A twelve-year-old boy was seen after he had sustained a knee injury two months earlier while playing basketball. On physical examination, he had a positive Lachman test and pivot shift test in the right knee. Magnetic resonance imaging (MRI) confirmed the diagnosis of an ACL tear. The patient's bone age was eleven years and six months according to the Greulich and Pyle standard atlas17. Since the time of the injury, he had been participating in physical therapy but complained of recurrent episodes of knee pain and sensations of instability. He had regained the preinjury range of motion, and otherwise the examination showed normal findings in both knees.
Three months after the injury, the patient underwent ACL reconstruction with an autologous hamstring graft, with use of the epiphyseal femoral tunnel described by Anderson and a transphyseal tibial tunnel15,16. A minimal notchplasty of the anterior to the distal portion of the lateral femoral condyle was performed for enhanced visualization. No bone or soft tissue was removed from the proximal notch to avoid any disturbance of the physis or perichondral ring. With use of fluoroscopic assistance, a 6-mm femoral tunnel was drilled from inside-out to the depth of 30 mm with the use of an Arthrex RetroDrill (Naples, Florida), leaving the lateral cortex intact. Additional confirmation of the lack of a breach of the physis was obtained by directly inspecting the inside walls of the tunnel. A 6-mm tunnel was drilled through the tibial physis, a step that differs from the technique described by Anderson. The graft was looped over two number-5 FiberWire sutures (Arthrex) and tied over a button on the lateral aspect of the distal part of the femur. On the tibial side, fixation was accomplished by placing a staple across the tensioned graft at the distal exit of the tibia. This was reinforced with a post-screw near the staple (Fig. 1). Postoperatively, the patient wore a brace and followed a standard physical therapy rehabilitation protocol.
Anteroposterior (A) and lateral (B) radiographs of the right knee two months after ACL reconstruction, demonstrating the location of the 6 × 30-mm epiphyseal femoral tunnel and the transphyseal tibial tunnel.
About six months following the ACL reconstruction, the patient slowly returned to sports without any difficulty. At nine months, he was noted to have pain on the lateral aspect of the right knee, directly over the femoral fixation site. He subsequently underwent removal of the button, with concurrent arthroscopy demonstrating no articular damage or meniscal injury and an intact ACL graft.
Fourteen and a half months following his initial ACL reconstruction, the patient sustained a noncontact injury while playing football. Physical examination and MRI confirmed a rerupture of the ACL. Standing hip-to-ankle anteroposterior radiographs were obtained (Fig. 2, A). As shown in Table I, there was symmetric mild valgus angulation in the left and right knees as demonstrated by the anatomic tibiofemoral angle (aTFA) and the hip-knee-ankle (HKA) angle18,19. However, the overall mechanical axis deviation (MAD) as well as the lateral distal femoral angle (LDFA) and the medial proximal tibial angle (MPTA) were all within normal limits. There was no clinical or radiographic evidence of leg-length discrepancy. An MRI acquired at this time revealed a small streak from the physis, but the lateral side of the distal femoral physis remained open with evidence of symmetric growth throughout (Fig. 3, A).
Standing hip-to-ankle anteroposterior radiographs made twenty-one months apart. A: Two months prior to revision of the ACL reconstruction. B: Nineteen months following revision of the ACL reconstruction, there was valgus deformity in the right knee. The LDFA and MPTA angles are shown.
A: Two months prior to the revision ACL reconstruction, a single sagittal image from a T1-weighted MRI demonstrated a physeal streak in the lateral condyle. This may represent a small physeal bar. B: An intraoperative photograph demonstrating no breach of the distal femoral physis. C and D: Sequential fast-spin-echo coronal MRI images with extended echo-train acquisition made five months after the revision ACL reconstruction also demonstrated the all-epiphyseal location of the femoral tunnel. Premature physeal closure on the lateral side compared with the medial side is evident in D.
