Staged Halo-Pelvic Traction Combined With Posterior Spinal Fusion versus One-Stage Surgery for Severe Rigid Spine Deformity Requiring Three-Column Osteotomy: A Comparative Study

Abstract

Study design

A retrospective cohort study.

Objective

Severe rigid spine deformity presents major challenges in surgical correction. Although three-column osteotomies (3-CO) provide powerful correction, they are associated with prolonged operative time, increased blood loss, and high neurological risk. This study aimed to compare the radiographic correction, surgical parameters, and complication profiles between staged halo-pelvic traction combined with posterior spinal fusion (HPT+PSF) and one-stage posterior spinal fusion (PSF) in patients with severe rigid spine deformity requiring 3-CO.

Methods

This retrospective cohort study included 110 patients (mean age, 25.7 ± 8.5 years) with severe rigid deformity (Cobb angle or maximal kyphosis >90°, flexibility <30%) who underwent either staged HPT+PSF (n = 60) or single-stage PSF (n = 50) between 2016 and 2023. Radiographic parameters, including the main curve Cobb angle and maximal kyphosis (MK), were measured preoperatively, post-traction, postoperatively, and at the latest follow-up. Operative time (ORT), estimated blood loss (EBL), osteotomy grade, intraoperative neuromonitoring (IONM) alerts, and complications were compared between groups. Pulmonary function tests (PFTs) were evaluated longitudinally.

Results

Following traction, the main curve Cobb angle and MK in the HPT+PSF group improved from 132.4° ± 23.3° to 65.1° ± 20.7° and from 122.3° ± 29.7° to 53.3° ± 24.1°, respectively. At the last follow-up, the correction rate of the main curve and MK were significantly higher in the HPT+PSF group than in the PSF group (62.9% vs 55.2%, P = 0.004; 67.7% vs 54.5%, P < 0.001). The HPT+PSF group demonstrated significantly reduced ORT (298.3 ± 64.7 vs 361.9 ± 76.3 min, P < 0.001) and EBL (490.5 ± 183.3 vs 1010.6 ± 252.1 mL, P < 0.001). The rate of IONM alerts (3.3% vs 20.0%, P = 0.005) and neurological deficits (3.3% vs 14.0%, P = 0.042) were markedly lower in the HPT+PSF group. Although traction-related complications occurred in 16.7% of patients, all were transient and resolved after adjustment of HPT devices.

Conclusion

Staged HPT followed by PSF provides a safe and effective strategy for the management of severe rigid spine deformities. In terms of radiographic correction and perioperative safety profile, this staged approach demonstrated significant advantages over one-stage 3-CO, notably reducing operative time, blood loss, and neurological risks. Preoperative traction effectively decreases deformity stiffness and magnitude, facilitating safer and less invasive surgical correction.

Keywords

halo-pelvic traction posterior spinal fusion severe rigid spine deformity three-column osteotomy intraoperative neuromonitoring

Introduction

Severe rigid spine deformity remains one of the most demanding conditions in spinal surgery, characterized by large angular curvature, minimal flexibility, and frequent cardiopulmonary compromise.^1,2 The primary goal of surgical management is to restore global spinal balance, achieve adequate deformity correction, and relieve cardiopulmonary restriction while minimizing neurological complications. However, due to the rigidity of the deformity and the proximity of neural elements, conventional posterior instrumentation or low-grade osteotomies often fail to provide sufficient correction.

Spine instrumentation and three-column osteotomies (3-CO), including pedicle subtraction osteotomy (PSO), bone–disc–bone osteotomy (BDBO), and vertebral column resection (VCR), have become standard techniques for such cases.^3-5 These procedures enable powerful correction but are associated with extensive operative time, increased blood loss, and a high risk of neurological complications, with reported rates of neurological deficits up to 30%. Thus, strategies that reduce surgical aggressiveness while maintaining correction efficacy are clinically desirable.

