ABSTRACT
Background: This trial reports the 2-year and immediate postremoval clinical outcomes of a novel posterior apical short-segment (PASS) correction technique allowing for correction and stabilization of adolescent idiopathic scoliosis (AIS) with limited fusion.
Methods: Twenty-one consecutive female AIS patients were treated at 4 institutions with this novel technique. Arthrodesis was limited to the short apical curve after correction with translational and derotational forces applied to upper and lower instrumented levels. Instrumentation spanned fused and unfused segments with motion and flexibility of unfused segments maintained. The long concave rods were removed at maturity. Radiographic data collected included preoperative and postoperative data for up to 2 years as well as after long rod removal.
Results: All 21 patients are beyond 2 years postsurgery. Average age at surgery was 14.2 years (11–17 years). A mean of 10.5 ± 1 levels per patient were stabilized and 5.0 ± 0.5 levels (48%) were fused. Cobb angle improved from 56.1° ± 8.0° to 20.8° ± 7.8° (62.2% improvement) at 1 year and 20.9° ± 8.4°, (62.0% improvement) at 2 years postsurgery. In levels instrumented but not fused, motion was 26° ± 6° preoperatively compared to 10° ± 4° at 1 year postsurgery, demonstrating 38% maintenance of mobility in nonfused segments. There was no report of implant-related complications.
Conclusions: PASS correction technique corrected the deformity profile in AIS patients with a lower implant density while sparing 52% of the instrumented levels from fusion through the 2-year follow-up.
INTRODUCTION
Common techniques for correction of adolescent idiopathic scoliosis (AIS) involve posterior arthrodesis of all motion segments that are instrumented, including the curve apex and several segments both proximal and distal to the apex in order to control the curve and optimize the correction. Constructs usually consist of vertically attached rigid rod-screw connections with a high density of pedicle screws, typically realigned with a combination of derotational and translational techniques.1–13 Although effective at correcting the deformity, this treatment immobilizes the entire length of the spine spanned by the instrumentation to assist with promoting fusion, resulting in permanent loss of motion of all instrumented levels. Long fusions carry the potential of earlier adjacent segment degeneration, instability, discomfort, or pain.14,15 These clinical observations have prompted investigation of newer, less aggressive methods that could achieve similar or better outcomes.16 In addition, cost and safety concerns, as well as other factors, have prompted considerations of lower implant-density constructs.17,18
The safety, efficacy, and motion preservation of this novel technique has already been reported through a 1-year follow-up.1 The purpose of this clinical pilot trial is to report 2-year performance and safety data for this new posterior instrumentation system for the surgical treatment of AIS. The surgical treatment under study addresses 4 basic principles: limited posterior muscle dissection, lower implant density, semiconstrained instrumentation with more limited arthrodesis, and preservation of motion at instrumented segments adjacent to the curve apex. The 2-year clinical and radiographic results using these techniques and a description of the novel instrumentation is presented.
MATERIALS AND METHODS
Study Design
From April 2012 to January 2013, 21 patients were enrolled in a prospective, nonrandomized, multi-center, post–Conformité Européenne approved clinical pilot trial employing a novel posterior technique that utilizes translation and derotation principles combined with a short apical arthrodesis and longer instrumentation of unfused motion segments for surgical treatment of AIS. Four study centers (Czech Republic, United Kingdom, Slovakia, and Turkey) participated. The ethics committee of each institution approved the clinical trial. Additionally, Turkish Ministry of Health approval was obtained, and the study received portfolio adoption by the National Institute for Health Research Clinical Research Network in the United Kingdom. Prior to enrollment, written informed consent was obtained from each subject. Subjects were enrolled in the trial based on established inclusion and exclusion criteria (Table 1). An independent surgeon advisory board conducted clinical trial safety oversight. The postoperative, 6-month, 12-month, and 24-month outcomes as well as post–system removal outcomes are presented in this manuscript.
Summary of key inclusion and exclusion criteria.
Clinical Outcomes
The presurgical patient health history, neurologic examination, rib prominence on forward bending, clinical photographs, and the Scoliosis Research Society 22r (SRS22r) questionnaire results were obtained for each subject.
