Partial Transphyseal All-Inside Anterior Cruciate Ligament Reconstruction in Skeletally Immature Adolescent Patients: Two-Year Minimum Follow-up Clinical and Radiologic Results

Abstract

Background:

Anterior cruciate ligament (ACL) injuries in pediatric and adolescent populations have seen an uptick due to increased competitive sports participation. Treatment paradigms have shifted from nonoperative management to early reconstruction to prevent meniscal and cartilage damage. However, no consensus exists on the optimal reconstruction technique, particularly in skeletally immature patients.

Purpose/Hypothesis:

The purpose of this study was to evaluate the efficacy and safety of a limited transphyseal all-inside ACL reconstruction (ACLR) technique, hypothesizing that it would yield favorable clinical results without growth disturbances, graft failure, or complications at minimum 2-year follow-up.

Study Design:

Case series; Level of evidence, 4.

Methods:

Consecutive patients aged between 10 and 18 years undergoing ACLR at Sant’Andrea University Hospital of Rome from January 2015 to June 2021 were enrolled. Exclusion criteria were closed physes, previous knee surgeries, and multiligament injuries. The limited transphyseal all-inside technique was used to minimize physeal damage during ACLR by carefully controlling tunnel size, orientation, and depth. Patients underwent standardized follow-ups at 2 and 6 weeks and at 3, 6, 12, and 24 months postoperatively. In addition, all patients were recalled for a final evaluation between February and August 2023. The final assessment included a physical examination (range of motion, pivot shift, Lachman test, Rolimeter-measured laxity, limb-length discrepancy, and deformity) and completion of patient-reported outcome measures (PROMs). Postoperative longstanding anteroposterior radiographs were obtained to evaluate radiographic growth disturbance. A comprehensive magnetic resonance imaging (MRI) assessment was performed by measuring the signal-to-noise quotient (SNQ), tunnel widening, graft healing, and graft maturity.

Results:

Overall, 24 patients met inclusion criteria, with a mean age of 14.2 years. At a mean follow-up of 53.1 months (range, 28-90), the graft rupture rate was 12.5%, and the contralateral ACL rupture rate was 8.3%. Overall, the incidence of growth disturbances was 15.8% in patients who underwent radiologic evaluation. These included cases of angular deformity of >5° or limb-length discrepancy of >10 mm. The mean limb-length discrepancy was 0.31 cm, with no significant differences in limb alignment. MRI analysis revealed a mean tibial tunnel widening of 5.1%, and the mean SNQ was 2.85 ± 0.91 (range, 1.5-4.4). Graft maturity, assessed using the Howell grading system, was classified as grade 1 in 50% of patients and grade 2 in the remaining 50%, while graft healing was graded as 1 in 60% and 2 in 40%; in both, lower grades indicated better results. PROMs showed high scores for Knee injury and Osteoarthritis Outcome Score, International Knee Documentation Committee, and Lysholm scales, with a postoperative median Tegner activity scale score of 7. All patients returned to sports activities.

Conclusion:

Limited transphyseal all-inside ACLR in skeletally immature patients demonstrated promising clinical outcomes and a low rate of significant growth disturbances, suggesting that it is an effective and safe technique for this population.

Anterior cruciate ligament (ACL) injuries in pediatric and adolescent populations have become more common, in concert with the higher level of participation in competitive sports.^2,34 There have been major changes in the management of ACL tears in this population over the past few decades, with a paradigm shift from nonoperative or delayed surgery to earlier ACL reconstruction (ACLR) in younger patients. ⁴⁴ This change is likely multifactorial, and one of the most convincing arguments for early ACLR in young patients is the risk of increased rates of meniscal damage and articular cartilage degeneration after nonoperative management of ACL tears. ⁴⁵ In addition to concerns about secondary injury after nonoperative management, the increase in the frequency of surgical management of pediatric ACL injuries may in part result from the better understanding of the risks associated with surgery. A primary concern with ACLR in pediatric and adolescent patients, compared with adults, is the significantly higher rate of ACL graft rupture, which can reach up to 25%.^13,53 There are also potential growth disturbances; however, the introduction of new physeal preserving techniques may help to reduce the risk of such complications.^11,30,46,55