Halo–pelvic traction (HPT) provides gradual deformity correction and soft-tissue adaptation before definitive fusion, potentially improving cardiopulmonary function and reducing deformity stiffness.^6-8 Although its use has declined with modern instrumentation, recent evidence suggests that staged HPT followed by posterior spinal fusion (PSF) may achieve satisfactory correction with improved safety profiles compared to one-stage high-grade osteotomy procedures.^9,10 However, direct comparative data between these 2 approaches remain scarce.

The present study aimed to compare the radiographic correction, surgical parameters, and complication profiles between staged HPT combined with PSF and one-stage PSF in patients with severe rigid spine deformities requiring 3-CO. We hypothesized that preoperative HPT could achieve comparable or superior correction while reducing operative time, blood loss, and neurological complications.

Methods

Patient Cohort

This is a single-center retrospective cohort study conducted from January 2016 to January 2023. Patients with severe rigid spine deformity who underwent either staged halo-pelvic traction combined with posterior spinal fusion (HPT+PSF) or single-stage PSF were retrospectively included. All surgical procedures were performed by the same senior spine surgeon. The inclusion criteria were as follows: 1. The coronal Cobb angle of the main curve or the maximal kyphosis greater than 90°; 2. Flexibility less than 30%; 3. Requiring three-column osteotomy (3-CO) based on surgical decision-making; 4. Minimum 2-year postoperative follow-up. Exclusion criteria included a history of active spinal infection, previous spine surgery and intolerance to HPT or PSF. The indications for 3-CO in our center were as follows: (1) fixed coronal and/or sagittal imbalance; (2) flatback deformity; (3) angular spine deformities with deformity angular ratio (DAR) > 20°^11,12; (4) congenital spine deformity with circumferential fusion; (5) accompanied spinal cord compression or myelopathy.¹³ Patients admitted before 2019 primarily underwent one-stage PSF. From January 2020 onwards, the staged HPT+PSF protocol was recommended for all patients meeting the inclusion criteria. However, to respect patient autonomy and ensure safety, patients in the post-2020 period were assigned to the one-stage PSF group if they: (1) presented with contraindications to HPT (e.g., severe osteoporosis, cervical instability, or skull/skin infections); (2) demonstrated intolerance to the traction device during the initial trial (in the first week); or (3) expressed a strong preference for a single-stage procedure after a detailed explanation of both options. Therefore, this cohort was divided into the HPT+PSF group and the PSF group. Ethical approval was obtained from the institutional review board at Beijing Chao-Yang Hospital (Approval No. 2025-HU-505), and all participants provided informed consent prior to enrollment.

Surgical Technique

Halo-Pelvic Traction

All patients underwent a standardized HPT protocol. Under general anesthesia, the pelvic fixation component was installed by placing 2 pins bilaterally into the iliac crest and the posterior superior iliac spine, waiting for the subsequent connection to a circular pelvic ring (Figure 1A-E). A standard halo ring was fixed to the skull with ten sharp-tipped pins, ensuring correct placement in safe zones around the skull (Figure 1F-G). Four threaded rods connected the halo ring to the pelvic ring, providing the foundation for gradual traction (Figure 2). Traction was initiated at a rate of 3-5 mm/day in the first week and reduced to 0.5-1 mm/day in the following weeks based on patient tolerance. Patients were closely monitored for neurological deterioration, pin-site infection, and mechanical complications. Daily neurological assessments and weekly radiographs were performed to evaluate the traction’s effects and detect potential complications. Neurological symptoms were treated by immediately reducing the traction distance by 10 mm and simultaneously administered methylprednisolone as well as mannitol via intravenous infusion. Gradual resumption could be tried if symptoms resolved. Once maximum safe correction was achieved via HPT, PSF surgery was performed with the HPT device sustained. The decision to terminate traction and proceed with definitive PSF was based on the following criteria: (1) achievement of the targeted radiographic correction (approximately 50% for the main curve or kyphosis); (2) observation of a correction plateau, indicated by no further significant improvement on sequential weekly radiographs; and (3) optimization of the patient’s nutritional status and pulmonary function.

