Aims. Although lumbosacral transitional vertebrae (LSTV) are well-documented, few large-scale studies have investigated thoracolumbar transitional vertebrae (TLTV) and spinal numerical variants. This study sought to establish the prevalence of numerical variants and to evaluate their relationship with clinical problems. Methods. A total of 1,179 patients who had undergone thoracic, abdominal, and pelvic CT scanning were divided into groups according to the number of thoracic and lumbar vertebrae, and the presence or absence of TLTV or LSTV. The prevalence of spinal anomalies was noted. The relationship of spinal anomalies to clinical symptoms (low back pain, Japanese Orthopaedic Association score, Roland-Morris Disability Questionnaire) and degenerative spondylolisthesis (DS) was also investigated. Results. Normal vertebral morphology (12 thoracic and five lumbar vertebrae without TLTV and LSTV) was present in 531 male (76.7%) and 369 female patients (75.8%). Thoracolumbar transitional vertebrae were present in 15.8% of males and 16.0% of females. LSTV were present in 7.1% of males and 9.0% of females. The prevalence of the anomaly of 16 presacral mobile vertebrae (total number of thoracolumbar vertebrae and TLTV) without LSTV was 1.0% in males and 4.1% in females, and that of the anomaly of 18 vertebrae without LSTV was 5.3% in males and 1.2% in females. The prevalence of DS was significantly higher in females with a total of 16 vertebrae than in those with normal morphology. There was no significant correlation between a spinal anomaly and clinical symptoms. Conclusion. Overall, 24% of subjects had anomalies in the thoracolumbar region: the type of anomaly differed between males and females, which could have significant implications for spinal surgery. A decreased number of vertebrae was associated with DS: numerical variants may potentially be a clinical problem. Cite this article: Bone Joint J 2021;103-B(7):1301–1308

Objectives. Cement augmentation of pedicle screws could be used to improve screw stability, especially in osteoporotic vertebrae. However, little is known concerning the influence of different screw types and amount of cement applied. Therefore, the aim of this biomechanical in vitro study was to evaluate the effect of cement augmentation on the screw pull-out force in osteoporotic vertebrae, comparing different pedicle screws (solid and fenestrated) and cement volumes (0 mL, 1 mL or 3 mL). Materials and Methods. A total of 54 osteoporotic human cadaver thoracic and lumbar vertebrae were instrumented with pedicle screws (uncemented, solid cemented or fenestrated cemented) and augmented with high-viscosity PMMA cement (0 mL, 1 mL or 3 mL). The insertion torque and bone mineral density were determined. Radiographs and CT scans were undertaken to evaluate cement distribution and cement leakage. Pull-out testing was performed with a material testing machine to measure failure load and stiffness. The paired t-test was used to compare the two screws within each vertebra. Results. Mean failure load was significantly greater for fenestrated cemented screws (+622 N; p ⩽ 0.001) and solid cemented screws (+460 N; p ⩽ 0.001) than for uncemented screws. There was no significant difference between the solid and fenestrated cemented screws (p = 0.5). In the lower thoracic vertebrae, 1 mL cement was enough to significantly increase failure load, while 3 mL led to further significant improvement in the upper thoracic, lower thoracic and lumbar regions. Conclusion. Conventional, solid pedicle screws augmented with high-viscosity cement provided comparable screw stability in pull-out testing to that of sophisticated and more expensive fenestrated screws. In terms of cement volume, we recommend the use of at least 1 mL in the thoracic and 3 mL in the lumbar spine. Cite this article: C. I. Leichtle, A. Lorenz, S. Rothstock, J. Happel, F. Walter, T. Shiozawa, U. G. Leichtle. Pull-out strength of cemented solid versus fenestrated pedicle screws in osteoporotic vertebrae. Bone Joint Res 2016;5:419–426

