- Email: [email protected]
Magnetic Resonance Imaging Evaluation of Physeal Violation in Adolescents After Transphyseal Anterior Cruciate Ligament Reconstruction
Magnetic Resonance Imaging Evaluation of Physeal Violation in Adolescents After Transphyseal Anterior Cruciate Ligament Reconstruction Joon-Ho Wang, M.D., Kang-Min Son, M.D., and Dae-Hee Lee, M.D.
Purpose: To quantify and compare the amount and location of physis violation of the distal femur and proximal tibia after transphyseal anterior cruciate ligament (ACL) reconstruction in skeletally immature patients. Methods: This study included 19 patients with open physes of the distal femur and proximal tibia who underwent ACL reconstruction with tibialis anterior allografts. Physeal tunnel volume and location on the growth plate, as well as obliquity to the growth plate, were measured by 3-dimensional postoperative magnetic resonance imaging of the distal femur and proximal tibia. Results: The percentage of physeal violation (ratio of the tunnel to the entire growth plate area) was similar for the distal femur and proximal tibia (3.95% vs 3.65%, P ¼ .582). There were no differences in tunnel obliquity to the growth plate in the coronal (56.1 vs 71.6 , P ¼ .061) and sagittal (85.9 vs 74.9 , P ¼ .092) planes. The distal femoral tunnel was located 6.2% (17.2% vs 23.4%, P ¼ .001) more peripherally in the anteroposterior direction and 9.7% (27.1% vs 36.8%, P < .001) more peripherally in the mediolateral direction than was the tibial tunnel. Conclusions: The mean percentages of physeal violation of tunnel creation during ACL reconstruction in adolescent patients were 3.95% for the distal femur and 3.65% for the proximal tibia. Moreover, femoral tunnels were located more peripherally on the growth plate than were tibial tunnels, in both the anteroposterior and mediolateral directions. Level of Evidence: Level IV, case series.
T
he incidence of anterior cruciate ligament (ACL) tears in skeletally immature adolescents has been increasing because of the increased intensity of adolescent sports activities and increased detection of these injuries resulting from the development of advanced diagnostic technologies, such as magnetic resonance imaging (MRI) and arthroscopy.1,2 Concerns about physeal injury during ACL reconstruction in skeletally immature patients, however, have resulted in conservative treatments, with braces or activity modification for ACL tears.3-5 Because nonsurgical treatments of ACL tears in skeletally immature patients have resulted in poor outcomes,6-8 physeal-sparing ACL
From the Department of Orthopaedic Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea. The authors report the following potential conflicts of interest or sources of funding: This study was supported by SMC-Ottogi Research Fund (SMX1162171). Received June 3, 2016; accepted December 5, 2016. Address correspondence to Dae-Hee Lee, M.D., Department of Orthopaedic Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Ilwon-ro, Gangnam-gu, Seoul 135-710, Republic of Korea. E-mail: [email protected] Ó 2016 by the Arthroscopy Association of North America 0749-8063/16509/$36.00 http://dx.doi.org/10.1016/j.arthro.2016.12.011
reconstruction techniques have been devised to overcome these poor outcomes, as well as concerns about physeal violations.9 However, these methods are technically demanding and do not entirely exclude the possibility of physeal injury, even in all-epiphyseal ACL reconstruction, which can accurately restore intraarticular anatomy.10 Transphyseal ACL reconstruction, which has been performed in adults, may therefore be a more practical surgical option, even in skeletally immature patients, although there are still concerns about limb-length discrepancy and angular deformity resulting from physis injury.11 Deformities after ACL reconstruction, including valgus or recurvatum, may be associated with physis injury, caused by reaming of the femoral or tibial tunnels during transphyseal ACL reconstruction.3,12-14 The development of these deformities may be correlated to the size and/or location of physis violations.15,16 Animal and simulation studies have suggested that the risk of limb-length inequality would be higher for physis violations of greater than 5% to 7% of the entire growth plate of the distal femur.17-21 The risk of angular deformity is also higher with more peripheral physis violations than with more central physis violations. To date, however, few in vivo studies have quantified the magnitude of injury to the growth plate or physis or
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol
-,
No
-
(Month), 2017: pp 1-8
1
2
J-H. WANG ET AL.
have compared the magnitude and location of physis violation to the distal femur and proximal tibia after transphyseal ACL reconstruction in skeletally immature patients. Furthermore, in vivo evaluation of physeal violation on postoperative MRI may yield results differing from those of imaging analysis for simulated surgery. This study was therefore designed to quantify and compare the amount and location of physis violation of the distal femur and proximal tibia after transphyseal ACL reconstruction in skeletally immature patients. It was hypothesized that physis violation would be greater and placed more peripherally in the distal femur than in the proximal tibia.
Methods This study identified potential subjects who had undergone ACL reconstruction during adolescence. The inclusion criteria were as follows: Patients must have undergone ACL reconstruction with the anteromedial (AM) portal technique and tibialis anterior allograft, they must have been aged less than 15 years with open physes of the distal femur and proximal tibia, they must have undergone immediate postoperative MRI, and they must have been followed up for more than 2 years. The exclusion criteria were patients who did not undergo immediate postoperative MRI or who had a follow-up duration of less than 2 years. After application of these criteria, this study enrolled 21 patients between March 2011 and November 2014. Two of these patients were excluded because one was followed up for less than 2 years and the other was not evaluated by postoperative MRI. The records of 19 patients were therefore retrospectively evaluated. All patients underwent postoperative MRI on postoperative day 1 or 2 after removal of the Hemovac drain. The diagnosis of an ACL tear was confirmed by MRI and physical examinations, including the positive anterior drawer, Lachman, and/or pivot-shift test (more than grade II). All procedures were performed by a single surgeon (JH.W.). The study was approved by the ethics committee of our institution.
to the lateral intercondylar ridge, or “resident ridge.”22 The center of the PL bundle footprint was positioned 5 mm anterior (higher) to the edge of the low cartilage on an imaginary line perpendicular to the tangent at the lowermost portion of the lateral femoral condyle at 90 of knee flexion (Fig 1). The femoral tunnel was created by reaming along the inserted guide pin with the knee in flexion greater than 100 . The single tibial tunnel was created at a point between the center of the AM and PL tibial footprints. The anatomic center of the tibial footprint was marked with a thermal device (Smith & Nephew Endoscopy, Andover, MA) and a curved Steadman awl (ConMed Linvatec, Largo, FL), and the tip of the guide was aimed at a point between the center of the AM and PL tibial footprints. A tibialis anterior allograft was folded and sutured together by use of No. 1 absorbable sutures to form a triple-strand graft. These grafts were fixed with an EndoButton (Smith & Nephew Endoscopy) on the femoral side and with a post tie with a cortical screw and washer on the tibial side.
Surgical Technique Standard ports were created, with an accessory AM port created 1.5 cm medial to the standard AM port just above the medial meniscus anterior horn. The diameters of the femoral and tibial tunnels were 8 mm each. The single femoral tunnel was created at a point between the center of the AM and posterolateral (PL) femoral footprints, which were assumed to be the centers of the anatomic ACL femoral footprint. On arthroscopy at 90 of flexion from the AM portal, the center of the AM bundle footprint was 6 to 7 mm shallower to the deep cartilage margin or 2 mm from the deep bony ridge of the lateral femoral condyle, called the “fellow ridge” because it was located posterior
MRI was performed at full knee extension with the Intera Achieva 3-T MRI system (Philips Healthcare, Andover, MA). In the coronal plane, MRI images were obtained by a 3-dimensional (3D) fast spin echo technique (echo time, 35 milliseconds; repetition time, 1,800 milliseconds; echo train length, 46; bandwidth, 101.5 kHz over entire frequency range; acquisition matrix, 320 318; number of excitations, 1; field to view, 160 mm; slice thickness, 0.5 mm with no gap; scan time, 6 minutes 49 seconds). DICOM (Digital Imaging and Communications in Medicine) data were exported with the Centricity RA-1000 system (GE
Fig 1. Location of center of femoral tunnel (red circle) between anteromedial (green circle) and posterolateral (blue circle) footprint centers, as well as neighboring anatomic structures.