At the age of fourteen years (bone age, thirteen years and six months), the patient underwent revision of the ACL reconstruction with use of an anterior tibialis allograft. At arthroscopy, the initial hamstring graft appeared to be torn in its midsubstance, closer to the femoral insertion. The tunnels were localized and then redrilled with an Arthrex RetroDrill, enlarging them to 7.5 mm to accommodate the size of the graft. Intraoperative fluoroscopy was again used to redrill the femoral tunnel to avoid the physis with subsequent arthroscopic confirmation of no physeal breach. The graft was fixed on the outside of the cortex of the lateral femoral condyle with a 15-mm Arthrex RetroButton and at the tibia with two staples. Postoperatively, the patient wore a brace and again followed the ACL reconstruction physical therapy protocol.
A year and a half following the ACL revision, standing hip-to-ankle anteroposterior radiographs demonstrated an increased valgus deformity in the right knee that was due to distal femoral valgus (Fig. 2, B). Whereas the values for overall valgus had normalized in the left knee, all of the overall alignment measurements indicated that the right genu valgum was present on the right. The MAD had increased to 32 mm, with the weight-bearing line now projecting lateral to the lateral tibial spine. Both the aTFA and the HKA angle indicated substantial genu valgum outside of the range of reported normal values18,19. Examination of the distal femoral and proximal tibial articular angles demonstrated the cause of this angular deformity to be the distal femoral valgus. The MPTA remained within normal limits, suggesting no tibial growth disturbance. However, the LDFA was outside the normal range and had decreased 6.2° in twenty-one months. An MRI from this time demonstrated a premature physeal closure of the lateral side of the femoral physis without a frank breach of the physis or a physeal bar (Fig. 3). The femoral physis remained open on the medial side. There was a minor leg-length discrepancy of 1.13 cm. As the patient was approaching skeletal maturity at this point, no procedure was undertaken to correct the valgus deformity in the right knee.
The apparent increase in ACL tears in skeletally immature patients and the belief that nonoperative treatment may lead to other intra-articular knee injuries has spurred interest in furthering ACL reconstructive techniques in the skeletally immature population. The technique described by Anderson15,16 theoretically avoids injury to the distal femoral physis. However, as was the case in our patient with distal femoral valgus associated with a revision ACL reconstruction done with use of this technique, the risk of a growth disturbance is still present. The technique for ACL reconstruction used in this patient does differ somewhat from that described by Anderson, but only with respect to the tibial tunnel; the femoral tunnel within the femoral epiphysis was as previously described and appears to be the cause of the premature physeal closure of the lateral aspect of the distal part of the femur and the subsequent growth disturbance in this patient.
The anatomy of the distal part of the femur is complicated by the undulations of the growth plate in both the coronal and the sagittal plane. It is difficult to accommodate for this anatomy when drilling an epiphyseal tunnel, especially because of the narrow margin of error (measured in millimeters) for placing the tunnel in the center of the anatomic ACL footprint while avoiding both the distal femoral physis and the articular surface. The use of fluoroscopy definitely aids in identifying the borders of the physis, but also underscores the need for better intraoperative imaging techniques because the success of an ACL repair with use of epiphyseal tunnels depends on maintaining pristine physeal borders. The intraoperative use of three-dimensional computed tomography, potentially even with computer navigation, could help to reduce the risk of tunneling too close to a healthy physis. This technique has recently been described20 and may become very useful to avoid physeal injury. Our case also highlights the need to continue to monitor pediatric patients with yearly evaluations until skeletal maturity is attained.
On the basis of the imaging history, it appears that the growth disturbance in our patient likely occurred primarily after the revision ACL reconstruction. The reasons why this happened are unclear. Standing hip-to-ankle anteroposterior radiographs made more than a year following the initial operation demonstrated symmetric mild genu valgum (Fig. 2, A, and Table I). In addition, while the MRI acquired after the patient sustained a retear did appear to show one small physeal bar, the physis of the lateral femoral condyle remained open and appeared equivalent to the physis in the medial femoral condyle. The patient's growth before the revision was symmetric. However, repeat standing hip-to-ankle anteroposterior radiographs made a year and a half following the revision surgery demonstrated an increase in the genu valgum of the right lower limb and a concurrent decrease in the previous physiologic genu valgum of the left leg (Fig. 2, B, and Table I). This suggests that the revision ACL reconstruction was responsible for the growth disturbance leading to the valgus deformity in the distal part of the femur. Despite this evidence, it is possible that the growth disturbance represents a delayed presentation of an insult to the physis during the initial ACL reconstruction.
