Preoperative Evaluation and Anterior Cruciate Ligament Reconstruction Technique for Skeletally Immature Patients in Tanner Stages 2 and 3

Transcription

1 /103/ $02.00/0 THE AMERICAN JOURNAL OF SPORTS MEDICINE, Vol. 31, No American Orthopaedic Society for Sports Medicine Preoperative Evaluation and Anterior Cruciate Ligament Reconstruction Technique for Skeletally Immature Patients in Tanner Stages 2 and 3 Vincenzo Guzzanti,* MD, Francesco Falciglia, MD, and Carl L. Stanitski, MD From the Orthopaedic Department, the Children s Hospital Bambino Gesù, Rome, Italy Background: Transphyseal anterior cruciate ligament reconstruction in skeletally immature patients carries the risk of adverse sequelae. Purpose: 1) To provide clinical and radiographic methods to identify skeletally immature patients with significant remaining lower limb growth. 2) To provide a method for calculating the percent area lesion created by a 6-mm distal femoral physeal tunnel. 3) To present a method of partial transphyseal intraarticular anterior cruciate ligament reconstruction with bipolar graft fixation in these patients. 4) To report on clinical, functional, and radiographic results at skeletal maturity. Methods: Fourteen adolescents with symptomatic anterior cruciate ligament instability who were identified as being in Tanner stages 2 and 3 underwent partial transphyseal intraarticular anterior cruciate ligament reconstruction with the use of hamstring tendon grafts (transphyseal only in the femur and through the epiphysis in the tibia). Results: At skeletal maturity, 10 patients were asymptomatic and fully active in sports. No patient had significant leg-length inequality or angular deformity after use of a 6-mm femoral physeal tunnel, which represented 1) less than 7% of the frontal plane and 2) less than 1% of the transverse plane cross-sectional femoral physeal areas. Conclusions: The choice of the exposed technique and the method used to select patients permitted us to avoid adverse sequelae American Orthopaedic Society for Sports Medicine Anterior cruciate ligament injury in the skeletally immature patient has been the subject of increasing attention during the past decade. 2,21,22,26,27,29,30 Unsatisfactory results of nonoperative treatment have been consistently reported for patients in this age group who return to high-level sports activity. 3,8,14,16,19,23 Surgical treatment to provide normal stability involves isometric graft placement and fixation. 4,7,12,13 Techniques to achieve this goal may lead to physeal injury and subsequent angular and longitudinal lower extremity deformities, particularly in adolescents in Tanner stage 2 and 3 with at least 5 cm of lower extremity growth remaining. Because of these potential consequences, some authors suggest that surgical management be delayed until skeletal maturity. 17 Others * Address correspondence and reprint requests to Prof. Vincenzo Guzzanti, MD, Children s Hospital Bambino Gesù, Orthopaedic Department, Piazza S. Onofrio 4, Rome, Italy. No author or related institution has received any financial benefit from research in this study. suggest that surgical correction during the final growth phase in adolescents, with the use of techniques similar to those used in adult reconstruction, produces satisfactory mechanical and biologic results. 5,21,27 It is not always easy to delay surgical treatment in truly skeletally immature emerging athletes. Patient compliance is poor during formative athletic years. Pressures from peers, parents, and coaches encourage return to play. In patients with an ACL-plus knee, that is, patients who have a concomitant meniscal tear, the issue becomes more exigent. In these patients and in the noncompliant patient, surgery is indicated to prevent progressive intraarticular damage. Many of these patients are at a high risk of possible growth disturbances after ACL reconstruction because of the significant remaining growth ( 5 cm) expected in the lower extremities. 19 Previous reports of ACL reconstruction in skeletally immature patients are few, and existing reports are flawed by limited follow-up and incomplete assessments of physiologic age, diagnosis, and lower extremity growth 941

