Long-Term Outcomes of the Sliding-Growing Rod Technique in Early-Onset Scoliosis: Deformity Control, Spinal Growth, and Implant Survival

Abstract

Study Design

Retrospective study.

Objectives

To evaluate the long-term outcomes of sliding-growing rod technique (SGRT) in early-onset scoliosis and assess rod translation through the domino connector as a radiographic marker of preserved sliding function during postoperative spinal growth.

Methods

Consecutive patients who underwent primary SGRT between 2012 and 2019 were reviewed. Inclusion required at least 5 years of follow-up and complete radiographic and revision data. Major Cobb angle, T1–S1 and T1–T12 lengths, and rod translation were measured serially. Correction-related length gain was analyzed separately from subsequent postoperative growth. Correlation analyses assessed the relationship between rod translation and T1–S1 growth, and revision-free implant survival was estimated using Kaplan–Meier analysis.

Results

Twenty-three patients were followed for a mean of 93.9 ± 16.8 months. The major Cobb angle improved from 78.9 ± 12.3° preoperatively to 19.3 ± 7.0° early postoperatively and was 24.6 ± 7.4° at final follow-up. Postoperative longitudinal length gain was 106.0 mm for T1–S1 and 72.9 mm for T1–T12, with annualized postoperative radiographic length-gain rates of 13.9 and 9.6 mm/year, respectively. Final rod translation was 56.4 ± 7.8 mm and correlated with postoperative T1–S1 growth (r = 0.73, P < .001). Four patients underwent unplanned revision. Revision-free implant survival was 95.7% at 2 years, 91.3% at 5 years, and 82.6% at 6 years. The mean final EOSQ-24 score was 79.9 ± 11.3.

Conclusions

SGRT treatment was associated with durable correction, continued postoperative spinal length gain, and acceptable revision-free implant survival without planned lengthening procedures. Serial rod translation may provide a reproducible radiographic marker of preserved sliding function during follow-up.

Introduction

Early-onset scoliosis (EOS) is defined as a spinal deformity diagnosed before 10 years of age and includes a heterogeneous spectrum of idiopathic, congenital, neuromuscular, and syndromic etiologies.^1,2 Because EOS occurs during a critical period of spinal, thoracic, and pulmonary development, progressive deformity may impair thoracic growth, respiratory function, sitting balance, and health-related quality of life. ³ The main goal of treatment is therefore to control curve progression while preserving spinal and thoracic growth and avoiding early definitive fusion whenever possible. ⁴

Growth-friendly surgical techniques were developed to address this balance between deformity correction and growth preservation.^4,5 Traditional growing rods can provide meaningful curve correction and spinal length gain, but repeated surgical lengthenings are associated with cumulative anesthetic exposure, wound morbidity, implant-related complications, anchor failure, and unplanned reoperations. ⁶ In addition, length gain may progressively decrease with repeated distractions, a phenomenon described as the “law of diminishing returns”. ⁷ Magnetically controlled growing rods reduce the need for repeated open lengthening procedures, but implant-related complications and revisions remain clinically relevant.^8,9

Growth-guidance and sliding constructs have been developed as alternatives to traditional distraction-based strategies, with the aim of maintaining deformity control while preserving spinal growth.^10,11 The bilateral sliding-growing rod technique (SGRT) used in the present study permits longitudinal rod translation through domino connectors, theoretically accommodating growth while maintaining deformity control. In this setting, rod translation through the connector may serve as a radiographic marker of continued construct motion and growth accommodation. However, long-term evidence specifically addressing sliding-growing constructs remains limited.^9,11

In contemporary EOS research, radiographic outcomes alone are insufficient to define treatment success. Implant survival, unplanned revision, functional status, and caregiver-reported quality of life are increasingly recognized as essential outcome measures.^12,13 The EOSQ-24 is a validated disease-specific instrument for evaluating health-related quality of life in children with EOS and their caregivers. ¹²

The purpose of this study was to evaluate the long-term radiographic, clinical, functional, and implant-related outcomes of SGRT in patients with EOS. We specifically assessed deformity correction, spinal and thoracic growth, rod translation through the domino connector, complications, revision-free implant survival, and patient-reported outcomes. We hypothesized that SGRT would provide sustained deformity correction and continued spinal growth with acceptable long-term implant survival, and that serial rod translation through the domino connector would remain measurable during follow-up as an indicator of preserved sliding function.

Materials and Methods

Study Design and Setting

This retrospective, single-center longitudinal cohort study evaluated the long-term clinical, radiographic, functional, and implant-related outcomes of the sliding-growing rod technique in patients with early-onset scoliosis treated between January 2012 and December 2019. Early-onset scoliosis was defined as a spinal deformity diagnosed before 10 years of age. The study protocol was approved by the institutional non-interventional clinical research ethics committee (approval details blinded for peer review) and was conducted in accordance with the Declaration of Helsinki. The study was reported in accordance with the STROBE guidelines for observational cohort studies.

Patient Selection

Eligibility criteria were diagnosis of EOS before 10 years of age, SGRT as the primary growth-friendly procedure, availability of preoperative, early postoperative, and final whole-spine radiographs, minimum follow-up of 5 years, and complete documentation of perioperative variables, complications, and revisions. Exclusion criteria were previous spinal instrumentation or definitive fusion, inadequate radiographic documentation, follow-up shorter than 5 years, or missing data precluding assessment of the primary outcome. Of 31 consecutive patients, 8 were excluded because of inadequate radiographic documentation (n = 4), previous spinal surgery (n = 2), or follow-up shorter than 5 years (n = 2), leaving 23 patients for analysis. For patients excluded because of incomplete radiographic documentation or follow-up shorter than 5 years, the available clinical records were reviewed separately to determine the duration of documented follow-up, implant-related complications, unplanned revisions, and last known construct status. Neither patient excluded because of follow-up shorter than 5 years had undergone an unplanned revision before the last available evaluation.

Surgical Indications

SGRT was considered in skeletally immature patients younger than 10 years with severe or progressive EOS when at least 2 high-risk features were present: a major Cobb angle typically approaching or exceeding 50°, documented progression despite casting or bracing, clinical or radiographic concern for impaired thoracic development, or failure of conservative treatment in rigid congenital, neuromuscular, or rapidly progressive deformities. ¹⁴ Surgical decision-making was individualized according to remaining growth potential, curve flexibility, sagittal deformity, pelvic obliquity, functional balance, and the feasibility of stable proximal and distal fixation. In infantile idiopathic scoliosis, increased rib–vertebral angle difference and phase-2 rib relationships were considered supportive markers of progression rather than universal surgical criteria. ¹⁵

Surgical Technique

All procedures were performed under general anesthesia with the patient positioned prone on a radiolucent operating table. A single preincision dose of a first-generation cephalosporin was administered. Posterior subperiosteal exposure was confined to the vertebral levels selected for pedicle screw insertion to minimize soft-tissue devascularization and unnecessary disruption of the posterior elements and facet joints.

