Derotation screws provide no advantage over polyaxial screws regarding coronal & sagittal correction in thoracic curves of AIS patients

We compared two techniques for thoracic apical derotation; one using conventional reduction screws (Single-Innie–SI) and one requiring special derotation screws that can be converted to monoaxial screws to enhance dorotation (Dual-Innie–DI) for coronal and sagittal correction and. A total of 200 patients with thoracic AIS have been included. In the SI-Group (n = 127) the convex rod was applied first. Vertebral derotation was done by translation to the concave rod with the convex rod being in place and center of rotation (COR). In the DI-Group (n = 73) correction started with translation on the concave side as well but now followed by derotation around the concave rod using the DI-mechanism. The mean rotation according to Raimondi and coronal correction was not sig. affected (72 (± 12) % in the SI-Group versus 68 (± 15) % in the DI-Group), even when flexibility was respected (Cincinnati Correction Index CCI was 2.9 (± 4.9) versus 3.5 (± 4.4). (p < 0.01). The gain of kyphosis was sig greater (2.7°) in the SI-group, but not clinical relevant. The use of DI screws for apical derotation did not provide an advantage for coronal correction or derotation in thoracic curves. Presumably after translation is performed in the DI-group, there was too much tension and friction in the construct impeding further derotation. Simultaneous translation and derotation in the SI-group, with the convex rod being the COR, yielded similar correction with better kyphosis and was faster and more economic.


Introduction
Over the past years many studies addressed apical derotation and direct vertebral derotation (DVD) with different techniques leading to heterogenous results.
In 2004 Lee et al. introduced a technique with monoaxial [1,2] pedicle screws reporting better apical rotation and coronal correction [3]. However, the effect on the sagittal profile remains controversial [4]. Some authors report a negative effect of DVD on the kyphosis while others did not [1,[5][6][7]. Monoaxial screws make the introduction of the contoured rod challenging [8] and might have adverse effects on the sagittal profile. In the aftermath screws were developed that combined the advantages of monoaxial screws with the easiness of polyaxial screws. In this paper we examined the Dual-Innie screws by DePuy, which can be locked after rod introduction converting them from polyaxial to monoaxial screws, so that the screw itself can be used for derotation.
In our paper two derotation techniques will be compared; one using Dual-Innie screws, as described above, and one using conventional Single-Innie (SI) polyaxial long-tab 1 3 screws. Here derotation was achieved by rod bending and translation in a technique described below.
The aim of our study is to find out if one technique is superior and whether special derotation screws are needed.

Study design & patient sample
This is a single surgeon retrospective comparative study. The data of 200 AIS patients with structural thoracic curves (Lenke I, II, III, IV) that have received thoracic spinal fusion by a single surgeon have been reviewed. The follow-up period was at least 1 year for all patients. Only patients that have received posterior spinal fusion with all-screw constructs and without additional anterior approach have been included. These were divided into two groups according to the derotation technique. In the SI-group of 127 patients common polyaxial reduction screws with a single-Innie have been used. In the DI-Group of 73 patients the screw head has a special Dual-Innie locking mechanism, allowing to convert them to monaxial screws.

Implants
All screws were part of the Expedium system (DePuy). So, rods, tools and other equipment were the same. In both groups long-tab reduction screws were placed on the concave side and short-tab screws on the convex side of the apex. The main difference was that in the DI-group the screws had a different head and dual locking Innies. Furthermore, the screw head had a favored angle so it could move to one side more than to the other. Regarding the Dual-Innie screws, two versions have been used. Initially DIFAR screws were used until the next generation of Dual-Innie screws, named 'Expedium VERSE,' were introduced in 2015. The main difference was that the head was smaller. 12 patients were operated with this later system.

Correction using the 'Derotation Screws' (Fig. 1)
Here the rod is applied in the Dual-Innie screws on the concave side, after being bend to the aspired sagittal profile and spinal alignment. Correction will then be performed by firstly translating the spine to the concave rod. The apical concave screws will then be locked, converting them into monoaxial screws. Hence the vertebral derotation is performed around the concave rod, being the center of rotation. The more flattened convex rod is then applied on the convex side, where a cantilever maneuver can be performed to support the vertebral derotation. The derotation screws will be finally locked by tightening the inner setscrews. Fine-tuning can be performed by distraction on the concave and compression on the convex side.

