The use of soft tissue grafts have shown to reduce the risk of growth disturbances and are preferred over bone-tendon grafts during ACL reconstruction in the skeletally immature patients [46–48]. Bone-tendon grafts are discouraged in younger patients to avoid harvest of bone close to physis and to avoid placement of bone plugs across open physis to minimize growth disturbances. Hamstring tendon graft is the most widely reported graft for ACL reconstruction in children. The outcomes of ACL reconstruction using hamstring graft have been reported for all-epiphyseal ACL reconstruction [49–53]. They are also commonly used for transphyseal ACL reconstruction [54–56].

Use of iliotibial band (ITB) for ACL reconstruction was first described by Macintosh and Darby in 1976 in adults [57]. This technique has been since modified and adopted for ACL reconstruction in younger patients (Tanner Stage 1 and 2) [58, 59]. As no tunnel is drilled, this technique is safer and would avoid growth disturbances from physeal violation, though tensioning of graft across the physis could, in theory, lead to physeal tethering and growth disturbances. The chapter on ACL reconstruction without bone tunnels outlines the details and outcomes of this technique.

Kohl et al. reported on the use of quadriceps tendon autograft without bone plug for transphyseal ACL reconstruction in 15 patients with mean age of 12.8 years [60]. Middle part of the quadriceps tendon, measuring 7–9 mm × 8 cm was used. At a mean follow-up for 4.1 years, no graft ruptures were reported. Of 49 patients treated with a quadriceps tendon autograft with bone plug using transphyseal technique, Mauch et al. described one growth disturbance in a 10.5 year old girl [61]. An all-epiphyseal technique using quadriceps tendon autograft with bone plug has also been described though outcomes have not been reported [62].

Bone-patellar tendon-bone (BTB) autograft could be safely used for ACL reconstruction in patients nearing skeletal maturity and when the physis are closing [63]. However, BTB graft has also been used in patients with open physis. Memeo et al. have reported on BTB autograft in 10 patients with mean age of 14.4 years (range 13–16 years) and mean follow-up of 24.9 months (range, 15–44 months) [64]. All patients were Tanner stage 3 with open physis. Slightly vertical tunnels measuring 7–9 mm were drilled across the proximal tibial and distal femoral physis. The author reported no growth disturbances [64]. Despite few reports of ACLR using BTB in pubescent patients with open physis, one has to be cautious as deleterious effects are possible from harvest of BTB graft and from placement of bone plugs across physis. Soft tissue autograft (compared to BTB) is a safer alternative in general. Bonnard et al. reported the results of ACL reconstruction in skeletally immature patients using periosteum-patellar tendon-periosteum autograft [65].

Allograft is a poor choice of graft in the young active patient population and should be discouraged. Multiple studies have reported on increased failure rates after allograft ACLR in younger patients. Engelman et al. performed a retrospective chart review on 73 patients [66]. They noted 15 graft failures, of which 11 were allograft. Analysis from the MOON data showed an alarmingly increased number of ACLR failures when using allograft in young patients [3]. The allograft rupture rate was 37.5% in the 10–19 year age group. As per their model, a 14 year old with ACLR using allograft had a 22% risk of re-rupture compared to 6.6% chance of re-rupture for ACLR using autograft [3]. Overall, the risk of graft failure has been reported to be 4.4–7.7 times higher for allograft in the young and active patient population when compared to autograft techniques [3, 66, 67]. Pallis et al. studied a cohort of US Military Academy recruits with an ACL reconstruction prior to matriculation [67]. One hundred twenty patients with 122 ACL reconstructions were followed. The grafts used were 61 BTB autografts, 45 HT autografts, and 16 allografts. Of the 16 allografts, 7 (44%) failed which was 7.7 times higher than their autograft counterparts. Thus allograft ALCR should be avoided for ACLR in younger patients unless in a revision or multi-ligament reconstruction setting.