MRI Protocol
PHYSEAL VIOLATION IN ADOLESCENTS
3
Fig 2. Measurement of surface areas and volumes of femoral and tibial growth plates (right knee). (A) Initial region of low signal intensity, including the growth plate, cortical bone, meniscus, knee around the tendons and ligaments, and external air in the environment, selected by setting the range of the threshold values. (B) Use of the “Edit Mask in 3D” function of Mimics to specify the rectangular area near the growth plate, separating the growth plate from other areas in the coronal and sagittal planes. (C) Use of manual editing on the growth plate to separate cortical bone areas bordering the growth plate. This process was performed on all slides.
Medical Systems, Milwaukee, WI), a type of picture archiving and communication system, and were used for measurement. Image Analysis Volumes of Growth Plate and Physeal Tunnel. DICOM files obtained from 3-T MRI images were imported with Mimics software (version 18.0; Materialise, Leuven, Belgium), and 3D reconstructions of the growth plate in the femur and tibia were performed by use of the “Thresholding,” “Edit Mask in 3D,” “Edit Mask,” and “Calculate 3D” functions in Mimics. An initial region was selected by setting the range of threshold values from 115 to 241 because of interindividual variations in patients (Fig 2A). The initial region included the growth plate, cortical bone, meniscus, tendons, ligaments, and external air. To separate the growth plate from other areas, a rectangular area, consisting of only the area near the growth plate, was specified by use of the “Edit Mask in 3D” function of Mimics in the coronal and sagittal planes (Fig 2B). The tunnel volume and the cortical bone bordering the growth plate were removed by manual editing, leaving only the growth plate (Fig 2C). Three-dimensional reconstructed models of the growth plate with and without the physeal tunnel were generated by the “Calculate 3D” function (Fig 3). Volumes of the 3D reconstructed models were measured by use of Mimics software, version 18.0. The physeal volume was calculated by subtracting the volume of the growth plate with the physeal tunnel from the volume of the entire growth plate without the physeal tunnel. Tunnel Location. Tunnel locations in the anteroposterior (AP) and mediolateral (ML) directions were analyzed by calculating the peripheral tunnel index (as a percentage) on the axial growth plates of the distal
femur and proximal tibia. The entire width of each growth plate and the distance between the center of the tunnel and the peripheral boundary of the growth plate in the AP and ML directions were measured on a 3D reconstruction model. The AP and ML peripheral tunnel indices were determined as the ratio between the distance from the center of the tunnel to the growth plate boundary and the entire width of the growth plate in the AP and ML directions, respectively (Fig 4). Tunnel Obliquity to Growth Plate. DICOM files were imported into OsiriX (version 7.0; Pixmeo, Geneva, Switzerland), an open-source software program. In the femur, the coronal plane, which showed the entire femoral tunnel, and the sagittal plane, which showed the length of growth plate violation by the femoral tunnel, were selected (Fig 5 A and B). In the tibia, the coronal and sagittal planes, which showed the entire tibial tunnel, were selected (Fig 5 C and D). In each plane, the angle between the long axis of the tunnel and the growth plate for tunnel obliquity was determined. All parameters, including growth plate and tunnel volumes, as well as locations and obliquity of tunnels, were independently evaluated by a single orthopaedic surgeon (D-H.L.) and a single research fellow (K-M.S.) with substantial experience in the field. Each evaluator twice measured all parameters in all 19 patients, with an interval of 2 weeks between measurements. Radiographic Assessment To evaluate angular deformity and leg-length discrepancy (LLD), long (130 cm) standing AP hip-toankle radiographs were taken preoperatively and at last follow-up. Leg length was measured from the center of the femoral head to the center of the distal tibial plafond. Differences between the injured and uninjured legs were measured preoperatively and postoperatively and compared to determine the net
4
J-H. WANG ET AL.
Fig 3. Generation of 3-dimensional (3D) reconstructed models of femoral (A) and tibial (B) growth plates with Mimics software (right knee). The 3D reconstructed model of the complete growth plate was obtained using the “Calculate 3D” function. After removal of the tunnel area by manual editing, the 3D reconstructed model of the growth plate was generated using the same function. Total surface area and volume were measured using the “3D Properties” function.
LLD. Angular deformity in the coronal plane was evaluated by measuring the angle between the mechanical axis of the femur and the mechanical axis of the tibia both preoperatively and postoperatively and by calculating the net change in angle. Knees were evaluated for recurvatum deformities on lateral radiographs. All parameters on radiographs and MRI scans were measured by 2 independent orthopaedic surgeons (D-H.L., K-M.S.) and averaged. Growth disturbance was defined as an LLD greater than 1 cm and angular deformity greater than 3 .12 Statistical Analysis At an a level of .05 and a power of 80%, a post hoc power analysis was performed to determine whether
the sample size of 19 patients was adequate to detect a mean difference in peripheral tunnel indices between the femur and tibia. Calculations showed that a sample of 19 patients had a power of 80.2% to detect a mean difference in the peripheral tunnel index in the AP direction of 6.2% and a power of 87.3% to detect a mean difference in the peripheral tunnel index in the ML direction of 9.7%. The sample size of 19 patients in this study showed a power of 78% to detect a mean difference of 15 in coronal tunnel obliquity and a power of 77% to detect a mean difference of 6 in sagittal tunnel obliquity. The volumes of the growth plate and tunnel, as well as the locations and obliquity of the femoral and tibial tunnels, were compared by paired Student t tests. LLD
Fig 4. Measurement of peripheral tunnel index in femoral and tibial growth plates (left knee). (A) On the femoral side, the peripheral tunnel index in the mediolateral direction was calculated as the ratio of the distance (b) between the center of the tunnel and the lateral boundary of the lateral femoral condyle, divided by the entire width (a) of the growth plate. In the anteroposterior direction, the peripheral tunnel index was calculated as the ratio of the distance (d) between the center of the tunnel and the posterior boundary of the lateral femoral condyle, divided by the entire width (c) of the growth plate. (B) On the tibial side, the peripheral tunnel index in the mediolateral direction was calculated as the ratio of the distance (f) between the center of the tunnel and the medial boundary of the growth plate, divided by the entire width (e) of the growth plate. In the anteroposterior direction, the peripheral tunnel index was calculated the ratio of the distance (h) between the center of the tunnel and the anterior boundary of the growth plate, divided by the entire width (g) of the growth plate.
5
PHYSEAL VIOLATION IN ADOLESCENTS
Fig 5. Measurements of tunnel obliquity in femur and tibia (right knee). After the DICOM (Digital Imaging and Communications in Medicine) file was imported, the best views for evaluation of tunnel obliquity were determined using the “3D Multiplanar Reconstruction (3D MPR)” function in OsiriX software. (A, B) In the femur, the coronal plane, which showed the entire femoral tunnel, and the sagittal plane, which had the longest violation length of the growth plate by the femoral tunnel, were selected. (C, D) In the tibia, the coronal and sagittal planes, which showed the entire tibial tunnel, were selected. In each plane, the angle between the long axis of the tunnel and the growth plate for the tunnel obliquity was measured.
and angular knee deformity from before to after surgery were also compared by paired Student t tests. The correlations between tunnel volume on the growth plate or percentage of physeal violation and tunnel angle to the growth plate both in the coronal plane and in the sagittal plane were estimated by Pearson correlation analyses. The reliabilities of parameter measurements were assessed by the intraclass correlation coefficient. Intraclass correlation coefficient values greater than 0.75 represent fair to good agreement, whereas values lower than 0.40 represent poor agreement. All analyses were performed with SPSS software (version 13.0; SPSS, Chicago, IL), with P < .05 considered statistically significant.
