Effect of Screw Length on Bioabsorbable Interference Screw Fixation in a Tibial Bone Tunnel*

Transcription

1 /101/ $02.00/0 THE AMERICAN JOURNAL OF SPORTS MEDICINE, Vol. 29, No American Orthopaedic Society for Sports Medicine Effect of Screw Length on Bioabsorbable Interference Screw Fixation in a Tibial Bone Tunnel* Jeffrey B. Selby, MD, Darren L. Johnson, MD, Peter Hester, MD, and David N. M. Caborn, MD From the University of Kentucky School of Medicine, Lexington, Kentucky ABSTRACT Initial tibial fixation strength is the weak link after anterior cruciate ligament reconstruction with a quadrupled hamstring tendon graft fixed with bioabsorbable interference screws. The purpose of this study was to determine the biomechanical differences between 28-mm and tapered 35-mm interference screws for tibial fixation of a soft tissue graft in 16 young cadaveric tibias. Failure mode, displacement before failure, and ultimate failure load were tested with a testing machine aligned with the tibial tunnel to simulate a worst-case scenario. The mode of failure was graft slippage past the screw in all but one of the specimens. The mean maximum load at failure of the 28-mm screw was N, with mean displacement at failure of mm. The mean maximum load at failure of the 35-mm screw was N, with a mean displacement to failure of mm. The 38% difference in mean maximal load at failure was significant. Important variables in hamstring tendon graft fixation within a bone tunnel include bone mineral density, dilatation, gap size, screw placement, and screw width and length. Attention to these variables will help to provide secure graft fixation during biologic incorporation throughout the rehabilitation period. Reconstruction of the ACL has become an accepted method of treatment for acute and chronic ACL tears, and many different graft types and fixation methods have been developed. The most commonly used graft is bone-patellar *Presented at the interim meeting of the AOSSM, Orlando, Florida, March Address correspondence and reprint requests to David N. M. Caborn, MD, Department of Orthopaedic Surgery, University of Kentucky School of Medicine, K401 Kentucky Clinic, Lexington, KY Funding was received from commercial parties related to products in this study. See Acknowledgment for funding information. tendon-bone autograft, and this has proven to have an acceptable success rate with a relatively low morbidity. 13,20 However, the problems of anterior knee pain, patellofemoral crepitus, patellar fracture, quadriceps muscle atrophy, and difficulty achieving knee motion that have been described are major reasons for the development of new graft materials. 2,4,18,23,26 Harvest of hamstring tendon grafts has proven to result in less morbidity than harvest of patellar tendon, and the incidence of extensor mechanism dysfunction and pain is negligible. 9,15,32 However, some questions remain regarding the quality of the bone-tunnel fixation of this soft tissue graft both immediately after the operation and in the long term. The benefit of early motion and intensive rehabilitation after ACL reconstruction makes initial fixation strength integral to the success of the procedure. Because interference screws have been used successfully in bone-patellar tendon-bone fixation, these screws have also been used with soft tissue grafts with little difference in functional outcome. 1,5,10,22 It was found that the metal interference screw, with its sharp threads, could injure the tissue and potentially cause failure of fixation. 7 New screws are available with softer threads, and bioabsorbable screws may be advantageous in cases where MRI or revision surgery is necessary. Studies have proven that, in the femur, fixation of soft tissue grafts with interference screws provides adequate strength for early rehabilitation, but the adequacy of interference screw fixation in the tibia has remained in question. 7,14,16 The tibial bone is less dense and the bone tunnel is inevitably longer than the screw. The screw must contact cortical bone as well as obtain cancellous purchase over a long area to provide good interference and fixation strength. This study attempts to prove the hypothesis that use of a longer screw in the tibial tunnel will improve initial fixation strength by retaining cortical contact with the anterior tibial cortex while filling the tunnel to provide for a better interference fit. Tunnel dilatation, screw placement, and meticulous graft suturing techniques are im- 614

