Despite the increasing prevalence of anterior cruciate ligament reconstructions, many nuances of anterior cruciate ligament reconstruction technique remain variable and at the discretion of the operating surgeon. The method of graft fixation on the tibial side has been heavily studied, and numerous techniques have been reported to provide secure graft fixation and low revision rates for both soft tissue and bone plugs in the tibial tunnel.1–7 Prior to biologic incorporation of the anterior cruciate ligament graft, the strength of the graft construct is primarily dependent on its fixation. Fixation failure, when it occurs, is more likely to happen on the tibial side.8,9
Although numerous reconstructive techniques, graft options, and fixation modalities have been reported with success, bone–patellar tendon–bone (BPTB) autograft secured with interference screws on both the tibial and femoral side remains a widely accepted anterior cruciate ligament reconstruction technique option in young, active patients.10 Bone plugs may be successfully secured with metal or bioabsorbable interference screws. No significant differences in functional outcomes have been reported between these fixation modalities, although some authors have cited concerns regarding increased knee effusions, screw breakage, and degradation with bioabsorbable screws.11–13 Metal interference screws remain a proven, effective, inexpensive option for graft fixation. In the tibial tunnel, the security of the graft–screw interface and the length of plug in contact with the screw must be evaluated and scrutinized intraoperatively. Multiple factors may influence the graft– screw interface, including tunnel lengths, graft lengths, and tunnel orientation.14,15
The orientation of the tibial bone plug and the position of the screw may also affect the overall construct integrity. The authors' institutional experience has been that placement of the screw along the cancellous surface of the graft provides more reliable and secure fixation than placement of the screw along the cortical side. This seems counterintuitive because cortical-based fixation is commonly used and preferred in fracture management over cancellous-based fixation. Additionally, the bone-to-bone healing provided by placing the cancellous portion of the graft against the cancellous tunnel wall would be optimal, as opposed to placing the cortical portion of the graft against the cancellous tunnel wall.16 The cortical side of the plug, however, is covered by the soft tissue envelope of the patellar tendon, so the screw does not directly engage cortical bone. Furthermore, there is a risk that the screw threads lacerate the tendon–bone interface and weaken the integrity of the insertion. This has been a point of controversy in previous publications.16,17
Rupp et al18 previously evaluated the tibial screw positioning in relation to the tibial bone plug in a porcine biomechanical study. They reported no difference in fixation strength between cortical and cancellous-sided screw placement, but there were differences in failure modality between the groups. Cortical-based screw placement failed via tendon rupture and bone plug fracture (70%), whereas cancellous-based screw placement failed via pullout (75%). Unfortunately, the authors did not describe their testing protocol in detail, and it was unclear if the specimens were precycled and preloaded prior to testing. Additionally, the authors used a 9-mm diameter standardized plug width in a 10-mm tunnel fixed with a 9×25-mm screw. It is possible that a wider plug would allow better press-fit in the tunnel and prevent pullout. Given the current authors' institutional experience with graft “slippage” with cortical-based screw fixation, they intended to replicate the experiment by Rupp et al in the setting of larger bone plugs and a formal loading protocol.
Thus, the purpose of this study was to use a porcine model to biomechanically evaluate 2 different tibial BPTB fixation constructs using a formal loading protocol. In the first group, the tibial interference screw would abut the cancellous surface of the graft. In the second group, the screw would abut the cortical and tendon insertion surface of the graft. On the basis of their institutional experience, the authors hypothesized that a cancellous screw interface would be biomechanically favorable while minimizing the risk for damage to the tendon–bone interface.
Materials and Methods
Forty fresh-frozen porcine tibias and BPTB specimens were purchased (Animal Technologies, Inc, Tyler, Texas). All donor specimens were harvested from 6- to 9-month-old Yorkshire pigs weighing less than 250 lb. Tissues were harvested within 6 hours of death and then immediately frozen for storage and transport.