Among the physeal-sparing techniques, all-inside, all-epiphyseal ACLR with soft tissue autograft has gained interest to minimize the risk to the growth plates while providing a robust ACLR, with excellent clinical results.^11,19,33 However, because the proper position of the tunnel must be verified by fluoroscopy, this technique increases operative time, cost, and radiation exposure. In addition, various technical issues must be considered, especially the fact that the button used for suspensory fixation remains in the joint over the lateral femoral condyle, requires an additional femoral incision, and can put important lateral extra-articular structures, such as the distal femoral physis, lateral collateral ligament, anterolateral ligament, popliteus, and articular cartilage, at risk. ⁵⁰ Several meta-analyses comparing transphyseal and physeal-sparing ACLR techniques found no clear superiority of one technique over the others in incidence of growth disturbances, graft survivorship, and subjective and objective scores.^38,43 Therefore, there is currently no clear guidance regarding the best technique to reconstruct the ACL in skeletally immature patients.

This study evaluates the efficacy and safety, assessed by clinical and radiologic outcomes, of ACLR in skeletally immature patients using an all-inside technique with limited transphyseal tunnel drilling. The hypotheses were that this approach would yield good clinical results at 2-year follow-up and that the use of a limited transphyseal all-inside technique would not result in significant growth failure or secondary complications.

Methods

Institutional review board approval was granted for this study. All participants and their legal guardians provided written informed consent to be enrolled in the study. A retrospective analysis of prospectively collected data of consecutive adolescent patients who underwent ACLR at the Sant’Andrea University Hospital of Rome between January 1, 2015, and June 30, 2021 was performed. The diagnosis of ACL rupture was formulated following clinical examination and magnetic resonance imaging (MRI), and all patients underwent knee radiographs before surgery. Tanner stage ¹⁵ was determined by a pediatrician, and patients with Tanner stage 1, 4, and 5 were excluded from the study, as were patients >18 years. Patients were also excluded if they had closed physes, previous ipsilateral knee surgery, concomitant fractures (except for Segond fractures), or multiligamentous injuries requiring reconstruction in addition to ACLR, except for lesions of the anterolateral ligament.

Surgical Procedures

ACLR was performed using a limited transphyseal approach, all-inside technique with an outside-in femoral and tibial tunnel retrodrilling.

This transphyseal approach has been defined as “limited” given the attempt to minimize the volume of physeal damage. This was based on different assumptions:

• - Avoiding empty tunnels: using a retrodrill, which could be adjusted to the desired diameter and used for a controlled depth, ensured no empty drill holes were left, reducing the risk of bone bridge formation and growth plate injury. ⁴⁹

• - Graft diameter: the graft diameter significantly affected physeal injury volume; hence, efforts were made to limit it to 8 mm, except in larger boys. ²⁸

• - Tunnel orientation: the femoral tunnel was positioned as perpendicular to the physis as possible, using a maximum 120° angle from the out-in guide, to reduce injury volume. ²⁸

• - Tibial tunnel length: a 15-mm tibial socket was produced to maintain safe distance from the tibial physis, considering the distance of 20 mm from the proximal tibial physis to tibial ACL footprint in early and middle adolescent patients. ²⁸

• - Femoral tunnel: based on the inclination in the 3 planes, the mean distance from the femoral footprint of the ACL to the distal femoral physeal plate ranged widely. ³¹ Therefore, it was not possible to standardize the technique for all patients. Strategies to minimize physeal damage included producing a 15-mm bone socket, setting the drill guide vertically, filling the tunnel with a soft tissue graft, and using appropriately sized grafts.

• - Graft tension control: to prevent tenoepiphysiodesis as per the Hueter-Volkmann principle, double cortical adjustable fixation was used, allowing for controlled graft tension during arthroscopy and avoiding excessive tension.^14,37,49

A single-bundle technique with quadrupled semitendinosus tendon autograft was used in all patients. The semitendinosus tendon was harvested and quadrupled to achieve a final graft length of 60 mm. The graft was prepared with 2 TightRope-RT adjustable loop cortical suspensory devices (Arthrex) and a high-strength, multistrand, long-chain, ultra–high molecular weight polyethylene core suture (No. 0 FiberWire; Arthrex) on each side. The femoral tunnel entry point was located in the center of the anatomic femoral footprint of the ACL, midway between the resident ridge and above the superior position, with the guide set at 120°. A tibial tunnel was drilled using a standard 65° tibial guide. Femoral and tibial sockets were drilled to a depth of 15 or 20 mm, depending on preoperative planning, using a retrodrill (Flipcutter II; Arthrex) in an outside-in fashion at low speed to avoid thermal damage to the physis. The graft was then introduced through the anteromedial portal using a shuttle suture on both sides and fixed first on the femoral side, then on the tibial side with the “flip-then-fill technique.” Both femoral and tibial fixation were performed with cortical suspensory devices.