2 Vol. 29, No. 5, 2001 Length of Bioabsorbable Screws for Interference Fixation 615 portant variables in increasing pull-out strength. In this study, the grafts were cycled at a low stress to decrease crimp in the system and were then pulled out at a low rate of 20 mm/min to place the maximum stress on the graftscrew interface. Young cadaveric bones were used to simulate the clinical situation. MATERIALS AND METHODS Sixteen paired fresh-frozen tibias from cadavers 24 to 45 years of age (average, 38.5) were used for the study. Half of the paired specimens were received with the majority of soft tissue attachments removed, and the other half comprised whole fresh-frozen knees. The hamstring tendons were harvested from the whole knees to be used for grafts, and allografts were used for those knees with the soft tissues removed. All the tibias were stripped of their soft tissue attachments for bone mineral density determination and pull-out testing. A DXA scanner (Hologic QDR-1000 whole-body x-ray Bone Densitometer; Hologic, Inc., Bedford, Massachusetts) was used to determine the bone mineral density of the proximal portion of each of the tibias after the bones were thawed in a refrigerator overnight. The bones were submerged in normal saline solution, and density determinations were obtained with use of the performance-spine mode. Measurements of the metaphyseal and diaphyseal portions of the bones were recorded. The metaphyseal density results are reported in this study because this is the area of interference screw fixation in the tibial tunnel. During graft preparation, care was taken to prevent desiccation of the tissue by using normal saline-solution spray. The grafts were folded over a No. 5 Ti-Cron suture (Ethicon, Inc., Somerville, New Jersey), and a 3 0 suture was used to connect the distal ends of the tendons together at about 100 mm to allow for equal tensioning. The length included 30 mm of sutured tendon on the femoral end for grasping by the testing machine, 40 mm on the tibial end, and 30 mm of intraarticular tissue. Grafts were presized before suturing, and the grafts were resized after suture placement because the dimensions might have been the same or might have been reduced by 0.5 mm. The grafts were tensioned to 44 N (4.5 kg) during preparation. A No. 5 Ti-Cron suture whipstitch was placed at both ends of the graft and pulled tight with each stitch (Fig. 1). The sutures were overlapped to interdigitate, giving the interference screws a path to follow without the risk of splitting the tendon fibers with their edges. The sutures on the end were tied to use for pull-through, and the graft was wrapped in moist gauze during tunnel preparation. The tunnels were prepared with use of a tibial-tunnel guide set to 50 and placed midway front-to-back on the anterior cortex, with the pin placed at the ACL footprint. The tunnel length was measured at about 40 mm, and the appropriate-sized reamer was then used to ream the tibial tunnel over the guide wire to a size 2 mm smaller than the desired tunnel diameter. Tunnel dilators were then used to sequentially dilate the tunnel by 0.5 mm to the desired diameter. The prepared grafts were then placed into the tunnel in the usual retrograde fashion and secured in the Figure 1. Sample of a prepared graft. Note the tight interdigitating sutures. tunnel with either a 28-mm screw or a tapered 35-mm screw (Arthrex, Naples, Florida), with a diameter 1 mm larger than the diameter of the tunnel (Fig. 2). For each pair of knees, one knee was randomly chosen to have the 28-mm screw and the other was fixed with the 35-mm screw. The folded end of the graft was always placed in the tibial tunnel and was left protruding approximately 5 mm. Each screw was placed posterior to the graft, to obtain maximum cortical compression at the distal aperture, and was inserted to the point where it was just flush with the cortex. This placement allowed for cortical and cancellous fixation of the graft specimen. Maximum insertion torque was measured with a custom-made torque-measuring screwdriver (Arthrex), which was read at the last turn of the screw. The specimen was then mounted on a servohydraulic testing machine (Instron model 1331 with an 850 controller; Instron Corp., Canton, Massachusetts) with a custom-made jig with six degrees of freedom to hold the tibia securely and allow pull in line with the tunnel. A tendon grasper was used to grasp the tendon tightly on the proximal end of the graft. A 0.45-mm wire was placed through Figure 2. Comparison of the two screw types used in this study. The top screw is 28 mm in length, and the bottom screw is 35 mm in length.