After thawing these specimens at room temperature for 24 hours, small BPTB grafts were discarded to standardize the graft pool and eliminate outliers. The 10×20-mm bone plugs were contoured from one side of the BPTB specimens (Figure 1). The authors' current institutional practice is to use a 10×25-mm plug on the tibial side; however, the porcine tibias were truncated at 20 mm length, necessitating modification of the plug dimensions. The authors used a 10-mm diameter to maximize press-fit in the tibial tunnel, which was verified with a sizing block. A 10-mm cylindrical reamer was used to retrograde ream a 10-mm diameter tibial tunnel in the specimens, oriented from the proximal medial tibia to the anterior cruciate ligament footprint. The femoral side of the graft was cut to 10 mm length and fashioned to facilitate placement into the testing apparatus.
Sample porcine bone–patellar tendon– bone graft specimen following preparation. Note the tibial bone plug fashioned to 10×20 mm with a single drill hole to facilitate placement. The femoral portion of the graft was tailored to fit the testing apparatus.
Suboptimal specimens were discarded secondary to small size, brittle tissue, or graft damage during preparation. The remaining intact specimens were then randomly divided into 2 groups. The grafts were then fixed within the tibial tunnels using a 9×25-mm titanium interference screw (Arthrex, Naples, Florida) (Figure 2).
Bone–patellar tendon–bone specimen secured in the tibial tunnel and ready for testing.
In group A, the interference screw was placed adjacent to the cancellous portion of the bone plug (n=13). In group B, the screw was placed adjacent to the cortical portion of the plug (n=14). Figure 3 provides a lateral view of the proximal tibia and the screw position in relation to the bone plug for groups A and B.
Illustration of a lateral view of the proximal tibia depicting interference screw placement anterior to the bone plug and the orientation of the cortical bone in the tibial tunnel (red=cortical side). In group A, the screw was placed adjacent to the cancellous surface. In group B, the screw was placed adjacent to the cortical surface. The tendon is depicted in blue. The screw is depicted in green.
All specimens were prepared in 1 day and then refrozen to await loading on a different day. They were rethawed during 24 hours in similar fashion prior to strength testing (total of 2 freeze–thaw cycles).
Testing was performed using an Insight 150SL loading machine (MTS Systems Corporation, Eden Prairie, Minnesota) (Figure 4). Specimens were placed through a sequence of pretensioning, cyclic loading, and load-to-failure testing. All loading was completed parallel to the long axis of the anterior cruciate ligament construct.
Graft construct loaded in the testing apparatus and ready for loading protocol. The testing apparatus and graft were oriented to pull directly in-line with the anterior cruciate ligament.
Considerable variability in loading protocols in previous studies precludes a widely accepted protocol in the porcine model. The loading protocol used included a brief precycling, cyclic loading, and ultimate load-to-failure and was modeled after several previous porcine studies.19–21 It began with a precycle of 20 N at 1 Hz for 5 minutes to pretension the construct. After pretensioning, the initial length of the graft was measured. Cyclic load testing consisted of 100 cycles between 50 and 250 N at a strain rate of 100 mm/min, 400 cycles between 50 and 100 N at a strain rate of 100 mm/min, and 500 cycles between 50 and 100 N at a strain rate of 100 mm/min. Finally, the construct underwent load-to-failure testing. Ultimate load-to-failure, elongation, stiffness, and peak stress were all recorded using Test Works V.4 software (MTS Systems Corporation). All modes of failure were visually inspected and recorded. Mode of failure was characterized as either high tendon, mid tendon, low tendon, or bone plug.