Results The study enrolled 21 patients, 2 of whom were excluded because of a follow-up period of less than 2 years and no postoperative MRI, respectively. Hence 19 patients were included in the final analysis. The mean age of the 19 subjects was 13.7 years (range, 12 to 15 years). Preoperative demographic data are presented in Table 1. The interobserver and intraobserver reliabilities for all parameters, including volumes of growth plate and tunnel physeal violation and tunnel locations and obliquities, were satisfactory, with a range of 0.787 to 0.866. Comparisons Between Distal Femur and Proximal Tibia The volume of the growth plate was significantly higher in the distal femur than in the proximal tibia (4,608 mm3 vs 3,366 mm3, P ¼ .009), but physeal tunnel volume did not differ significantly (151 mm3 vs 119 mm3, P ¼ .106). The ratio of the physeal tunnel to the entire growth plate volume, termed the “percentage
of physeal violation,” was also similar for the distal femur and proximal tibia (3.95% vs 3.65%, P ¼ .582), with both percentages being less than 4%. The tunnels of the distal femur had a mean 15 more acute angle to the growth plate in the coronal plane (56.1 vs 71.6 , P ¼ .061) and were 11 more vertical to the growth plate in the sagittal plane (85.9 vs 74.9 , P ¼ .092) than were the tunnels of the proximal tibia, but these differences were not statistically significant. The femoral tunnels in the AP and ML directions were located 6.2% and 9.7%, respectively, more peripherally on the growth plate than were the tibial tunnels. Table 2 summarizes all details of these measured parameters. Correlation Studies The volume of the distal femoral tunnel did not correlate with the femoral tunnel angles in the coronal and sagittal planes. The percentage of physeal violation by the femoral tunnel on the growth plate of the distal femur was found to correlate negatively with the femoral tunnel angle to the growth plate in the coronal plane (r ¼ 0.504, P ¼ .046) but not in the sagittal plane. At the proximal tibia, neither the tunnel volume nor the physeal violation of the tibial tunnel on the Table 1. Patient Demographic Characteristics and Tanner Stage Parameter Patients, n Age, yr Height, cm Weight, kg Body mass index Tanner stage, n
Data 13 male and 6 female 13.7 (12-15) 165 (151-178) 65.3 (55-88) 24.2 (19.7-35.1) 11 III and 8 IV
NOTE. Data are presented as median (range) unless otherwise indicated.
6
J-H. WANG ET AL.
Table 2. Areas of Physis and Tunnels, as well as Obliquity and Peripheral Tunnel Indices of Tunnels, of Distal Femur and Proximal Tibia Volume of growth plate, mm3 Volume of physeal tunnel, mm3 % of physeal violation Coronal tunnel obliquity, Sagittal tunnel obliquity, Anteroposterior peripheral tunnel index, % Mediolateral peripheral tunnel index, %
Distal Femur 4,608.3 2,280.2 150.7 81.5 3.96 2.26 56.1 20.3 85.9 20.8 17.2 4.4 27.1 4.3
Proximal Tibia 3,366.6 1,083.9 118.8 57.8 3.65 1.96 71.6 17.3 74.9 11.4 23.4 3.7 36.8 4.1
P Value .009* .106 .582 .061 .092 .001* <.001*
NOTE. Data reported as mean standard deviation. *P < .05.
growth plate correlated with the tibial tunnel angle in the coronal plane. However, both the physeal tunnel volume (r ¼ 0.570, P ¼ .021) and physeal violation of the tibial tunnel (r ¼ 0.710, P ¼ .002) were negatively correlated with the tibial tunnel angle in the sagittal plane (Table 3). LLD and Angular Deformity At a mean follow-up of 25.8 months (range, 24 to 34 months), the mean net LLD was 1 mm shorter (range, 6 to 5 mm) when the affected and unaffected limbs were compared (Table 1). The mean preoperative LLD was 0.4 mm (range, 5 to 8 mm), and the mean postoperative LLD at last follow-up was 1.4 mm (range, 4 to 8 mm). None of the 19 patients had a change in leg length greater than 8 mm and none showed an angular deformity on standing hip-to-ankle radiographs at last follow-up. The mean net angular deformity in the coronal plane was 0.5 varus (range, 1.3 valgus to 2.5 varus). The cutoff values of growth disturbance, which were 1 cm of limb-length discrepancy and 3 of angular deformity,12 were not exceeded in any patients, and no patients showed recurvatum deformity on lateral radiographs.
Discussion The main findings of this study were that the mean volume of physeal violation during transphyseal ACL reconstruction was less than 4% in both the distal femur and proximal tibia and that femoral tunnels were located more peripherally on the growth plate than were tibial tunnels in both the AP and ML directions. Because the volume of physeal violation constitutes the main risk factor of growth plate disturbance, it is important to determine the threshold for physeal arrest. In an animal study by Makela et al.,19 who conducted transphyseal ACL reconstruction in 44 rabbits with open physes, it was found that destruction of 7% of the cross-sectional area of the growth plate led to growth disturbance and shortening of the femur. Guzzanti et al.17 similarly performed transphyseal ACL reconstruction in 21 skeletally immature rabbits. They showed that the mean physeal violation percentages
were 3% of the femoral and 4% of the tibial crosssectional growth plate. Two rabbits had valgus deformities, and one showed shortening of the tibia. These conflicting results in animals could not be directly applied to humans because the physeal growth rate is much slower in humans than in animal models,15 reducing the risk of growth disturbances in humans. Three-dimensional virtual models have been used to determine the “threshold of injury,” defined as the maximal amount of physeal violation that does not produce growth arrest. Simulated ACL reconstruction using 3D computer models created from MRI scans of 10 pediatric knees found that the mean percentages of removed physeal volume for the whole cross-sectional area of the growth plate were 2.94% for the distal femur and 4.20% for the proximal tibia when an 8-mmdiameter reamer was used.20 Simulation of ACL drill tunnels crossing the physes with an 8-mm graft resulted in a 2.4% injury to the distal femoral physis and a 2.5% injury to the proximal tibia.18 However, the simulation surgery environment was not equivalent to an arthroscopic environment, in which reconstruction techniques, such as transtibial or AM portal reconstruction, are normally performed. In addition, these simulation studies could not determine whether growth disturbances actually occur after transphyseal ACL reconstruction in patients with open physes. Our study reported physeal violations based on a computer model created from in vivo MRI scans of pediatric patients Table 3. Correlations Between Tunnel Obliquities to Growth Plate in Coronal and Sagittal Plane, Tunnel Volumes, and Percentages of Physeal Violation Coronal Angle Femoral tunnel Tunnel volume % of physeal violation Tibial tunnel Tunnel volume % of physeal violation
Sagittal Angle
r
P Value
r
P Value
0.180 0.504
.505 .046*
0.379 0.361
.148 .169
0.054 0.021
.842 .938
0.570 0.710
.021* .002*
r, Pearson correlation coefficient. *P < .05.
7
PHYSEAL VIOLATION IN ADOLESCENTS
who underwent actual ACL reconstruction with the AM portal technique and an 8-mm reamer. We found that physeal violation percentages in the distal femur (3.95%) and proximal tibial (3.65%) were less than 4% with no growth disturbance, suggesting that this small magnitude of physeal violation may be the main reason for lack of growth disturbance, including LLD and angular deformity, after ACL reconstruction in the skeletally immature patients assessed in this study. However, it is difficult to directly compare the results of different studies, owing mainly to the different ages of pediatric patients and the differences in surgical environments between simulation and in vivo arthroscopic surgery. Subjects in the aforementioned simulation studies18,20 ranged in age from 5 to 10 years, whereas subjects in our study ranged in age from 12 to 15 years. Tunnel location in the growth plate is also predictive of the magnitude of physeal violation with subsequent growth disturbance, especially in patients with angular deformity. Physeal growth disturbance is more frequent after peripheral physeal injury than after central physeal injury.21 Our results showed that femoral tunnels were located more peripherally on axial growth plates than were tibial tunnels. In the AP direction of axial growth plates, femoral tunnels were located 6% more peripherally from the center of growth plates than tibial tunnels. In the ML direction, femoral tunnels were positioned approximately 10% more peripherally from the center of growth plates than were tibial tunnels. These situations may be attributable to the more peripheral location of the ACL femoral footprint compared with the relatively central position of the tibial footprint and may indicate that an angular deformity, such as valgus or procurvatum (flexion deformity), of the distal femur is more likely to develop than varus or recurvatum of the proximal tibia. These findings suggest that the femoral tunnel should be higher in the low-to-high direction and shallower in the deep-to-shallow direction,11 from the point of view of the arthroscopist,23,24 in the flexed knee. Tunnel obliquity to the growth plate should also be considered during ACL reconstruction in immature patients. Tunnels more perpendicular to the physes have less potential for injury because the apertures of these tunnels have a smaller cross section on growth plates. As the angle of the drill hole deviates from the perpendicular to the physis, the volume of physis damage increases. A previous simulation study showed that increasing the tunnel drill angle from 45 to 70 reduced the volume percentage removed from 4.1% to 3.1%, with a mean 0.2% decrease in physeal volume damaged for every 5 increase in tunnel angle.17,21 Our study also showed a correlation between the percentage of physeal violation of the distal femur and tunnel obliquity in the coronal plane, as well as between the percentage of physeal violation of the proximal tibia
and tunnel obliquity in the sagittal plane. These findings suggest that a more vertical tunnel obliquity to the physis of the distal femur and proximal tibia was associated with a reduction in physis violation of the distal femur and proximal tibia. Limitations This study had several limitations. The computer model used to calculate the percentage of physeal violation was a composite of MRI files. MRI scanning loses a small amount of precision because of signal averaging between slices.25 In addition, the methods required to extract desired information from the MRI images are both difficult and less objective. Another limitation of this study was that all subjects underwent ACL reconstruction with the AM portal technique. Therefore, these results may not be directly applicable to pediatric patients undergoing ACL reconstruction with the transtibial or outside-in technique,26,27 although a recent simulation study showed that the percentage of physeal violation was similar using the transtibial and AM portal techniques.11 Directly comparing the magnitude of physeal violations between studies is difficult because the ages of patients vary widely among studies or are skewed. Finally, the statistically nonsignificant association between tunnel obliquity and physeal tunnel volume observed in this study may have been a type II error resulting from the small sample size.