Figure 1.

Illustration of the operating room phase of halo-pelvic traction devices setup. (a) patient positioning; (b-c) pelvic pin insertion in one side; (d) patient repositioning to the opposite side; (e) pelvic pin insertion in the opposite side; (f-g) halo ring installation

Figure 2.

Illustration of the non-operating room phase of halo-pelvic traction devices setup. (a-b) Pelvic ring assembly by combining the 2 semi-circular rings; (c) fixing pelvic pins to the pelvic ring; (d) installing vertical threaded rods to connect pelvic ring and halo ring

Posterior Spinal Fusion

All patients underwent PSF under general anesthesia. The spine was exposed through a posterior midline approach, and pedicle screws were inserted at the planned levels. The osteotomy technique—Smith–Petersen osteotomy (SPO), Ponte osteotomy, pedicle subtraction PSO, BDBO, VCR, or a combination of osteotomies—was selected according to surgical planning. Appropriately contoured rods were connected to the screws, and correction was achieved through vertebral derotation, concave distraction, convex compression maneuvers and thoracoplasty. Fluoroscopy confirmed satisfactory implant positioning and balanced shoulder–pelvic alignment before final tightening. Posterolateral fusion was completed using autologous and allogeneic bone grafts placed along the decorticated lamina and facet joints. Titanium cage with bone grafts was placed for interbody fusion in patients underwent VCR. Representative case examples were presented in Figures 3 and 4. Continuous intraoperative neuromonitoring (IONM) including somatosensory evoked potentials (SSEPs) and transcranial motor evoked potentials (MEPs) was applied throughout the procedure, with MEPs serving as the standard for detecting neurological compromise. An actionable IONM alert was strictly defined as a ≥80% decrease in MEPs amplitude, or a complete loss of the MEPs signal, compared to the established baseline.

Figure 3.

A 38-year-old female with congenital spine deformity who underwent 19-week staged halo-pelvic traction combined with posterior spinal fusion (PSF). The main curve Cobb angle improved from 147.5° preoperatively to 74.2° post-traction, and further to 65.3° at the last follow-up after PSF. The maximal kyphosis (MK) improved from 129.7° preoperatively to 65.0° post-traction, and further to 52.8° at the last follow-up after PSF. The correction rate of main curve Cobb angle and MK was 55.7% and 59.3%, respectively. The deformity angular ratio (DAR) was markedly reduced from 21.1 preoperatively to 10.6 post-traction, and ultimately to 9.3 at the last follow-up after PSF

Figure 4.

A 35-year-old female with congenital spine deformity who underwent one-stage posterior spinal fusion (PSF) and apical vertebral column resection. The main curve Cobb angle was corrected from 132.6° preoperatively to 60.7° at the last follow-up after PSF. The maximal kyphosis (MK) was corrected from 125.9° preoperatively to 56.8° at the last follow-up after PSF. The correction rate of main curve Cobb angle and MK was 54.2% and 54.9%, respectively. The deformity angular ratio (DAR) was markedly reduced from 22.1 preoperatively to 10.1 at the last follow-up after PSF

Evaluation of Radiographic Deformity Correction

Radiographic parameters were assessed using standing posteroanterior and lateral full-length radiographs. In the HPT+PSF group, measurements were obtained at 4 time points: pre-traction, post-traction, postoperative (2 weeks after surgery), and at the latest follow-up. In the PSF group, evaluations were conducted preoperatively, postoperatively, and at the latest follow-up. The measured parameters included the main curve Cobb angle, maximal kyphosis (MK) angle, coronal balance distance (CBD), and sagittal vertical axis (SVA). The deformity correction rate was calculated as the difference between the first and the last measurement points. All radiographic measurements were performed independently by 2 spine surgeons who were not involved in the surgical treatment of the patients. Measurements were conducted on digital radiographs using a standard Picture Archiving and Communication System. The mean values of the 2 measurements were calculated and used for the final statistical analysis.