Bilateral instrumentation was performed using polyaxial pedicle screws and 5-mm Ti6Al4V-ELI titanium-alloy rods from the Karspine® Spinal Stabilisation System (Karmed Sağlık Group, Istanbul, Türkiye). Standard side-to-side (“domino”) connectors compatible with the same system were used; the construct did not incorporate a purpose-built low-friction gliding or growth-guidance bearing. The proximal foundation extended to T2–T3 to improve control of the long-segment deformity and provide stable upper fixation. The distal instrumented level was selected according to curve pattern, underlying etiology, pelvic obliquity, and the quality of available distal fixation. Iliac fixation was added when additional distal stability was required, particularly in patients with inadequate lumbosacral support or clinically relevant pelvic imbalance.

Pedicle screws were inserted bilaterally under intermittent anteroposterior and lateral fluoroscopic guidance. After preparation of the pedicle tract, its integrity was confirmed with a ball-tipped feeler, and screws were inserted to the planned diameter and depth. Pulsed low-dose fluoroscopy was used to limit radiation exposure.

The rods were contoured according to the intended coronal and sagittal correction. Gentle, nonaggressive reduction maneuvers were applied. After distal rod seating, the set screws were sequentially tightened in a distal-to-proximal direction to translate the spine toward the rods and obtain the desired alignment. During the initial correction phase, all screw heads were rigidly secured to the rods.

After correction had been achieved, rigid fixation was maintained at the 2 most proximal vertebral levels and at the distal terminal fixation segments to create stable proximal and distal foundations. At the intervening vertebral levels, the set screws were intentionally backed off sufficiently to permit longitudinal rod translation while maintaining screw–rod engagement; they were not completely removed. Thus, the intermediate screw heads functioned as guided sliding points rather than rigid fixation points.

Domino connectors were incorporated bilaterally. The distal portion of each connector was rigidly secured to the distal fixation foundation, whereas the rod segment extending toward the thoracic region was allowed to translate longitudinally through the connector. The proximal rod segments were intentionally left sufficiently long beyond the sliding interface to accommodate anticipated longitudinal growth. This configuration allowed continued longitudinal translation between the stable terminal foundations while maintaining coronal and sagittal deformity control. Figure 1

Figure 1.

Surgical configuration and early radiographic appearance of the bilateral sliding-growing rod technique

No kyphectomy, corpectomy, vertebral column resection, or deformity-level osteotomy was performed. No intentional fusion or bone grafting was undertaken along the intervening mobile spinal segments. A Hemovac drain was placed before standard layered wound closure.

Routine planned lengthening procedures were not performed. Subsequent surgery was reserved for rod exhaustion or fracture, anchor loosening or pull-out, loss of fixation, clinically relevant deformity progression associated with construct dysfunction, infection, or another implant-related complication requiring operative management.

Radiographic Assessment

Standardized whole-spine posteroanterior and lateral radiographs were obtained preoperatively, within 7 days after surgery, at approximately 2 and 5 years, and at final follow-up. Coronal measurements included major Cobb angle, coronal balance, shoulder imbalance, pelvic obliquity, T1–S1 length, T1–T12 length, and rod translation through the domino connector. Sagittal-plane assessment included T5–T12 thoracic kyphosis, L1–S1 lumbar lordosis, C7–S1 sagittal vertical axis, and proximal junctional angle. Thoracic kyphosis was measured from the superior endplate of T5 to the inferior endplate of T12, and lumbar lordosis from the superior endplate of L1 to the superior endplate of S1. Sagittal vertical axis was defined as the horizontal distance between the C7 plumb line and the posterosuperior corner of S1. The proximal junctional angle was measured between the caudal endplate of the upper instrumented vertebra and the cephalad endplate of the vertebra two levels above. Radiographic proximal junctional kyphosis was defined as a proximal junctional angle ≥10° together with an increase of ≥10° from the preoperative value. ¹⁶

Linear measurements were calibrated using a radiographic marker or a known implant dimension. To verify longitudinal calibration consistency, all postoperative radiographs contributing to the growth analysis were re-audited using the known 5-mm rod diameter as a uniform internal calibration reference. The same anatomical landmarks were used at each assessment, with T1–S1 and T1–T12 measurements repeated on the early postoperative, approximately 2-year, 5-year, and final follow-up radiographs. This standardized approach was used to minimize the potential influence of serial magnification and calibration differences on longitudinal measurements.

Serial radiographs were additionally reviewed for mechanical abnormalities at the sliding interfaces, including set-screw migration or back-out, loss of screw–rod engagement, focal rod deformation or notching, and connector-related abnormalities. For patients who underwent revision surgery, operative records were also reviewed for gross screw–rod interface wear, metallosis, or abnormal tissue discoloration. Formal implant-retrieval analysis, microscopic wear assessment, histologic evaluation, and metal-ion testing were not performed.

Correction Metrics

Initial correction rate was calculated for each patient as:

[(preoperative Cobb angle − early postoperative Cobb angle)/preoperative Cobb angle] × 100.

The proportion of correction maintained at final follow-up was calculated as:

[(preoperative Cobb angle − final follow-up Cobb angle)/(preoperative Cobb angle − early postoperative Cobb angle)] × 100.

Cohort values were then summarized as the mean of the individual patient-level percentages.

Growth Definitions

To avoid misclassifying correction-related length gain as biological growth, immediate operative length gain was defined as the early postoperative minus preoperative length. Postoperative longitudinal growth was defined as final minus early postoperative length. Annualized growth velocity was calculated for each patient using the exact interval between the early postoperative and final radiographs. Preoperative-to-final length change was reported only as total observed length gain.

Rod Translation Through the Domino Connector

Rod translation was defined as the longitudinal displacement of the sliding rod relative to the fixed edge of the domino connector. On each calibrated radiograph, the distance between the fixed connector edge and the free end of the sliding rod was measured. Interval rod translation was calculated relative to the early postoperative radiograph. Final rod translation represented the cumulative change from the early postoperative radiograph to the final follow-up examination. Measurements were obtained bilaterally and averaged to derive a patient-level value. Annualized rod-translation velocity was calculated using each patient’s exact follow-up duration. Complete construct jamming was defined as the absence of further measurable rod translation on at least 2 consecutive examinations despite continued measurable spinal growth and without rod exhaustion.