Conclusions The mean percentages of physeal violation observed by creating tunnels during ACL reconstruction in adolescent patients were 3.95% in the distal femur and 3.65% in the proximal tibia, suggesting that physeal violations of less than 4% were not associated with growth disturbances. In addition, femoral tunnels were located more peripherally on the growth plate than were tibial tunnels, in both the AP and ML directions.
References 1. Davis DL, Chen L, Ehinger M. A study of epiphyses in the young prepubescent knee using magnetic resonance imaging: Evaluation of parameters for anterior cruciate ligament reconstruction. Orthop J Sports Med 2014;2: 2325967114530090. 2. Domzalski M, Karauda A, Grzegorzewski A, Lebiedzinski R, Zabierek S, Synder M. Anterior cruciate ligament reconstruction using the transphyseal technique in prepubescent athletes: Midterm, prospective evaluation of results. Arthroscopy 2016;32:1141-1146. 3. Fabricant PD, Jones KJ, Delos D, et al. Reconstruction of the anterior cruciate ligament in the skeletally immature athlete: A review of current concepts: AAOS exhibit selection. J Bone Joint Surg Am 2013;95:e28. 4. Higuchi T, Hara K, Tsuji Y, Kubo T. Transepiphyseal reconstruction of the anterior cruciate ligament in
8
5. 6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
J-H. WANG ET AL. skeletally immature athletes: An MRI evaluation for epiphyseal narrowing. J Pediatr Orthop B 2009;18: 330-334. Jung YB. Recent evolution of cruciate ligament surgery of the knee. Clin Orthop Surg 2012;4:103-106. Behr CT, Potter HG, Paletta GA Jr. The relationship of the femoral origin of the anterior cruciate ligament and the distal femoral physeal plate in the skeletally immature knee. An anatomic study. Am J Sports Med 2001;29:781-787. Aichroth PM, Patel DV, Zorrilla P. The natural history and treatment of rupture of the anterior cruciate ligament in children and adolescents. A prospective review. J Bone Joint Surg Br 2002;84:38-41. Ramski DE, Kanj WW, Franklin CC, Baldwin KD, Ganley TJ. Anterior cruciate ligament tears in children and adolescents: A meta-analysis of nonoperative versus operative treatment. Am J Sports Med 2014;42:2769-2776. Xerogeanes JW, Hammond KE, Todd DC. Anatomic landmarks utilized for physeal-sparing, anatomic anterior cruciate ligament reconstruction: An MRI-based study. J Bone Joint Surg Am 2012;94:268-276. Nawabi DH, Jones KJ, Lurie B, Potter HG, Green DW, Cordasco FA. All-inside, physeal-sparing anterior cruciate ligament reconstruction does not significantly compromise the physis in skeletally immature athletes: A postoperative physeal magnetic resonance imaging analysis. Am J Sports Med 2014;42:2933-2940. Kachmar M, Piazza SJ, Bader DA. Comparison of growth plate violations for transtibial and anteromedial surgical techniques in simulated adolescent anterior cruciate ligament reconstruction. Am J Sports Med 2016;44:417-424. Collins MJ, Arns TA, Leroux T, et al. Growth abnormalities following anterior cruciate ligament reconstruction in the skeletally immature patient: A systematic review. Arthroscopy 2016;32:1714-1723. Frosch KH, Stengel D, Brodhun T, et al. Outcomes and risks of operative treatment of rupture of the anterior cruciate ligament in children and adolescents. Arthroscopy 2010;26:1539-1550. Vavken P, Murray MM. Treating anterior cruciate ligament tears in skeletally immature patients. Arthroscopy 2011;27:704-716. Chudik S, Beasley L, Potter H, Wickiewicz T, Warren R, Rodeo S. The influence of femoral technique for graft placement on anterior cruciate ligament reconstruction using a skeletally immature canine model with a rapidly growing physis. Arthroscopy 2007;23:1309-1319.e1. Shifflett GD, Green DW, Widmann RF, Marx RG. Growth arrest following ACL reconstruction with hamstring autograft in skeletally immature patients: A review of 4 cases. J Pediatr Orthop 2016;36:355-361.
17. Guzzanti V, Falciglia F, Gigante A, Fabbriciani C. The effect of intra-articular ACL reconstruction on the growth plates of rabbits. J Bone Joint Surg Br 1994;76:960-963. 18. Kercher J, Xerogeanes J, Tannenbaum A, Al-Hakim R, Black JC, Zhao J. Anterior cruciate ligament reconstruction in the skeletally immature: An anatomical study utilizing 3-dimensional magnetic resonance imaging reconstructions. J Pediatr Orthop 2009;29:124-129. 19. Makela EA, Vainionpaa S, Vihtonen K, Mero M, Rokkanen P. The effect of trauma to the lower femoral epiphyseal plate. An experimental study in rabbits. J Bone Joint Surg Br 1988;70:187-191. 20. Shea KG, Belzer J, Apel PJ, Nilsson K, Grimm NL, Pfeiffer RP. Volumetric injury of the physis during singlebundle anterior cruciate ligament reconstruction in children: A 3-dimensional study using magnetic resonance imaging. Arthroscopy 2009;25:1415-1422. 21. Shea KG, Grimm NL, Belzer JS. Volumetric injury of the distal femoral physis during double-bundle ACL reconstruction in children: A three-dimensional study with use of magnetic resonance imaging. J Bone Joint Surg Am 2011;93:1033-1038. 22. Wang JH, Kim JG, Ahn JH, Lim HC, Hoshino Y, Fu FH. Is femoral tunnel length correlated with the intercondylar notch and femoral condyle geometry after double-bundle anterior cruciate ligament reconstruction using the transportal technique? An in vivo computed tomography analysis. Arthroscopy 2012;28:1094-1103. 23. Lee DH, Kim HJ, Ahn HS, Bin SI. Comparison of femur tunnel aperture location in patients undergoing transtibial and anatomical single-bundle anterior cruciate ligament reconstruction. Knee Surg Sports Traumatol Arthrosc 2016;24:3713-3721. 24. Shin YS, Ro KH, Lee JH, Lee DH. Location of the femoral tunnel aperture in single-bundle anterior cruciate ligament reconstruction: Comparison of the transtibial, anteromedial portal, and outside-in techniques. Am J Sports Med 2013;41:2533-2539. 25. Shea KG, Apel PJ, Pfeiffer RP, Traughber PD. The anatomy of the proximal tibia in pediatric and adolescent patients: Implications for ACL reconstruction and prevention of physeal arrest. Knee Surg Sports Traumatol Arthrosc 2007;15:320-327. 26. Nha KW, Han JH, Kwon JH, Kang KW, Park HJ, Song JG. Anatomical single-bundle anterior cruciate ligament reconstruction using a freehand transtibial technique. Knee Surg Relat Res 2015;27:117-122. 27. Sohn OJ, Lee DC, Park KH, Ahn HS. Comparison of the modified transtibial technique, anteromedial portal technique and outside-in technique in ACL reconstruction. Knee Surg Relat Res 2014;26:241-248.