Evaluation of Surgical Variables and Complications

In the HPT+PSF group, the duration of traction was recorded. Surgical variables, including operative time (ORT), estimated blood loss (EBL), SRS–Schwab osteotomy grade, and intraoperative neurological monitoring (IONM) alerts, were compared between the 2 groups. Traction-related and surgery-related complications were recorded. Neurological deficit was classified based on the duration of symptoms (transient or permanent) and anatomical level of involvement (spinal cord or nerve root). A transient deficit was defined as a new-onset neurological deficit that resolved completely within the follow-up period, whereas a permanent deficit was defined as one persisting at the last follow-up. Nerve root deficits were defined as unilateral motor or sensory disturbances consistent with a specific dermatome or myotome. Spinal cord deficits were characterized by bilateral sensorimotor deficits, long tract signs, bowel/bladder dysfunction, or paraplegia indicative of myelopathy.

Pulmonary Function Test

Pulmonary function tests (PFT) were conducted pre-traction, post-traction, and at the last follow-up in the HPT+PSF group. In the PSF group, tests were performed preoperatively and at the latest follow-up. Parameters including forced vital capacity (FVC), percentage of predicted FVC (FVC%), forced expiratory volume in one second (FEV₁), and percentage of predicted FEV₁ (FEV₁%) were recorded.

Statistical Analysis

All statistical analysis were performed using IBM SPSS Statistics, version 27 (IBM Corporation, Armonk, NY, USA). The Shapiro-Wilk test was performed to determine whether continuous variables had a normal distribution. Continuous variables with a normal distribution are presented as the mean ± standard deviation; otherwise, the median and interquartile range are used. The counts and percentages are presented for categorical variables. Repeated-measures analysis of variance (ANOVA) and post hoc tests were used to compare differences in continuous variables between groups and across different measurement time points; for categorical variables, the Pearson chi-square test or Fisher’s exact test was applied. A two-sided P value of less than 0.05 was considered statistically significant.

Results

Baseline Characteristics

A total of 110 patients (36 males and 74 females) with severe rigid spine deformity were retrospectively recruited in this study. The mean age at surgery was 25.7 ± 8.5 years. Among the patients, 60 were in the HPT+PSF group and 50 were in the PSF group. There were no significant differences in age, sex, BMI, and etiology between the groups. The mean duration of HPT was 20.0 ± 4.9 weeks. The mean follow-up duration was 32.8 ± 8.5 months (Table 1). In preoperative computed tomography evaluations, spontaneous vertebral autofusion at the apical region was identified in 38 patients (63.3%) in the HPT+PSF group and 33 patients (66.0%) in the PSF group (P = 0.771). Additionally, congenital spinal anomalies were observed in 36 patients (60.0%) in the HPT+PSF group and 27 patients (54.0%) in the PSF group (P = 0.526). There were no statistically significant differences in these morphological characteristics between the groups.

Table 1.

Baseline Characteristics

	HPT+ PSF group	PSF group	P
Number of patients	60	50
Age (year)	26.6 ± 9.6	24.6 ± 6.8	0.218
Gender			0.882
Male	20	16
Female	40	34
BMI	19.7 ± 3.2	19.9 ± 2.8	0.729
Etiology			0.540
Congenital	36	27
Idiopathic	19	21
Neuromuscular	4	2
Syndromic	1	0
Follow-up time (months)	30.2 ± 7.4	35.9 ± 8.7	<0.001

HPT indicates halo-pelvic traction; PSF, posterior spinal fusion; BMI, body mass index