Patients with high-grade pivot shift (≥2+) or who were willing to return to high-risk sports (including pivot and/or contact sports) received an additional lateral extra-articular procedure. When a lateral extra-articular tenodesis (LET) was performed, it was using the Arnold-Coker modification of the MacIntosh technique. A 10- to 12-cm incision was made from the lateral femoral condyle to the Gerdy tubercle, and a 1 cm–wide, 13 cm–long strip of the iliotibial band was harvested, preserving its distal attachment. The strip was passed under the fibular collateral ligament, reflected and sutured under tension with the knee flexed at 90°. This permits the performance of LET without additional bone tunnel and risks of physeal damage.

Rehabilitation

Immediately after surgery, patients were allowed to bear weight on crutches, as tolerated. If a concomitant meniscal repair had been performed, patients remained nonweightbearing for 2 weeks. All patients wore a knee extension brace for 4 weeks. During the first 2 weeks, the brace was removed only to perform physical therapy, which began on the second postoperative day. At 2 weeks postoperatively, the brace was adjusted to allow a range of motion of 0° to 90°. The early focus of rehabilitation was to regain full range of motion and full weightbearing within 4 weeks. Beginning in the second postoperative month, a more intensive muscle strengthening program was prescribed. Patients began gradual noncontact sports activities and sport-specific training at 3 months. Return to full sports activities was allowed after completion of rehabilitation, between 9 and 12 months.

Follow-up

Patients were evaluated at 2 weeks, 6 weeks, 3 months, 6 months, 12 months, and 24 months postoperatively. All patients were then invited to undergo a clinical and radiologic evaluation for final follow-up between February 2023 and August 2023.

The final outpatient evaluation was standardized and included a physical examination performed by an attending surgeon (E.M.) to assess range of motion, pivot-shift grade, Lachman test, side-to-side anteroposterior (AP) laxity difference using a Rolimeter (Aircast), and clinical assessment of limb-length discrepancy and deformity. Longstanding radiographs and MRIs were also performed as part of the final follow-up. The patients were scheduled for the same day, before clinical examination, lower extremity scanograms, and knee MRI. Patients were asked to complete patient-reported outcome measures (PROMs), including the Knee injury and Osteoarthritis Outcome Score (KOOS), International Knee Documentation Committee (IKDC), and Lysholm Knee Scale questionnaires.^24,47,52

Any complications, reinjury, reoperation, or symptoms of instability, effusion, or pain were recorded. A review of medical records was used to extract data on demographics, Tanner grade at the time of surgery, and level of preoperative sports participation.

Endpoints

Graft Failure

A diagnosis of graft failure was defined by the following criteria and/or MRI confirmation: side-to-side maximal manual difference of >5 mm using the Rolimeter arthrometer and/or a pivot-shift grade ≥2+.

Postoperative Longstanding AP View Radiograph Examination

Bilateral limb-length and alignment measurements were performed independently by a sports medicine fellow (A.C.) and a fellowship-trained musculoskeletal radiologist (N.C.B.) on full-length AP radiographs. Analyses were performed on the mean values of the measurements taken by the 2 observers^9,48 (Figure 1).

Figure 1.

Illustration of the measurements performed on radiograph. MAD, mechanical axis deviation; LDFA, mechanical lateral distal femoral angle; MPTA, medial proximal tibial angle.

• - Limb length was measured as the distance between the center of the femoral head (considered the center of the best-fitting circle drawn on the head) and the midpoint of the talar trochlea at its superior contour.

• - The mechanical lateral distal femoral angle (LDFA) was defined as the angle between the femoral mechanical axis and the articular surface of the distal femur (line tangent to the distal femoral condyles).

• - The medial proximal tibial angle (MPTA) was defined as the angle between the tibial mechanical axis and the articular surface of the proximal tibia (line tangent to the tibial plateau).