3 616 Selby et al. American Journal of Sports Medicine the graft several times and wrapped around it to allow the grasper to obtain a tight hold. This method has been shown in our laboratory to be effective enough to hold more than 1000 N of force and, theoretically, does not alter the chemical properties of the graft. The graft was then preloaded to 25 N of force and cycled from 0 to 50 N of force, after which it was subjected to 20 mm/min of traction force parallel to the axis of the graft (Fig. 3). StatView (SAS Institute, Inc., Cary, North Carolina) software program was used to formulate the loading curve, and the results of the maximum load at failure, the maximum displacement at failure, and the stiffness of the graft construct were recorded. Analysis of variance of all recorded parameters was performed with StatView software. Significance was set at P Simple regression analysis was used to correlate bone mineral density with insertion torque and pull-out strength. RESULTS Table 1 lists the ages of the cadavers from which the bone specimens were taken, the tunnel size, and the results of bone mineral density scans for each paired set of knees. It is notable that the bone mineral density in these specimens was approximately 0.20 to 0.30 g/cm 3 greater than that previously reported by our laboratory for older cadaveric bones. 3,7,8 When compared with the density measurements of other studies, however, the measurements of the present study are very close to those of bones from TABLE 1 Age of Cadaveric Bone Specimen, Tunnel Size, and Bone Mineral Density for All Specimens a Age (years) Tunnel size (mm) Bone-mineral density (g/cm 3 ) 28-mm screws 35-mm screws Mean a Screw diameter was 1 mm larger than tunnel size for each screw unless the tunnel was a half size, in which case the screw was 1.5 mm larger than tunnel size. older cadavers and lower than those from young cadavers (Table 2). Our method for bone mineral density scanning was consistent with these different studies. The mean bone mineral density for the bones in which 28-mm screws were used was g/cm 3, and that for 35-mm screws was g/cm 3. There were no significant differences between tunnel size and bone mineral density between the two groups. The mechanism of failure in all of the specimens except one was slippage of the graft past the screw in the tunnel. The graft with the highest pull-out strength failed at the proximal grip. It was observed that any laxity of the construct was removed with the initial cycling, and there were no failures at this low load. No reflective markers were placed on the bone or grafts. Thus, the exact mechanism of displacement could not be determined, but observation during the pull-out test revealed a combination of graft deformation and slippage past the screw at ultimate failure. Examination of the screws at the end of the tests, with the grafts removed, revealed that they were all partially in the tunnel and partially out of the tunnel and had not moved from their initial position. The grafts were stretched but not torn, and all sutures remained intact. The data from the pull-out tests are presented in Table 3. Mean ultimate pull-out strength in the 35-mm screw group was N and in the 28-mm screw group TABLE 2 Bone Mineral Density of Proximal Tibias From Various Studies Figure 3. Sample preparation of the tested construct loaded in the Instron testing machine. Study Tibial bone mineral density (g/cm 3 ) Age of specimens (years) Range (Average) Present (38.5) Brand et al., 3 Caborn et al (69.4) Stadelmaier (62.6) Vouri et al a Madsen et al (69) McKeon et al McKeon et al McKeon et al McKeon et al McKeon et al a Female specimens only.

4 Vol. 29, No. 5, 2001 Length of Bioabsorbable Screws for Interference Fixation 617 Screw length TABLE 3 Results of Load-to-Failure Pull-Out Testing Comparing 28-mm versus 35-mm Length Screws Specimen Insertion torque (N m) Displacement (mm) Stiffness (N/mm) Pull-out strength (N) 28 mm 1L L R L R L L L Mean mm 1R R L R L R R R Mean it was N. These results were statistically significantly different at P (Fig. 4). The displacement at failure was mm in the 35-mm screw group and mm in the 28-mm screw group, and these results were statistically significantly different at P (Fig. 5). The insertion torque was N m for the 35-mm screws and N m inthe 28-mm screw group, which was also statistically significantly different at P Other measurements that were not significantly different between screws were graft-construct stiffness, bone age, tunnel size, and bone mineral density. There was no significant difference in the pull-out strengths between the allograft and autograft tendon constructs. DISCUSSION The initial fixation strength required for ACL grafts in bone tunnels has been widely studied, but it remains a debated topic with regard to soft tissue grafts because tibial-tunnel fixation has not been shown to be satisfactory in the laboratory. 