Radiographic Parameters

The preoperative main curve Cobb angle was 132.4° ± 23.3° in the HPT+PSF group and 124.8° ± 17.3° in the PSF group. There were no significant differences in preoperative main curve Cobb angle between the groups (P = 0.058). Following HPT, the main curve Cobb angle was significantly improved to 65.1° ± 20.7° in the HPT+PSF group. After the PSF, the main Cobb angle was significantly corrected to 49.7° ± 20.3° in the HPT+PSF group and 57.9° ± 19.3° in the PSF group. At the last follow-up, the correction rate of main curve Cobb angle in the HPT+PSF group was significantly greater than the PSF group (62.9 ± 15.4% vs 55.2 ± 11.7%, P = 0.004). The preoperative MK was 122.3° ± 29.7° in the HPT+PSF group and 115.2° ± 27.6° in the PSF group. There were no significant differences in preoperative MK between the groups (P = 0.197). Following HPT, the MK was significantly improved to 53.3° ± 24.1° in the HPT+PSF group. After the PSF, the MK was significantly corrected to 38.7° ± 16.5° in the HPT+PSF group and 52.5° ± 19.9° in the PSF group. At the last follow-up, the correction rate of MK in the HPT+PSF group was significantly greater than the PSF group (67.7 ± 10.9% vs 54.5 ± 15.2%, P < 0.001). There were no significant differences in preoperative DAR between the groups (21.2 ± 6.4 vs 19.8 ± 5.9, P = 0.183). Following HPT, the DAR was significantly reduced to 10.5 ± 4.8 in the HPT+PSF group (P < 0.001). There were no significant differences in CBD and SVA between the groups at each measurement time point (Table 2).

Table 2.

Comparison of Radiographic Parameters

	HPT+ PSF group	PSF group	P
	(n = 60)	(n = 50)	P
Main curve Cobb angle (°)
Preoperative	132.4 ± 23.3	124.8 ± 17.3	0.058
Post-traction	65.1 ± 20.7^a
Postoperative	49.7 ± 20.3^a,b	57.9 ± 19.3^a	0.035
Follow-up	49.2 ± 19.3^a,b	56.5 ± 18.5^a	0.046
Correction rate (%)	62.9 ± 15.4	55.2 ± 11.7	0.004
MK (°)
Preoperative	122.3 ± 29.7	115.2 ± 27.6	0.197
Post-traction	53.3 ± 24.1^a
Postoperative	38.7 ± 16.5^a,b	52.5 ± 19.9^a	< 0.001
Follow-up	39.3 ± 16.1^a,b	51.7 ± 19.1^a	< 0.001
Correction rate (%)	67.7 ± 10.9	54.5 ± 15.2	< 0.001
DAR
Preoperative	21.2 ± 6.4	19.8 ± 5.9	0.183
Post-traction	10.5 ± 4.8^a
Postoperative	8.0 ± 3.8^a,b	9.2 ± 3.5^a	0.088
Follow-up	8.0 ± 3.7^a,b	9.0 ± 3.5^a	0.137
CBD (mm)
Preoperative	26.5 ± 26.2	23.6 ± 20.1	0.525
Post-traction	21.5 ± 20.8
Postoperative	26.3 ± 22.3	27.4 ± 23.0	0.791
Follow-up	23.1 ± 17.6	24.3 ± 16.0	0.720
SVA (mm)
Preoperative	21.2 ± 22.7	22.3 ± 19.8	0.784
Post-traction	27.6 ± 34.9
Postoperative	26.3 ± 24.3	25.9 ± 28.2	0.937
Follow-up	25.7 ± 17.7	22.9 ± 21.7	0.449

^asignificant difference compared to preoperative meausurement.

^bsignificant difference compared to post-traction meausurement.

MK indicates maximal kyphosis; DAR, deformity angular ratio; CBD, coronal balance distance; SVA, sagittal vertical axis; HPT, halo-pelvic traction; PSF, posterior spinal fusion. Bold indicate statistically significant.

Pulmonary Function Outcomes

There were no significant differences in preoperative FVC, FVC%, FEV1 and FEV1% between groups (P > 0.05). Following HPT, no significant improvement in the outcomes of pulmonary function tests was observed in the HPT+PSF group. At the last follow-up, there were no significant differences in FVC, FVC%, FEV1 and FEV1 between groups (P > 0.05). However, in both groups, the outcomes of pulmonary function tests were significantly improved at the last follow-up compared to preoperative measurement (Table 3).