Aims. The aim of this study was to investigate the incidence and characteristics of instrumentation failure (IF) after total en bloc spondylectomy (TES), and to analyze risk factors for IF. Methods. The medical records from 136 patients (65 male, 71 female) with a mean age of 52.7 years (14 to 80) who underwent TES were retrospectively reviewed. The mean follow-up period was 101 months (36 to 232). Analyzed factors included incidence of IF, age, sex, BMI, history of chemotherapy or radiotherapy, tumour histology (primary or metastasis; benign or malignant), surgical approach (posterior or combined), tumour location (thoracic or lumbar; junctional or non-junctional), number of resected vertebrae (single or multilevel), anterior resection line (disc-to-disc or intravertebra), type of bone graft (autograft or frozen autograft), cage subsidence (CS), and local alignment (LA). A survival analysis of the instrumentation was performed, and relationships between IF and other factors were investigated using the Cox regression model. Results. A total of 44 patients (32.4%) developed IF at a median of 31 months (interquartile range 23 to 74) following TES. Most IFs were rod fractures preceded by a mean CS of 6.1 mm (2 to 18) and LA kyphotic enhancement of 10.8° (-1 to 36). IF-free survival rates were 75.8% at five years and 56.9% at ten years. The interval from TES to IF peaked at two to three years postoperatively and continued to occur over a period of time thereafter; the early IF-developing group had greater CS at one month postoperatively (CS1M) and more lumbar TES. CS1M ≥ 3 mm and sole use of frozen autografts were identified as independent risk factors for IF. Conclusion. IF is a common complication following TES. We have demonstrated that robust spinal reconstruction preventing CS, and high-quality bone grafting are necessary for successful reconstruction. Cite this article: Bone Joint J 2023;105-B(2):172–179

Introduction. Osteoporotic fracture reduces vertebral stiffness, and alters spinal load-sharing. Vertebroplasty partially reverses these changes at the fractured level, but is suspected to increase deformations and stress at adjacent levels. We examined this possibility. Methods. Twelve pairs of three-vertebra cadaver spine specimens (67-92 yr) were loaded to induce fracture. One of each pair underwent vertebroplasty with PMMA, the other with a resin (Cortoss). Specimens were then creep-loaded at 1.0kN for 1hr. In 15 specimens, either the uppermost or lowest vertebra was fractured, so that compressive stress distributions could be determined in the disc between adjacent non-fractured vertebrae. Stress was measured in flexion and extension, at each stage of the experiment, by pulling a pressure-transducer through the disc whilst under 1.0kN load. Stress profiles quantified intradiscal pressure (IDP), stress concentrations in the posterior annulus (SP. P. ), and compressive load-bearing by the neural arch (F. N. ). Elastic deformations in adjacent vertebrae were measured using a MacReflex tracking system during 1.0kN compressive ramp loading. Results. No differences were found between Cortoss and PMMA so data was pooled. Following fracture, IDP fell by 27% in extension (P=0.009), and SP. P. increased by 277% in flexion (P=0.016). F. N. increased from 17% to 30% of the applied load in flexion, and from 23% to 37% in extension (P<0.05). Vertebroplasty partially reversed these changes without inducing any increase in elastic deformation of the adjacent vertebrae. Conclusion. Vertebral fracture increases stress concentrations acting on the vertebral bodies and neural arches of adjacent (non-fractured) vertebrae, and these increases can be partially reversed by vertebroplasty

We studied 52 patients, each with a lumbosacral transitional vertebra. Using MRI we found that the lumbar discs immediately above the transitional vertebra were significantly more degenerative and those between the transitional vertebrae and the sacrum were significantly less degenerative compared with discs at other levels. We also performed an anatomical study using 70 cadavers. We found that the iliolumbar ligament at the level immediately above the transitional vertebra was thinner and weaker than it was in cadavers without a lumbosacral transitional vertebra. Instability of the vertebral segment above the transitional vertebra because of a weak iliolumbar ligament could lead to subsequent disc degeneration which may occur earlier than at other disc levels. Some stability between the transitional vertebra and the sacrum could be preserved by the formation of either an articulation or by bony union between the vertebra and the sacrum through its transverse process. This may protect the disc from further degeneration in the long term