To evaluate changes in sliding efficiency over time, interval rod-translation velocity was additionally calculated for the early postoperative-to-2-year, 2-to-5-year, and 5-year-to-final follow-up periods using the actual elapsed time between radiographs.

Measurement Reliability

Final Cobb angle, final T1–S1 length, and final rod translation through the domino connector were independently assessed by two observers. Interobserver reliability was evaluated using a two-way random-effects, absolute-agreement, single-measure intraclass correlation coefficient [ICC(2,1)] with 95% confidence intervals.

To further assess the reproducibility of longitudinal spinal-length measurements, T1–S1 and T1–T12 measurements at the early postoperative, approximately 2-year, 5-year, and final follow-up examinations were independently repeated by the same two observers using consistent anatomical landmarks. In addition, postoperative radiographs contributing to the growth analysis were re-audited using the known 5-mm rod diameter as a uniform internal calibration reference. Agreement between the original measurements and the standardized rod-calibrated measurements was assessed using the same ICC model together with absolute measurement differences.

Complications and Survival

Complications were categorized as implant-related, anchor-related, junctional, infectious, neurological, or other. An unplanned revision was any unanticipated return to the operating room related to the SGRT construct, fixation foundations, or associated complications. Revision-free implant survival was calculated from the index procedure to the first unplanned revision. Patients who underwent definitive fusion without a prior unplanned revision were censored at the date of definitive fusion, whereas those who retained the SGRT construct were censored at their last available follow-up. Definitive fusion performed after completion or near-completion of growth was recorded separately from unplanned construct-related revision. The indication for definitive fusion was categorized as planned treatment completion or progression of deformity outside the originally instrumented region.

Patient-Reported and Functional Outcomes

EOSQ-24 scores obtained at final follow-up were transformed to a 0–100 scale, with higher scores indicating better health-related quality of life. Complete final follow-up EOSQ-24 data were available for 21 patients; two incomplete questionnaires were excluded without imputation. Caregiver satisfaction was assessed at final follow-up using a 0–100 numerical rating scale. In the neuromuscular subgroup, sitting ability was categorized using a four-level functional classification, and improvement was defined as progression by at least one category.

Statistical Analysis

Continuous variables were assessed using histograms, Q-Q plots, and the Shapiro-Wilk test. Normally distributed data are presented as mean ± standard deviation; categorical variables as number and percentage. Paired continuous outcomes were assessed using paired-samples t tests. Cobb angle across three time points was evaluated using repeated-measures analysis with Holm-adjusted pairwise comparisons. Final follow-up EOSQ-24 and caregiver satisfaction scores were summarized descriptively and were not subjected to longitudinal hypothesis testing. Associations were assessed using Pearson correlation. Sensitivity analyses evaluated annualized values and partial correlation adjusted for follow-up duration. Kaplan-Meier analysis was used for revision-free implant survival, with Greenwood 95% confidence intervals. Because only four revision events occurred, multivariable regression was not performed. All tests were two-sided with p<0.05. Because of the small number of patients within each etiologic category, no formal statistical comparisons between etiologic subgroups were performed. To assess whether the pooled outcomes were disproportionately influenced by a single etiologic group, key radiographic, growth, implant-translation, and revision outcomes were summarized descriptively for patients with neuromuscular, congenital, and juvenile idiopathic scoliosis. These subgroup analyses were considered exploratory and underpowered.

Results

Patient Characteristics

The cohort included 23 patients: 13 females (56.5%) and 10 males (43.5%). Mean age at index surgery was 6.7 ± 1.4 years (range, 4.0-9.3), and mean follow-up was 93.9 ± 16.8 months (range, 68-122). Etiology was neuromuscular in 9 patients, congenital in 8, and juvenile idiopathic in 6. Baseline demographic and operative characteristics are summarized in Table 1.

Table 1.

Baseline and operative characteristics

Characteristic	Value
Patients, n	23
Female sex, n (%)	13 (56.5)
Age at index surgery, years	6.7 ± 1.4 (4.0-9.3)
Follow-up, months	93.9 ± 16.8 (68-122)
Neuromuscular scoliosis, n (%)	9 (39.1)
Congenital scoliosis, n (%)	8 (34.8)
Juvenile idiopathic scoliosis, n (%)	6 (26.1)
Operative time, min	241.0 ± 38.0
Estimated blood loss, mL	278.0 ± 115.0
Preoperative hemoglobin, g/dL	11.8 ± 1.2
Postoperative hemoglobin, g/dL	10.3 ± 1.3
Number of vertebral levels spanned by the construct	14.7 ± 2.1

Data are presented as mean ± standard deviation, mean ± standard deviation (range), or number (%), as appropriate.

The 4 patients excluded because of inadequate serial radiographic documentation had a median last documented clinical follow-up of 80 months (range, 72–91 months). None had experienced a documented implant-related complication or undergone an unplanned revision before the last available evaluation, and no clinically apparent construct failure was recorded. Their exclusion was therefore based on insufficient longitudinal radiographic documentation rather than documented treatment failure. These patients were not included in the radiographic or implant-survival analyses because they did not meet the prespecified imaging requirements.

Operative Outcomes

Mean operative time was 241 ± 38 minutes and mean estimated blood loss was 278 ± 115 mL. Hemoglobin decreased by 1.5 g/dL (95% CI, 1.26 to 1.74; p<0.001). No persistent neuromonitoring change or postoperative neurological deficit occurred.

Coronal Correction and Truncal Balance

Mean major Cobb angle decreased from 78.9 ± 12.3° preoperatively to 19.3 ± 7.0° early postoperatively and was 24.6 ± 7.4° at final follow-up. Although the 5.3° increase between the early postoperative and final assessments was statistically significant, its absolute magnitude was small, and substantial deformity correction was maintained throughout follow-up. This corresponded to a mean initial correction rate of 75.5%. At final follow-up, 90.2 ± 5.9% of the correction achieved at the index procedure was maintained. In parallel, coronal balance, shoulder symmetry, and pelvic obliquity improved significantly at final follow-up, indicating that radiographic improvement extended beyond correction of the major curve to overall truncal alignment. Detailed coronal and truncal balance outcomes are presented in Table 2, whereas the longitudinal Cobb angle measurements and Holm-adjusted pairwise comparisons are provided in Table 3.

