Preoperative Halo Traction for Severe Scoliosis

Chang Ju Hwang, MD, PhD; Dong Gyun Kim, MD; Choon Sung Lee, MD, PhD; Dong-Ho Lee, MD, PhD; Jae Hwan Cho, MD, PhD; Jae-Woo Park, MD; Jong Min Baik, MD; Kwan Bum Lee, MD


Spine. 2020;45(18):E1158-E1165. 

In This Article


Treating severe scoliosis is difficult and involves multiple risks. In the past, severe scoliosis was treated by an anterior release followed by posterior correction and instrumented fusion.[7] HT was usually applied following anterior release for good correction in a staged operation. Several studies have reported that the combined anterior and posterior procedure has increased the length of hospital stay and has had a negative impact on pulmonary function.[8] These traditional surgical techniques have changed with the development of modern surgical instrumentation. Luhmann et al[9] reported that 84 patients, who had from 70° to 100° coronal Cobb angles, were treated with pedicle screw-only constructs, and presented similar results who treated anterior–posterior spinal fusion. Recently, Crostelli et al[10] reported that 25 patients, with 95° coronal Cobb angles were treated by posterior spinal fusion with pedicle screw-only instrumentation and corrected with a 37° Cobb angle postoperatively. In severe scoliosis, the use of HT is decreasing due to the reduction of anterior release, but HT is still necessary because excessive and sudden correction of severe spinal curvature with modern instrumentation increases the risk of spinal cord injury and paralysis.

Since Perry and Nickel introduced the halo device in 1959,[11] several studies have reported on the possible safety and effectiveness of treating severe spinal curvature through preoperative HT.[3,12,13] In a study by Nenami et al,[14] the mean preoperative coronal Cobb angle of 29 patients was 136°, but it decreased to 90° after halo traction and 57° after final corrective surgery (56% correction). Sink et al[15] performed preoperative halo-gravity traction for 6 to 21 weeks on 19 patients. The Cobb angle improved from 84° preoperatively to 55° (35% correction) after HT and finally to 51° after the final corrective surgery. Tigran et al[16] corrected 21 patients with a mean preoperative angle of 101° to 69° after traction and 58° after the final surgery. Saeedreza et al[17] corrected 23 patients with severe scoliosis from a mean preoperative angle of 99.9° to 75.5° after halo-gravity or halo-femoral traction and to 49.5° after one-stage posterior fusion operation. Our study also revealed effective curvature correction, as found in previous studies. The mean preoperative angle of 96.9° was corrected to 63.3° after HT and to 32.5° postoperatively, with a correction rate of 65.9%. No serious complications were noted during the traction or surgery.

The mean preoperative Cobb angle was 96.9°, and curves of this degree are not regularly treated with traction when modern surgical instrumentation is available. Although some patients with severe scoliosis are now treated without HT following the development of modern surgical methods using pedicle screws, we believe that preoperative traction must be used for safety even if the Cobb angle is <90° when the deformity is very rigid or has a kyphotic component. Most of the patients had a severe deformity with a Cobb angle >100° in this study.

It is necessary to consider several factors for safe and effective HT, although not many studies have been conducted in this area. First, in a study about the appropriate duration of HT, Park et al[18] measured the correction rate of the curvature every week in 20 patients with severe scoliosis who underwent preoperative HT with a mean angle of 84.7°. They reported that most of the correction occurred within the first 2 weeks of the procedure. Letts et al[4] reported that, in 10 patients with a mean of 81° scoliosis, halo or skull tong and femoral traction were performed for 2 to 3 weeks, and most of the curvature correction occurred within the first week. Bogunovic et al[12] reported that the curvature of 33 patients with a mean angle of 96.4° was corrected to 66% after 2 weeks and to >91% after 3 weeks. They reported that the maximum correction was achieved in an average of 42.6 days. In our study, most of the correction occurred within 1 week (28.2%), and statistically significant changes in the curvature occurred up to 2 weeks (34%). These findings suggest that traction for ≥3 weeks is not necessary for optimal results. However, because the purpose of preoperative HT is correction of curvature and restoration of pulmonary function and nutritional status before corrective surgery, various other factors should be considered to determine the appropriate duration of traction.

Studies on factors related to halo traction are rare. Sink et al[15] reported that the effect of traction was independent of the diagnosis, age, and previous surgery history in their study on halo-gravity traction performed on 19 patients for 6 to 21 weeks. Koller et al[13] reported that the duration of traction did not affect the total correction rate during HT. Our results agree with this observation. Additionally, patients in group B (PBC < PTC) had lower height, lower body weight, higher TBWR, and more halo-femoral traction than patients in group A (PBC ≒ PTC). This suggests that for shorter patients and those who weigh less, halo-femoral traction with a heavy traction weight was effective in treating severe scoliosis. The authors found that these results may be related to the mechanical properties associated with spinal curvature. Each vertebral joint acts as a fulcrum in traction, and tall patients are thought to require more force due to the longer lever arm from the apex of the spinal curvature. Also, the taller and heavier the weight, the larger the size of each vertebral body, and the facet joint can become resistance to traction.

It is well known that the traction force of the pelvis and femur is greater than that of gravity.[2,12,15] Applying more traction force can be effective in improving spinal curvature; accordingly, TBWR should be larger, and the method of traction should be stronger. However, several studies recommend that TBWR should not exceed 50%.[3,13] These studies suggest that high tractional forces can cause problems, such as cranial nerve palsy, loss of cervical lordosis, and degenerative changes of the apophyseal joint.[19–21] Thus, our study adhered to these principles for all patients, and we could not confirm the association between patients with complications and high TBWR. Additionally, pelvic and femoral traction are considered more invasive than gravity traction, and complications, such as pin site problems, are very common.[22] Therefore, careful selection of the traction method is necessary. In this study, halo-femoral traction resulted in fear and discomfort among patients, and it was performed only in the early days of the study period.

This study had some limitations. First, this retrospective study was conducted using medical records and radiographic analysis. Second, there was no comparison control group of patients who had not undergone traction. However, failure to apply HT in treating severe scoliosis may increase the perioperative risk among patients. Thus, no appropriate control group could be selected in our center. Third, patients belonging to the neuromuscular scoliosis group and the syndromic scoliosis group were grouped in one category, despite the differences in their characteristics and medical history. This was inevitable because the number of patients with severe scoliosis and a single disease was small. These limitations can be overcome through multi-institutional research in the future. Fourth, no data related to preoperative HT and pulmonary function were included. An advantage of preoperative HT is that it improves pulmonary function before surgery to mitigate the inherent risks of surgery, but our study did not consider preoperative lung function. However, our study is the first comprehensive study of HT-associated factors, which have only been partially suggested in the past, and can be supplemented by further studies. Additionally, considering the lack of related studies and data owing to the small number of patients with scoliosis who require HT, our findings may help guide surgeons who are struggling with the treatment of rigid scoliosis.

In conclusion, safer and more effective spinal curvature correction can be achieved in severe scoliosis cases through preoperative HT, and most of this correction can occur within 2 weeks. The most appropriate traction period should be determined considering the possible complications of preoperative HT. In patients with low height and weight, halo-femoral traction with a heavy traction weight was effective in treating severe scoliosis. These factors must be considered when using HT for patients with rigid and severe scoliosis.