Table 3.

Comparison of the Pulmonary Function

	HPT+PSF group	PSF group	P
	(n = 60)	(n = 50)	P
FVC (L)
Preoperative	1.6 ± 0.7	1.6 ± 0.6	0.790
Post-traction	1.6 ± 0.6
Follow-up	2.0 ± 0.7^a,b	2.1 ± 0.6^a	0.460
FVC (%)
Preoperative	48.2 ± 18.4	44.1 ± 14.9	0.216
Post-traction	45.9 ± 15.2
Follow-up	62.2 ± 15.7^a,b	58.5 ± 14.4^a	0.205
FEV1 (L)
Preoperative	1.3 ± 0.6	1.3 ± 0.7	0.766
Post-traction	1.4 ± 0.6
Follow-up	1.7 ± 0.6^a,b	1.7 ± 0.6^a	0.575
FEV1 (%)
Preoperative	46.2 ± 17.7	43.2 ± 15.6	0.355
Post-traction	47.9 ± 16.0
Follow-up	62.2 ± 16.1^a,b	59.2 ± 14.1^a	0.309

^asignificant difference compared to preoperative meausurement.

^bsignificant difference compared to post-traction meausurement.

FVC indicates forced vital capacity; FEV1, forced expiratory volume in one second; HPT, halo-pelvic traction; PSF, posterior spinal fusion.

Surgical Variables

The ORT in the HPT+PSF group was significantly shorter than that in the PSF group (298.3 ± 64.7 vs 361.9 ± 76.3 min, P < 0.001). Similarly, EBL in the HPT+PSF group was significantly lower than that in the PSF group (490.5 ± 183.3 vs 1010.6 ± 252.1 mL, P < 0.001). Only 26.7% of patients in the HPT+PSF group underwent osteotomy and 6.7% underwent 3-CO, both proportions being significantly lower than those in the PSF group (P < 0.001). Among patients undergoing osteotomy in the HPT+PSF group, 12 underwent Ponte osteotomy, one underwent pedicle subtraction osteotomy, and 3 underwent vertebral column resection. The incidence of IONM alert in the HPT+PSF group was 3.3%, which was significantly lower than that of 20.0% in the PSF group (P = 0.005) (Table 4). Furthermore, subgroup analysis revealed that patients who triggered IONM alerts during surgery achieved significantly lower final correction rates for both the main curve Cobb angle (51.5% vs 60.4%, P = 0.002) and MK (54.3% vs 62.6%, P = 0.011) compared to those without alerts.

Table 4.

Surgical Variables

	HPT+PSF group	PSF group	P
	(n = 60)	(n = 50)	P
Operating time (minutes)	298.3 ± 64.7	361.9 ± 76.3	<0.001
Estimated blood loss (mL)	490.5 ± 183.3	1010.6 ± 252.1	<0.001
Osteotomy	16 (26.7%)	50 (100.0%)	<0.001
SRS-Schwab grade of osteotomy			<0.001
2	12 (20.0%)	0 (0.0%)
3	1 (1.7%)	3 (6.0%)
4	0 (0.0%)	2 (4.0%)
5	3 (5.0%)	42 (84.0%)
6	0 (0.0%)	3 (6.0%)
IONM alert	2 (3.3%)	10 (20.0%)	0.005
Fusion levels (n)	13.8 ± 1.6	14.1 ± 1.3	0.219

IONM indicates intraoperative neurological monitoring; HPT, halo-pelvic traction; PSF, posterior spinal fusion

