Growing rods are widely used to treat EOS, but few studies have been conducted to investigate characteristics and limitations of the proximal anchor points. Our results showed that fixing 3 proximal vertebral bodies with 6 anchors provides better outcomes than fixing 2 vertebral bodies with 4 anchors when using growing rods for the treatment of EOS, and is associated with a lower rate of instrument-related complications. Importantly, use of 6 anchors resulted in better outcomes of severe scoliosis and kyphosis without affecting growth of the spine.
A study of 107 patients with EOS treated with growing rod by Helenius et al. showed that patients with severe scoliosis (mean main Cobb angle = 101°) had a higher complication rate than those with moderate scoliosis (mean main Cobb angle = 67°), and they concluded that severe EOS can be effectively treated with growing rods but with a high risk of complications.14 About 30% of patients with EOS have combined thoracic kyphosis, which is the main risk factor for internal fixation-related complications and increases the overall risk of complications 3-fold.15,16 While various complications are associated with growing rods for the treatment of EOS, rates are reduced with careful patient selection.10
In our practice we also observed that patients with severe scoliosis combined with kyphosis treated with growing rods had a higher anchor point-related complication rate. For this reason, for children with large coronal major scoliosis, especially for children with large thoracic scoliosis, we added 2 proximal anchor points for fixation to reduce the possibility of complications and hence increase the effectiveness of the surgery. We chose to add anchor points at the proximal end rather than the distal end of the rods because combined proximal kyphosis will cause screws pullout because the distraction force and the screws are in the same direction. This problem does not exist at the distal end of the rods (in the thoracic-lumbar or lumbar spine). In addition, the distal vertebral bodies are larger than the proximal, thus thicker and longer pedicle screws can be used at the distal end of the rods to increase the anchoring force.
Our results showed that at the last follow-up the Cobb angle of the main curve in both groups was significantly reduced compared with the preoperative values, and the height of T1-S1 was significantly increased compared with the preoperative values. These results suggest that the 2 fixation methods aree both effective in controlling coronal scoliosis, while allowing for appropriate spinal growth. The results also showed that patients who received fixation of 3 proximal vertebral bodies received a greater benefit with respect to correction than those who received fixation of 2 proximal 2 vertebral bodies.
At the last follow-up, the Cobb angle of thoracic kyphosis was significantly decreased as compare to the preoperative value in the Proximal 6 group but not in the Proximal 4 group. These results indicate that fixing 3 proximal vertebral bodies can improve the sagittal TK, but this cannot be guaranteed when only 2 vertebral bodies are fixed. Furthermore, the changes in the Cobb angle of the main curve and the Cobb angle of thoracic kyphosis from preoperative to last follow-up were significantly greater in the Proximal 6 group than the Proximal 4 group. Taken together, the findings suggest that fixing 3 proximal vertebral bodies provides greater corrective ability for coronal main scoliosis and sagittal thoracic kyphosis. The improved correction of fixing 3 proximal 3 vertebral bodies may be attributed to the fact that the corrective force at the proximal end of the rods is dispersed to 6 pedicle screws, generating a greater distraction force during the lengthening procedures.
At the last follow-up, the rate of instrument complications, and the rates of proximal and distal screw loosening, displacement, and pullout were all significantly lower in the Proximal 6 group than the Proximal 4 group. These results suggest that for patients with severe kyphosis, the proximal 3 vertebral body fixation method can effectively reduce screw- and instrument-related complications, thus maintaining a constant distraction force. It is worth noting that among anchor point abnormalities, proximal anchor point abnormalities account for 56.52% (13/23).
None of the patients in this study had evidence of PJK (increase of PJA ≥ 10°). Some studies, however, have reported PJA increases of 20° or higher.17 Pan et al. reported that in patients with EOS treated with growing rods the incidence of PJK was 28%, and the independent risk factor was locating the proximal anchor points at the T2 spinal segment.18 Wantanabe et al.17 reported risk factors for PJK included a lower instrumented vertebra at or cranial to L3 (odds ratio [OR] = 3.32), a proximal thoracic scoliosis of ≥ 40° (OR = 2.95), and a main thoracic kyphosis of ≥ 60° (OR = 5.08). These findings indicate that although increasing the number of proximal anchor points can reduce anchor point-related complications, it also increases the risk of spinal cord injury, and in addition proximal screw placement is relatively difficult because proximal pedicle screws are thin. Thus, in addition to increasing the number of anchor points, an appropriate proximal spine segment and pedicle screws should be selected for the anchor points. Consistent with a report by Li et al.,19 our results showed that treatment with growing rods has no effect on coronal imbalance.
There are limitations to this study that should be considered. The numbers of patients were relatively small, and the length of follow-up was relatively short for the treatment of EOS. In addition, radiographic parameters were determined by a single observer and no analyses of inter-observer or intra-observer reliability were performed.