Radiographic Outcomes
Preoperative standing posteroanterior (PA) and lateral, left/right lateral supine bending, and hand radiographs were obtained. The PA radiograph analysis included main thoracic curve Cobb angle, coronal balance (C7–central sacral vertical line) and clavicle angle (angle between horizontal reference line and superior margins of clavicles). Standing lateral radiographs were utilized for measuring thoracic kyphosis (Cobb angle T5-12), lumbar lordosis (Cobb angle L1-S1), and sagittal balance (C7–posterior superior corner of sacrum). Right and left side-bending supine radiographs assessed curve flexibility as well as motion of the segments that were instrumented but not fused. Vertebral rotation was assessed using a preoperative low-dose computed tomography (CT) scan of the thoracic curve apex (angle between the midsagittal line and the line connecting the vertebral centroid and posterior junction of the vertebral arch). An independent imaging core lab performed all radiographic analyses. Radiographic measurement utilized the validated Quantitative Motion Analysis software (Medical Metrics Inc., Houston, Texas).19–21
Safety and Efficacy Outcomes
The primary safety objective was avoidance of serious device- and/or procedure-related adverse events. The primary safety endpoint was measured by the percentage of patients who experienced one or more serious device- and/or procedure-related adverse events postoperatively. Secondary safety objectives included collecting the frequency and severity of all adverse events at time of procedure and over the follow-up period.
The primary efficacy objective was measured by the mean percentage of correction of the main thoracic curve at the postoperative assessments compared to preoperative assessments. A mean Cobb angle reduction of 50% was hypothesized. Secondary efficacy objectives included evaluation of patient outcomes utilizing the SRS22r questionnaire. Exploratory objectives included blood loss, operative time, length of hospital stay, sagittal and coronal alignment, kyphosis, lordosis, clavicle angle, rib prominence, and apical vertebral rotation.
Statistical Methods
Device- and procedure-related adverse events were expressed as a percentage and reported along with the corresponding 2-sided exact 95% binomial confidence interval. Power analysis for the main thoracic curve correction determined that a sample size of at least 13 patients was required to detect a 15% difference between the mean reduction of the main thoracic Cobb angle and the hypothesized population value of at least 50% reduction with 80% power and alpha of 0.05 using a 2-sided 1-sample Wilcoxon rank sum test. The mean, median, standard deviation, and minimum and maximum percentages of change were calculated. The significance of differences between the observed mean percentage of change and the hypothesized value of 50% reduction was evaluated using a 2-sided 1-sample Wilcoxon rank sum test, with P ≤ .05 being statistically significant. Differences between baseline and follow-up measurements of other variables were evaluated using a Wilcoxon matched pairs signed ranks test. Unless stated otherwise, 2-sided statistical tests were performed with alpha set at 0.05. P values ≤ .05 indicate statistical significance.
Surgical Procedure
Implantation was performed utilizing one midline skin incision from the upper instrumented vertebra (UIV) to the lowest instrumented vertebra (LIV) and 3 subfascial windows at the UIV, LIV, and apical region (ie, 3 to 5 vertebrae adjacent to curve apex) to minimize the amount of soft tissue disruption (Figure 1).
Depiction of 3-window surgical technique.
The system utilizes pedicle screws to anchor the instrumentation to the spine. The pedicles were prepared and screws implanted using the standard technique. The system is comprised of novel mechanically articulating Heim anchors implanted at the UIV and LIV and L-shaped transverse members implanted at the apical region. The articulating Heim anchors substantially block lateral translation while allowing angle and alignment changes to facilitate and control correction from the apex while retaining mobility at either end of the construct (Figure 2). The L-shaped fixation members traverse a vertebral body and avoid spanning motion segments vertically such that when a correction force is applied to the end of the L-shaped member, simultaneous translation and derotational forces are applied to the vertebrae (Figure 3).
Mobile Heim anchors utilized at upper instrumented vertebra and lowest instrumented vertebra allow relative motion of the rod (eg, bending, flexion-extension, axial slide).
(a) Depiction of the deformity correction forces applied through L-shaped member. (b–e) Photographs showing axial correction during tightening of L-shaped member, done at initial surgery.
Once the concave side rod was attached via the articulating Heim anchors and the L-shaped members were secured to the convex apical pedicle screws, deformity correction was obtained by pulling the L-shaped members with a horizontal force toward the concave side rod. The L-shaped member has a wire that comes out horizontally and attaches to a wind-up component on the concave rod. The wire tightens in this apparatus and brings the vertebra over to the rod using both derotation and translation forces. After achieving visually improved alignment, the apical region was bilaterally stabilized with a short rod-screw construct connecting the screws on the convex side. The end result was bilateral rigid fixation of the apical segments, with only the concave side rod extending proximally and distally across nongrafted segments to the UIV and LIV (Figure 4). Local autograft was utilized for fusion of the immobilized apical motion segments with cortico-cancellous allograft filler as needed according to surgeon preference. The motion segments adjacent to the apical region were not prepared for fusion.
Radiograph image of typical postcorrection construct.