Goddard et al. reported on 32 children with mean age of 13 years, who underwent transphyseal ACL reconstruction using living donor (parental) hamstring tendon allografts. Excellent subjective and clinical outcomes were reported with 93% return to strenuous activities. Two children sustained ACL graft rupture within 2 years after surgery [68]. Hui et al. reported on 14 prepubertal patients who received living donor (parental) hamstring tendon allograft during transphyseal ACL reconstruction. At minimum 2 years follow-up, there were no graft re-ruptures and all patients had returned to strenuous activities [69].

It is controversial whether small diameter autograft would be appropriate for patient’s size or would increase the risk for subsequent graft failure. Weight less than 50 kg, height less than 140 cm, thigh circumference less than 37 cm, and a body mass index less than 18 are factors associated with a quadrupled hamstring graft diameter less than 7 mm [70, 71]. In adults, 8 mm is considered to be the minimum required diameter for ACLR. Grafts smaller than 8 mm in diameter are at a 6.8 times greater risk of failure and every 0.5 mm increase in diameter from 7 to 9 mm decreases the likelihood of graft rupture by 0.82 [72–74]. The minimum or optimal graft diameter for ACL reconstruction in children is not established yet.

Guzzanti et al. reported on eight patients with mean age of 11.2 years and Tanner Stage I [49]. All patients underwent physeal sparing ACL reconstruction using hamstring tendon autograft and 6 mm tunnel in the tibial epiphysis [49]. All patients were able to return to full activities without any growth disturbances or graft failures. Anderson reported on 12 patients with mean age of 12.9 years (Tanner stages I-III), who underwent all-epiphyseal ACL reconstruction using quadrupled hamstring autograft. The minimum diameter of the graft in the report was 6 mm [53]. At a mean 4.1 year follow-up there were no graft re-ruptures.

Bollen et al. reported on the fate of quadrupled hamstring graft used for transphyseal ACL reconstruction in five patients (3 Tanner I, 2 Tanner II) at mean of 34.6 months [75]. The diameter of the quadrupled graft (6–7.5 mm) did not change despite 42% average increase in length. Most of the gain in length was on the femoral side. The study indicates neogenesis of the graft and not mere stretching of the graft, as stretching would likely decrease the diameter as length increases. The findings of this study were refuted by Astur et al., who reported decrease in the diameter of quadrupled hamstring graft by average 25.3% and postulated that to be the reason for higher graft rupture rate in children [76].

In 103 patients who underwent all-epiphyseal ACL reconstruction using predominantly hamstring autograft, there was no difference in graft rupture rates between those with graft diameters of ≥8 mm and those with diameters <8 mm [77]. In this large series, one patient had a 6 mm graft, 16 patients had 7 mm graft, 15 patients had 7.5 mm graft, and rest had graft diameter of ≥8 mm. The authors preferred to tailor the graft size to individual patient without a minimum or optimal graft size. It appears from literature review that less than 8 mm diameter graft is acceptable for all-epiphyseal ACL reconstruction which is typically performed in prepubescent patients [53]. It is not known if less than 8 mm diameter graft would be optimal for transphyseal ACL reconstruction which is typically performed in pubescent patients.

If the harvested graft diameter is <8 mm, and ≥8 mm graft diameter is desirable, then few options exist. Augmentation of autograft with an allograft (Hybrid graft) is one option, but it is controversial. Jacobs et al. reported hamstring graft augmentation with allograft in patients whose graft size was less than 8 mm and were able to achieve a graft size of nearly 10 mm [78]. It reduced the graft failure rate from 28.3 to 11.9%. In contrast, Pennock et al. demonstrated that augmented grafts with a final size 8.9 mm had a re-rupture rate of 30% when compared to 5% re-rupture rate with nonaugmented autografts of 6.4 mm diameter [79]. It is not known if augmentation of autograft with an allograft can lead to an immunologic reaction, graft absorption, or can affect tendon-bone healing. Another technique for increasing graft diameter is by increasing the number of folds to make a 5 or 6 strand hamstring graft, provided adequate graft length is present [80, 81].