2 942 Guzzanti et al. American Journal of Sports Medicine potential. 2,18,19,25 There is only one report on bipolar graft fixation, 18 and it did not include data on physiologic age or specific radiographic information on the results of the femoral and tibial physeal and epiphyseal status of the reconstruction. In the skeletally immature patient, the placement of physeal and epiphyseal drill holes and the passage of tendon grafts through them raises biologic issues. 2,8,14,19,27 Some reports suggest that the proximal tibial physis can be drilled without consequence, provided the lesion is central In one study, the authors indicated that the distal femoral epiphysis can be drilled transversely with impunity. 18 These studies in animals do not provide a directly transferable approach to the human distal femoral and proximal tibial physes and epiphyses. 10,15,28 There are still uncertainties concerning drill hole magnitude and orientation relative to the ratio of the tunnel area to the total physeal area. Experimental data from rabbits and dogs indicate that the tolerable ratio of physeal transgression remains a concern. 10,15,28 We believe that the initial step in surgical management of these young patients should be identification of the critical tolerance of human physeal transgression by drill and tendon graft that allows stable, isometric graft fixation in patients who are in Tanner stages 2 and 3 and who have a lower limb growth prediction of at least 5 cm. The goals of this study were 1) to use clinical and radiographic evaluation to characterize skeletally immature patients at high and intermediate risks of adverse sequelae of transphyseal ACL reconstruction, 2) to provide methods to calculate the percent area lesion created by a 6-mm distal femoral physeal tunnel, 3) to describe a partial transphyseal intraarticular ACL reconstruction method with bipolar graft fixation, and 4) to report on clinical, functional, and radiographic results at skeletal maturity in patients who have undergone ACL reconstruction with this technique. Between 1989 and 2000, 47 preadolescents and adolescents underwent surgical treatment for symptomatic ACL insufficiency by one surgeon (VG). Most of the patients injuries had occurred during vigorous sports participation. In all of the patients, nonoperative treatment had failed because of noncompliance. The patients were evaluated preoperatively for physiologic maturity by assessment of Tanner stage, 31 bone age, 9 and lower extremity remaining growth. 1,6 The total growth prediction was based on assessment of chronologic age, bone age, parents heights, patient s standing height, and patient s sitting height. In Tanner stage 1 patients, lower extremity growth projection is considered to be 50% of total projected height. In Tanner stage 2 and 3 patients, lower extremity growth potential represents 40% of total projected height and, in Tanner stage 4 and 5 patients, lower extremity growth potential is 15% to 20% of the total eventual height. The patients were stratified as being at high, intermediate, or low risk of developing longitudinal and angular deformities after transphyseal ACL reconstruction. High-risk patients were preadolescents who were in Tanner stage 1, had a bone age of 11 years for girls and 12 years for boys, and whose lower extremity growth potential exceeded 7 cm. These patients had not started their accelerated phase of adolescent growth and radiographs showed wide open distal femoral and proximal tibial physes 19 (Fig. 1). The intermediate-risk patients were in Tanner stage 2 or 3, had a bone age of 13 years for girls and 15 years for boys, and lower extremity growth potential of 1.5 to 7 cm. They were at varying stages of their adolescent rapid growth phase and radiographs showed distal femoral and proximal tibial physes that were still open 18,21,27 (Fig. 2). Low-risk patients were in Tanner stage 4 or 5, had a bone age of 14.5 years in girls and 16 in boys, were past their adolescent rapid growth phase, and had a lower extremity growth potential of less than 1.5 cm. Radiographic images in these patients showed almost closed or closed distal femoral and proximal tibial physes 5,18,21,27 (Fig. 3). Preoperative AP knee radiographs were obtained with a standard 100-cm beam source distance. Distal femoral CT scans at the level of the distal femoral physis were obtained as well. The theoretical percent of the total physeal area transgressed by a 6-mm transphyseal femoral tunnel was calculated from the transverse plane of the CT scan and the frontal plane of the AP knee radiographs, as MATERIALS AND METHODS Figure 1. Radiograph showing a wide open physis.