Complications

The incidence of traction-related complications was 16.7% (Table 5). Three patients experienced pin loosening during the distraction period. Pin site infection occurred in 3 patients and was successfully managed after dressing and antibiotics. Four patients experienced temporary neurological symptoms during traction: One patient experienced dysarthria and choking on water at 8 weeks of traction, 2 patients complained of numbness in the upper limbs at 10 weeks and 12 weeks of traction, and one patient experienced a decrease in muscle strength in the lower extremity at 16 weeks of traction. Symptoms were relieved in all patients with reduced traction distance. The incidence of surgery-related complications in the HPT+PSF group was significantly lower than the PSF group (P = 0.012). Eleven patients in the HPT+PSF group developed surgery related complications with an incidence of 18.3%. Among them, 2 patients experienced temporary neurological deficit, one patient experienced proximal junctional failure, and 2 patients experienced distal adding-on requiring revision surgery. Twenty patients in the PSF group developed surgery related complications with an incidence of 40.0%. Among them, 5 patients experienced temporary neurological deficit, 2 patients experienced permanent neurological deficit, one patient experienced proximal junctional failure, and 3 patients experienced rod fracture. The incidence of neurological deficit in the HPT+PSF group was significantly lower than that in the PSF group (3.3% vs 14.0%, P = 0.042). It should be explicitly noted that all reported neurological deficits were perioperative events directly associated with the index surgical procedure; no neurological deficits occurred secondary to delayed mechanical complications. The detailed information of neurological deficit in both groups was presented in Table 6. In the PSF group, the complications rate was 37.8% (14/37) during the early period (before 2020) and 46.2% (6/13) during the late period (2020-2023); the incidence of neurological deficits was 13.5% (5/37) during the early period and 15.4% (2/13) during the late period. There were no statistically significant differences in the complication rate (P = 0.867) and neurological deficit rate (P = 0.598) between the 2 time periods.

Table 5.

Complications in Both Groups

	HPT+PSF group	PSF group	P
Traction-related complications	10 (16.7%)
Pin site infection	3 (5.0%)
Pins loosening	3 (5.0%)
Cranial nerve neurapraxia	1 (1.7%)
Brachial plexus palsy	2 (3.3%)
Weakness oflower extremities	1 (1.7%)
Surgery-related complications	11 (18.3%)	20 (40.0%)	0.012
Surgical site infection	1 (1.7%)	2 (4.0%)	0.454
Neurological deficit	2 (3.3%)	7 (14.0%)	0.042
Instrumentation related	3 (5.0%)	4 (8.0%)	0.521
Duar tear	1 (1.7%)	3 (6.0%)	0.227
Intraspinal hematoma	0 (0.0%)	1 (2.0%)	0.271
Pulmonary related	3 (5.0%)	3 (6.0%)	0.818
Acute gastritis	1 (1.7%)	0 (0.0%)	0.359

HPT, halo-pelvic traction; PSF, posterior spinal fusion.

Table 6.

Detailed Information of Neurological Deficits in Both Groups

Group	Anatomical origin	Clinical presentation	Outcome	No. of cases	Management & Prognosis
HPT+PSF	Spinal cord	Bilateral lower extremity weakness	Transient	1	Recovered with conservative therapy (3 mo).
HPT+PSF	Nerve root	Transient radiculopathy (L4 weakness)	Transient	1	Recovered with conservative therapy (2 weeks).
PSF	Spinal cord	Paraplegia	Permanent	2	Incomplete recovery at last follow-up
	Spinal cord	Bilateral lower extremity weakness	Transient	4	Recovered with conservative therapy (6 month to 2 year)
	Nerve root	Transient radiculopathy (L2 weakness)	Transient	1	Recovered with conservative therapy (6 month)

HPT, halo-pelvic traction; PSF, posterior spinal fusion; mo, month; yr, year.