Purpose: To quantify and compare the amount and location of physis violation of the distal femur and proximal tibia after transphyseal anterior cruciate ligament (ACL) reconstruction in skeletally immature patients. Methods: This study included 19 patients with open physes of the distal femur and proximal tibia who underwent ACL reconstruction with tibialis anterior allografts. Physeal tunnel volume and location on the growth plate, as well as obliquity to the growth plate, were measured by 3-dimensional postoperative magnetic resonance imaging of the distal femur and proximal tibia. Results: The percentage of physeal violation (ratio of the tunnel to the entire growth plate area) was similar for the distal femur and proximal tibia (3.95% vs 3.65%, P ¼ .582). There were no differences in tunnel obliquity to the growth plate in the coronal (56.1 vs 71.6 , P ¼ .061) and sagittal (85.9 vs 74.9 , P ¼ .092) planes. The distal femoral tunnel was located 6.2% (17.2% vs 23.4%, P ¼ .001) more peripherally in the anteroposterior direction and 9.7% (27.1% vs 36.8%, P < .001) more peripherally in the mediolateral direction than was the tibial tunnel. Conclusions: The mean percentages of physeal violation of tunnel creation during ACL reconstruction in adolescent patients were 3.95% for the distal femur and 3.65% for the proximal tibia. Moreover, femoral tunnels were located more peripherally on the growth plate than were tibial tunnels, in both the anteroposterior and mediolateral directions. Level of Evidence: Level IV, case series.
T
he incidence of anterior cruciate ligament (ACL) tears in skeletally immature adolescents has been increasing because of the increased intensity of adolescent sports activities and increased detection of these injuries resulting from the development of advanced diagnostic technologies, such as magnetic resonance imaging (MRI) and arthroscopy.1,2 Concerns about physeal injury during ACL reconstruction in skeletally immature patients, however, have resulted in conservative treatments, with braces or activity modification for ACL tears.3-5 Because nonsurgical treatments of ACL tears in skeletally immature patients have resulted in poor outcomes,6-8 physeal-sparing ACL
From the Department of Orthopaedic Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea. The authors report the following potential conflicts of interest or sources of funding: This study was supported by SMC-Ottogi Research Fund (SMX1162171). Received June 3, 2016; accepted December 5, 2016. Address correspondence to Dae-Hee Lee, M.D., Department of Orthopaedic Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Ilwon-ro, Gangnam-gu, Seoul 135-710, Republic of Korea. E-mail: [email protected] Ó 2016 by the Arthroscopy Association of North America 0749-8063/16509/$36.00 http://dx.doi.org/10.1016/j.arthro.2016.12.011
reconstruction techniques have been devised to overcome these poor outcomes, as well as concerns about physeal violations.9 However, these methods are technically demanding and do not entirely exclude the possibility of physeal injury, even in all-epiphyseal ACL reconstruction, which can accurately restore intraarticular anatomy.10 Transphyseal ACL reconstruction, which has been performed in adults, may therefore be a more practical surgical option, even in skeletally immature patients, although there are still concerns about limb-length discrepancy and angular deformity resulting from physis injury.11 Deformities after ACL reconstruction, including valgus or recurvatum, may be associated with physis injury, caused by reaming of the femoral or tibial tunnels during transphyseal ACL reconstruction.3,12-14 The development of these deformities may be correlated to the size and/or location of physis violations.15,16 Animal and simulation studies have suggested that the risk of limb-length inequality would be higher for physis violations of greater than 5% to 7% of the entire growth plate of the distal femur.17-21 The risk of angular deformity is also higher with more peripheral physis violations than with more central physis violations. To date, however, few in vivo studies have quantified the magnitude of injury to the growth plate or physis or
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol
-,
No
-
(Month), 2017: pp 1-8
1
2
J-H. WANG ET AL.
have compared the magnitude and location of physis violation to the distal femur and proximal tibia after transphyseal ACL reconstruction in skeletally immature patients. Furthermore, in vivo evaluation of physeal violation on postoperative MRI may yield results differing from those of imaging analysis for simulated surgery. This study was therefore designed to quantify and compare the amount and location of physis violation of the distal femur and proximal tibia after transphyseal ACL reconstruction in skeletally immature patients. It was hypothesized that physis violation would be greater and placed more peripherally in the distal femur than in the proximal tibia.
Methods This study identified potential subjects who had undergone ACL reconstruction during adolescence. The inclusion criteria were as follows: Patients must have undergone ACL reconstruction with the anteromedial (AM) portal technique and tibialis anterior allograft, they must have been aged less than 15 years with open physes of the distal femur and proximal tibia, they must have undergone immediate postoperative MRI, and they must have been followed up for more than 2 years. The exclusion criteria were patients who did not undergo immediate postoperative MRI or who had a follow-up duration of less than 2 years. After application of these criteria, this study enrolled 21 patients between March 2011 and November 2014. Two of these patients were excluded because one was followed up for less than 2 years and the other was not evaluated by postoperative MRI. The records of 19 patients were therefore retrospectively evaluated. All patients underwent postoperative MRI on postoperative day 1 or 2 after removal of the Hemovac drain. The diagnosis of an ACL tear was confirmed by MRI and physical examinations, including the positive anterior drawer, Lachman, and/or pivot-shift test (more than grade II). All procedures were performed by a single surgeon (JH.W.). The study was approved by the ethics committee of our institution.
to the lateral intercondylar ridge, or “resident ridge.”22 The center of the PL bundle footprint was positioned 5 mm anterior (higher) to the edge of the low cartilage on an imaginary line perpendicular to the tangent at the lowermost portion of the lateral femoral condyle at 90 of knee flexion (Fig 1). The femoral tunnel was created by reaming along the inserted guide pin with the knee in flexion greater than 100 . The single tibial tunnel was created at a point between the center of the AM and PL tibial footprints. The anatomic center of the tibial footprint was marked with a thermal device (Smith & Nephew Endoscopy, Andover, MA) and a curved Steadman awl (ConMed Linvatec, Largo, FL), and the tip of the guide was aimed at a point between the center of the AM and PL tibial footprints. A tibialis anterior allograft was folded and sutured together by use of No. 1 absorbable sutures to form a triple-strand graft. These grafts were fixed with an EndoButton (Smith & Nephew Endoscopy) on the femoral side and with a post tie with a cortical screw and washer on the tibial side.
Surgical Technique Standard ports were created, with an accessory AM port created 1.5 cm medial to the standard AM port just above the medial meniscus anterior horn. The diameters of the femoral and tibial tunnels were 8 mm each. The single femoral tunnel was created at a point between the center of the AM and posterolateral (PL) femoral footprints, which were assumed to be the centers of the anatomic ACL femoral footprint. On arthroscopy at 90 of flexion from the AM portal, the center of the AM bundle footprint was 6 to 7 mm shallower to the deep cartilage margin or 2 mm from the deep bony ridge of the lateral femoral condyle, called the “fellow ridge” because it was located posterior
MRI was performed at full knee extension with the Intera Achieva 3-T MRI system (Philips Healthcare, Andover, MA). In the coronal plane, MRI images were obtained by a 3-dimensional (3D) fast spin echo technique (echo time, 35 milliseconds; repetition time, 1,800 milliseconds; echo train length, 46; bandwidth, 101.5 kHz over entire frequency range; acquisition matrix, 320 318; number of excitations, 1; field to view, 160 mm; slice thickness, 0.5 mm with no gap; scan time, 6 minutes 49 seconds). DICOM (Digital Imaging and Communications in Medicine) data were exported with the Centricity RA-1000 system (GE
Fig 1. Location of center of femoral tunnel (red circle) between anteromedial (green circle) and posterolateral (blue circle) footprint centers, as well as neighboring anatomic structures.