Initially, the authors calculated means and standard deviations for continuous variables and frequencies and proportions for categorical variables. Independent t tests were used to evaluate between-group differences in ultimate load-to-failure. The mode of failure was expressed as a proportion of the total tendons in each group to facilitate descriptive comparisons between the 2 groups. Because several specimens failed prior to load-to-failure testing, the authors calculated Kaplan–Meier survival estimates by group to compare time to failure during the cyclic loading procedure. The non-parametric log-rank test for equality of survivor functions was used to compare time to failure between groups. Cox proportional hazards regression analysis was used to estimate hazard ratios and 95% confidence intervals between the groups. All statistical analyses were completed using Stata SE version 10.1 software (StataCorp, College Station, Texas). The alpha level was set at P<.05 a priori for all between-group comparisons.
The ultimate load to construct failure was significantly higher in specimens where the screw was placed adjacent to the cancellous portion of the bone plug (group A) when compared with specimens where the screw was placed adjacent to the cortical portion (group B). The mean ultimate load to construct failure in group A was 493±245 N compared with 304±145 N in group B (t=2.83, P=.008). Mean elongation and stiffness were not reported because multiple tendons from each group failed before the completion of the cyclic loading protocol.
The mode of failure and proportion of failure by group are presented in Figure 5. Specimens in group A (cancellous-sided screw placement) most commonly experienced high or mid-tendon failure, whereas specimens in group B (cortical-sided screw placement) most commonly failed low on the tendon adjacent to the tendon–plug interface or by pullout. None of the group A specimens failed via plug pullout, whereas 14% of group B specimens failed via plug pullout.
Graph showing the proportion of failure modes by group (high tendon, mid tendon, low tendon, bone plug). Group A failures are depicted in blue. Group B failures are depicted in red.
Kaplan–Meier survival estimates by group are presented in Figure 6. Specimens with fixation placed adjacent to the cortical portion of the bone plug (group B) failed at a greater rate when compared with group A (chi-square=4.43, P=.035). All specimens that failed during the cyclic loading protocol did so within the first 500 cycles (Table). Overall, fixation construct failure was nearly 4 times more likely (hazard ratio, 3.80; 95% confidence interval, 1.00–14.40; P=.050) during the cyclic loading protocol for specimens in group B.
Kaplan–Meier survival curve depicting graft survival during cyclic loading during 1000 cycles. Group A (cancellous) is depicted in blue. Group B (cortical) is depicted in red.
Survival Functions by Group at Specific Time Points in the Cyclic Loading Procedure
The results of this biomechanical study reveal stronger and more durable tibial fixation for BPTB grafts fixed with an interference screw adjacent to the cancellous portion of the bone plug as opposed to adjacent to the cortical portion. Cancellous-based screw placement produced higher ultimate failure loads and increased survival during cyclic loading in comparison with cortical-based screw placement. The authors also identified a different mode of failure between the 2 groups, with 85% of group A failing at the mid tendon or higher as opposed to 100% of group B failing at the mid tendon or lower.
These results differ from those of Rupp et al,18 who investigated a similar porcine tibial BPTB fixation model. They found no significant difference in fixation strength between the 2 groups, whereas the current study revealed significant differences in ultimate load-to-failure and cyclic loading survival between the cortical and cancellous groups. This difference may be explained by the smaller-diameter bone plugs (9 mm) used in their study as well as the difference in their loading protocol. The current authors used a loading protocol, with precycling to reduce graft creep, that has been described and used in recent biomechanical studies, although there is significant variability among those studies that precludes use of a single, widely accepted protocol.19–21 Rupp et al18 also reported that most of the cancellous-based screw constructs failed via pullout. Considering that none of the specimens in the current cancellous group failed via pullout, it would be reasonable to conclude that use of the largest-diameter plug possible (10 mm) in the tibial tunnel (10 mm) provides optimal press-fit and construct resistance to pullout.