Table 2.

Coronal and sagittal radiographic alignment outcomes

Variable	Preoperative	Final follow-up	Mean change*	95% CI	P value
Coronal plane
Coronal imbalance, mm	32.4 ± 14.2	11.7 ± 7.5	−20.7	−28.0 to −13.4	<0.001
Shoulder imbalance, mm	18.1 ± 9.2	6.4 ± 4.3	−11.7	−16.4 to −7.0	<0.001
Pelvic obliquity,	14.6 ± 6.1	7.1 ± 3.8	−7.5	−10.9 to −4.1	<0.001
Sagittal plane
T5–T12 thoracic kyphosis,	41.2 ± 13.7	40.1 ± 12.9	−1.1	−5.9 to 3.7	0.64
L1–S1 lumbar lordosis,	40.8 ± 13.8	39.6 ± 12.7	−1.2	−6.9 to 4.5	0.67
C7–S1 sagittal vertical axis, mm	31.4 ± 22.6	23.8 ± 19.7	−7.6	−17.8 to 2.6	0.14
Proximal junctional angle,	6.2 ± 4.2	8.1 ± 6.1	+1.9	−0.9 to 4.7	0.17

Values are presented as mean ± standard deviation. *Mean change represents final follow-up minus the preoperative value; negative values indicate a reduction and positive values indicate an increase. Pairwise comparisons are presented with 95% confidence intervals. P values are two-sided.

Table 3.

Major Cobb angle and prespecified pairwise comparisons

Assessment or comparison	Mean ± SD or mean difference, °	95% CI	Holm-adjusted P value
Preoperative	78.9 ± 12.3	—	—
Early postoperative	19.3 ± 7.0	—	—
Final follow-up	24.6 ± 7.4	—	—
Preoperative vs early postoperative	59.6	52.5 to 66.7	<0.001
Preoperative vs final follow-up	54.3	47.2 to 61.4	<0.001
Early postoperative vs final follow-up	5.3	4.3 to 6.3	<0.001

Values at each assessment are presented as mean ± standard deviation. Pairwise comparisons are presented as mean differences with 95% confidence intervals. Differences represent reduction for the first 2 comparisons and loss of correction for the final comparison. P values were adjusted using the Holm method.

Sagittal-Plane Outcomes

Sagittal alignment remained broadly preserved during follow-up. Mean T5–T12 thoracic kyphosis was 41.2 ± 13.7° preoperatively, decreased to 35.6 ± 11.8° in the early postoperative period, and was 40.1 ± 12.9° at final follow-up. Mean L1–S1 lumbar lordosis was 40.8 ± 13.8°, 36.5 ± 11.9°, and 39.6 ± 12.7°, respectively. C7–S1 sagittal vertical axis decreased from 31.4 ± 22.6 mm preoperatively to 22.1 ± 18.5 mm early postoperatively and was 23.8 ± 19.7 mm at final follow-up. Mean proximal junctional angle was 6.2 ± 4.2° preoperatively, 5.4 ± 3.9° early postoperatively, and 8.1 ± 6.1° at final follow-up. No clinically relevant deterioration in thoracic kyphosis, lumbar lordosis, or global sagittal alignment was observed during follow-up. One patient met this radiographic definition of proximal junctional kyphosis; in this patient, the proximal junctional angle increased from 7° preoperatively to 21° at 38 months and was 19° at final follow-up. The patient remained asymptomatic and was managed with observation without revision. Detailed sagittal-plane outcomes are presented in Table 2.

Spinal and Thoracic Growth

Mean T1–S1 and T1–T12 lengths increased significantly during follow-up. After separating the immediate correction-related length gain from subsequent growth, mean postoperative longitudinal growth was 106.0 mm for T1–S1 and 72.9 mm for T1–T12, corresponding to annualized growth rates of 13.9 and 9.6 mm/year, respectively. Detailed spinal and thoracic length measurements are presented in Table 4.

Table 4.

Spinal and thoracic growth

Outcome	T1-S1	T1-T12
Preoperative length, mm	261.4 ± 24.8	169.3 ± 18.7
Early postoperative length, mm	286.8 ± 25.1	184.6 ± 19.2
Final length, mm	392.8 ± 31.9	257.5 ± 29.6
Immediate operative gain, mm	25.4 (21.7 to 29.1)	15.3 (12.7 to 17.9)
Postoperative longitudinal growth, mm	106.0 (98.7 to 113.3)	72.9 (65.7 to 80.1)
Annualized postoperative growth, mm/year	13.9 ± 2.9	9.6 ± 2.9

Length measurements and annualized growth rates are presented as mean ± standard deviation. Immediate operative gain and postoperative longitudinal growth are presented as mean change (95% confidence interval).

Rod Translation Through the Domino Connector

Mean rod translation through the domino connector was 18.9 ± 4.1 mm at 2 years, 37.8 ± 6.7 mm at 5 years, and 56.4 ± 7.8 mm at final follow-up. Final rod translation was associated with postoperative T1–S1 length gain (r = 0.73, P < 0.001). The association remained significant when annualized values were analyzed (r = 0.82, P < 0.001) and after adjustment for follow-up duration (partial r = 0.69, P < 0.001). Descriptive group-level interval translation velocity was 9.5 mm/year from the early postoperative assessment to 2 years, decreased to 6.3 mm/year between 2 and 5 years, and remained at 6.6 mm/year from 5 years to final follow-up. Thus, translation velocity decreased after the first 2 postoperative years but remained measurable during subsequent follow-up. No patient met the prespecified radiographic definition of complete construct jamming. Detailed rod-translation measurements, interval velocities, and correlation analyses are presented in Table 5.

Table 5.