Patients were ambulatory without bracing the day following surgery. Activity restrictions after surgery varied with each site, with the general trend toward release to normal activities by 3 months postoperation. Patients were discouraged from participating in strenuous sports during the study period.
Surgical data collection included operative time, blood loss, levels fused, and levels instrumented. All device- and/or procedure-related adverse events were reported.
In this trial, it is hypothesized that motion of the instrumented but unfused segments will be maintained by virtue of the degrees of freedom afforded by UIV and LIV articulating Heim anchors at each end of the rod. By design, the articulating Heim anchors should allow axial compression, rotation, flexion-extension, and lateral bending of the instrumented but unfused segments. Thus, the system is postulated to be a load-sharing device that will make the spine the weakest stabilization link and not the implant itself.
RESULTS
Demographics and Operative Outcomes
The demographics for the 21 subjects in this trial are summarized in Table 2. Operative time averaged 258.0 ± 56.3 minutes (range: 158–410 minutes) with blood loss of 933.3 ± 429.9 mL (range: 300–2000 mL). Intraoperative blood transfusions, done at the discretion of anesthesiology, averaged 1.9 units (range: 1–4 units) of packed red blood cells per patient. Due to AIS surgical management routine protocol of the partaking centers, 19 out of 21 patients received blood transfusion intraoperatively. Postoperatively, 5 units of blood were transfused in 3 patients, each being reported as an adverse event since the transfusions were done after surgery. One of the 3 patients did not have any transfusions during the implant procedure. Intraoperative electrophysiological testing was stable in all monitored patients (19 patient), and there were no neurologic deficits at any point after surgery in the 21 patients. Hospital stay averaged 8.2 ± 2.9 days (range: 4–13days).
Demographics (n = 21).
A mean of 11.3 ± 1.8 pedicle screws (range: 9–14 screws) were used per patient. The average number of pedicle screws per vertebra within the construct (screw density) was 0.97 ± 0.15 screws (range: 0.82–1.27 screws/vertebra). The UIV levels varied: T2 (9 patients), T3 (9), T4 (3); LIV also varied: T11 (1 patient), T12 (3), L1 (7), L2 (8), and L3 (2). Independent CT assessment of all 238 screws placed showed 19 screws (7.9%) had breached pedicle margins to some degree. None were symptomatic, though 1 screw (0.4%) in 1 patient was removed due to a lateral position near the aorta.
The posterior arthrodesis averaged 5.0 ± 0.5 motion segments (range: 4–6), with an average of 10.5 motion segments instrumented per patient. Fusion was accomplished at 104 levels (47% of instrumented levels), and 116 levels were instrumented without attempt at fusion.
Clinical Outcomes
There were no neurologic deficits, deep wound infections, medical complications, loosening of screws, failures of the implant components, or any other device-related complications. Analgesic use at 2 weeks postoperatively was increased over preoperative use (P = .01), but improved compared to preoperative use at 3 months, 6 months, and 12 months postsurgery (Table 2).
The average SRS22r total scores in this trial were 3.7 (range: 1.9–4.5) at baseline, 4.1 (range: 2.4–5.0) at 6 months, 4.2 (range: 2.9–5.0) at 12 months, and 4.2 (range: 2.6–5.0) at 24 months. At 12 months postsurgery, 19/21 patients (91%) stated they were either very satisfied or satisfied and 2/21 noted they were neither satisfied nor unsatisfied. Patients were also asked if they would have the same procedure again, 19/21 patients (91%) were definitely or probably sure they would have this treatment again, 2/21 were unsure.
Radiographic Outcomes
A representative series of anteroposterior (AP) standing radiographs is in Figure 5. The main thoracic Cobb angle and percentage of improvement in Cobb angle are summarized in Table 3. Overall, the instrumentation and correction methodology utilized in this trial were able to obtain and maintain a 62% improvement in the main thoracic Cobb angle.
Standing AP radiographs. (a) preoperative, (b) postoperative, (c) 3 months postoperative, (d) 6 months postoperative, (e) 12 months postoperative, (f) 24 months postoperative, (g) 4 months postremoval, (h) 10 months postremoval.
Cobb angle of main thoracic curve. The values given are mean ± SD (range).