• - Mechanical axis deviation (MAD) was measured as the perpendicular distance between the mechanical axis of the limb (a line from the center of the femoral head to the center of the talar dome) and the center of the intercondylar notch. Varus axes have medial locations, indicated by positive numbers, and valgus axes have lateral locations, indicated by negative numbers. Radiographic growth disturbance was defined as a 10-mm leg-length discrepancy, a 10-mm MAD, or a 5° difference in LDFA or MPTA compared with the nonoperative side.^3,9 Angular deformity was also defined by MAD, LDFA, or MPTA outside the established range of reference values (MAD, 1-15 mm from the weightbearing line; LDFA, 85°-90°; and MPTA, 85°-90°^39,48).

Postoperative MRI Examination

MRI scans were performed on a 1.5-Tesla device (Sonata Maestro; gradient 40 mT; Syngo A35 software; Siemens). Three-dimensional proton density–weighted turbo spin-echo (PD-TSE) and sagittal PD–weighted fat-suppressed (PD-FS) sequences were used to evaluate the primary endpoints. Each MRI was reviewed to assess 4 imaging features to evaluate ligamentous and bone-tendon healing. All the analyses were conducted using Horos medical image viewer for MacOS (Version 3.3.0).

Signal-to-Noise Quotient

The signal-to-noise ratio (SNQ) is an indicator of the mechanical strength of the ligament, with lower values indicating better results, and a lower SNQ indicates a signal more comparable with that of a healthy posterior cruciate ligament (PCL), which is used as a reference in the calculation.^{1,21,25,36,54}

Circular regions of interest of 0.05 cm² on oblique sagittal PD-FS slices were used for MRI analysis. Graft signal intensity was assessed intra-articularly at 3 locations (proximal, mid-, and distal) and averaged to be used for analysis. Background signal was measured 2 cm anterior to the patellar tendon.

The SNQ was then calculated according to the calculation stated by Weiler et al ⁵⁴ as follows:

$$\mathrm{SNQ}=\frac{\mathrm{Graft}\mathrm{signal}-\mathrm{PCL}\mathrm{signal}}{\mathrm{Background}\mathrm{signal}}$$

Tunnel Widening

The cross-sectional areas (CSAs; in cm²) of the superior portion of the tibial bone tunnel were measured on 3-dimensional PD-TSE sequences (Figure 2).

Figure 2.

Example magnetic resonance imaging (MRI) of tunnel-widening calculations 12 months after anterior cruciate ligament reconstruction. The mean area at each tibial tunnel entrance was measured on oblique MRI perpendicular to the tunnel cross section.

Tibial and femoral tunnel CSA changes after ACLR were calculated using the following formula ¹⁸ :

$$\mathrm{CSA}\mathrm{increase}(\%)=\frac{\mathrm{Measured}\mathrm{CSA}-\mathrm{Drilled}\mathrm{CSA}}{\mathrm{Drilled}\mathrm{CSA}}\times 100$$

Graft Healing

Graft healing was defined on PD-FS sequences in sagittal oblique images. On the basis of the signal intensity at the bone-graft interface, patients were assigned a grade of 1 of 3 ²⁰ :

1. Low intensity, no fibrosis at the bone-graft interface, complete attachment.

2. High intensity over a portion of the interface.

3. High intensity over the entire bone-graft interface, poor attachment.

Graft Maturity

According to Howell et al, ²⁵ a 4-grade system based on the MRI signal of the graft in the tibial tunnel was used to determine graft maturity:

1. Homogeneous, low-intensity signal indistinguishable from the PCL and patellar tendon.

2. Normal ligament signal over ≥50% of its volume, interspersed with areas of increased signal intensity.

3. Increased signal intensity over ≥50% of its volume intermingled with areas of standard ligament signal.

4. Diffuse increased signal intensity without strands with normal ligament appearance.

Statistical Analysis

Analyses were performed using SPSS Statistics software (Version 28.0.0.1; IBM SPSS). Statistical significance was set at P < .05. Descriptive statistics were reported according to the type of variable. For continuous variables, we reported the number of observations, mean, standard deviation, median, and minimum and maximum values. Categorical variables were presented as frequencies and percentages.