3 Vol. 31, No. 6, 2003 ACL Reconstruction in Skeletally Immature Patients 943 Figure 2. Radiograph showing an open physis (Case 8). described in a previous experimental study in animals. 10 Plans for surgical partial transphyseal reconstruction were performed only if the femoral physeal tunnel percentage area was less than 7.3% in the frontal plane and 1% or less in the transverse plane (Table 1), based on experimental data from outcomes in lesions 11% in the frontal plane and 3% in the transverse plane. 10 Patients in the high-risk group were not treated with a transphyseal method because the calculated physeal area of the drill hole was thought to be too great and would have created a major risk of abnormal physeal growth consequences. The patients in this group were treated with a physeal-sparing technique. 11 The low-risk group of patients was treated with the same standard complete transphyseal methods used in adults because the potential for postoperative growth abnormalities in this group was minimal or nonexistent. The intermediate-risk group underwent the surgical procedure outlined here (see Surgical Technique ) for patients with symptomatic ACL deficiency, a 3 positive Lachman and pivot shift test, and poor end points on examination. Because of noncompliance, all patients in the intermediate-risk group had experienced failure of a 4-month nonoperative rehabilitation program that included activity modification. Additional inclusion criteria for this group were clinical and radiologic follow-up data at skeletal maturity and a follow-up of at least 24 months. Four patients in this group had associated medial meniscal injuries. Figure 3. Radiographs showing a closing physis. Above, Tanner stage 5 male (M) patient with a bone age of 15 years; Below, Tanner stage 4 female (F) patient with a bone age of 15 years. Surgical Technique After arthroscopic confirmation of the intraarticular lesions, the hamstring tendons (semitendinosus alone in case 1 and semitendinosus with gracilis in the remaining cases) were released proximally. An anterior miniarthrotomy incision was used. An eccentric 6-mm drill hole was made through the proximal tibial epiphysis between the physis and articular surface with exit at the tibial eminence. The hole was drilled once under fluoroscopic guidance. The 6-mm femoral tunnel was drilled transphyseally through the distal femoral physis via the intercondylar notch with use of fluoroscopic control 29 (Fig. 4). Notchplasty was not performed. The grafted tendons were not doubled. The tendon graft bundle was passed through both tunnels and attached to the distal femoral lateral metaphysis with care taken to avoid the distal femoral physis, which was identified by multiple fluoroscopic images. The ligament graft was tensioned with the knee in 30 of flexion before femoral fixation. 18 Femoral attachment of the graft was performed via a small lateral incision. Periosteal suture was used in one case and staple fixation was used in all of the other cases. The tendon graft was sutured to the tibial periosteum at the tibial metaphysis in all cases. Postoperatively, a hinged knee brace was used to control knee motion from 20 to 70 of flexion. Active assisted and passive knee range of motion was begun immediately postoperatively and incorporated use of a passive range of

4 944 Guzzanti et al. American Journal of Sports Medicine Case Sex Tanner stage Age (years) TABLE 1 Preoperative Data on 14 Skeletally Immature Patients Bone age (Greulich-Pyle) Growth prediction of lower limb (cm) % damage on radiographs (AP) distal femoral physis % damage on CT distal femoral physis Type of lesion 1 M Midsubstance 2 M Midsubstance 3 M Midsubstance 4 M Midsubstance 5 M Midsubstance 6 M Midsubstance 7 M Midsubstance 8 M Femoral insertion 9 M Midsubstance 10 M Midsubstance 11 M Midsubstance 12 M Midsubstance 13 M Midsubstance 14 F Midsubstance motion machine at home. Partial weightbearing gait with the protection of crutches was used for the first 4 weeks. RESULTS Ten patients (the first 10 patients listed in Table 1) of 14 adolescent patients with an average chronologic age of 13.6 years (range, 13.2 to 14.0) and average bone age of 13.5 years (range, 13.5 to 14) had review of all clinical and radiographic records before and after partial physealsparing intraarticular ACL reconstruction. At final review, four patients (cases 11 through 14, Table 1) were excluded because they had not reached skeletal maturity (Table 2). Follow-up averaged 40.1 months (range, 24 to 108), with a median of 35 months. Three patients were in Tanner stage 2 and seven patients were in Tanner stage 3 at the time of surgery. Preoperative predictions of lower extremity growth averaged 5.6 cm (range, 5 to 7). As derived from AP radiographs, the preoperatively calculated percent projected invasion of the distal femoral physis ranged from 6.38% to 6.8% and averaged 6.6%. As derived from the CT scans, the preoperatively calculated percent transgression of the femoral distal physes averaged 0.91% (range, 0.78% to 1.0%). No patient had preoperative KT-2000 arthrometer (Medmetric Corp., San Diego, California) testing because of the unavailability of the arthrometer at that time. At final evaluation, KT-2000 arthrometer scores averaged 1.2 mm at 30 pounds of force (range, 2 to 0 mm). Orthopadische Arbeitsgruppe Knie (OAK) 24 final scores averaged 97.5 (range, 94 to 100), with seven patients scoring equal to or greater than 97. Three of these seven patients scored 100. Any value over 90 is considered an excellent clinical result. The radiographic follow-up at the end of the skeletal growth period showed no significant limb-length discrepancies or angular deformities (Table 2) (Figs. 5 and 6). The total average gain in height was 12.8 cm (range, 10.5 to 15). The only significant differences between the Figure 4. Schematic drawing of the partial transphyseal procedure. Semitendinosus and gracilis tendons, proximally detached, pass though 6-mm tunnels, one in the tibial epiphysis and the other in the femoral physis and epiphysis.