Fifty-seven percent of the group B specimens failed low on the tendon or due to bone plug pullout, and 100% of group B specimens failed at or below mid tendon. Theoretically, the metal screw threads compressing the tendon attachment against the cortical side of the plug may act to lacerate or weaken the tendon interface and predispose the construct to failure. However, this has previously been a point of controversy. Nogalski and Bach17 recommended placement of the screw on the cortical side of the graft to provide cortico-cancellous fixation, which they viewed as optimal for stability. However, Kurzweil and Jackson16 advocated placement of the screw adjacent to the cancellous side to avoid risk to the tendon interface. Rupp et al18 recognized the potential for tendon laceration and the “unpredictability” of cortical-sided placement; however, their results indicated no difference between the groups in terms of construct strength. The results of this biomechanical study support the findings of Kurzweil and Jackson16 in that cancellous-sided screw placement may be optimal.
This study had several limitations, which were largely related to the storage and handling of the large amount of biologic material needed to conduct testing. It is possible that graft integrity in this study was influenced by multiple freeze– thaw cycles. During the course of graft preparation and testing, the specimens underwent 2 freeze–thaw cycles, and the authors had concerns regarding the brittleness of some of the specimens after the second cycle. Lee et al22 investigated the effect of freeze–thaw cycles on porcine bone–tendon–bone grafts tested under cyclic loading. They subjected their specimens to 2 freeze–thaw cycles and found no difference in strength during testing. However, considering that a large number of the specimens failed to complete the cyclic loading protocol intact, the authors theorize that the freeze–thaw process may have negatively influenced the specimens' structural integrity nonetheless. The differences noted are unlikely to be a function of the authors' model, as porcine specimens are well validated for assessing biomechanical properties of bone and ligament.23,24 Additionally, modifications to the loading protocol, such as increasing the precycling load and reducing the strain rate, may have allowed more specimens to complete load-to-failure testing. This study would be optimally performed in fresh tissue; however, the logistical limitations of transport, storage, and specimen preparation make a study of this nature difficult. The ages and weights of the donor animals also influence the specimen quality (ie, BPTB graft size), and this was apparent during handling and preparation. The authors attempted to mitigate these differences by eliminating the small or brittle grafts and standardizing the total pool prior to randomization, but this certainly may have affected their results. Finally, the Insight 150SL loading machine used for load testing in this study is designed ideally for high-strength materials testing and not specifically for biologic purposes, although “biomedical” use is listed as a potential application in the product guide.25 To detect subtle differences in weaker biologic tissue, a loading device specifically designed for biologic use would have been ideal. This may also partially explain why a large number of the specimens did not complete the cyclic loading protocol.
Secure graft fixation is paramount to the success of an anterior cruciate ligament reconstruction.26,27 The authors' results indicate that interference screw positioning adjacent to the cancellous surface of the BPTB plug is biomechanically superior compared with the cortical surface. This has clinical relevance in the early postoperative period prior to biologic incorporation of the graft. However, these results must be interpreted with caution due to both concerns about specimen quality and initial fixation strength not being the only factor a surgeon must consider in selecting a graft construct. Multiple factors affect graft construct choice, including patient age, activity level, ligamentous laxity, concurrent injuries, and available fixation devices. The authors did not evaluate bioabsorbable screws, and the biomechanical findings of this study may not apply to that clinical scenario. Further study investigating whether there is a clinical difference in biologic incorporation between the 2 screw positions or the clinical outcomes would also be useful.