Purpose. To study the relationship between Zygoapophysial Joint Tropism and pathologic fractures affecting lumbar and thoracic vertebrae in elderly patients. Methods. The sagittal plane orientation of the Zygoapophysial joints (facets or ZAJ) of 324 vertebrae of 63 patients were measured on MRI scans, stratified into lumbar and thoracic, fractured and non-fractured, and then classified according to the presence of tropism. The correlation between tropism and fractures, demographics pertaining to age, spinal level, and morbidity were studied. Results. From the 415 ZAJ pairs studied, 23 were excluded because of insufficient imaging leaving 388 ZAJ pairs. In 155 Thoracic ZAJ pairs, there 92 with a fracture, 39% demonstrating tropism; and there were 63 with no fracture, 43% demonstrating tropism. In 237 Lumbar ZAJ pairs, there were 144 with a fracture, 73% demonstrating tropism; there were 93 with no fracture, 42% demonstrating tropism. Conclusion. Our study suggests a correlation between the emergence of pathological fractures in the spine and tropism which is statistically significant in the lumbar but not in the thoracic. The authors confirm that this abstract has not been previously published in whole or substantial part nor has it been presented previously at a national meeting. Conflicts of interest: No conflicts of interest. Sources of funding: No funding obtained

We studied MR images of the spine in a consecutive series of 100 patients with acute compression of the spinal cord due to metastases. All patients had documented neurological deficit and histologically proven carcinoma. MRI was used to localise bony metastatic involvement and soft-tissue impingement of the cord. A systematic method of documenting metastatic involvement is described. A total of 43 patients had compression at multiple levels; 160 vertebral levels were studied. In 120 vertebrae (75%), anterior, lateral and posterior bony elements were involved. Soft-tissue impingement of the spinal cord often involved more than one quadrant of its circumference. In 69 vertebrae (43%) there was concomitant anterior and posterior compression. Isolated involvement of a vertebral body was observed in only six vertebrae (3.8%). We have shown that in most cases of acute compression of the spinal cord due to metastases there is coexisting involvement of both anterior and posterior structures

Transitional vertebrae (TV) may be one of the risk factors for lumbar disc herniation. It is not clear, however, whether the presence of TV can affect the development of nerve-root symptoms. Our aim was to clarify this relationship. A total of 501 patients with lumbar degenerative disease and nerve-root symptoms was studied in respect of their level and the presence of TV. As a control group, 508 patients without low back pain or nerve-root symptoms were studied to establish the incidence of TV. In patients with disc herniation, the incidence was statistically higher and the mean age lower in patients with TV than in those without. In most patients, the symptomatic disc level was just above the TV. Similarly, in those with stenosis of the spinal canal without spondylolisthesis, the symptomatic disc level was most commonly just above the TV

Aims

The aim of this study was to report the long-term prognosis of patients with multiple Langerhans cell histiocytosis (LCH) involving the spine, and to analyze the risk factors for progression-free survival (PFS).

Introduction. Vertebral osteoporotic fracture increases both elastic and time-dependent (‘creep’) deformations of the fractured vertebral body during subsequent loading. This is especially marked in central and anterior regions of the vertebral body, and could explain the development of kyphotic deformity in life. We hypothesise that vertebroplasty can reduce these creep deformations. Methods. Twelve pairs of spine specimens, each comprising three vertebrae and the intervening soft tissue, were obtained from cadavers aged 67-92 yr. They were compressed to failure, after which one of each pair underwent vertebroplasty with polymethylmethacrylate cement, the other with a resin (Cortoss). A 1kN compressive force was applied for 1 hour before fracture, after fracture, and after vertebroplasty, while creep deformation was measured in the anterior, middle, and posterior region of each vertebral body using a MacReflex optical tracking system. Results. Cement type had little influence on creep deformation, so data from all 24 specimens were pooled. After fracture, creep in the anterior vertebral body increased from 4,863 (STD 5060) to 57,674 (STD 57,663) microstrains (P<0.001), and creep in the central vertebral body increased from 738 (STD 5,643) to 37,025 (STD 43,519) microstrain (P<0.001). Vertebroplasty reduced creep deformations by 67% (P=0.002) and 66% (P=0.011) in the anterior and central regions respectively. Conclusion. Vertebroplasty reduces progressive creep deformations of a fractured vertebral body. These findings suggest that vertebroplasty may slow or prevent the development of kyphotic deformity following vertebral osteoporotic fracture