The values for thoracic kyphosis, lumbar lordosis, sagittal balance, coronal balance, clavicle angle, rib prominence, and apical rotation are summarized in Table 4. Thoracic kyphosis (T5–T12) improved an average 7.6° at 12 months and 8.8° at 24 months compared to baseline with an average increase of 6.2° and 7.3° in lumbar lordosis. Changes in mean coronal and sagittal balance were not substantially different at any of the assessments. Clavicle angle was defined as the angle between the horizontal reference line (HRL) and the line connecting the superior margins of the left and right clavicles. A depressed right clavicle yields a positive angle, while a right clavicle above the HRL will give a negative angle. Clavicle angle changed 4.6° at 12 months (−0.8° preoperatively to 3.8° at 12 months) and 2.8° at 24 months (−0.8° preoperatively to 2.0° at 24 months). Mean rib prominence decreased 38% (16.7° versus 10.4°), the mean apical vertebral body rotation as measured on CT increased 2.5° at 12 months compared with preoperative values, from 13.3° ± 5.8° (range: 4°–28°) to 15.8° ± 5.2° (range: 7°–29°).
Sagittal, rotational, and balance measurements. The values given are mean ± SD (range).
In levels instrumented but not fused, motion measured based on the description of Hosseini et al1 was 26° preoperatively compared to 10° at 1-year postsurgery demonstrating 38% maintenance of mobility in the nonfused segments.
Implant Safety and Adverse Events
There was one serious adverse event and 17 nonserious adverse events observed during the first 12 months. The serious adverse event required revision surgery for removal of an asymptomatic malpositioned pedicle screw identified on CT scan. There were no additional revision surgeries and there were no implant failures reported. Nonserious events included a single superficial wound infection and a urinary tract infection healed with antibiotics. One patient had a scab remaining on the surgical incision at 3 months, which spontaneously resolved. Three patients developed postoperative anemia, which resolved after blood transfusion. One patient developed short-term self-resolving dizziness 6 months postoperatively for no identifiable reason. One patient developed a self-resolving back bruise after falling from a bike.
Nine of 21 patients reported back pain at some point during the 12 months follow-up. Minor pain began around 3 months post-operatively in 2 patients, associated with the ends of the construct, one of which resolved spontaneously. Two patients had nonfocal back pain before surgery and used analgesics. Both noted slightly worsened pain after surgery and continued using analgesics at a low or moderate frequency at 12 months. Temporary rib discomfort was described in one patient 4 months postoperatively, which resolved by 12 months. At 9 months postsurgery, 3 additional patients developed non–procedure/non–implant-related back pain after falling, after extensive coughing, and after being “kneed” in the back by a friend. At 12 months postsurgery, only the cough-related pain remained. Prior to surgery, 16/21(76%) of patients described no pain. At 12 months, 17/21(81%) were pain free.
DISCUSSION
Long-instrumented fusions have been the standard surgical treatment for AIS since the introduction of Harrington rods in the 1960s.22–24 Improved understanding of the forces acting on the deformity2,25 has led to refinements of correction strategies, fixation devices, and implant materials.4,7,26–33 More recently, the possibilities of shortening the required arthrodesis33–35 using less soft tissue dissection,16,36 and using fewer implants18,37–40 have been investigated in an effort to decrease fusion-related morbidity,13,14 retain maximum spinal motion, and decrease costs.17 The concept of dynamic segmental fixation to guide spinal growth without fusion has also been explored.41
This pilot trial prospectively evaluated a novel correction process and an improved spinal instrumentation system that utilized decreased soft and bony tissue dissection, lowered implant density, allowed apical derotation and translational deformity correction, and limited apical arthrodesis in AIS. Surgical correction was achieved using unique rod-screw connections, which facilitated derotational and translational correction of the apical vertebral bodies. The correction was maintained while allowing motion at the instrumented but unfused segments outside the apex, like the Shilla technique used in juvenile idiopathic scoliosis.42 Though not a true minimally invasive surgery (MIS),36 subperiosteal dissection for these patients was limited only to the apical motion segments, UIV, and LIV intended for fusion. Also, unlike MIS techniques, all pedicle screws were placed under direct landmark visualization without the need for extensive fluoroscopy. Since the technique described here keeps half of the motion segments beneath the instrumentation unfused and mobile, the option of removal of the instrumented but unfused levels after maturation is encouraged. Regarding the indication for implant removal in this new technique, it is important to mention that at the time of current study the regulations climate was not in favor of leaving any implant in situ without fusion at skeletal maturity. In addition, at the time of the study, the authors believed that leaving implants in without fusion would both decrease range of motion of spine and increase the possibility of implant failure, which both will be reflected in results of the new system. Longer-term results will certainly answer whether implant removal at skeletal maturity will improve the results or not.