One potential explanation for the growth disturbance could be that the physis was entered by increasing the diameter of the femoral tunnel by 1.5 mm. While there is extensive radiographic and intraoperative evidence that the physis was not directly violated at either the primary or the revision reconstruction, the proximity of the tunnel to the physis may have directly or indirectly induced the growth disturbance. For instance, thermal injury at the time of drilling may have created a zone of injury greater than the actual tunnel diameter. Alternatively, the proximity of the tunnel could have altered the blood supply to the epiphyseal side of the physis between the tunnel and the physis. However, these possibilities seem less likely since they would be thought to have produced a localized insult to the physis, thus creating more of a physeal bar effect.
The MRI evidence instead suggests a global arrest of the entire lateral side of the femoral physis at areas distant from the femoral tunnel (Fig. 3). To explain this occurrence, another possible reason for the observed growth arrest is excessive pressure on the surrounding physis because the area overlying the tunnel may have been relatively unloaded, a so-called trampoline effect. Furthermore, it is possible that the segment of physeal arrest could have been subject to shear stresses arising from the initial anchoring of the graft on the lateral aspect of the lateral femoral condyle. However, once the graft becomes incorporated in the tunnel, since it is placed in the center of the femoral ACL footprint, it should not transmit forces to the lateral femoral condyle any differently than the native ACL does in a skeletally immature knee. Another explanation could be that the valgus deformity resulted from an indirect, double insult to the physis.
While the technique for ACL reconstruction described here very closely reproduces the natural anatomic form and function of the ACL, it is not the only approach that can be used to reconstruct ACL tears in skeletally immature children. A physeal sparing, combined intra-articular and extra-articular reconstruction with use of an autogenous iliotibial band graft, whereby the graft remains attached to the tibia and the proximal part of the graft is routed over the back of the lateral femoral condyle and into the notch of the knee, has been described21,22. The graft is then brought through the knee and under the intermeniscal ligament and eventually is anchored to the anterior aspect of the tibia21. Many good results have been reported with this technique, although growth disturbances have also been noted12,23. Transphyseal reconstructions have also been described in this age group; however, animal models have elucidated the dangers associated with this technique, and clinical examples of growth disturbances have been reported5-12,14,15. In addition, many of the reported series of transphyseal reconstructions were in patients with open physes but approaching skeletal maturity, and the reconstructions were performed with a vertically placed femoral tunnel. With a more vertical tunnel, the physeal injury is more central in the femur and is less likely to cause a growth disturbance, as was demonstrated with the tibial tunnel in this case. However, as our understanding of the anatomic ACL footprint has evolved, the ideal position for the femoral tunnel has changed from 11 o'clock to further down the wall of the notch, in the anatomic center of the ACL footprint. Placing the center of the femoral tunnel in this position increases the obliquity and eccentricity of the tunnel as it passes through the physis. As the obliquity of the tunnel through the physis increases, the area of physeal damage increases. In addition, as the eccentricity of the physeal injury increases, the risk of any growth disturbance leading to a clinically relevant effect also increases.
Finally, while the primary goal of early ACL reconstruction in patients with substantial growth remaining is stabilization of the knee in the hope of preventing further intra-articular damage, most ACL injuries in this age group are in patients who want to return to sports. Unfortunately, as Shelbourne et al. recently highlighted, young patients who return to high-level sports participation are at increased risk of rerupture24. The present case draws attention to the consequences associated with that increased risk. Thus, since the potential consequences of revision surgery may be increased in this population, physicians should consider this risk of rerupture when counseling young athletes and their families about the child returning to sports following ACL reconstruction.