Correction using the 'Single-Innie long-arm screws' (Fig. 2)
Again, the rod is introduced in the long-arm Single-Innie screws on the concave side, followed by derotation of the rod and initial translating the spine to the concave rod. The flatter convex rod is then applied and a cantilever maneuver is done to correct the rib hump. By translation to the overcontoured concave rod the vertebra rotates around the convex rod, which thus is the center of rotation. Distraction on the concave and compression on the convex side can be done if needed.
The main difference between the two techniques, is that in the DI-Group the whole correction is done from the concave side. In the SI-Group vertebral derotation will be performed around the convex rod. Derotation and translation are done

Radiographic measures
The radiographic outcome measures were thoracic and lumbar Cobb angle, thoracic kyphosis, apical vertebral translation (AVT), apical vertebral rotation according to Raimondi [11], and coronal imbalance with truncal shift defined as the lateral deviation of the C7 plumbline from the central sacral vertical line (CSVL), in millimeters. All these measures were done pre-and postoperatively by an independent observer. Percentage postoperative correction (POC), preoperative flexibility (PF) and the Cincinnati Correction Index (CCI) which is the percentage of correction in relation to the flexibility [5] were calculated.

Implant and surgical details
The types of screws (Single-or Dual-Innie), the material of the rods, number of fused segments, the UIV and LIV have been recorded. Intraoperative Neuromonitoring (IOM) with motor evoked potentials (MEP), was used in all cases to control neurological function. All patients received postoperative epidural anesthesia.

Complications
Surgery related complications were recorded. These included perioperative and late wound infections, wound healing disturbances, instrumentation-related complications such as screw loosening, and rod breakage.

Statistical methods
Data were statistically described in terms of mean ± standard deviation (± SD), median and range, or frequencies (number of cases) and percentage when appropriate. Numerical data were tested for normal assumption using Kolmogorov Smirnov test. Comparison of numerical variables between the study groups was done using Student t test for independent samples. For comparing categorical data, Chi-square (χ 2 ) test was performed. Exact test was used instead when the expected frequency is less than 5. Two-sided p values less than 0.05 was considered statistically significant. All statistical calculations were done using computer program IBM SPSS (Statistical Package for the Social Science; IBM Corp, Armonk, NY, USA) release 22 for Microsoft windows.

Demographic data and curve patterns
200 patients were included in the study. Regarding the perioperative data including age, gender, length of operation, length of hospital stay, Lenke classification [12], including lumbar & sagittal modifiers, Risser score [13] and curve patterns no significant differences between the two groups were found. The data are presented in Tables 1 and 2. There were slightly more kyphotic and less lordotic patients in the DI-group, without reaching significance.

Coronal correction
The mean postoperative correction (POC) was 71% (± 13%), and the mean CCI was 3.1 (± 4.7) for all patients.  No sig. difference was found in preoperative flexibility between the SI-Group (38 ± 19%) and the DI-Group (33 ± 22%). The postoperative correction (POC) was slightly higher in the SI-Group (72% ± 12%) than in the DI-Group (68 ± 15%). In order to compare the correction in correlation to the flexibility the CCI was calculated. It was 2.9 (± 4.9) in the SI-Group and 3.5 (± 4.4) in the DI-Group. The postoperative change in rotation in relation to the preoperative value was nearly similar in both groups (11 ± 8 in the SIgroup, 10 ± 8 in the DI-group) The differences between the 2 groups in all the preoperative and postoperative coronal parameters were statistically insignificant (Table 3).

Sagittal correction
Sagittal kyphosis was lower in the SI-group preoperatively, but was restored to similar postoperative values (Table 3). While the kyphosis did not change postoperatively in the DI-group (− 0.4), it increased in the SI-group by 2.3°. The difference was statistically significant (p = 0.013), however, so small that it was of no clinical relevance.

Implant and surgical details
Thoracoplasty for rib hump correction was not performed, but a release of the costotransversal joints in curves exceeding 60° was done. The mean length of fusion was 9.1 (± 1.9) segments. The differences between the 2 groups regarding length of fusion, the use of connectors and the need for rib head release were not statistically significant.

Complications
17 out of the 200 patients had one complication. This included 3 wound infection, 4 wound healing disturbances, 9 screw loosenings and 1 case of rod breakage. No major complications have been observed. The differences in complications' rate between the two groups was not statistically significant.