7,14,16 Rodeo and colleagues 21 showed in a dog model that grafts evaluated at 2, 4, and 8 weeks all failed at the bone-tunnel interface, whereas Figure 4. Box plot comparing the mean pull-out strength at failure of the long (35-mm) screws and the short (28-mm) screws. The points are the maximum and minimum values for the group, and the box represents 75% of the results. Mean difference, 230 N; critical difference, N; P Figure 5. Box plot comparing the mean displacement at failure of the long (35-mm) screws and the short (28-mm) screws. The points are the maximum and minimum values for the group, and the box represents 75% of the results. Mean difference, 3.4 mm; critical difference, 2.34 mm; P

5 618 Selby et al. American Journal of Sports Medicine those evaluated at 12 and 24 weeks failed primarily at the midsubstance. According to at least one study, an arbitrary initial fixation strength needed to withstand the force of rehabilitation has been estimated to be around 450 N. 20 The fixation strength of the interference fit in the femoral tunnel is satisfactory to withstand the 450-N force needed for rehabilitation; however, in the tibial tunnel, it has been less than satisfactory, according to the results of a previous study. 7 We have shown that the fixation strength of an interference screw fixation in the tibial tunnel can be increased to over 800 N by using a longer interference screw. Many of the previous studies of graft fixation have used bones from older cadavers that have low bone mineral density. 6 8,25,27 It has been hypothesized that increased bone mineral density leads to increased pull-out strength. 7 To simulate the clinical situation more closely, bones from young cadavers were used in this study. The bone mineral density of the tibial metaphyseal bone, where the tunnel traverses, was tested and calculated. The bone mineral density results for the bones used in our study were similar to those reported in studies using older cadaveric bones, but the pull-out strengths were much greater in our study (Table 2). Our bone mineral density results, however, were much higher than those published by our laboratory in other studies using older cadavers. It is likely that the methods and machines used to measure bone mineral density are not standardized, and thus the results cannot be compared directly with one another. The tunnels in this study were prepared by dilating the tunnel diameter 2 mm beyond that drilled in an attempt to increase the strength of the cancellous bone surrounding the tunnel. In a recent study, tunnel dilatation was shown to increase pull-out strength in the tibial tunnel by 63% (B. B. Phillips et al., unpublished data, 1999). Dilatation is a simple technique that is minimally time-consuming and may be worth the extra time and instrumentation if strength of fixation is increased. Weiler et al. 31 have shown in a histologic study that tunnel dilatation may delay bone healing. More controlled studies are necessary to determine the exact histologic effects of tunnel dilatation on bone healing. One study of the use of differing screw lengths showed no statistically significant difference between a 25-mm and 40-mm screw after cycling the grafts to the point of slippage and then pulling them to their ultimate pull-out strength. 25 This study used older cadaveric bones with a mean age of 62.6 years, and the screws were placed anterior to the graft and in the subchondral bone. The graft slippage began at 238 N, and the ultimate pull-out strength was 336 N in both groups. Use of the two techniques, tunnel dilatation and increased screw length, have not been studied together, and, in the present study, we optimized fixation with posterior screw placement at the distal end of the tunnel. The pitch of the screw used in this study was smaller than that of the RCI screw (Smith & Nephew DonJoy, Carlsbad, California) or other similar screws, so there were more threads in contact with the graft. Also, the taper of the screw allows for easier insertion, and the screw can act as a wedge to prevent graft pullout. Increasing the length of the screw allows for cortical contact of the screw distally while providing for interference over a larger distance, thus increasing stiffness of the graft construct. Little emphasis is placed on preparation of the graft in the technique of soft tissue graft ACL reconstruction. This is another step in the process in which tunnel fixation can be maximized and grafts can be made stiffer. It is important to have tension on the graft during preparation to decrease elongation once the graft is implanted. 