Normality of continuous variables was assessed using the Kolmogorov-Smirnov test. According to distribution, either chi-square or Fisher exact test was used to compare categorical variables.

For comparisons of continuous variables between independent groups, we used the Student t test for normally distributed data and the Mann-Whitney U test for nonnormally distributed data.

For paired comparisons of continuous variables (eg, limb-length discrepancy, LDFA, MPTA, or MAD between the operated and contralateral limbs), we used the paired t test for normally distributed data and the Wilcoxon signed-rank test for nonnormally distributed data.

Survival analysis for graft failure was assessed using the Kaplan-Meier method, and survival distributions were compared using the log-rank test.

PROMs were analyzed by calculating mean scores and standard deviations. The proportion of patients achieving Patient Acceptable Symptom State (PASS) for each PROM was determined using previously published thresholds, according to Muller et al ³⁵ : 75 for the IKDC Subjective Knee Form, 88.9 for the KOOS Pain, 57.1 for the KOOS Symptoms, 100.0 for the KOOS Activities of Daily Living, 75.0 for the KOOS Sport and Recreation, and 62.5 for the KOOS Quality of Life subscales.

Results

Patients and Clinical Characteristics

Overall, 28 skeletally immature patients underwent ACLR with limited transphyseal, all-inside technique during the study period. The inclusion criteria were met by 24 patients, and none were lost to follow-up. The final study population, therefore, comprised 24 patients who underwent clinical examination at a mean follow-up of 53.1 ± 21.1 (range, 28-90) months.

The study flowchart is presented in Figure 3.

Figure 3.

The study flowchart in line with the Strengthening the Reporting of Observational Studies in Epidemiology statement (http://www.strobe-statement.org). ACL, anterior cruciate ligament; ACLR, ACL reconstruction.

The mean age of the overall population was 14.2 ± 1.2 years, and 20 (83.3%) were male (Table 1).

Table 1.

Patient Demographics ^a

	All Patients (N = 24)
Age, y, mean ± SD (range)	14.2 ± 1.2 (12.9-16.5)
Tanner stage, n (%)
3	9 (37.5)
4	15 (62.5)
Sex, n (%)
Male	20 (83.3)
Female	4 (16.7)
Side, n (%)
Dominant	13 (54.2)
Nondominant	11 (45.8)
Time from injury to surgery, d, mean ± SD (range)	199.8 ± 201 (70-1497)
Tegner activity scale, median (range)	7 (4-10)

^aACLR, anterior cruciate ligament reconstruction.

All patients underwent the same intra-articular ACLR; 20 of these 24 patients also received LET because of additional risk factors for recurrent ACL graft rupture. Intraoperative characteristics are described in Table 2.

Table 2.

Intraoperative Findings and Surgical Characteristics ^a

	All Patients (N = 24)
Pivot-shift grade, n (%)
0 or 1+	7 (29.2)
2+or 3+	17 (70.8)
Preoperative AP laxity, mm, mean ± SD (range)	6.2 ± 1.7 (3.4-9)
Meniscal injuries, n (%)
None	20 (83.3)
Medial meniscus	3 (12.5)
Lateral meniscus	1 (4.2)
Graft diameter, mm, mean ± SD (range)	8.6 ± 0.6 (8-10)
Additional LET, n (%)	20 (83.3)

^aAP, anteroposterior; LET, lateral extra-articular tenodesis. Only patients with the presence of ≥1 of the following specific additional risk factors for graft rupture (pivot-shift grade 2 or 3, high level of sporting activity defined as Tegner activity scale ≥7, participation in pivoting sports, and those with Segond fractures) were offered a LET in addition to anterior cruciate ligament reconstruction. This was performed using the Arnold-Coker modification of the MacIntosh procedure. ³⁵

ACL Graft Ruptures and Contralateral Injuries

At a mean follow-up of 53.1 months, the overall graft failure rate was 12.5% (3 of 24 patients). One patient sustained atraumatic ACL graft failure and underwent a revision ACLR with patellar tendon 7 months after the initial surgery. The second patient who experienced ACL graft failure resumed soccer 12 months after the index surgery. Two years after the index ACLR, he experienced ACL graft rupture with direct trauma to his knee during a soccer game. He underwent ACLR with quadriceps tendon autograft and anterolateral ligament reconstruction.