MRI Protocol
PHYSEAL VIOLATION IN ADOLESCENTS
3
Fig 2. Measurement of surface areas and volumes of femoral and tibial growth plates (right knee). (A) Initial region of low signal intensity, including the growth plate, cortical bone, meniscus, knee around the tendons and ligaments, and external air in the environment, selected by setting the range of the threshold values. (B) Use of the “Edit Mask in 3D” function of Mimics to specify the rectangular area near the growth plate, separating the growth plate from other areas in the coronal and sagittal planes. (C) Use of manual editing on the growth plate to separate cortical bone areas bordering the growth plate. This process was performed on all slides.
Medical Systems, Milwaukee, WI), a type of picture archiving and communication system, and were used for measurement. Image Analysis Volumes of Growth Plate and Physeal Tunnel. DICOM files obtained from 3-T MRI images were imported with Mimics software (version 18.0; Materialise, Leuven, Belgium), and 3D reconstructions of the growth plate in the femur and tibia were performed by use of the “Thresholding,” “Edit Mask in 3D,” “Edit Mask,” and “Calculate 3D” functions in Mimics. An initial region was selected by setting the range of threshold values from 115 to 241 because of interindividual variations in patients (Fig 2A). The initial region included the growth plate, cortical bone, meniscus, tendons, ligaments, and external air. To separate the growth plate from other areas, a rectangular area, consisting of only the area near the growth plate, was specified by use of the “Edit Mask in 3D” function of Mimics in the coronal and sagittal planes (Fig 2B). The tunnel volume and the cortical bone bordering the growth plate were removed by manual editing, leaving only the growth plate (Fig 2C). Three-dimensional reconstructed models of the growth plate with and without the physeal tunnel were generated by the “Calculate 3D” function (Fig 3). Volumes of the 3D reconstructed models were measured by use of Mimics software, version 18.0. The physeal volume was calculated by subtracting the volume of the growth plate with the physeal tunnel from the volume of the entire growth plate without the physeal tunnel. Tunnel Location. Tunnel locations in the anteroposterior (AP) and mediolateral (ML) directions were analyzed by calculating the peripheral tunnel index (as a percentage) on the axial growth plates of the distal
femur and proximal tibia. The entire width of each growth plate and the distance between the center of the tunnel and the peripheral boundary of the growth plate in the AP and ML directions were measured on a 3D reconstruction model. The AP and ML peripheral tunnel indices were determined as the ratio between the distance from the center of the tunnel to the growth plate boundary and the entire width of the growth plate in the AP and ML directions, respectively (Fig 4). Tunnel Obliquity to Growth Plate. DICOM files were imported into OsiriX (version 7.0; Pixmeo, Geneva, Switzerland), an open-source software program. In the femur, the coronal plane, which showed the entire femoral tunnel, and the sagittal plane, which showed the length of growth plate violation by the femoral tunnel, were selected (Fig 5 A and B). In the tibia, the coronal and sagittal planes, which showed the entire tibial tunnel, were selected (Fig 5 C and D). In each plane, the angle between the long axis of the tunnel and the growth plate for tunnel obliquity was determined. All parameters, including growth plate and tunnel volumes, as well as locations and obliquity of tunnels, were independently evaluated by a single orthopaedic surgeon (D-H.L.) and a single research fellow (K-M.S.) with substantial experience in the field. Each evaluator twice measured all parameters in all 19 patients, with an interval of 2 weeks between measurements. Radiographic Assessment To evaluate angular deformity and leg-length discrepancy (LLD), long (130 cm) standing AP hip-toankle radiographs were taken preoperatively and at last follow-up. Leg length was measured from the center of the femoral head to the center of the distal tibial plafond. Differences between the injured and uninjured legs were measured preoperatively and postoperatively and compared to determine the net
4
J-H. WANG ET AL.
Fig 3. Generation of 3-dimensional (3D) reconstructed models of femoral (A) and tibial (B) growth plates with Mimics software (right knee). The 3D reconstructed model of the complete growth plate was obtained using the “Calculate 3D” function. After removal of the tunnel area by manual editing, the 3D reconstructed model of the growth plate was generated using the same function. Total surface area and volume were measured using the “3D Properties” function.
LLD. Angular deformity in the coronal plane was evaluated by measuring the angle between the mechanical axis of the femur and the mechanical axis of the tibia both preoperatively and postoperatively and by calculating the net change in angle. Knees were evaluated for recurvatum deformities on lateral radiographs. All parameters on radiographs and MRI scans were measured by 2 independent orthopaedic surgeons (D-H.L., K-M.S.) and averaged. Growth disturbance was defined as an LLD greater than 1 cm and angular deformity greater than 3 .12 Statistical Analysis At an a level of .05 and a power of 80%, a post hoc power analysis was performed to determine whether
the sample size of 19 patients was adequate to detect a mean difference in peripheral tunnel indices between the femur and tibia. Calculations showed that a sample of 19 patients had a power of 80.2% to detect a mean difference in the peripheral tunnel index in the AP direction of 6.2% and a power of 87.3% to detect a mean difference in the peripheral tunnel index in the ML direction of 9.7%. The sample size of 19 patients in this study showed a power of 78% to detect a mean difference of 15 in coronal tunnel obliquity and a power of 77% to detect a mean difference of 6 in sagittal tunnel obliquity. The volumes of the growth plate and tunnel, as well as the locations and obliquity of the femoral and tibial tunnels, were compared by paired Student t tests. LLD
Fig 4. Measurement of peripheral tunnel index in femoral and tibial growth plates (left knee). (A) On the femoral side, the peripheral tunnel index in the mediolateral direction was calculated as the ratio of the distance (b) between the center of the tunnel and the lateral boundary of the lateral femoral condyle, divided by the entire width (a) of the growth plate. In the anteroposterior direction, the peripheral tunnel index was calculated as the ratio of the distance (d) between the center of the tunnel and the posterior boundary of the lateral femoral condyle, divided by the entire width (c) of the growth plate. (B) On the tibial side, the peripheral tunnel index in the mediolateral direction was calculated as the ratio of the distance (f) between the center of the tunnel and the medial boundary of the growth plate, divided by the entire width (e) of the growth plate. In the anteroposterior direction, the peripheral tunnel index was calculated the ratio of the distance (h) between the center of the tunnel and the anterior boundary of the growth plate, divided by the entire width (g) of the growth plate.
5
PHYSEAL VIOLATION IN ADOLESCENTS
Fig 5. Measurements of tunnel obliquity in femur and tibia (right knee). After the DICOM (Digital Imaging and Communications in Medicine) file was imported, the best views for evaluation of tunnel obliquity were determined using the “3D Multiplanar Reconstruction (3D MPR)” function in OsiriX software. (A, B) In the femur, the coronal plane, which showed the entire femoral tunnel, and the sagittal plane, which had the longest violation length of the growth plate by the femoral tunnel, were selected. (C, D) In the tibia, the coronal and sagittal planes, which showed the entire tibial tunnel, were selected. In each plane, the angle between the long axis of the tunnel and the growth plate for the tunnel obliquity was measured.
and angular knee deformity from before to after surgery were also compared by paired Student t tests. The correlations between tunnel volume on the growth plate or percentage of physeal violation and tunnel angle to the growth plate both in the coronal plane and in the sagittal plane were estimated by Pearson correlation analyses. The reliabilities of parameter measurements were assessed by the intraclass correlation coefficient. Intraclass correlation coefficient values greater than 0.75 represent fair to good agreement, whereas values lower than 0.40 represent poor agreement. All analyses were performed with SPSS software (version 13.0; SPSS, Chicago, IL), with P < .05 considered statistically significant.