For hamstring graft preparation , Anderson recommended both ends of the hamstring tendon graft to be sutured together rather than suturing each end of the graft separately to minimize the size of the graft and therefore the size of the required tunnel [53, 82]. Along with pretensioning of the graft to remove creep prior to implantation, Cruz et al. recommended circumferential compression of the graft using a standard cylindrical sizing block [83]. In a laboratory study using hamstring tendon allograft, the authors were able to decrease the diameter of the graft from mean of 8.28 mm at baseline to 7.38 mm after tensioning and compression. Since the collagen content of the graft does not change, a compressed graft is desirable as it would require a smaller size tunnel for pediatric ACL reconstruction.

For the all-epiphyseal ACL reconstruction, Guzzanti et al. reported their technique of keeping the hamstring grafts attached to the tibia and rerouting the detached proximal end of the graft through an epiphyseal tibial tunnel [49]. This would obviate any fixation in the tibial tunnel but the graft is tensioned across the proximal tibial physis. Anderson described similar metaphyseal fixation distal to the proximal tibial physis using screw and post [53]. For ITB extra-physeal technique, the tibial fixation is performed by suturing the graft to the periosteum on the anterior aspect of tibia and below the level of tibial physis [58]. Patients are subsequently immobilized to allow for healing of the graft. Edwards et al. studied the effect of tensioned (80 N) iliotibial band graft across open growth plates in canine model [84]. They found a substantial rate of distal femoral valgus deformity and proximal tibial varus deformity despite no evidence of a bony bar. Thus, there are concerns with fixation across (not through) the physis, including tethering of the physis, compression on the periochondrial ring, and growth disturbances according to the Heuter-Volkmann principle, but clinical outcomes have not shown any significant growth disturbances [82, 84, 85].

During all-epiphyseal ACL reconstruction, there are concerns about the use of interference screw in the tibial epiphysis, which include tunnel dilatation and injury to either the physis or the articular surface. Interference screw do require a certain minimum tibial tunnel size. Twenty millimeters of minimum tibial tunnel has been suggested [86]. Current methods of tibial tunnel fixation include suspensory fixation, split tibial tunnels with interposed bone bridge (Fig. 11.3), and metaphyseal fixation across the tibial physis [87, 88].

For femoral fixation during all-epiphyseal technique, the most common methods used are suspensory fixation or interference screw [53, 87]. When outside-in epiphyseal tunnels are drilled, the suspensory device may need an extension or washer to prevent the device from slipping into the tunnel [53]. There is possibility of irritation of the hardware by overlying iliotibial band. Kohl et al. used femoral metaphyseal screw as a post to tie the graft in 15 skeletally immature patients and reported one case of progressive valgus [60].

Meller et al. performed transphyseal ACL reconstruction in 4-month-old skeletally immature sheep and found no angular deformity or leg-length discrepancies [89]. The surgery was performed using several key principles of transphyseal technique that are recommended to avoid growth disturbances, i.e., the tibial tuberosity was spared to prevent genu recurvatum, thermal damage to the physis was avoided, small transphyseal drill holes were made in the center of the growth plate, soft tissue graft was used, graft fixation was achieved away from the physis, and the graft was moderately pretensioned before fixation.

Animal studies have shown that for transphyseal ACLR, the prevalence of physeal arrest increases when physeal damage from tunnel drilling involves >7% of the total physeal volume [90]. Kercher et al. performed MRI-based computational study and noted that graft radius (drill size) was the most critical parameter affecting the volume of physeal injury [91]. Based on computer modeling, Shea et al. demonstrated that increasing the drill diameter from 6 to 9 mm would increase the percentage of physeal volume removed from 1.6 to 3.8% for the tibial physis and 2.4 to 5.4% for the femoral physis [92].

Besides the tunnel diameter, the orientation and location of the tunnel is an important determinant of physeal injury. For tibial transphyseal drilling, tunnels that start more medial, distal, and with a steeper angle of inclination reduced the amount of tibial physeal and apophyseal violation compared to tunnels that start more lateral, proximal, and with a shallow angle of inclination [93]. Increasing tunnel drill angle from 45° to 70° would decrease the physeal volume percent injury from 4.1 to 3.1%. The average angle to maintain a distance of 20 mm from the proximal tibial physis was 65° (range, 40–85°) [91].