The third patient had resumed his sport, rugby, and 4 years after the index surgery had a traumatic failure after a tackle resulting in ACL graft failure and grade 3 medial collateral ligament injury. The patient underwent ACLR revision with quadriceps tendon autograft, medial collateral ligament repair, and LET.

The contralateral ACL rupture rate was 8.3% (2 patients). One patient resumed her sport (volleyball) and sustained a noncontact contralateral ACL rupture after 4 years. The other patient resumed competitive basketball and sustained a contralateral ACL rupture 2 years after the first rupture.

Kaplan-Meier cumulative survivorship is reported in Figure 4.

Figure 4.

Kaplan-Meier survival plot demonstrating graft survivorship. x-axis, follow-up time in months; y-axis, cumulative survival.

Knee Laxity and PROMS

The mean side-to-side difference was 1.5 ± 0.43 mm. Two patients postoperatively demonstrated a positive pivot shift, classified as low grade (1+).

At the last follow-up, the mean KOOS score was 96 ± 3.8, the mean IKDC score was 91.5 ± 7.6, and the mean Lysholm score was 93.9 ± 6.3. Figure 5 shows the percentage of patients who achieved PASS for the KOOS subscales and IKDC. Patients with ACL graft rupture were excluded from PROM collection.

Figure 5.

Percentage (blue) of patients who achieved the Patient Acceptable Symptom State according to Muller et al ³⁵ and mean values (orange) of Knee injury and Osteoarthritis Outcome Score (KOOS) subscales and International Knee Documentation Committee (IKDC).

Return to Sport

All patients returned to sport. The postoperative median Tegner activity scale was 7 (4-10). Twenty patients (83.3%) maintained the same or reached a higher sports level, while 4 (16.7%) resumed sports but at a lower level than before the injury. Two patients reached the level of professional soccer player, 1 at international level.

Surgical Complications

One patient developed an infection and underwent arthroscopic washout, and no other patients required subsequent non–ACL related surgery (eg, meniscectomy). No patient experienced complications such as knee stiffness or arthrofibrosis. One patient complained of tibial button discomfort and 1 patient complained of knee squeaking, an uncommon complication consisting of a rubbing noise caused by friction between the articular surface and high-strength wires. Both chose not to undergo hardware removal.

Radiographic Outcomes

At the final follow-up, 19 patients completed the radiographic evaluation, which consisted of longstanding AP view radiographs. The mean difference in length of the operated limb compared with the healthy limb was 0.31 cm ± 0.58 cm. The difference in MPTA was −0.05°± 1.68°, in LDFA 1.8°± 2.29°, and in MAD 5.5 mm ± 8.96 mm. Significant changes in LDFA and MPTA occurred in 10.5% and 0%, but none of the mean comparisons showed statistical significant differences in the analyzed characteristics of the operated limb compared with the unaffected limb. When paired sample analyses were performed, the LDFA showed a significant decrease in the operated knees (mean 87.83° and 86.03° in unoperated and operated knees, respectively; mean difference, 1.80°; P = .035).

Two patients developed an angular deformity of >5°, with an increase in MAD, and one patient reported a limb-length discrepancy of >10 mm. One patient exhibited a 5.6° difference in LDFA (operated vs nonoperated limb, 82.6° vs 88.2°), resulting in genu valgum (MAD, 26.1 mm). Another patient had a 6° decrease in LDFA resulting in mild genu valgum (MAD, 9.2 mm). One patient had a limb-length discrepancy of 1.4 cm with no significant change in angle or MAD.

The mean values, mean differences, ranges, and comparisons of these different outcomes are represented in Table 3.

Table 3.

Description and Comparisons of Radiographic Outcomes ^a

		Mean	SD	Minimum	Maximum	P
Limb length, cm	Healthy	90.79	4.16	84.10	96.35	.87
	ACL-reconstructed	90.48	4.34	84.10	96.37
	Intrapatient difference	0.31	0.58	−0.6	1.4	.13
MPTA, deg	Healthy	87.39	2.13	84.05	90.80	.96
	ACL-reconstructed	87.44	2.37	83.25	90.40
	Intrapatient difference	−0.05	1.68	−2.49	3.37	.93
LDFA, deg	Healthy	87.83	2.16	84.76	91.40	.12
	ACL-reconstructed	86.03	2.68	81.15	89.30
	Intrapatient difference	1.80	2.29	−0.46	6.02	.035
MAD, mm	Healthy	10.27	6.45	2.98	24.30	.15
	ACL-reconstructed	4.76	9.45	−18.40	16.20
	Intrapatient difference	5.50	8.96	−4.6	26.21	.08