Aims

This systematic review aims to identify 3D predictors derived from biplanar reconstruction, and to describe current methods for improving curve prediction in patients with mild adolescent idiopathic scoliosis.

Aims. To evaluate the perioperative complications associated with total en bloc spondylectomy (TES) in patients with spinal tumours, based on the extent and level of tumour resection. Methods. In total, 307 patients who underwent TES in a single centre were reviewed retrospectively. There were 164 male and 143 female patients with a mean age at the time of surgery of 52.9 years (SD 13.3). A total of 225 patients were operated on for spinal metastases, 34 for a malignant primary tumour, 41 for an aggressive benign tumour, and seven with a primary of unknown origin. The main lesion was located in the thoracic spine in 213, and in the lumbar spine in 94 patients. There were 97 patients who underwent TES for more than two consecutive vertebrae. Results. Major and minor perioperative complications were observed in 122 (39.7%) and 84 (27.4%) patients respectively. The breakdown of complications was as follows: bleeding more than 2,000 ml in 60 (19.5%) patients, hardware failure in 82 (26.7%), neurological in 46 (15.0%), surgical site infection in 23 (7.5%), wound dehiscence in 16 (5.2%), cerebrospinal fluid leakage in 45 (14.7%), respiratory in 52 (16.9%), cardiovascular in 11 (3.6%), digestive in 19 (6.2%)/ The mortality within two months of surgery was four (1.3%). The total number of complications per operation were 1.01 (SD 1.0) in the single vertebral resection group and 1.56 (SD 1.2) in the group with more than two vertebral resections. Cardiovascular and respiratory complications, along with hardware failure were statistically higher in the group who had more than two vertebrae resected. Also, in this group the amount of bleeding in patients with a lumbar lesion or respiratory complication in patients with a thoracic lesion, were statistically higher. Multivariate analysis showed that using a combined anterior and posterior approach, when more than two vertebral resections were significant independent factors. Conclusion. The characteristics of perioperative complications after TES were different depending on the extent and level of the tumour resection. In addition to preoperative clinical and pathological factors, it is therefore important to consider these factors in patients who undergo en bloc resection for spinal tumours. Cite this article: Bone Joint J 2021;103-B(5):976–983

Aims. To clarify the asymmetrical ossification of the epiphyseal ring between the convex and concave sides in patients with adolescent idiopathic scoliosis (AIS). Patients and Methods. A total of 29 female patients (mean age, 14.4 years; 11 to 18) who underwent corrective surgery for AIS (Lenke type 1 or 2) were included in our study. In all, 349 vertebrae including 68 apical vertebrae and 87 end vertebrae in the main thoracic (MT) curve and thoracolumbar/lumbar (TL/L) curve were analysed. Coronal sections (anterior, middle and posterior) of the vertebral bodies were reconstructed from pre-operative CT scans (320-row detector; slice thickness, 0.5 mm) and the appearances of the ossification centre in the epiphyseal ring at four corners were evaluated in three groups; all vertebrae excluding end vertebrae, apical vertebrae and end vertebrae. The appearance rates of the ossification centre at the concave and convex sides were calculated and compared. Results. The appearance rates of the ossification centres in all vertebrae excluding end vertebrae and apical vertebrae were significantly lower on the concave side than on the convex side in both MT and TL/L curves irrespective of curve flexibility. There was no significant difference in the rate of appearance of the ossification centres on the concave or convex sides in end vertebrae. Conclusion. The asymmetric bony growth of vertebral body came into existence at both structural and non-structural curves, and was more apparent around the apical vertebrae. Evaluation of the ossification centre in the epiphyseal ring could be a measure of the effectiveness of brace treatment. Take home message: The ossification of the epiphyseal ring in patients with AIS was delayed or absent on the concave side particularly around the apical vertebrae. Cite this article: Bone Joint J 2016;98-B:666–71