5 Vol. 31, No. 6, 2003 ACL Reconstruction in Skeletally Immature Patients 945 Case TABLE 2 Clinical and Radiographic Data for Study Group Patients at Skeletal Maturity Graft Follow-up (months) KT-2000 arthrometer (mm) OAK Evaluation a Femoral length (cm) Limb-length discrepancies Tibial length (cm) 1 Semitendinosus Semitendinosus and gracilis Semitendinosus and gracilis Semitendinosus and gracilis Semitendinosus and gracilis Semitendinosus and gracilis Semitendinosus and gracilis Semitendinosus and gracilis Semitendinosus and gracilis Semitendinosus and gracilis a Orthopadische Arbeitsgruppe Knie. Height gained (cm) Figure 5. Case 1: Tanner 2 male, bone age 14 years, 6 cm projected lower extremity growth. Postoperative radiograph at 6 months showing symmetric development of the physis and the epiphysis of the operated (A) and nonoperated (B) side. On the operated side, a small trace of calcification parallel to the direction of the tendon in the lateral part of physis can be observed. C, radiograph 42 months postoperatively at skeletal maturity. Tanner stage 2 and Tanner stage 3 groups were in average follow-up time (59.3 versus 31.8 months), average lower limb predicted growth (6.56 versus 5.28 cm), and average total height gained (14.2 versus 12.3 cm). The latter two values for height were not unexpected given that the Tanner stage 2 patients were of lesser maturity at the outset. Magnetic resonance imaging studies done at 1 and 2 years postoperatively did not show evidence of focal epiphyseodesis at the graft sites (Fig. 7). The MRI signals were symmetric and homogeneous within the physis in the frontal (Fig. 7A) and sagittal planes (Fig. 7B). No evidence of graft calcification or ossification was seen on MRI or CT images (Fig. 8). All patients returned to high-level sports activity equivalent to their preinjury status. DISCUSSION Growth characterization by preoperative clinical evaluation (history, physical examination, Tanner staging) and radiographic imaging methods (bone age, teleroentgenograms, radiographs, CT scans, and MRI) allowed identification of patients in Tanner stages 2 and 3 with significant remaining lower extremity growth ( 5 cm). The calculated values of the percent area of physeal involvement produced by a transphyseal approach also aided in decisions about tunnel sites, sizes, and graft size. Choice of a partial transphyseal intraarticular reconstruction method with hamstring tendon grafts placed through 6-mm femoral physeal and epiphyseal and tibial epiphyseal tunnels was based on these data.

6 946 Guzzanti et al. American Journal of Sports Medicine Figure 6. Case 6: Tanner 3 male, bone age 14 years, projected 5.5 cm lower extremity growth. Radiograph at 30 months follow-up at skeletal maturity shows no lower extremity deformity. Reviews of CT scans and standard radiographic knee images showed that the injury produced from violating the distal femoral physis with a drill hole and graft insertion would be acceptable if the injury area was less than 7% in the frontal and less than 1% in the transverse planes of the total physeal area. These calculations and the relative reduced exposure time of physeal growth of these patients aided in choosing the appropriate surgical technique used in the ACL reconstruction. 29 Stable, isometric graft fixation was achieved without adverse sequelae of physeal transgression. A femoral rather than tibial physeal tunnel was chosen because experimental data from rabbits showed 2 of 18 cases of tibia valga and 1 of 18 cases of shortened tibia after transphyseal tibial tunnel formation. 10 We also believed that creation of an isometric graft was technically easier with this technique than with a transphyseal tibial tunnel and over-the-top femoral approach, as previously reported. 18 If we had used a 9-mm femoral physeal tunnel, the percent physeal damage would have been quite close to experimental values reported to cause deformity. 17 Creation of an intraepiphy- Figure 7. Case 6 at 24 months follow-up, MRI scan (T1- weighted) shows symmetric and homogeneous signal of the femoral growth plate transgressed by the neoligament. No signs of focal epiphyseodesis are seen in either femoral or tibial physes. A, frontal plane; B, sagittal plane.