^aACL, anterior cruciate ligament; HKA, hip-knee-ankle angle; LDFA, lateral distal femoral angle; MPTA, medial proximal tibial angle. Two statistical comparisons were made for each radiographic parameter. The first P value compares the mean values of the operated and contralateral limbs across all patients (interpatient analysis). The second P value assesses the mean intrapatient difference (operated versus nonoperated limb within the same individual). Boldface indicates statistical significance (P < .05).

MRI Evaluation

Patients who experienced graft failure were excluded from the MRI evaluations, and 1 patient did not have an MRI at the 1-year follow-up visit. A total of 20 patient MRIs were therefore evaluated. At 1 year, the mean SNQ was 2.85 ± 0.91, ranging from 1.5 to 4.4. Graft maturity was graded 1 in 50% of patients and 2 in the remaining 50%. Graft healing was graded 1 in 60% of patients and 2 in 40% of patients, with a mean of 1.4 ± 0.5.

The mean tibial tunnel widening, measured with the CSA, was 5.06% ± 44.6%, ranging from −52% to +69%.

Discussion

The main finding of the present study was that limited transphyseal all-inside ACLR in skeletally immature adolescent patients results in favorable clinical outcomes without clinically significant growth disturbances.

Several surgical ACLR techniques have been described and, given the complex anatomy of the pediatric knee, each technique has its own risks and complications; all can potentially cause secondary growth abnormalities.^10,17,32 The technique described in the present investigation resulted in limb-length discrepancy in 1 of 19 patients (5.3%) who underwent longstanding AP radiographs, and significant MAD in 5.3% of cases (1 of 19). Significant changes in LDFA and MPTA occurred in 10.5% and 0%, respectively. Overall, 15.8% of patients who underwent radiographic evaluation evidenced growth disturbances. However, all of these changes were small, and no clinically relevant growth disturbance was reported.

Although 10 mm was chosen as the threshold for limb-length discrepancy after ACLR in this and other studies, up to 25% of normal asymptomatic individuals have a limb-length discrepancy of ≥10 mm. ²⁹ Limb-length discrepancies are not clinically relevant until the difference reaches 20 or 25 mm, including studies of gait, low back pain, and hip and knee osteoarthritis.^23,29,51,55

Also, side-to-side differences in alignment, albeit moderate, are not uncommon in the general population.^26,27,55 The patients with limb-length discrepancies and angular deformities in this case series were asymptomatic, as reported in the vast majority of patients from different systematic reviews studying complications after ACLR in pediatric patients, raising the question of the clinical relevance of the growth disturbances identified at imaging.^5,10,55

While consensus has been reached on the surgical management of this type of lesion, there is still no clear evidence as to which reconstruction technique provides the best functional results. ² The technique used in this study is transphyseal and tries to respect the open physes as much as possible, with the goal to reduce the volume of the drilled physes, minimizing thermal damage, filling the sockets with soft tissue autograft, and not placing screws inside the tunnels. This technique was chosen because the various physeal-sparing ACLR techniques do not fully address the issue of induced growth discrepancies, which can occur with any reconstruction technique. Collins et al ¹⁰ conducted a systematic review on 12 studies reporting on 39 patients with growth anomalies: the most common angular deformity was genu valgum (81%; n = 13; mean, 6.5°). Physeal-sparing techniques were used in 25% of patients with angular deformity and 47% of cases of limb-length discrepancy: this technique is supposed to reduce the risk of growth disturbance by not violating the physis. Another systematic review demonstrated similar incidence rates of limb-length discrepancy and angular deformity in transphyseal and physeal-sparing cohorts. ⁴³ The incidence of growth disturbances found in the current study does not appear to be different from that reported in studies describing clinical and imaging outcomes of ACLR using a hybrid approach, with a physeal-sparing technique in the femur and the transphyseal technique in the tibia. In fact, using radiographic analysis and thresholds similar to those used in the current study, Chambers et al ⁹ reported a 20.8% incidence of growth disturbances, most of which consisted of limb shortening and lateral translation of the MAD. Faunø et al, ¹⁶ using the same technique, also reported an incidence of growth disturbances of 18%. Half of these resulted from limb shortening and half from increased valgus angulation.