29 The sutures must be placed through all arms of the graft and pulled tightly as they are placed. When suturing the other side of the graft, it is important to interdigitate the sutures so that a path will be made for the screw, decreasing the chance of damaging the graft with the threads. The graft is sized after completion and before tunnel drilling so that sizing to the nearest 0.5 mm can maximize the pullout strength (J. C. Brand et al., unpublished data, 1999). The technique used for mechanical pullout in this study was the same as that used in other studies in our laboratory so that the results could be compared with each other. All grafts were pulled in line with the graft to simulate a worst-case scenario of pullout. The grafts were pulled at 20 mm/min to maximize the stress on the fixation site instead of within the graft. A soft tissue grip was used proximally and was seen to achieve good fixation. There was a significant amount of graft displacement before failure, and a significant difference between the two groups. The exact mechanism for the larger displacement with the long screw could have been determined with reflective markers measured during displacement, but this was not done in this study. It did appear by observation only, however, that displacement in the 35-mm screw group was larger because of graft elongation. The goal of the present study was to determine the fixation strength of the graft-tendon interface, and more study is needed to determine if there is intrinsic damage to the tissue during the displacement. Graft tensioning on the back table may help to decrease this mode of displacement. Recent studies have focused on cyclic loading of soft tissue grafts in an attempt to more closely mimic the clinical situation of rehabilitation (Refs. 11, 25, 28; M. T. Havig et al., unpublished data, 1999). Studies of cyclic loading of grafts would theoretically detect a lower load needed to pull the graft out of the tunnel because repeated stress would weaken the construct over time. It has been shown in an in vivo goat model that maximum forces on the ACL during trotting are 150 N and in an in vivo human model during rehabilitation that maximum forces during walking are only 84 N (Ref. 12; K. Shino et al, unpublished data). Both bone-patellar tendon-bone and semitendinosus-gracilis tendon grafts have been shown to slip with cyclic loads of less than 300 N, which is higher than the loads reported in the in vivo studies (Refs. 24, 25; M. T. Havig et al., unpublished data, 1999). The grafts are also much stiffer after this cycling and the pull-out strengths are greater. It is difficult to tell if this type of testing mimics the clinical situation, because, after these large cyclic loads, the graft has already elongated 5 to 10 mm, which mimics a grade II or III ACL tear. Both bone-

6 Vol. 29, No. 5, 2001 Length of Bioabsorbable Screws for Interference Fixation 619 patellar tendon-bone and hamstring tendon grafts have been shown to elongate to 10 mm with loads as low as 350 N. 7,22,27 From in vitro experiments it is known that active extension against gravity induces forces in the graft between 200 and 250 N. A cyclic load in this range may be a realistic test. Optimization of graft suturing, graft-to-tunnel size matching, screw oversizing, tunnel dilatation, and bone mineral density are all important factors affecting pullout strength. We have shown that increasing the length of the screw while keeping these other variables constant increases the ultimate failure load by 38%. ACKNOWLEDGMENT Funding for this study was received from Arthrex, Naples, Florida. REFERENCES 1. Aglietti P, Buzzi R, Zaccherotti G, et al: Patellar tendon versus doubled semitendinosus and gracilis tendons for anterior cruciate ligament reconstruction. Am J Sports Med 22: , Bonatus TJ, Alexander AH: Patellar fracture and avulsion of the patellar ligament complicating arthroscopic anterior cruciate ligament reconstruction. Orthop Rev 20: , Brand J Jr, Hamilton D, Selby J, et al: Biomechanical comparison of quadriceps tendon fixation with patellar tendon bone interference fixation in cruciate ligament reconstruction. Arthroscopy 16: , Breitfuss H, Fröhlich R, Povacz P, et al: The tendon defect after anterior cruciate ligament reconstruction using the midthird patellar tendon a problem for the patellofemoral joint? Knee Surg Sports Traumatol Arthrosc 3: , Brown CH Jr, Hecker AT, Hipp JA, et al: The biomechanics of interference screw fixation of patellar tendon anterior cruciate ligament grafts. Am J Sports Med 21: , Brown GA, Peña F, Grøntvedt T, et al: Fixation strength of interference screw fixation in bovine, young human, and elderly