For femoral transphyseal drilling, tunnels drilled through the anteromedial portal would violate the femoral physis more lateral and more obliquely creating a larger percentage of physeal damage, than those drilled using the trans-tibial technique [94, 95] (Fig. 11.4). Tunnel through the anteromedial portal, however, could be placed more vertical and perpendicular to the physis if desired [69, 96]. Outside-in transphyseal drilling of the femoral tunnel produces more damage to the femoral physis than trans-tibial drilling (4.9% vs. 2.1%), as the outside-in transphyseal tunnels are placed at a more oblique angle (72.8° vs. 32.1°) [51].

Transphyseal ACL reconstruction has been reported to be safe in prepubertal patients (Tanner I and II) [69, 96, 97]. However, most surgeons would prefer to avoid drilling through the physis in this age group. The two widely accepted and safer surgical options are the extraphyseal technique (over-the-top on femur and anterior to tibia without drilling tunnels) and all-epiphyseal ACL reconstruction (tunnels with in femoral and tibial epiphysis) [53, 58]. The extraphyseal technique is safe from physeal injury standpoint as no tunnels are drilled, but the graft is in a nonanatomic position. The all-epiphyseal technique is more anatomic but there is concern about physeal injury. In the absence of comparative clinical study, which technique should be considered?

One of the limiting factors for the all-epiphyseal technique is the size/height of the tibial epiphysis [98]. In order to establish a tibial tunnel, there probably needs to be a threshold under which drilling through tibial epiphysis is neither safe nor feasible. Normative data suggests that the height of the tibial epiphysis is 15–16 mm in adolescents [98, 99]. The maximum oblique depth (along the distal aspect of epiphyseal tibial tunnel) was ~30 mm, occurring at a mean angle of 50° regardless of age or sex [99]. Davis et al. simulated a “safe” tibial tunnel placement which was about 5 mm proximal to the maximum oblique depth; it measured 19 mm in younger patients and 21 mm in older patients (Fig. 11.5). According to Anderson, the more important factor is the height of the epiphysis relative to the size of the tunnel that is drilled; smaller the child, smaller the size of tibial epiphysis but there would be corresponding decrease in the size of harvested hamstring graft which would require smaller tunnel to be drilled [82].

We recommend the extraphyseal technique in the very young child (skeletal age ≤10 years) as the small tibial epiphysis may not permit safe and adequate tibial tunnel placement. For children >10 years age, an all-epiphyseal technique would be used (Fig. 11.6).

It has been recognized that young age is a risk factor for ACL tear and ACL graft rupture after reconstruction [3, 100]. The increased rate of failure of pediatric ACL reconstruction may be secondary to increased activities, decreased compliance, and/or altered neuromuscular control. There has been interest in trying to reduce the rate of ACL failure. In adults, DB ACL reconstruction has shown to better restore anterior and rotational stability compared to single bundle ACL reconstruction, and may help to decrease ACL graft failure rates [101]. This principle could be applied to pediatric ACL reconstruction (Fig. 11.7). Salzmann et al. reported a case of a 14 year old female who underwent DB partial transphyseal ACL reconstruction using hamstring tendons, transphyseal tibial tunnels, epiphyseal femoral tunnels, and interference screw fixation in all four tunnels [102]. Siebold et al. compared 16 skeletally immature patients with DB ACL reconstruction with 17 skeletally immature patients with single-bundle ACL reconstruction [103]. There were no differences in the subjective scores though rotational stability was better restored with DB ACL reconstruction and patients were more satisfied with it. There was a significantly higher rate (25.7%) of graft re-rupture for single-bundle ACL reconstruction, compared to 14.3% re-rupture rate after DB ACL reconstruction. Using computer modeling of knee MRI, Shea et al. recommended a cautious approach during DB transphyseal ACL reconstruction. The combined physeal damage from 9 mm transphyseal anteromedial and posterolateral tunnel placement was 6.5%, which approached the threshold of 7% to prevent growth disturbances [104].