Rod translation through the domino connector and its association with spinal growth

Outcome	Estimate	95% CI	P value
Rod translation at 2 years, mm	18.9 ± 4.1	—	—
Rod translation at 5 years, mm	37.8 ± 6.7	—	—
Rod translation at final follow-up, mm	56.4 ± 7.8	—	—
Interval translation velocity, early postoperative to 2 years, mm/year	9.5	—	—
Interval translation velocity, 2 to 5 years, mm/year	6.3	—	—
Interval translation velocity, 5 years to final follow-up, mm/year	6.6	—	—
Annualized rod translation, mm/year	7.4 ± 1.2	—	—
Patients meeting the prespecified radiographic definition of complete construct jamming, n (%)	0 (0)	—	—
Final rod translation vs postoperative T1–S1 growth	r = 0.73	0.46 to 0.88	<0.001
Annualized rod translation vs annualized growth	r = 0.82	0.62 to 0.92	<0.001
Partial correlation adjusted for follow-up duration	r = 0.69	0.39 to 0.86	<0.001

Rod translation values and annualized rod translation are presented as mean ± standard deviation. Correlation coefficients are Pearson r values with 95% confidence intervals. The partial correlation was adjusted for follow-up duration. Complete construct jamming was defined as the absence of further measurable rod translation on at least two consecutive examinations despite continued measurable spinal growth and without rod exhaustion. All P values are two-sided.

Complications and Revision-Free Survival

Six patients (26.1%) experienced one complication each: rod fracture in 2, distal anchor loosening in 2, proximal junctional kyphosis in 1, and superficial infection in 1. The median time to revision was 55 months (range, 24-64). The type, timing, management, and subsequent course of each complication are summarized in Table 6. Kaplan–Meier analysis showed revision-free implant survival of 95.7% at 2 years, 91.3% at 5 years, and 82.6% at 6 years. The median revision-free survival was not reached (Figure 2).

Table 6.

Complications and unplanned revisions

Complication	n	Time to event	Management	Subsequent course
Rod fracture	2	24 and 61 months	Rod exchange	No further revision
Distal anchor loosening	2	49 and 64 months	Anchor revision	Stable fixation
Proximal junctional kyphosis	1	38 months	Observation	Stable and asymptomatic
Superficial infection	1	3 weeks	Local wound care and oral antibiotics	Resolved without implant removal

Four patients underwent unplanned revision for rod fracture or distal anchor loosening. Proximal junctional kyphosis and superficial infection were managed without revision surgery.

Figure 2.

Kaplan–Meier analysis of revision-free implant survival following the sliding-growing rod technique

No set-screw back-out or migration, loss of screw–rod engagement, focal rod notching, or connector-related deformation was identified on serial radiographs. No gross screw–rod interface wear, metallosis, or abnormal tissue discoloration was documented in the available operative records. The 2 rod fractures occurred at 24 and 61 months, respectively, and neither was associated with set-screw back-out, disengagement of the sliding interface, or other identifiable mechanical abnormality at the screw–rod or connector interfaces.

Descriptive Outcomes by Etiology

Descriptive analysis by etiology showed continued postoperative spinal growth and measurable rod translation across all three groups. Mean annualized T1–S1 growth was 13.2 mm/year in the neuromuscular group, 13.8 mm/year in the congenital group, and 15.1 mm/year in the juvenile idiopathic group. Final rod translation was 53.8, 56.0, and 60.9 mm, respectively. Unplanned revision occurred in 2 of 9 patients with neuromuscular scoliosis, 1 of 8 patients with congenital scoliosis, and 1 of 6 patients with juvenile idiopathic scoliosis. Although numerical differences were observed among etiologic groups, the subgroup sizes were too small for reliable inferential comparisons, and these findings should be regarded as descriptive and exploratory.

Treatment Status at Final Follow-up

By final follow-up, definitive fusion had been performed in 9 of 23 patients (39.1%). Seven patients underwent planned definitive fusion after reaching or approaching skeletal maturity. In 2 additional patients, definitive fusion was undertaken after adulthood because of progressive lumbar deformity distal to the original surgical construct Figure 3. The remaining 14 patients (60.9%) continued to be followed with the SGRT construct in situ.

Figure 3.

Long-term clinical and radiographic course of a patient with juvenile idiopathic scoliosis who underwent bilateral SGRT at 7 years of age

Patient-Reported and Functional Outcomes

At final follow-up, the mean EOSQ-24 score among the 21 patients with complete questionnaire data was 79.9 ± 11.3, and mean caregiver satisfaction was 87.4 ± 9.3. Sitting balance improved by at least one functional category in 7 of 9 patients with neuromuscular scoliosis and remained stable in the other 2. No ambulatory patient deteriorated in functional category (Table 7).

Table 7.

Patient-reported and functional outcomes at final follow-up

Outcome	Final follow-up result
EOSQ-24 score, n = 21	79.9 ± 11.3
Caregiver satisfaction, n = 23	87.4 ± 9.3
Improved sitting balance, neuromuscular subgroup, n = 9	7/9 (77.8%)
Stable sitting balance, neuromuscular subgroup, n = 9	2/9 (22.2%)
Maintained ambulatory category, n = 14	14/14 (100%)

Data are presented as mean ± standard deviation or number (%), as appropriate. EOSQ-24 and caregiver satisfaction were assessed at final follow-up. Higher EOSQ-24 scores indicate better health-related quality of life.

Measurement Reliability

Interobserver reliability was excellent for final Cobb angle, final T1–S1 length, and final rod translation through the domino connector. Serial spinal-length measurements also demonstrated excellent reproducibility, with ICC(2,1) values of 0.996 (95% CI, 0.994–0.998) for T1–S1 and 0.994 (95% CI, 0.991–0.997) for T1–T12. Agreement between the original and standardized rod-calibrated measurements was similarly excellent, with ICC values of 0.998 (95% CI, 0.997–0.999) for T1–S1 and 0.997 (95% CI, 0.995–0.998) for T1–T12. The corresponding mean absolute differences were 1.6 ± 1.2 mm and 1.4 ± 1.0 mm, respectively. No progressive increase in calibration-related measurement differences was observed across postoperative time points. Standardized recalibration did not materially alter the annualized estimates, which remained 13.9 ± 2.9 mm/year for T1–S1 and 9.6 ± 2.9 mm/year for T1–T12. Detailed reliability and calibration-agreement estimates are presented in Table 8.

Table 8.

Radiographic measurement reliability and calibration agreement

Measurement	Analysis	ICC(2,1)	95% CI	Mean absolute difference	Interpretation
Final Cobb angle	Interobserver reliability	0.964	0.916–0.984	—	Excellent
Final T1–S1 length	Interobserver reliability	0.997	0.993–0.999	—	Excellent
Final rod translation	Interobserver reliability	0.983	0.960–0.993	—	Excellent
Serial T1–S1 length	Interobserver reliability	0.996	0.994–0.998	—	Excellent
Serial T1–T12 length	Interobserver reliability	0.994	0.991–0.997	—	Excellent
T1–S1 length, original vs standardized calibration	Calibration agreement	0.998	0.997–0.999	1.6 ± 1.2 mm	Excellent
T1–T12 length, original vs standardized calibration	Calibration agreement	0.997	0.995–0.998	1.4 ± 1.0 mm	Excellent

ICC(2,1): two-way random-effects, absolute-agreement, single-measure intraclass correlation coefficient. Serial T1–S1 and T1–T12 reliability estimates were derived from measurements obtained at the early postoperative, approximately 2-year, 5-year, and final follow-up examinations. Calibration agreement compares the original measurements with measurements obtained after standardized recalibration using the known 5-mm rod diameter as an internal reference. Mean absolute differences are presented as mean ± standard deviation.