Results The study enrolled 21 patients, 2 of whom were excluded because of a follow-up period of less than 2 years and no postoperative MRI, respectively. Hence 19 patients were included in the final analysis. The mean age of the 19 subjects was 13.7 years (range, 12 to 15 years). Preoperative demographic data are presented in Table 1. The interobserver and intraobserver reliabilities for all parameters, including volumes of growth plate and tunnel physeal violation and tunnel locations and obliquities, were satisfactory, with a range of 0.787 to 0.866. Comparisons Between Distal Femur and Proximal Tibia The volume of the growth plate was significantly higher in the distal femur than in the proximal tibia (4,608 mm3 vs 3,366 mm3, P ¼ .009), but physeal tunnel volume did not differ significantly (151 mm3 vs 119 mm3, P ¼ .106). The ratio of the physeal tunnel to the entire growth plate volume, termed the “percentage
of physeal violation,” was also similar for the distal femur and proximal tibia (3.95% vs 3.65%, P ¼ .582), with both percentages being less than 4%. The tunnels of the distal femur had a mean 15 more acute angle to the growth plate in the coronal plane (56.1 vs 71.6 , P ¼ .061) and were 11 more vertical to the growth plate in the sagittal plane (85.9 vs 74.9 , P ¼ .092) than were the tunnels of the proximal tibia, but these differences were not statistically significant. The femoral tunnels in the AP and ML directions were located 6.2% and 9.7%, respectively, more peripherally on the growth plate than were the tibial tunnels. Table 2 summarizes all details of these measured parameters. Correlation Studies The volume of the distal femoral tunnel did not correlate with the femoral tunnel angles in the coronal and sagittal planes. The percentage of physeal violation by the femoral tunnel on the growth plate of the distal femur was found to correlate negatively with the femoral tunnel angle to the growth plate in the coronal plane (r ¼ 0.504, P ¼ .046) but not in the sagittal plane. At the proximal tibia, neither the tunnel volume nor the physeal violation of the tibial tunnel on the Table 1. Patient Demographic Characteristics and Tanner Stage Parameter Patients, n Age, yr Height, cm Weight, kg Body mass index Tanner stage, n
Data 13 male and 6 female 13.7 (12-15) 165 (151-178) 65.3 (55-88) 24.2 (19.7-35.1) 11 III and 8 IV
NOTE. Data are presented as median (range) unless otherwise indicated.
6
J-H. WANG ET AL.
Table 2. Areas of Physis and Tunnels, as well as Obliquity and Peripheral Tunnel Indices of Tunnels, of Distal Femur and Proximal Tibia Volume of growth plate, mm3 Volume of physeal tunnel, mm3 % of physeal violation Coronal tunnel obliquity, Sagittal tunnel obliquity, Anteroposterior peripheral tunnel index, % Mediolateral peripheral tunnel index, %
Distal Femur 4,608.3 2,280.2 150.7 81.5 3.96 2.26 56.1 20.3 85.9 20.8 17.2 4.4 27.1 4.3
Proximal Tibia 3,366.6 1,083.9 118.8 57.8 3.65 1.96 71.6 17.3 74.9 11.4 23.4 3.7 36.8 4.1
P Value .009* .106 .582 .061 .092 .001* <.001*
NOTE. Data reported as mean standard deviation. *P < .05.
growth plate correlated with the tibial tunnel angle in the coronal plane. However, both the physeal tunnel volume (r ¼ 0.570, P ¼ .021) and physeal violation of the tibial tunnel (r ¼ 0.710, P ¼ .002) were negatively correlated with the tibial tunnel angle in the sagittal plane (Table 3). LLD and Angular Deformity At a mean follow-up of 25.8 months (range, 24 to 34 months), the mean net LLD was 1 mm shorter (range, 6 to 5 mm) when the affected and unaffected limbs were compared (Table 1). The mean preoperative LLD was 0.4 mm (range, 5 to 8 mm), and the mean postoperative LLD at last follow-up was 1.4 mm (range, 4 to 8 mm). None of the 19 patients had a change in leg length greater than 8 mm and none showed an angular deformity on standing hip-to-ankle radiographs at last follow-up. The mean net angular deformity in the coronal plane was 0.5 varus (range, 1.3 valgus to 2.5 varus). The cutoff values of growth disturbance, which were 1 cm of limb-length discrepancy and 3 of angular deformity,12 were not exceeded in any patients, and no patients showed recurvatum deformity on lateral radiographs.
Discussion The main findings of this study were that the mean volume of physeal violation during transphyseal ACL reconstruction was less than 4% in both the distal femur and proximal tibia and that femoral tunnels were located more peripherally on the growth plate than were tibial tunnels in both the AP and ML directions. Because the volume of physeal violation constitutes the main risk factor of growth plate disturbance, it is important to determine the threshold for physeal arrest. In an animal study by Makela et al.,19 who conducted transphyseal ACL reconstruction in 44 rabbits with open physes, it was found that destruction of 7% of the cross-sectional area of the growth plate led to growth disturbance and shortening of the femur. Guzzanti et al.17 similarly performed transphyseal ACL reconstruction in 21 skeletally immature rabbits. They showed that the mean physeal violation percentages
were 3% of the femoral and 4% of the tibial crosssectional growth plate. Two rabbits had valgus deformities, and one showed shortening of the tibia. These conflicting results in animals could not be directly applied to humans because the physeal growth rate is much slower in humans than in animal models,15 reducing the risk of growth disturbances in humans. Three-dimensional virtual models have been used to determine the “threshold of injury,” defined as the maximal amount of physeal violation that does not produce growth arrest. Simulated ACL reconstruction using 3D computer models created from MRI scans of 10 pediatric knees found that the mean percentages of removed physeal volume for the whole cross-sectional area of the growth plate were 2.94% for the distal femur and 4.20% for the proximal tibia when an 8-mmdiameter reamer was used.20 Simulation of ACL drill tunnels crossing the physes with an 8-mm graft resulted in a 2.4% injury to the distal femoral physis and a 2.5% injury to the proximal tibia.18 However, the simulation surgery environment was not equivalent to an arthroscopic environment, in which reconstruction techniques, such as transtibial or AM portal reconstruction, are normally performed. In addition, these simulation studies could not determine whether growth disturbances actually occur after transphyseal ACL reconstruction in patients with open physes. Our study reported physeal violations based on a computer model created from in vivo MRI scans of pediatric patients Table 3. Correlations Between Tunnel Obliquities to Growth Plate in Coronal and Sagittal Plane, Tunnel Volumes, and Percentages of Physeal Violation Coronal Angle Femoral tunnel Tunnel volume % of physeal violation Tibial tunnel Tunnel volume % of physeal violation
Sagittal Angle
r
P Value
r
P Value
0.180 0.504
.505 .046*
0.379 0.361
.148 .169
0.054 0.021
.842 .938
0.570 0.710
.021* .002*
r, Pearson correlation coefficient. *P < .05.
7
PHYSEAL VIOLATION IN ADOLESCENTS
who underwent actual ACL reconstruction with the AM portal technique and an 8-mm reamer. We found that physeal violation percentages in the distal femur (3.95%) and proximal tibial (3.65%) were less than 4% with no growth disturbance, suggesting that this small magnitude of physeal violation may be the main reason for lack of growth disturbance, including LLD and angular deformity, after ACL reconstruction in the skeletally immature patients assessed in this study. However, it is difficult to directly compare the results of different studies, owing mainly to the different ages of pediatric patients and the differences in surgical environments between simulation and in vivo arthroscopic surgery. Subjects in the aforementioned simulation studies18,20 ranged in age from 5 to 10 years, whereas subjects in our study ranged in age from 12 to 15 years. Tunnel location in the growth plate is also predictive of the magnitude of physeal violation with subsequent growth disturbance, especially in patients with angular deformity. Physeal growth disturbance is more frequent after peripheral physeal injury than after central physeal injury.21 Our results showed that femoral tunnels were located more peripherally on axial growth plates than were tibial tunnels. In the AP direction of axial growth plates, femoral tunnels were located 6% more peripherally from the center of growth plates than tibial tunnels. In the ML direction, femoral tunnels were positioned approximately 10% more peripherally from the center of growth plates than were tibial tunnels. These situations may be attributable to the more peripheral location of the ACL femoral footprint compared with the relatively central position of the tibial footprint and may indicate that an angular deformity, such as valgus or procurvatum (flexion deformity), of the distal femur is more likely to develop than varus or recurvatum of the proximal tibia. These findings suggest that the femoral tunnel should be higher in the low-to-high direction and shallower in the deep-to-shallow direction,11 from the point of view of the arthroscopist,23,24 in the flexed knee. Tunnel obliquity to the growth plate should also be considered during ACL reconstruction in immature patients. Tunnels more perpendicular to the physes have less potential for injury because the apertures of these tunnels have a smaller cross section on growth plates. As the angle of the drill hole deviates from the perpendicular to the physis, the volume of physis damage increases. A previous simulation study showed that increasing the tunnel drill angle from 45 to 70 reduced the volume percentage removed from 4.1% to 3.1%, with a mean 0.2% decrease in physeal volume damaged for every 5 increase in tunnel angle.17,21 Our study also showed a correlation between the percentage of physeal violation of the distal femur and tunnel obliquity in the coronal plane, as well as between the percentage of physeal violation of the proximal tibia
and tunnel obliquity in the sagittal plane. These findings suggest that a more vertical tunnel obliquity to the physis of the distal femur and proximal tibia was associated with a reduction in physis violation of the distal femur and proximal tibia. Limitations This study had several limitations. The computer model used to calculate the percentage of physeal violation was a composite of MRI files. MRI scanning loses a small amount of precision because of signal averaging between slices.25 In addition, the methods required to extract desired information from the MRI images are both difficult and less objective. Another limitation of this study was that all subjects underwent ACL reconstruction with the AM portal technique. Therefore, these results may not be directly applicable to pediatric patients undergoing ACL reconstruction with the transtibial or outside-in technique,26,27 although a recent simulation study showed that the percentage of physeal violation was similar using the transtibial and AM portal techniques.11 Directly comparing the magnitude of physeal violations between studies is difficult because the ages of patients vary widely among studies or are skewed. Finally, the statistically nonsignificant association between tunnel obliquity and physeal tunnel volume observed in this study may have been a type II error resulting from the small sample size.