7 Vol. 31, No. 6, 2003 ACL Reconstruction in Skeletally Immature Patients 947 Figure 8. A, case 1 MR image (T1- and T2-weighted). B, no calcification or osseous metaplasia of the grafted tendons was seen in the tunnel imaged at skeletal maturity, 4 years postoperatively. C, case 6 MR image (T1-weighted) showed no osseous metaplasia of the grafted tendons in the tunnel 2 years postoperatively. seal tibial tunnel of that size would have not been technically possible. Physeal and epiphyseal drill lesions and graft presence always create local inflammatory reaction. This response is not of eventual consequence if limits are not overcome. The thin cylinders of bone formed about the tendons within the tunnels are fractured by distraction forces during growth. Physeal growth has been experimentally demonstrated to be reduced initially after drilling, with later evidence of catch-up growth to allow equal length at maturity. 20 Data from rabbit and dog models that show no consequenses of physeal transgression must be reviewed with caution because these animals reach maturity at much faster rates than humans, and total exposure times are much longer in the skeletally immature patient than in the laboratory models. 14 Our preoperative assessment methods allowed a choice of surgery that led to no sequelae of physeal arrest in adolescents in Tanner stages 2 and 3. In previously reported series without postreconstruction deformities, even in those series without rigorous preoperative evaluation, the graft and tunnel sizes chosen may have fortunately been small enough to not exceed the tolerable invasive area. In many reported series, the majority of the patients were at or well on the way to skeletal maturity, with little risk present from remaining growth. Such factors may account for the lack of postreconstruction deformities in these studies. Previous recommendations to involve or avoid femoral or tibial chondroepiphyses during reconstruction should be reviewed with caution when such recommendations are based on preoperative assessment of only bone age or menarchal absence. Such incomplete evaluation could result in the use of complete transphyseal methods with later physeal growth disturbance. Inaccurate preoperative assessment can also suggest the need for partial transphyseal techniques when, in reality, a complete transphyseal technique would be more appropriate based on factors previously discussed. Thus, a more complete preoperative assessment is necessary. It must be emphasized that the complications of growth arrest reported in the literature can be attributed to two types of iatrogenic deficiencies: 1) inadequate preoperative clinical and radiographic assessment of physiologic maturity and growth, and 2) faulty surgical judgment and technical errors; for instance, screw fixation across the physis, bone plugs within the transphyseal tunnels, or excessively large physeal tunnels. In summary, our study has defined several issues, including the clinical and radiographic characterization of risk levels for development of physeal consequences from an ACL reconstruction during skeletal immaturity. A threshold value, that is, the percentage of eccentric femoral damage from drilling and graft presence that did not cause growth consequences in these at-risk patients, was determined. A surgical technique was presented that allows isometric partial transphyseal intraarticular ACL reconstruction with the use of hamstring tendon grafts placed through distal femoral epiphyseal and physeal tunnels and a proximal tibial epiphyseal tunnel. This technique does not result in adverse physeal sequelae in this patient group, which, as discussed in this study, has welldefined growth characteristics. A limitation of this study is the relatively small number of patients. Thorough follow-up to skeletal maturity demonstrated excellent clinical and radiographic short-term results in this selected group. These outcomes validated the choice of the partial transphyseal technique, which has not been previously reported in skeletally immature patients. Further investigation is needed to identify the percent of physeal invasion allowable in preadolescents, that is, in those patients who are significantly skeletally immature, with significant growth remaining and major exposure over time to the effects of a transphyseal bony tunnel and tendon graft. REFERENCES 1. Anderson M, Green WT, Messner MB: Growth and predictions of growth in the lower extremities. J Bone Joint Surg 45A: 1 14, Andrews M, Noyes FR, Barber-Westin SD: Anterior cruciate ligament allograft reconstruction in the skeletally immature athlete. Am J Sports Med 22: 48 54, Angel KR, Hall DJ: Anterior cruciate ligament injury in children and adolescents. Arthroscopy 5: , 1989