Discussion

The principal findings of this study were that SGRT provided substantial and durable deformity correction, accommodated continued spinal and thoracic growth, and demonstrated acceptable revision-free implant survival during long-term follow-up. Progressive rod translation through the domino connector accompanied postoperative longitudinal spinal growth and was reproducibly measurable, supporting its potential use as a construct-specific radiographic marker of preserved sliding function. These outcomes were accompanied by final follow-up EOSQ-24 data, high caregiver satisfaction, and preserved or improved functional status. Collectively, the results suggest that SGRT may maintain deformity control while allowing growth without routine planned lengthening procedures.

The cohort included neuromuscular, congenital, and juvenile idiopathic etiologies, reflecting the heterogeneity of EOS encountered in clinical practice. Outcomes of growth-friendly surgery are influenced by age, etiology, curve rigidity, sagittal alignment, pelvic involvement, and remaining growth potential.^17,18 The relatively young age at index surgery meant that the construct was required to maintain correction throughout a prolonged period of growth and repetitive mechanical loading. Although no persistent neurological deficit occurred, the limited sample size precludes firm conclusions regarding neurological safety.

The descriptive etiologic breakdown provides additional context for interpretation of the pooled results. Continued postoperative growth and measurable rod translation were observed across all etiologic groups, although juvenile idiopathic patients showed numerically greater longitudinal growth and rod translation, whereas unplanned revisions were numerically more frequent in the neuromuscular subgroup. These differences should be interpreted cautiously because of the small subgroup sizes and the absence of formal between-group testing. Accordingly, the pooled outcomes should not be interpreted as representing a uniform treatment response across etiologies.

SGRT achieved marked initial deformity correction, approximately 90% of which was maintained at final follow-up. Although the mean Cobb angle increased by 5.3° from the early postoperative assessment to final follow-up, the magnitude of this change was small in relation to the initial correction and should not be interpreted as clinically meaningful deterioration. Rather, the overall degree of correction remained largely preserved during long-term follow-up. Durable correction in growth-friendly surgery depends not only on longitudinal guidance but also on adequate control of the apex and intermediate segments^19-21 Enercan et al demonstrated that apical and intermediate anchors without fusion improved both coronal correction and thoracic kyphosis, supporting the concept that local deformity control may enhance correction without eliminating growth potential. ¹⁹ Recent apex-control and modified growing-rod constructs have similarly emphasized the importance of maintaining control of the major curve during continued growth.^20,21 The use of a bilateral construct in the present cohort is also consistent with recent European expert consensus, in which bilateral rods were recommended for growth-friendly management of progressive non-idiopathic EOS to reduce the risk of rod fracture. ²²

These findings also extend previous experience with sliding-growing constructs. ¹¹ In the present cohort, correction remained durable over a mean follow-up approaching 8 years, while serial connector translation and implant survival were additionally evaluated. Although differences in construct configuration, curve characteristics, and patient selection preclude direct comparison, the findings support the durability of the sliding-growth concept beyond the early postoperative period.

Improvement was not limited to Cobb angle. Coronal imbalance, shoulder imbalance, and pelvic obliquity decreased significantly, indicating better overall truncal alignment. These parameters are particularly relevant in neuromuscular EOS, in which pelvic leveling and sitting balance may be more clinically meaningful than isolated curve correction. ²³ Sagittal-plane assessment showed that thoracic kyphosis, lumbar lordosis, and global sagittal alignment were broadly preserved during follow-up. Thoracic kyphosis and lumbar lordosis decreased modestly in the early postoperative period and subsequently returned toward their preoperative values at final follow-up, while C7–S1 sagittal vertical axis remained stable. One patient met the radiographic criteria for proximal junctional kyphosis; this patient remained asymptomatic and did not require revision. Given the single event, no conclusions regarding PJK risk factors can be drawn from the present cohort.

A major methodological strength of this study was the separation of immediate correction-related length gain from subsequent postoperative growth. Recent European expert consensus also identified serial T1–T12 and T1–S1 measurements as the preferred radiographic method for monitoring growth gain during growth-friendly treatment. ²² Substantial longitudinal spinal and thoracic growth continued throughout follow-up after the initial operative length gain had been excluded. This distinction is important because preoperative-to-final length change may otherwise misclassify correction-related spinal straightening as biological growth. By using the early postoperative measurement as the reference point, the present analysis provides a more clinically meaningful estimate of postoperative growth.

The annualized postoperative T1–S1 length gain of 13.9 mm/year in the present cohort was higher than that reported in many growth-friendly series and therefore warrants careful interpretation. Dimeglio and Canavese described normal T1–S1 growth of approximately 10 mm/year between 5 and 10 years of age, followed by an increase to approximately 18 mm/year from 10 years to skeletal maturity. ²⁴

In the present cohort, treatment began at a mean age of 6.7 years and continued for a mean of 7.8 years, corresponding to an average final age of approximately 14.5 years. Thus, the annualized value observed in this study spans both the relatively slower childhood growth period and the subsequent period of accelerated adolescent spinal growth rather than representing a purely prepubertal growth rate.

Published postoperative growth rates after growing-rod treatment are generally lower, although substantial methodological heterogeneity exists among studies. In a case-matched comparison, Akbarnia et al. reported annual T1–S1 gains of 8.1 mm/year with magnetically controlled growing rods and 9.7 mm/year with traditional growing rods. ²⁵ In a systematic review of growth-friendly systems, Wijdicks et al. similarly emphasized that reported growth rates are strongly influenced by how spinal length gain is defined. ²⁶ Notably, only four traditional growing-rod studies specifically isolated growth occurring after the initial surgical correction and before final fusion, yielding a weighted T1–S1 growth rate of approximately 6 mm/year. The same review showed that initial surgical correction alone accounted for a mean T1–S1 length gain of approximately 39 mm across published series. This distinction is particularly relevant to the present study because the early postoperative radiograph, rather than the preoperative radiograph, was used as the baseline for longitudinal growth, thereby excluding the 25.4-mm immediate T1–S1 length gain attributable to index correction.