The only procedure-related adverse event was the single malpositioned pedicle screw, which required revision. There were no associated sequelae with the removal surgery, and deformity correction minimally worsened. The 17 nonserious adverse events that were noted during 12 months of follow-up reflected the investigators' close patient scrutiny to include even minor and unrelated back complaints not normally reported as complications. It was also noted that blood loss and hospital stay were longer than expected, which might be due to the learning curve that occurs in any new technique. In addition, in this study PASS was performed in different centers in 4 different countries. Longer hospital stays can be attributed to different hospital stay protocols. The blood loss issue was a determinant decided by each center and they attempted to keep blood count above a level that was more numeric than symptomatic; anesthetic control techniques may also have varied between centers.
The primary efficacy objective was the correction of the main thoracic curve compared to baseline. Curve correction in this series (62.2% at 12 months and 62.0% at 24 months postsurgery) was similar to the deformity correction achieved using standard surgical techniques,12,42–45 but was achieved while limiting fusion at 4 to 6 levels; arthrodesis of 9 to 12 levels is common with standard AIS surgery. As it is shown in Table 4, there were changes in coronal balance, sagittal balance, and spinal height (T1–S1 height). However, as is supported by P values, these changes are not statistically significant (P > .05). In addition, postoperative thoracic hypokyphosis as reported by others 43,46–50 was not noted in this clinical trial. Thoracic kyphosis in this clinical trial increased an average 7.6° at 12 months and 8.8° at 24 months with the described technique.
A secondary efficacy objective was the difference in domains in the SRS22r 51,52 questionnaire from baseline. In the evaluation of scoliosis the SRS22r is a disease-specific instrument to assess health-related quality of life. It is composed of function, pain, self-image, mental health, and satisfaction. The average score of these 5 components showed improvement in this study, which was in line with what others have noted.8
The data reported in this clinical trial are sufficient to address safety and efficacy through the first 24 months and immediately following system removal. Future study is needed to monitor deformity correction stability as well as to compare spinal column motion of the instrumented but unfused segments compared to the motion of these segments after the long stabilizing rod is removed.
In the big picture, this concept of correction may be advanced to allow for apical correction, curve stabilization, and removal of implants at maturation in AIS patients as the system allows for more biomechanical freedom when implants are in place. This is accomplished by having horizontal implant capture and connection at several points compared to the present-day vertical segmental fixation methods spanning multiple levels. The technique reported is designed to allow more freedom for potential structural remodeling of bone and soft tissue (disc) that could lead to correction, stabilization, and greater maintenance of spinal mobility. In the future, it may be found that deformity correction with changes in load sharing and continued spinal motion may lead to bone and soft tissue remodeling and force redistribution of imposed alignment. This may mimic a limb-growth stapling procedure or the Ilizarov deformity correction methodology.
CONCLUSION
Safety and efficacy outcomes data over 24 months suggests this novel posterior technique for the treatment of AIS achieved similar deformity correction as standard surgical techniques but required fewer surgical implants and a shorter arthrodesis. Quality-of-life total scores as measured by SRS22r were improved at 6, 12, and 24 months compared to baseline. There was one device/procedure–related adverse event (asymptomatic malpositioned pedicle screw) that required revision as a safeguard. The results of this clinical trial through 24 months present a safe and effective new posterior technique for the treatment of AIS that has great potential to enhance and evolve the biomechanical solutions presently available for treating AIS.
Footnotes
Disclosures and COI: K-Spine Inc. funded this study from 2012-14 prior to being acquired by K2M in 2015. L. Rehák received grants/research support from DePuy Synthes Spine and Kspine. M. Repko received grants/research support from Kspine. M. Grevitt received grants/research support from DePuy Synthes Spine and Medtronic and acted as a consultant for DePuy Synthes Spine. U. Aydinli received grants/research support from Kspine. A. Carl served on an advisory board or panel for K2M and Kspine and acted as a consultant and received other financial or material support (royalties, patents, stock options, etc.) from Kspine. C. Nnadi has no conflicts. B.A. Akbarnia was a stock or shareholder for Alphatec and Nuvasive; received grants/research support from DePuy Synthes Spine and Nuvasive; was a consultant for Ellipse Technology, K2M, and Kspine; served on an advisory board or panel for Kspine; and received other financial or material support (royalties, patents, stock options, etc.) from Kspine, Nuvasive, K2M, Ellipse Technology, and Depuy Synthes Spine. D. Crandall was a consultant for Ellipse Technology, Kspine, Medtronic, Spinewave, Stryker, and Zyga; served on an advisory board or panel for Kspine and on a speaker's bureau for Stryker; and received other financial or material support (royalties, patents, stock options, etc.) from Kspine and Medtronic. For the rest of coauthors nothing was disclosed. This study was Institutional Review Board–approved at each participating institution prior to initiation. The medical device stated in this manuscript is not cleared by US Food and Drug Administration.
- ©International Society for the Advancement of Spine Surgery
REFERENCES
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.