Conclusions The mean percentages of physeal violation observed by creating tunnels during ACL reconstruction in adolescent patients were 3.95% in the distal femur and 3.65% in the proximal tibia, suggesting that physeal violations of less than 4% were not associated with growth disturbances. In addition, femoral tunnels were located more peripherally on the growth plate than were tibial tunnels, in both the AP and ML directions.
References 1. Davis DL, Chen L, Ehinger M. A study of epiphyses in the young prepubescent knee using magnetic resonance imaging: Evaluation of parameters for anterior cruciate ligament reconstruction. Orthop J Sports Med 2014;2: 2325967114530090. 2. Domzalski M, Karauda A, Grzegorzewski A, Lebiedzinski R, Zabierek S, Synder M. Anterior cruciate ligament reconstruction using the transphyseal technique in prepubescent athletes: Midterm, prospective evaluation of results. Arthroscopy 2016;32:1141-1146. 3. Fabricant PD, Jones KJ, Delos D, et al. Reconstruction of the anterior cruciate ligament in the skeletally immature athlete: A review of current concepts: AAOS exhibit selection. J Bone Joint Surg Am 2013;95:e28. 4. Higuchi T, Hara K, Tsuji Y, Kubo T. Transepiphyseal reconstruction of the anterior cruciate ligament in
8
5. 6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
J-H. WANG ET AL. skeletally immature athletes: An MRI evaluation for epiphyseal narrowing. J Pediatr Orthop B 2009;18: 330-334. Jung YB. Recent evolution of cruciate ligament surgery of the knee. Clin Orthop Surg 2012;4:103-106. Behr CT, Potter HG, Paletta GA Jr. The relationship of the femoral origin of the anterior cruciate ligament and the distal femoral physeal plate in the skeletally immature knee. An anatomic study. Am J Sports Med 2001;29:781-787. Aichroth PM, Patel DV, Zorrilla P. The natural history and treatment of rupture of the anterior cruciate ligament in children and adolescents. A prospective review. J Bone Joint Surg Br 2002;84:38-41. Ramski DE, Kanj WW, Franklin CC, Baldwin KD, Ganley TJ. Anterior cruciate ligament tears in children and adolescents: A meta-analysis of nonoperative versus operative treatment. Am J Sports Med 2014;42:2769-2776. Xerogeanes JW, Hammond KE, Todd DC. Anatomic landmarks utilized for physeal-sparing, anatomic anterior cruciate ligament reconstruction: An MRI-based study. J Bone Joint Surg Am 2012;94:268-276. Nawabi DH, Jones KJ, Lurie B, Potter HG, Green DW, Cordasco FA. All-inside, physeal-sparing anterior cruciate ligament reconstruction does not significantly compromise the physis in skeletally immature athletes: A postoperative physeal magnetic resonance imaging analysis. Am J Sports Med 2014;42:2933-2940. Kachmar M, Piazza SJ, Bader DA. Comparison of growth plate violations for transtibial and anteromedial surgical techniques in simulated adolescent anterior cruciate ligament reconstruction. Am J Sports Med 2016;44:417-424. Collins MJ, Arns TA, Leroux T, et al. Growth abnormalities following anterior cruciate ligament reconstruction in the skeletally immature patient: A systematic review. Arthroscopy 2016;32:1714-1723. Frosch KH, Stengel D, Brodhun T, et al. Outcomes and risks of operative treatment of rupture of the anterior cruciate ligament in children and adolescents. Arthroscopy 2010;26:1539-1550. Vavken P, Murray MM. Treating anterior cruciate ligament tears in skeletally immature patients. Arthroscopy 2011;27:704-716. Chudik S, Beasley L, Potter H, Wickiewicz T, Warren R, Rodeo S. The influence of femoral technique for graft placement on anterior cruciate ligament reconstruction using a skeletally immature canine model with a rapidly growing physis. Arthroscopy 2007;23:1309-1319.e1. Shifflett GD, Green DW, Widmann RF, Marx RG. Growth arrest following ACL reconstruction with hamstring autograft in skeletally immature patients: A review of 4 cases. J Pediatr Orthop 2016;36:355-361.
17. Guzzanti V, Falciglia F, Gigante A, Fabbriciani C. The effect of intra-articular ACL reconstruction on the growth plates of rabbits. J Bone Joint Surg Br 1994;76:960-963. 18. Kercher J, Xerogeanes J, Tannenbaum A, Al-Hakim R, Black JC, Zhao J. Anterior cruciate ligament reconstruction in the skeletally immature: An anatomical study utilizing 3-dimensional magnetic resonance imaging reconstructions. J Pediatr Orthop 2009;29:124-129. 19. Makela EA, Vainionpaa S, Vihtonen K, Mero M, Rokkanen P. The effect of trauma to the lower femoral epiphyseal plate. An experimental study in rabbits. J Bone Joint Surg Br 1988;70:187-191. 20. Shea KG, Belzer J, Apel PJ, Nilsson K, Grimm NL, Pfeiffer RP. Volumetric injury of the physis during singlebundle anterior cruciate ligament reconstruction in children: A 3-dimensional study using magnetic resonance imaging. Arthroscopy 2009;25:1415-1422. 21. Shea KG, Grimm NL, Belzer JS. Volumetric injury of the distal femoral physis during double-bundle ACL reconstruction in children: A three-dimensional study with use of magnetic resonance imaging. J Bone Joint Surg Am 2011;93:1033-1038. 22. Wang JH, Kim JG, Ahn JH, Lim HC, Hoshino Y, Fu FH. Is femoral tunnel length correlated with the intercondylar notch and femoral condyle geometry after double-bundle anterior cruciate ligament reconstruction using the transportal technique? An in vivo computed tomography analysis. Arthroscopy 2012;28:1094-1103. 23. Lee DH, Kim HJ, Ahn HS, Bin SI. Comparison of femur tunnel aperture location in patients undergoing transtibial and anatomical single-bundle anterior cruciate ligament reconstruction. Knee Surg Sports Traumatol Arthrosc 2016;24:3713-3721. 24. Shin YS, Ro KH, Lee JH, Lee DH. Location of the femoral tunnel aperture in single-bundle anterior cruciate ligament reconstruction: Comparison of the transtibial, anteromedial portal, and outside-in techniques. Am J Sports Med 2013;41:2533-2539. 25. Shea KG, Apel PJ, Pfeiffer RP, Traughber PD. The anatomy of the proximal tibia in pediatric and adolescent patients: Implications for ACL reconstruction and prevention of physeal arrest. Knee Surg Sports Traumatol Arthrosc 2007;15:320-327. 26. Nha KW, Han JH, Kwon JH, Kang KW, Park HJ, Song JG. Anatomical single-bundle anterior cruciate ligament reconstruction using a freehand transtibial technique. Knee Surg Relat Res 2015;27:117-122. 27. Sohn OJ, Lee DC, Park KH, Ahn HS. Comparison of the modified transtibial technique, anteromedial portal technique and outside-in technique in ACL reconstruction. Knee Surg Relat Res 2014;26:241-248.