authors established that direct visualization helps to ensure optimal FCB placement.
The authors examined postoperative radiographs from patients who had undergone an ACL reconstruction with an FCB to evaluate its position and placement. A previous technique described by Sonnery-Cottet et al9 involved directly examining the cortical button through an arthrotomy in the superior lateral capsule. Those authors stated that since the implementation of this technique, no additional misplacements were noted. The current study's evidence supports the use of this described direct visualization method.
Another technique by Harato et al8 described marking the passing suture at a precise position equal to lateral femur length and using the flip sutures and back tension to ensure the button is correctly positioned. They reported that 100 cases performed using this technique demonstrated proper placement on the lateral femoral cortex. This method is also the described technique provided by the manufacturer's guide. Of the 76 patients reviewed in the current study, the authors found only 1 patient to have a type 4 placement (Figure 5). In this instance, the button was positioned inside the femoral tunnel. Position in the femoral tunnel was based off of the standard postoperative anteroposterior radiograph, but the position was further confirmed after obtaining multiple internal rotation views. This patient was in the indirect visualization group. Karaoglu et al5 reported a similar case where an FCB device was found to be lodged in the femoral tunnel. Neither the current authors' patient who was found to have a type 4 placement nor the patient mentioned in the above-referenced study had graft failure at final follow-up. The postoperative rehabilitation protocol for the current patient who was found to have a type 4 placement was slowed for the first 2 months due to the suboptimal button placement. It was thought that the button became wedged in the tunnel intraoperatively, simulating the feeling of fixation when tested by pulling counter-traction. The authors believe that these types of complications can be successfully avoided by directly visualizing FCB placement.
Postoperative anteroposterior knee radiograph showing a femoral cortical button as a type 4. Notice the button is positioned within the femoral tunnel.
The FCB has been shown to be a successful option for ACL graft fixation.1 However, several reports have indicated that the FCB may migrate, resulting in loss of fixation.4,6 Mae et al6 analyzed postoperative radiographs of their ACL reconstructions secured with FCB, looking specifically at FCB position changes at each follow-up. They determined that any portion of button elevated off the femoral cortex on radiographs was considered to have soft tissue interposition between the button and cortical bone. Their findings noted that buttons with soft tissue inter-position had significantly higher rates of migration compared with buttons deemed flush against bone. This supports the current authors' belief that it is of the utmost importance to achieve and confirm proper button placement to limit fixation migration and potential graft loosening.
A secondary goal of the current study was to create a categorical grading system to describe button placement. Two previous studies analyzed postoperative FCB placement on radiographs and used the measurement from the lateral femoral cortex and angle of the button compared with the axis of the femur.3,6 Uchida et al3 stated that exact anteroposterior and lateral radiographs must be performed to accurately measure positioning. The current authors' scale allows an evaluator to quickly and accurately grade an FCB placement without measurement tools or software and without an exact anteroposterior radiograph. The interobserver reliability of the authors' scale demonstrated moderate agreement between the 2 graders. There was no disagreement between the graders regarding FCB placements determined to be type 3 or type 4, which were deemed suboptimal.
All patients included in this study had radiographs obtained at follow-up, and, in some instances, the knee was rotated in such a way that the button was not visualized parallel to the lateral cortex. Although the authors note that exact anteroposterior and lateral radiographs are not required, grades may have been affected by the obliquity of the provided radiographs. The authors' postoperative protocol dictated that patients receive knee radiographs at the 1-week postoperative visit. The FCB position could possibly change within the first week, given the previous reports of button migration.3,6
This study was a retrospective case review. Only patients who underwent this procedure at the authors' institution were included, thereby limiting the number available for analysis. The authors speculate that increasing the number of patients in each group would increase the significance of their results. However, given their experience of more optimal placement with the direct visualization technique, the primary surgeon is reluctant to return to the previously used indirect techniques to create a prospective, randomized study.
Several studies have shown that button misplacement and migration are complications associated with FCB fixation of an ACL graft. This study demonstrated that direct arthroscopic visualization while placing an FCB during ACL reconstruction with adjustable loop cortical fixation buttons can help ensure optimal button placement.
- Kousa P, Järvinen TLN, Vihavainen M, Kannus P, Järvinen M. The fixation strength of six hamstring tendon graft fixation devices in anterior cruciate ligament reconstruction: Part I. Femoral site. Am J Sports Med. 2003;31(2):174–181. doi:10.1177/03635465030310020401 [CrossRef] PMID:12642249
- Lubowitz JH, Ahmad CH, Anderson K. All-inside anterior cruciate ligament graft-link technique: second-generation, no-incision anterior cruciate ligament reconstruction. Arthroscopy. 2011;27(5):717–727. doi:10.1016/j.arthro.2011.02.008 [CrossRef]
- Uchida R, Mae T, Matsumoto N, Kuroda S, Toritsuka Y, Shino K. The effect of cortical button location on its post-operative migration in anatomical double-bundle anterior cruciate ligament reconstruction. Knee Surg Sports Traumatol Arthrosc. 2014;22(5):1047–1054. doi:10.1007/s00167-013-2458-3 [CrossRef] PMID:23462955
- Yanmis I, Tunay S, Oguz E, Yildiz C, Ozkan H, Kirdemir V. Dropping of an EndoButton into the knee joint 2 years after anterior cruciate ligament repair using proximal fixation methods. Arthroscopy. 2004;20(6):641–643. doi:10.1016/j.arthro.2004.03.016 [CrossRef] PMID:15241318
- Karaoglu S, Halici M, Baktir A. An unidentified pitfall of Endobutton use: case report. Knee Surg Sports Traumatol Arthrosc. 2002;10(4):247–249. doi:10.1007/s00167-002-0287-x [CrossRef] PMID:12172720
- Mae T, Kuroda S, Matsumoto N, et al. Migration of EndoButton after anatomic double-bundle anterior cruciate ligament reconstruction. Arthroscopy. 2011;27(11):1528–1535. doi:10.1016/j.arthro.2011.06.024 [CrossRef] PMID:21924859
- Nag HL, Gupta H. Seating of TightRope RT button under direct arthroscopic visualization in anterior cruciate ligament reconstruction to prevent potential complications. Arthrosc Tech. 2012;1(1):e83–e85. doi:10.1016/j.eats.2012.03.003 [CrossRef] PMID:23766982
- Harato K, Niki Y, Toyoda T, et al. Self-flip technique of the TightRope RT button for soft-tissue anterior cruciate ligament reconstruction. Arthrosc Tech. 2016;5(2):e391–e395. doi:10.1016/j.eats.2016.01.022 [CrossRef] PMID:27462539
- Sonnery-Cottet B, Rezende FC, Martins Neto A, Fayard JM, Thaunat M, Kader DF. Arthroscopically confirmed femoral button deployment. Arthrosc Tech. 2014;3(3):e309–e312. doi:10.1016/j.eats.2014.01.007 [CrossRef] PMID:25126492
- McHugh ML. Interrater reliability: the kappa statistic. Biochem Med (Zagreb).2012;22(3):276–282. doi:10.11613/BM.2012.031 [CrossRef] PMID:23092060