8 948 Guzzanti et al. American Journal of Sports Medicine 4. Brand J Jr, Weiler A, Caborn DNM, et al: Graft fixation in cruciate ligament reconstruction. Am J Sports Med 28: , DeLee JC: Ligamentous injury of the knee, in Stanitski CL, DeLee JC, Drez WB (eds): Pediatric and Adolescent Sports Medicine. WB Saunders Company, 1994, pp Dimeglio A: La Croissance en Orthopédie. EMC Paris, Appareil locomoteur , A10, 3, 1984, pp Fu FH, Bennett CH, Lattermann C, et al: Current trends in anterior cruciate ligament reconstruction. Part 1: Biology and biomechanics of reconstruction. Am J Sports Med 27: , Graf BK, Lange RH, Fujisaki CK, et al: Anterior cruciate tears in skeletally immature patients: Meniscal pathology at presentation and after attempted conservative treatment. Arthroscopy 8: , Greulich WW, Pyle SI: Radiographic Atlas of Skeletal Development of the Hand and Wrist. Second edition. Stanford, CA, Stanford University Press, Guzzanti V, Falciglia F, Gigante A, et al: The effect of intra-articular ACL reconstruction on the growth plates of rabbits. J Bone Joint Surg 76B: , Guzzanti V, Falciglia F, Stanitski CL: Physeal-sparing intraarticular anterior cruciate ligament reconstruction in preadolescents. Am J Sports Med 31: , Hefzy MS, Grood ES: Sensitivity of insertion locations on length patterns of anterior cruciate ligament fibers. J Biomech Eng 108: 73 82, Hefzy MS, Grood ES, Noyes FR: Factors affecting the region of most isometric femoral attachments. Part II: The anterior cruciate ligament. Am J Sports Med 17: , Janarv PM, Nystrom A, Werner S, et al: Anterior cruciate ligament injuries in skeletally immature patients. J Pediat Orthop 16: , Janarv PM, Wikstrom B, Hirsch G: The influence of transphyseal drilling and tendon grafting on bone growth: An experimental study in the rabbit. J Pediatr Orthop 18: , Kannus P, Jarvinen M: Knee ligament injuries in adolescents: Eight years follow-up of conservative management. J Bone Joint Surg 70B: , Koman JD, Sanders JO: Valgus deformity after reconstruction of the anterior cruciate ligament in a skeletally immature patient. A case report. J Bone Joint Surg 81A: , Lipscomb AB, Anderson AF: Tears of the anterior cruciate ligament in adolescents. J Bone Joint Surg 68A: 19 28, Lo IK, Kirkley A, Fowler PJ, et al: The outcome of operatively treated anterior cruciate ligament disruptions in the skeletally immature child. Arthroscopy 13: , Makela EA, Vainionpaa S, Vihtonen K, et al: The effect of trauma to the lower femoral epiphyseal plate: An experimental study in rabbits. J Bone Joint Surg 70B: , McCarroll JR, Shelbourne KD, Porter DA, et al: Patellar tendon graft reconstruction for midsubstance anterior cruciate ligament rupture in junior high school athletes. An algorithm for management. Am J Sports Med 22: , Micheli LJ, Rask B, Gerberg L: Anterior cruciate ligament reconstruction in patients who are prepubescent. Clin Orthop 364: 40 47, Mizuta H, Kubota K, Shiraishi M, et al: The conservative treatment of complete tears of the anterior cruciate ligament in skeletally immature patients. J Bone Joint Surg 77B: , Muller W, Biedert R, Hefti F, et al: OAK knee evaluation: A new way to assess knee ligament injuries. Clin Orthop 232: 37 50, Nakhostine M, Bollen SR, Cross MJ: Reconstruction of mid-substance anterior cruciate rupture in adolescents with open physes. J Pediatr Orthop 15: , Parker AW, Drez D Jr, Cooper JL: Anterior cruciate ligament injuries in patients with open physes. Am J Sports Med 22: 44 47, Pressman AE, Letts RM, Jarvis JC: Anterior cruciate ligament tears in children: An analysis of operative versus nonoperative treatment. J Pediatr Orthop 17: , Stadelmaier DM, Arnoczky SP, Dodds J, et al: The effect of drilling and soft tissue grafting across open growth plates. A histologic study. Am J Sports Med 23: , Stanitski CL: Anterior cruciate ligament injury in the skeletally immature patient: Diagnosis and treatment. J Am Acad Orthop Surg 3: , Stanitski CL, Harvell JC, Fu F: Observations on acute knee hemarthrosis in children and adolescents. J Pediatr Orthop 13: , Tanner JM: Foetus into Man: Physical Growth from Conception to Maturity. Second edition. Ware, Castlemead Productions, 1989