- Aga C, Rasmussen MT, Smith SD, et al. Biomechanical comparison of interference screws and combination screw and sheath devices for soft tissue anterior cruciate ligament reconstruction on the tibial side. Am J Sports Med. 2013; 41(4):841–848. doi:10.1177/0363546512474968 [CrossRef]
- Harilainen A, Sandelin J. A prospective comparison of 3 hamstring ACL fixation devices— Rigidfix, BioScrew, and Intrafix—randomized into 4 groups with 2 years of follow-up. Am J Sports Med. 2009; 37(4):699–706. doi:10.1177/0363546508328109 [CrossRef]
- Xie X, Liu X, Chen Z, Yu Y, Peng S, Li Q. A meta-analysis of bone-patellar tendon-bone autograft versus four-strand hamstring tendon autograft for anterior cruciate ligament reconstruction. Knee. 2015; 22(2):100–110. doi:10.1016/j.knee.2014.11.014 [CrossRef]
- Prentice HA, Lind M, Mouton C, et al. Patient demographic and surgical characteristics in anterior cruciate ligament reconstruction: a description of registries from six countries. Br J Sports Med. 2018; 52(11):716–722. doi:10.1136/bjsports-2017-098674 [CrossRef]
- Chivot M, Harrosch S, Kelberine F, Pithioux M, Argenson JN, Ollivier M. Pull-out strength of four tibial fixation devices used in anterior cruciate ligament reconstruction. Orthop Traumatol Surg Res. 2018; 104(2):203–207. doi:10.1016/j.otsr.2017.12.011 [CrossRef]
- Kousa P, Järvinen TL, Vihavainen M, Kannus P, Järvinen M. The fixation strength of six hamstring tendon graft fixation devices in anterior cruciate ligament reconstruction: Part II. Tibial site. Am J Sports Med. 2003; 31(2):182–188. doi:10.1177/03635465030310020501 [CrossRef]
- Adam F, Pape D, Schiel K, Steimer O, Kohn D, Rupp S. Biomechanical properties of patellar and hamstring graft tibial fixation techniques in anterior cruciate ligament reconstruction: experimental study with roentgen stereometric analysis. Am J Sports Med. 2004; 32(1):71–78. doi:10.1177/0095399703258608 [CrossRef]
- Scheffler SU, Südkamp NP, Göckenjan A, Hoffmann RF, Weiler A. Biomechanical comparison of hamstring and patellar tendon graft anterior cruciate ligament reconstruction techniques: the impact of fixation level and fixation method under cyclic loading. Arthroscopy. 2002; 18(3):304–315. doi:10.1053/jars.2002.30609 [CrossRef]
- Campbell KA, Looze C, Bosco JA, Strauss EJ. The weak link in anterior cruciate ligament reconstruction: what is the evidence for graft fixation devices?Bull Hosp Jt Dis (2013). 2016; 74(1):12–23.
- Chen NC, Brand JC Jr, Brown CH Jr, . Biomechanics of intratunnel anterior cruciate ligament graft fixation. Clin Sports Med. 2007; 26(4):695–714. doi:10.1016/j.csm.2007.06.009 [CrossRef]
- Debieux P, Franciozi CE, Lenza M, et al. Bioabsorbable versus metallic interference screws for graft fixation in anterior cruciate ligament reconstruction. Cochrane Database Syst Rev. 2016; 7:CD009772.
- Mascarenhas R, Saltzman BM, Sayegh ET, et al. Bioabsorbable versus metallic interference screws in anterior cruciate ligament reconstruction: a systematic review of overlapping meta-analyses. Arthroscopy. 2015; 31(3):561–568. doi:10.1016/j.arthro.2014.11.011 [CrossRef]
- Shen C, Jiang SD, Jiang LS, Dai LY. Bioabsorbable versus metallic interference screw fixation in anterior cruciate ligament reconstruction: a meta-analysis of randomized controlled trials. Arthroscopy. 2010; 26(5):705–713. doi:10.1016/j.arthro.2009.12.011 [CrossRef]
- Boddu CK, Arif SK, Hussain MM, Sankaranarayanan S, Hameed S, Sujir PR. Prevention of graft-tunnel mismatch during anatomical anterior cruciate ligament reconstruction using a bone-patellar tendon-bone graft. Bone Joint J. 2015; 97-B(3):324–328. doi:10.1302/0301-620X.97B3.34653 [CrossRef]
- Shaffer B, Gow W, Tibone JE. Graft-tunnel mismatch in endoscopic anterior cruciate ligament reconstruction: a new technique of intraarticular measurement and modified graft harvesting. Arthroscopy. 1993; 9(6):633–646. doi:10.1016/S0749-8063(05)80499-2 [CrossRef]
- Kurzweil PR, Jackson DW. Chronic anterior cruciate ligament injuries. In: Fu FH, Harner CD, Vince KG, eds. Knee Surgery. Baltimore, MD: Williams & Wilkins; 1994:731–748.