Aims. Anchorage of pedicle screw rod instrumentation in the elderly spine with poor bone quality remains challenging. Our study aims to evaluate how the screw bone anchorage is affected by screw design, bone quality, loading conditions, and cementing techniques. Methods. Micro-finite element (µFE) models were created from micro-CT (μCT) scans of vertebrae implanted with two types of pedicle screws (L: Ennovate and R: S. 4. ). Simulations were conducted for a 10 mm radius region of interest (ROI) around each screw and for a full vertebra (FV) where different cementing scenarios were simulated around the screw tips. Stiffness was calculated in pull-out and anterior bending loads. Results. Experimental pull-out strengths were excellently correlated to the µFE pull-out stiffness of the ROI (R. 2. > 0.87) and FV (R. 2. > 0.84) models. No significant difference due to screw design was observed. Cement augmentation increased pull-out stiffness by up to 94% and 48% for L and R screws, respectively, but only increased bending stiffness by up to 6.9% and 1.5%, respectively. Cementing involving only one screw tip resulted in lower stiffness increases in all tested screw designs and loading cases. The stiffening effect of cement augmentation on pull-out and bending stiffness was strongly and negatively correlated to local bone density around the screw (correlation coefficient (R) = -0.95). Conclusion. This combined experimental, µCT and µFE study showed that regional analyses may be sufficient to predict fixation strength in pull-out and that full analyses could show that cement augmentation around pedicle screws increased fixation stiffness in both pull-out and bending, especially for low-density bone. Cite this article: Bone Joint Res 2021;10(12):797–806

Introduction. Lumbosacral transitional vertebrae (LSTV) are increasingly recognized as a common anatomical variant and is the most common congenital anomaly of the lumbosacral spine. Patients can have symptomatic LSTV, known as Bertolotti's Syndrome, where transitional anatomy can cause back, L5 distribution leg, hip, and groin pain. We propose an outline for diagnosis and treatment of Bertolotti's Syndrome. Methods. We retrospectively reviewed over 500 patients presenting to the primary author with low back, buttock, hip, groin and/or leg pain from April 2009 through April 2024. Patients with radiographic findings of an LSTV and clinical presentation underwent diagnostic injections to confirm diagnosis of Bertolotti's syndrome. Treatment was determined based on patient's LSTV classification. 157 patients with confirmed Bertolotti's syndrome underwent surgical treatment. Results. Over 500 patients presented with an appropriate clinical presentation and radiographic findings of an LSTV. Diagnostic injections were targeted into the transitional anatomy confirming the LSTV as the primary pain generator to make the diagnosis of Bertolotti's syndrome. The decision in the type of surgical intervention, resection or fusion, was made based on patient's LSTV anatomy. 157 patients with confirmed Bertolotti's Syndrome underwent surgical treatment (121 fusions (77%), 36 resections (23%)). The classification system and surgical outcomes, in part, have been previously published in World Neurosurgery. Conclusion. We have outlined the best practice of diagnosis and treatment selection for Bertolotti's syndrome. We have shown significant improvement in outcomes based on this method. We hope to aid in both patient education and provide an outline on how clinicians can become knowledgeable on Bertolotti's syndrome. Conflicts of interest. No conflicts of interest. Sources of funding. No funding obtained