Because small differences in radiographic calibration could accumulate over a prolonged follow-up period, we also re-audited the serial postoperative measurements using the known 5-mm rod diameter as a uniform internal calibration reference. Serial T1–S1 and T1–T12 measurements demonstrated excellent interobserver reproducibility, and agreement between the original and standardized rod-calibrated measurements was similarly excellent. Importantly, standardized recalibration did not materially alter the annualized T1–S1 or T1–T12 estimates, making cumulative calibration drift an unlikely explanation for the observed growth rate. The use of a known rod diameter as an internal calibration reference is supported by previous pediatric spinal deformity work demonstrating excellent agreement among linear radiographic measurements obtained with different calibration methods. ²⁷ Nevertheless, the 13.9 mm/year value should be interpreted as annualized postoperative radiographic T1–S1 length gain rather than as a direct measure of biological vertebral growth.

Traditional growing rods, magnetically controlled growing rods, VEPTR, and growth-guidance constructs can all increase spinal height, although their mechanical limitations and treatment burdens differ.^28,29 Magnetically controlled systems avoid repeated open distractions but remain associated with diminishing distraction efficiency, actuator dysfunction, metallosis, rod failure, and unplanned revision during prolonged treatment.^30-32 In the present cohort, substantial longitudinal growth occurred without routine planned lengthening, suggesting that passive sliding may represent an effective alternative to serial distraction. Nevertheless, increased T1–T12 length should not be equated with improved pulmonary function because thoracic height alone does not reliably predict respiratory performance. ³³

Progressive rod translation through the domino connector was observed throughout follow-up, increasing from 18.9 mm at 2 years to 37.8 mm at 5 years and 56.4 mm at final follow-up. The association between rod translation and postoperative T1–S1 length gain should be interpreted in the context of the construct design. Because the proximal and distal foundations are rigidly fixed, longitudinal growth of the intervening spine necessarily requires relative translation of the rod through the sliding interface; therefore, these two measurements are mechanically coupled rather than biologically independent. The observed correlation (r = 0.73), although substantial, was not expected to approach unity because rod translation and T1–S1 length represent different radiographic measurements. T1–S1 length reflects global spinal length and is influenced by changes in spinal alignment and radiographic projection, whereas rod translation reflects motion at a specific implant interface. In addition, friction at the screw–rod and connector interfaces, rod contour, soft-tissue constraint, and measurement variability may contribute to differences between the two measures. Accordingly, the clinical value of rod translation lies not in predicting spinal growth, but in providing a serial, reproducible radiographic marker that the sliding mechanism remains functional during growth.

These findings support the mechanical rationale of SGRT. Continued rod translation through the domino connector indicates that the sliding interface remains active and capable of accommodating longitudinal change. Contemporary growth-guidance research has emphasized preservation of gliding capacity, reduction of friction, and avoidance of mechanical obstruction at the rod–implant interface.^29,34 Newer multidirectional sliding systems have also shown that guided translation can maintain deformity control while permitting spinal height gain without scheduled distraction. ³⁴ However, differences in connector design, rod material, and apical fixation limit direct comparison with SGRT.

No patient met the prespecified radiographic definition of complete construct jamming; however, this finding should not be interpreted as evidence that sliding efficiency remained constant throughout treatment. Interval rod-translation velocity decreased from approximately 9.5 mm/year during the first 2 postoperative years to 6.3 mm/year between 2 and 5 years and 6.6 mm/year thereafter. This pattern suggests a reduction in translation efficiency over time and may be consistent with a diminishing-returns phenomenon, although measurable translation persisted throughout follow-up. Cessation or reduction of translation may reflect connector obstruction, rod deformation, rod exhaustion, soft-tissue tethering, or autofusion.^35,36 Importantly, autofusion was not systematically assessed with CT in the present cohort; therefore, continued rod translation cannot exclude partial or subclinical autofusion, and the absence of complete radiographic jamming should not be equated with the absence of biological restriction. Serial rod translation should therefore be considered a surveillance marker of preserved sliding function rather than an independent measure or predictor of longitudinal spinal growth.

The SGRT construct used standard posterior spinal instrumentation rather than a purpose-built gliding bearing, with the intermediate set screws intentionally backed off while remaining engaged and bilateral domino connectors serving as sliding interfaces. No set-screw back-out, loss of screw–rod engagement, focal rod notching, connector-related deformation, or gross interface wear or metallosis was identified, and neither rod fracture was associated with failure of the sliding interface. However, microscopic fretting or wear cannot be excluded because formal implant-retrieval, histologic, and metal-ion analyses were not performed.

The overall complication and unplanned revision rates were within the range expected for prolonged growth-friendly treatment. Mechanical failure and anchor-related complications accounted for all revision procedures, whereas the remaining adverse events were managed without additional surgery. These findings emphasize the importance of distinguishing complications from revision-requiring events because not every adverse event imposes an additional operative burden.

Kaplan–Meier analysis demonstrated acceptable revision-free implant survival during long-term follow-up, and median revision-free survival was not reached. Time-to-event analysis is more informative than a crude revision percentage because it accounts for unequal follow-up and demonstrates the timing of failure. Contemporary series of magnetically controlled and traditional growing rods continue to report mechanical complications and unplanned reoperations during prolonged follow-up.^30-32,37 Although differences in etiology, implant generation, follow-up, and event definitions preclude direct numerical comparison, the survival profile observed with SGRT appears acceptable within the context of growth-friendly surgery.

The survival curve reached a plateau after the early-to-middle treatment period, which may suggest that constructs susceptible to fatigue failure or foundation-related complications tend to fail before later follow-up. However, this pattern should be considered exploratory because only four events occurred and fewer patients remained at risk at later time points. The later survival estimates should therefore be interpreted as cohort-specific.