- Nogalski MP, Bach BR. Acute anterior cruciate ligament injuries. In: Fu FH, Harner CD, Vince KG, eds. Knee Surgery. Baltimore, MD: Williams & Wilkins; 1994:679–730.
- Rupp S, Seil R, Krauss PW, Kohn DM. Cortical versus cancellous interference fixation for bone-patellar tendon-bone grafts. Arthroscopy. 1998; 14(5):484–488. doi:10.1016/S0749-8063(98)70076-3 [CrossRef]
- Walsh MP, Wijdicks CA, Parker JB, Hapa O, LaPrade RF. A comparison between a retrograde interference screw, suture button, and combined fixation on the tibial side in an all-inside anterior cruciate ligament reconstruction: a biomechanical study in a porcine model. Am J Sports Med. 2009; 37(1):160–167. doi:10.1177/0363546508323747 [CrossRef]
- Smith PA, DeBerardino TM. Tibial fixation properties of a continuous-loop ACL hamstring graft construct with suspensory fixation in porcine bone. J Knee Surg. 2015; 28(6):506–512.
- Monaco E, Fabbri M, Lanzetti RM, Del Duca A, Labianca L, Ferretti A. Biomechanical comparison of four coupled fixation systems for ACL reconstruction with bone socket or full-tunnel on the tibial side. Knee. 2017; 24(4):705–710. doi:10.1016/j.knee.2017.05.003 [CrossRef]
- Lee GH, Kumar A, Berkson E, Verma N, Bach BR Jr, Hallab N. A biomechanical analysis of bone-patellar tendon-bone grafts after repeat freeze-thaw cycles in a cyclic loading model. J Knee Surg. 2009; 22(2):111–113. doi:10.1055/s-0030-1247734 [CrossRef]
- Pearce AI, Richards RG, Milz S, Schneider E, Pearce SG. Animal models for implant biomaterial research in bone: a review. Eur Cell Mater. 2007; 13:1–10. doi:10.22203/eCM.v013a01 [CrossRef]
- Fuss FK. Anatomy and function of the cruciate ligaments of the domestic pig (Sus scrofa domestica): a comparison with human cruciates. J Anat. 1991; 178:11–20.
- MTS Systems Corporation. MTS Insight Electromechanical Testing Systems. https://www.upc.edu/sct/documents_equipament/d_69_id-423.pdf. Accessed April 2, 2018.
- Dargel J, Gotter M, Mader K, Pennig D, Koebke J, Schmidt-Wiethoff R. Biomechanics of the anterior cruciate ligament and implications for surgical reconstruction. Strategies Trauma Limb Reconstr. 2007; 2(1):1–12. doi:10.1007/s11751-007-0016-6 [CrossRef]
- Hapa O, Barber FA. ACL fixation devices. Sports Med Arthrosc Rev. 2009; 17(4):217–223. doi:10.1097/JSA.0b013e3181bf668c [CrossRef]
Survival Functions by Group at Specific Time Points in the Cyclic Loading Procedure
|Loading Cycles||Fixation Construct, Kaplan–Meier Survival Estimate (95% Confidence Interval)|
|Group A||Group B|
|100||0.846 (0.512–0.959)||0.571 (0.284–0.779)|
|250||0.769 (0.442–0.919)||0.500 (0.229–0.722)|
|500||0.692 (0.373–0.872)||0.214 (0.052–0.448)|
|750||0.692 (0.373–0.872)||0.214 (0.052–0.448)|
|1000||0.692 (0.373–0.872)||0.214 (0.052–0.448)|