Aims. A total of 30 patients with thoracolumbar/lumbar adolescent idiopathic scoliosis (AIS) treated between 1989 and 2000 with anterior correction and fusion surgery using dual-rod instrumentation were reviewed. . Patients and Methods. Radiographic parameters and clinical outcomes were compared among patients with lowest instrumented vertebra (LIV) at the lower end vertebra (LEV; EV group) (n = 13) and those treated by short fusion (S group), with LIV one level proximal to EV (n = 17 patients). . Results. The allocation of the surgical technique was determined by the flexibility of the TL/L curves and/or neutral vertebrae located one level above LEV as determined on preoperative radiographs. If these requirements were met a short fusion was performed. The mean follow-up period was 21.4 years (16 to 27). The mean correction rate at final follow-up was significantly lower in the S group (74 . sd. 11%) than in the EV group (88 . sd. 13%) (p = 0.004).Coronal and sagittal balance, thoracic kyphosis, lumbar lordosis, and clinical outcomes evaluated by the Scoliosis Research Society-22 questionnaire scores were equivalent between the two groups. . Conclusion. Short fusion strategy, which uses LIV one level proximal to LEV can be considered as an alternative to the conventional strategy, which includes LEV in the fusion, when highly flexible TL/L curves are confirmed and/or neutral vertebrae are located one level above LEV in patients with thoracolumbar/lumbar AIS curves. Take home message: Short fusion strategy can be considered as an alternative to the conventional strategy in patients with thoracolumbar/lumbar AIS curves undergoing anterior spinal fusion with dual-rod instrumentation. Cite this article: Bone Joint J 2016;98-B:402–9

In patients with osteoporosis there is always a strong possibility that pedicle screws will loosen. This makes it difficult to select the appropriate osteoporotic patient for a spinal fusion. The purpose of this study was to determine the correlation between bone mineral density (BMD) and the magnitude of torque required to insert a pedicle screw. To accomplish this, 181 patients with degenerative disease of the lumbar spine were studied prospectively. Each underwent dual-energy x-ray absorptiometry (DEXA) and intra-operative measurement of the torque required to insert each pedicle screw. The levels of torque generated in patients with osteoporosis and osteopenia were significantly lower than those achieved in normal patients. Positive correlations were observed between BMD and T-value at the instrumented lumbar vertebrae, mean BMD and mean T-value of the lumbar vertebrae, and mean BMD and mean T-value of the proximal femur. The predictive torque (Nm) generated during pedicle screw insertion was [-0.127 + 1.62 × (BMD at the corresponding lumbar vertebrae)], as measured by linear regression analysis. The positive correlation between BMD and the maximum torque required to insert a pedicle screw suggests that pre-operative assessment of BMD may be useful in determining the ultimate strength of fixation of a device, as well as the number of levels that need to be fixed with pedicle screws in patients who are suspected of having osteoporosis

Aims. The surgical treatment of tuberculosis (TB) of the spine consists of debridement and reconstruction of the anterior column. Loss of correction is the most significant challenge. Our aim was to report the outcome of single-stage posterior surgery using bone allografts in the management of this condition. Patients and Methods. The study involved 24 patients with thoracolumbar TB who underwent single-stage posterior spinal surgery with a cortical bone allograft for anterior column reconstruction and posterior instrumentation between 2008 and 2015. A unilateral approach was used for 21 patients with active TB, and a bilateral approach with decompression and closing-opening wedge osteotomy was used for three patients with healed TB. Results. A median of 1.25 vertebrae were removed (interquartile range (IQR) 1 to 1.75) and the median number of levels that were instrumented was five (IQR 3 to 6). The median operating time was 280 minutes (IQR 230 to 315) and the median blood loss was 700 ml (IQR 350 to 900). The median postoperative kyphosis was 8.5° (IQR 0° to 15°) with a mean correction of the kyphosis of 71.6%. Good neurological recovery occurred, with only two patients (8%) requiring assistance to walk at a mean follow-up of 24 months (9 to 50), at which time there was a mean improvement in disability, as assessed by the Oswestry Disability Index, of 83% (90% to 72%). Conclusion. The posterior-only approach using cortical allografts for anterior column reconstruction achieved good clinical and radiological outcomes. Differentiation should be made between flexible (active) and rigid (healed) TB spine. Cite this article: Bone Joint J 2018;100-B:1208–13