The risk of subsequent injuries is another important issue associated with ACLR in pediatric and adolescent patients. Indeed, a second ACL injury, due to either a retear of the ACL graft or an injury to the contralateral knee, occurs in up to 35% of children and adolescents.^12,40 Five patients, 3 ipsilateral and 2 contralateral, experienced a second ACL injury in the current study. Of the three patients who reinjured the same knee, 1 patient sustained early ACLR failure 7 months after surgery and without significant trauma. This patient had undergone an isolated intra-articular ACLR. In contrast, the other 2 patients experienced recurrences after sports-related knee trauma. In addition to ACLR reconstruction, these 2 patients also underwent LET. These good results probably arise from both the anatomic positioning of the tunnels using this technique and the addition of a LET in most patients. The latter has been shown to provide significant benefits with regard to ACL graft ruptures in pediatric and adolescent patients.^6,22,34,42

In the current study, MRI 1 year after ACLR showed mean SNQ values of 2.85 ± 0.91. Using the same graft in skeletally mature population, at 1 year follow-up Cavaignac et al ⁷ reported an SNQ of 5.9 ± 3.7. However, there is a wide range of SNQ values, and, in patients with open physes, native ACL and ACL graft appear to have different mean values for Howell grade and SNQ. Pauvert et al ⁴¹ reported a mean SNQ of 3.87 for native ACL and 9.2 for soft tissue autograft at 1 year after ACLR in a pediatric population. Also, higher signal intensity on contrast-enhanced MRI corresponds to lower mechanical strength of the graft; therefore, the SNQ is inversely proportional to the tensile strength of the graft. ⁵⁴

The Howell score has also been used to provide an indication of ACL graft maturation. Pauvert et al ⁴¹ reported that in a pediatric population 1 year after ACLR, only 3% of grafts were graded Howell 1, 54.5% were graded 2, 36.4% were graded 3, and 6.1% were graded 4. In the current population, 50% were graded as Howell 1, 50% as 2, and none as 3 or 4.

An additional parameter examined was the incorporation of the ACL graft and its attachment to the bone within the tibial tunnel. ²⁰ In the current study, when analyzing graft healing, 60% were graded as 1, 40% as 2, and none as 3 or 4, with a mean of 1.4 ± 0.5. Cavaignac et al ⁸ reported a mean graft maturity of 2.0 ± 0.6 in patients who received a quadruple semitendinosus tendon autograft and a mean graft maturity of 1.7 ± 0.6 in patients who also received a LET.

The presence of an additional LET in most patients probably contributed to the good results in the different parameters of graft maturation and integration on MRI investigated in the current study. In a previous study, the MRI appearance of quadrupled semitendinosus tendon ACL grafts at 1 year postoperatively showed better incorporation and maturation when combined with LET. ⁸

In the current cohort, the mean tunnel enlargement was 5.06% ± 44.6%, ranging from −52% to +69%. This is consistent with a previous report of low levels of tunnel enlargement using 4-strand semitendinosus tendon grafts with adjustable cortical suspensory fixation for ACLR. Biset et al ⁴ recently reported a mean of 13% of tunnel widening in 149 patients receiving an all-inside ACLR with semitendinosus tendon autograft.

Limitations

The present study has several limitations. First, there was no comparison group. Also, it is a retrospective study design without the recording of preoperative variables further studied as PROMs or baseline radiographic characteristics on alignment or limb-length comparisons. Preoperative radiographs were not used to determine bone age, but instead a Tanner stage assessment was performed.

The small sample size did not allow for subgroup analysis, which affected the generalizability of the findings.

In addition, some patients did not undergo radiological follow-up but were asymptomatic and had no clinical signs of deformity. Finally, the PASS was based on previously published data rather than being derived from the cohort studied. ³⁴

Conclusion

The limited transphyseal technique used in this study demonstrated good functional outcomes and acceptably low graft failure and reoperation rates. Although angular deformities and limb-length discrepancies occurred in 15.8% of patients, the severity of growth disturbance was mild and asymptomatic, and no patients required guided growth intervention or corrective osteotomy.