Avoidance of routine planned lengthening remains a potential advantage of SGRT. The total treatment burden of traditional growing rods includes not only unplanned revision but also repeated scheduled procedures, anesthesia exposure, hospitalization, school absence, and caregiver burden. Therefore, comparisons based solely on unplanned revision may underestimate the overall burden of distraction-based systems. Management after completion of growth remains less clearly defined. Evidence from growing-rod graduates suggests that routine definitive fusion may not be necessary in every well-balanced patient, although residual deformity, symptoms, implant condition, and remaining growth should guide individualized decision-making. ³⁸ By final follow-up, 39.1% of the cohort had undergone definitive fusion, whereas the SGRT construct remained in situ in the remaining 60.9%. Most fusion procedures represented planned treatment completion at or near skeletal maturity; only 2 patients required extension to definitive fusion because of progressive lumbar deformity distal to the original construct. These findings indicate that SGRT may serve as a long-term growth-preserving strategy, although definitive fusion remains necessary in a proportion of patients after growth completion or distal curve progression.

Patient-reported outcomes provided complementary information beyond the radiographic findings. The mean final EOSQ-24 score of 79.9 should, however, be interpreted in the context of the etiologic heterogeneity of the cohort. Previous studies have consistently shown that EOSQ-24 scores differ according to underlying diagnosis, with patients with neuromuscular and syndromic EOS generally reporting lower health-related quality-of-life scores than those with idiopathic or congenital EOS. Matsumoto et al. demonstrated significantly lower scores in several functional and health-related domains in patients with neuromuscular EOS, and a subsequent multicenter study of 610 patients confirmed that neuromuscular etiology was associated with the lowest EOSQ scores overall, whereas congenital and idiopathic patients had broadly similar profiles. More recently, these etiology-related differences were shown to persist at 5 years after growth-friendly instrumentation, with congenital and idiopathic cohorts demonstrating higher EOSQ domain scores than neuromuscular and syndromic cohorts.^12,39,40 Accordingly, the pooled score of 79.9 in the present cohort, which included neuromuscular, congenital, and juvenile idiopathic etiologies, should not be interpreted as a universal benchmark of treatment success or as evidence of improvement attributable to SGRT. Because EOSQ-24 data were available only at final follow-up, no within-patient change from baseline could be assessed, and the absence of a contemporaneous comparator further limits causal interpretation. The EOSQ-24 result should therefore be regarded as a cross-sectional description of health-related quality of life at final follow-up rather than evidence that the radiographic or mechanical outcomes of SGRT produced an improvement in quality of life.

Radiographic improvement alone, however, does not fully explain health-related quality of life. Baird et al identified substantial heterogeneity in the measurement and reporting of EOS-related quality of life and emphasized the importance of validated disease-specific instruments. ¹³ HRQoL after growth-friendly treatment is influenced by diagnosis, functional status, respiratory health, complications, and total treatment burden rather than implant type alone. ⁴¹ Accordingly, the final follow-up EOSQ-24 findings should be interpreted as complementary to the radiographic outcomes rather than as evidence that radiographic correction directly resulted in improved health-related quality of life.

Functional outcomes were particularly relevant in the neuromuscular subgroup. Sitting balance improved in 7 of 9 patients and remained stable in the other 2, while no ambulatory patient deteriorated in functional category. These findings were concordant with improvements in pelvic obliquity and coronal balance. In neuromuscular EOS, trunk stability, pelvic leveling, sitting tolerance, mobility, and ease of care may represent more meaningful goals than maximal Cobb correction. Nevertheless, the functional classification used was broad and was not a validated performance-based instrument.

Interobserver reliability was excellent for final Cobb angle, T1–S1 length, and rod translation through the domino connector. The high reliability of rod-translation measurements is especially relevant because the usefulness of a construct-specific marker depends on reproducibility.

This study has several limitations. Its retrospective, single-center design introduces potential selection and treatment bias, and the relatively small, etiologically heterogeneous cohort limits generalizability and precluded reliable inferential subgroup comparisons or multivariable analyses. Although outcomes were summarized descriptively by etiology, these subgroup estimates were underpowered and should not be interpreted as evidence of etiology-specific treatment effects. The requirement for a minimum of 5 years of follow-up together with complete longitudinal radiographic documentation also introduces the possibility of selection and survivorship bias. In particular, exclusion of patients with incomplete serial imaging could preferentially retain patients with more consistent follow-up and potentially better-performing constructs. Although review of the available records of the excluded patients did not identify a documented excess of implant-related complications or unplanned revisions, events occurring after their last available assessment cannot be excluded. Accordingly, the reported long-term radiographic and implant-survival outcomes may somewhat overestimate construct durability. There was no contemporaneous comparison group, and all comparisons with other growth-friendly systems were indirect. Pulmonary function was not assessed systematically, and thoracic growth cannot be assumed to indicate respiratory improvement. The number at risk decreased during later follow-up, resulting in wide confidence intervals around survival estimates. Baseline EOSQ-24 data were not collected prospectively; therefore, longitudinal changes in health-related quality of life could not be evaluated. In addition, the etiologically heterogeneous cohort and absence of a contemporaneous comparison group limit interpretation of the pooled final EOSQ-24 score. Final follow-up EOSQ-24 data were incomplete for 2 patients, and caregiver satisfaction was assessed only at the final evaluation. Because rod translation and longitudinal spinal lengthening are mechanically coupled by the construct design, their correlation should not be interpreted as evidence that rod translation independently predicts spinal growth. Rather, serial rod translation should be regarded as a construct-specific surveillance marker of preserved sliding function. The observed reduction in interval translation velocity further indicates that continued rod translation did not necessarily reflect constant sliding efficiency over time. Finally, autofusion was not systematically assessed with CT or other advanced imaging; therefore, partial or subclinical autofusion cannot be excluded despite continued measurable rod translation, and the absence of complete radiographic jamming should not be interpreted as evidence of unrestricted biological motion.

Conclusions

In conclusion, SGRT was associated with durable deformity control, continued postoperative spinal and thoracic growth, and acceptable revision-free implant survival during long-term follow-up. Serial rod translation through the domino connector was reproducibly measurable throughout follow-up and may serve as a practical radiographic marker of preserved sliding function during growth. Larger multicenter comparative studies incorporating pulmonary assessment, validated functional outcomes, serial assessment of rod translation, and total treatment-burden measures are needed to clarify the role of SGRT relative to established growth-friendly techniques.

ORCID iDs

Muhammed Enes Karataş https://orcid.org/0000-0003-0995-0953

Bekir Yavuz Uçar https://orcid.org/0000-0002-5175-3091

Ethical Considerations

This study was approved by the Institutional Review Board of Istanbul Medipol University Faculty of Medicine (Approval No: E-10840098-202.3.02-1004). The study was conducted in accordance with the principles of the Declaration of Helsinki.

Consent to Participate

Written informed consent was obtained from all patients (or their legal guardians) prior to inclusion in the study.