human cadaver knees: Influence of insertion torque, tunnel-bone block gap, and interference. Knee Surg Sports Traumatol Arthrosc 3: , Caborn DNM, Coen M, Neef R, et al: Quadrupled semitendinosus-gracilis autograft fixation in the femoral tunnel: A comparison between a metal and a bioabsorbable interference screw. Arthroscopy 14: , Caborn DNM, Urban WP Jr, Johnson DL, et al: Biomechanical comparison between BioScrew and titanium alloy interference screws for bonepatellar tendon-bone graft fixation in anterior cruciate ligament reconstruction. Arthroscopy 13: , Carter TR, Edinger S: Isokinetic evaluation of anterior cruciate ligament reconstruction: Hamstring versus patellar tendon. Arthroscopy 15: , Corry IS, Webb JM, Clingeleffer AJ, et al: Arthroscopic reconstruction of the anterior cruciate ligament: A comparison of patellar tendon autograft and four-strand hamstring tendon autograft. Am J Sports Med 27: , Hamner DL, Brown CH Jr, Steiner ME, et al: Hamstring tendon grafts for reconstruction of the anterior cruciate ligament: Biomechanical evaluation of the use of multiple strands and tensioning techniques. J Bone Joint Surg 81A: , Holden JP, Grood ES, Korvick DL, et al: In vivo forces in the anterior cruciate ligament: Direct measurements during walking and trotting in a quadruped. J Biomech 27: , Jomha NM, Pinczewski LA, Clingeleffer A, et al: Arthroscopic reconstruction of the anterior cruciate ligament with patellar-tendon autograft and interference screw fixation. The results at seven years. J Bone Joint Surg 81B: , Kohn D, Rose C: Primary stability of interference screw fixation. Influence of screw diameter and insertion torque. Am J Sports Med 22: , Kohn D, Sander-Beuermann A: Donor-site morbidity after harvest of a bone-tendon-bone patellar tendon autograft. Knee Surg Sports Traumatol Arthrosc 2: , Kurosaka M, Yoshiya S, Andrish JT: A biomechanical comparison of different surgical techniques of graft fixation in anterior cruciate ligament reconstruction. Am J Sports Med 15: , Madsen KL, Adams WC, Van Loan MD: Effects of physical activity, body weight and composition, and muscular strength on bone density in young women. Med Sci Sports Exerc 30: , Marder RA, Raskind JR, Carroll M: Prospective evaluation of arthroscopically assisted anterior cruciate ligament reconstruction: Patellar tendon versus semitendinosus and gracilis tendons. Am J Sports Med 19: , McKeon EW, Wang J, Pierson RN Jr, et al: Dual photon absorptiometry in obesity: Effects of massive weight loss. Basic Life Sci 55: , Noyes FR, Butler DL, Grood ES, et al: Biomechanical analysis of human ligament grafts used in knee-ligament repairs and reconstructions. J Bone Joint Surg 66A: , Rodeo SA, Arnoczky SP, Torzilli PA, et al: Tendon-healing in a bone tunnel. A biomechanical and histological study in the dog. J Bone Joint Surg 75A: , Rowden NJ, Sher D, Rogers GJ, et al: Anterior cruciate ligament graft fixation. Initial comparison of patellar tendon and semitendinosus autografts in young fresh cadavers. Am J Sports Med 25: , Saddemi SR, Frogameni AD, Fenton PJ, et al: Comparison of perioperative morbidity of anterior cruciate ligament autografts versus allografts. Arthroscopy 9: , Simonian PT, Williams RJ, Deng XH, et al: Hamstring and patellar tendon graft response to cyclical loading. Am J Knee Surg 11: , Stadelmaier DM, Lowe WR, Ilahi OA, et al: Cyclic pull-out strength of hamstring tendon graft fixation with soft tissue interference screws: Influence of screw length. Am J Sports Med 27: , Steen H, Tseng KF, Goldstein SA, et al: Harvest of patellar tendon (bone-tendon-bone) autograft for ACL reconstruction significantly alters surface strain in the human patella. J Biomech Eng 121: , Steiner ME, Hecker AT, Brown CH Jr, et al: Anterior cruciate ligament graft fixation: Comparison of hamstring and patellar tendon grafts. Am J Sports Med 22: , Straight CB, France EP, Paulos LE, et al: Soft tissue fixation to bone: A biomechanical analysis of spiked washers. Am J Sports Med 22: , Tohyama H, Yasuda K: Significance of graft tension in anterior cruciate ligament reconstruction. Basic background and clinical outcome. Knee Surg Sports Traumatol Arthrosc 6 (Suppl 1): S30 S37, Vuori I, Heinonen A, Sievanen H, et al: Effects of unilateral strength training and detraining on bone mineral density and content in young women: A study of mechanical loading and deloading on human bones. Calcif Tissue Int 55(1): 59 67, Weiler A, Windhagen HJ, Raschke MJ, et al: Biodegradable interference screw fixation exhibits pull-out force and stiffness similar to titanium screws. Am J Sports Med 26: , Yasuda K, Tsujino J, Ohkoshi Y, et al: Graft site morbidity with autogenous semitendinosus and gracilis tendons. Am J Sports Med 23: , 1995