Aims. The aim of this study was to compare the peak pull-out force (PPF) of pedicle-lengthening screws (PLS) and traditional pedicle screws (TPS) using instant and cyclic fatigue testing. Materials and Methods. A total of 60 lumbar vertebrae were divided into six groups: PLS submitted to instant pull-out and fatigue-resistance testing (groups A1 and A2, respectively), TPS submitted to instant pull-out and fatigue-resistance testing (groups B1 and B2, respectively) and PLS augmented with 2 ml polymethylmethacrylate, submitted to instant pull-out and fatigue-resistance testing (groups C1 and C2, respectively). The PPF and normalized PPF (PPFn) for bone mineral density (BMD) were compared within and between all groups. Results. In all groups, BMD was significantly correlated with PPF (r = 0.83, p < 0.001). The PPFn in A1 was significantly less than in B1 (p = 0.006) and C1 (p = 0.002). The PPFn of A2 was significantly less than in B2 (p < 0.001) and C2 (p < 0.001). The PPFn in A1, B1, and C1 was significantly greater than in A2 (p = 0.002), B2 (p = 0.027), and C2 (p = 0.003). There were no significant differences in PPFn between B1 and C1, or between B2 and C2. Conclusion. Pedicle lengthening screws with cement augmentation can provide the same fixation stability as traditional pedicle screws and may be a viable clinical option. Cite this article: Bone Joint J 2018;100-B:516–21

Aims. The direct posterior approach with subperiosteal dissection of the paraspinal muscles from the vertebrae is considered to be the standard approach for the surgical treatment of adolescent idiopathic scoliosis (AIS). We investigated whether or not a minimally-invasive surgery (MIS) technique could offer improved results. Methods. Consecutive AIS patients treated with an MIS technique at two tertiary centres from June 2013 to March 2016 were retrospectively included. Preoperative patient deformity characteristics, perioperative parameters, power of deformity correction, and complications were studied. A total of 93 patients were included. The outcome of the first 25 patients and the latter 68 were compared as part of our safety analysis to examine the effect of the learning curve. Results. In the first 25 cases, with a mean follow-up of 5.6 years (standard deviation (SD) 0.4), the mean preoperative major Cobb angle was 57.6° (SD 9.8°) and significantly corrected to mean 15.4° (SD 5.6°, 73% curve correction). The mean preoperative T5-T12 was 26.2 (SD 12.8) and significantly increased to mean 32.9 (SD 8.3). Both frontal and sagittal plane correction was conserved two years after surgery. The rate of perioperative complications was 12% and three further complications occurred (three deep delayed infection). In the latter cases, 68 patients were included with a mean follow-up time of three years (SD 0.6). The mean preoperative major Cobb angle was 58.4° (SD 9.2°) and significantly corrected to mean 20.4° (SD 7.3°).The mean preoperative T5-T12 kyphosis was 26.6° (SD 12.8°) and was significantly increased to mean 31.4° (SD 8.3°). Both frontal and sagittal correction was conserved two years after surgery. The perioperative (30 day) complication rate was 1.4%. Two (2.9%) additional complications occurred in two patients. Conclusion. MIS for AIS is associated with a significant correction of spine deformity in the frontal and sagittal planes, together with low estimated blood loss and short length of stay. The perioperative complication rate seems to be lower compared with the standard open technique based on the literature data. The longer-term safety of MIS for AIS needs to be documented with a larger cohort and compared with the standard posterior approach. Cite this article: Bone Joint J 2020;102-B(4):506–512