Discussion
The aim of scoliosis surgery is to correct not only the sagittal plane, but to restore the thoracic kyphosis and derotate the apex. A technique for apical derotation was first introduced by Lee et al. [3] Since then, a variety of techniques for apical derotation was described using monoaxial, uniplanar or poliaxial screws with special derotators [1,4,[14][15][16][17].
The senior author of this paper started early with derotation techniques [8] hoping that this would solve the problem of the three-dimensional correction. As this technique which needed special derotation screws was more expensive and some publications suggested that derotation around the concave rod would cause a flattening of the thoracic kyphosis [4], it was abandoned after several years. An alternative approach with conventional polyaxial reduction screws and a different correction strategy was established in which rotation was done around the convex rod with simultaneous translation to the concave side. In a preliminary series [18] we found more kyphosis than by rotating around the concave rod with the DI screws. As conflicting publications about the pros and cons of direct vertebral derotation exist, we now compare these two groups in a larger series which were otherwise comparable and were all operated by the same surgeon.
We found no significant difference in coronal correction when using the special Dual-Innie screws compared to the long-tab polyaxial screws. The postoperative coronal correction (POC) was 68 (± 15) % in the DI-Group and 72 (± 12) % in the SI-Group. These results were comparable to those of Vallespir et al. [9] who reported coronal correction of 73.3% and Qiu et al. [19] who reported coronal correction of 71.8%.
As the amount of correction depends on the preoperative stiffness of the curve, the Cincinnati Index (CCI) was used. It compares the percentage of correction related to the preoperative flexibility with a higher CCI indicating more relative correction. It was 3.5 (± 4.4) in the DI-Group compared to 2.9 (± 4.9) in the SI-Group, with the difference not being statistically significant.
Few studies address the amount of derotation achieved as pre-and postoperative CT scans are hard to justify for research issues. Especially, as Ruchton and Grevitt [20] who examined clinical factors found no effect on rib hump correction or patient satisfaction. We therefore measured the rotation of the apex according to Raimondi [11] and could not find a difference in the two groups but are aware that this is not as exact as a CT scan would have been. Di Silvestre et al. [2] performed rod derotation followed by DVD around the concave rod as done in our DI-group and measured rotation with pre-and postoperative CTs. They observed a better coronal correction and derotation with DVD compared to correction by concave rod rotation alone. However, worsening of thoracic kyphosis was seen.
As other authors indicate that derotation will lead to a flattening of the spine as well [7,21], we looked for the sagittal plane in our patients. Here a minimal decrease in thoracic kyphosis in the DI-Group, from 32.3 (± 16) preoperatively to 31.9 (± 10) postoperatively; and a small increase in thoracic kyphosis in the SI-Group from 27° (± 11) preoperatively to 29° (± 10) degrees postoperatively, was seen. The difference was statistically significant, but is not clinically relevant and within the margin of error. The results were superior to those provided by Kim et al. [21] who reported 14 degrees decrease in thoracic kyphosis and Lowenstein et al. [7] who reported 10 degrees decrease in thoracic kyphosis. Watanabe et al. [15] also showed that vertebral derotation caused hypokyphosis of the thoracic spine. This was partly attributed to the wedge deformity of the vertebral bodies, being taller on the convex side and taller ventrally; resulting in shifting the taller convex wall and the taller anterior wall ventrally during the vertebral derotation, thus elongating the anterior column and decreasing the thoracic kyphosis [15]. On the other hand, Urbanski and colleagues [15] showed contradicting results with en-block DVD maintaining, or even improving, the thoracic kyphosis [22]. We believe that this heterogeneity can be explained by different techniques used to perform DVD. Tsirikos et al. [23] for example describes a technique with DI reduction screws, only on the convex side, for both thoracic and lumbar curves, without using screws on the concave side of the apex. He reported a thoracic coronal correction of approximately 71% with an increase in kyphosis of 13°.
The first rod applied will be the center of rotation (COR). If the COR is the concave rod, a derotation will lead to a decrease in kyphosis. A derotation around the convex rod can increase kyphosis as shown in our patients. Furthermore, the sequence of correction steps is important. In the DI-group we started with translation to create kyphosis and correct the coronal curve and then locked the screws to perform derotation afterward. However, after translation is performed, there is a lot of tension on the construct making any further derotation, difficult; thus, minimizing the effect of vertebral derotation that could be obtained. In the SIgroup the convex rod was established as COR with concave translation and convex derotation being performed simultaneously. By connecting both rods with cross links the stiffness of the whole construct was increased before the final correction started. Thus, a slightly better sagittal correction could be achieved.
Although it was disappointing that our initial technique was not as superior as we hoped, it helped us to achieve a better understanding. It must be emphasized that this study not only compares different screw designs but also two correction strategies, as the DI design allows derotation only after the rod was fixed with the first set screw. The same DI mechanism might help in lumbar curves where lordosis has to be increased.
The main limitations of the study would be the retrospective design, and the lack of clinical correlation with the radiographic outcomes (e.g., with SRS-30 and SF-36 questionnaires). The strengths are the number of patients with similar curves and that all surgeries were done by a single surgeon and the same system so that other variables did not influence the results.

Conclusion
According to our experience, segmental apical derotation with Dual-Innie screws does not provide any advantage, regarding coronal or sagittal correction, when correcting thoracic curves in AIS patients. Similar results could be achieved with regular reduction screws using a different technique, which was faster and easier. A possible explanation is that after translation is performed in the DI-group, there is so much tension and friction in the construct that further derotation was impeded.
When performing DVD, a technique in which derotation is done before applying translation could result in better rotational correction.