Discussion

The management of severe rigid spine deformities remains one of the most formidable challenges in spine surgery, largely due to the technical complexity of deformity correction and the elevated risk of neurological injury. Since its introduction in 1971, HPT has served as an important modality in the preoperative management of complex spinal deformities. Nonetheless, its clinical use has progressively declined with the advancement and popularization of modern internal fixation techniques. Patients with severe rigid spine deformities frequently exhibit cardiopulmonary insufficiency, gastrointestinal disturbances, and poor nutritional status, all of which exacerbate the risks associated with one-stage corrective procedures. Such operations are often characterized by prolonged operative duration, excessive intraoperative blood loss, and a potential for irreversible spinal cord injury. Moreover, achieving satisfactory correction through a one-stage procedure and 3-CO remains technically demanding and is often associated with disproportionately high surgical risks and morbidity when the Cobb angle exceeds 100°. The findings of this study suggest that a staged strategy incorporating HPT followed by PSF provides a more effective and safer alternative for addressing the multifaceted challenges.

The primary goal of surgical treatment for spine deformity is to achieve satisfactory correction and restore global spinal balance. However, in cases of severe rigid spine deformities, relying solely on spinal instrumentation and low-grade osteotomies often yields suboptimal outcomes due to limited correction potential. To address this challenge, 3-CO techniques have been developed to achieve more powerful corrections. Among these, VCR provides the greatest corrective capability. A study by Riely et al reported that the major radiographic deformity was reduced by 61.6% in the pediatric cohort and by 53.9% in the adult cohort.¹⁴ A systematic review by Yang et al suggested that the mean amount of correction by PVCR was 64.1° in scoliosis and 58.9° in kyphosis, accounting a correction rate of 61.2% and 63.1%, respectively.¹⁵ Despite these impressive outcomes, 3-CO is technically demanding and associated with substantial surgical morbidity. Preoperative HPT offers an alternative strategy by gradually reducing the magnitude and rigidity of the deformity before definitive fusion, thereby decreasing the amount of correction required in the second-stage surgery. However, studies investigating the efficacy of staged HPT followed by PSF remain limited. Zhou et al reported that preoperative HPT combined with multi-level Ponte osteotomies and PSF could yield an overall correction rate of 58.3% in the coronal plane and 63.4% in the sagittal plane, which is comparable to previous studies reported for 3-CO.² A recent comparative study by Zhao et al reported that there were no significant differences in correction effects between HPT+PSF and one-stage PSF with PVCR for severe rigid spine deformity patients.⁹ In the current study, even combining apical 3-CO and Ponte osteotomies to maximize release in the PSF group, patients in the HPT+PSF group achieved a more greater correction rate of main curve Cobb angle and MK. Following HPT, the main curve Cobb angle and MK were corrected to 65.1° ± 20.7° and 53.3° ± 24.1°, respectively. Although definitive PSF had not yet been performed, the correction achieved showed no significant difference compared with that observed in the PSF group after surgery (main curve Cobb angle, P = 0.062; MK, P = 0.856). Therefore, low-grade osteotomies, rather than apical 3-CO, were performed to further improve correction in the HPT+PSF group. These findings indicate that staged HPT combined with PSF can achieve superior deformity correction to one-stage 3-CO procedures, while potentially reducing surgical aggressiveness and technique requirements. It must be noted that the occurrence of IONM alerts inherently limits the ability to achieve maximal correction. When an alert is triggered, corrective maneuvers must be immediately halted or reversed to restore neural conduction and prioritize neurological safety, thereby compelling the surgeon to accept a suboptimal radiographic alignment. Our subgroup analysis confirmed this clinical compromise, demonstrating significantly lower final correction rates in patients who experienced intraoperative alerts compared to those who did not.

The choice between HPT and Halo-Gravity Traction (HGT) remains a topic of debate. While HGT is widely favored for its less-invasive nature, our center employs a strict stratification strategy based on patient age, etiology, and deformity rigidity. HGT is primarily indicated as the first-line treatment for pediatric patients, particularly those with severe early onset scoliosis aged under 10 years who are candidates for growing rod implantation, as well as for non-ambulatory patients with severe neuromuscular spine deformities. However, for the specific cohort of severe and rigid deformities addressed in this study, the passive, gravity-dependent force of HGT is often insufficient to overcome rigid bony ankylosis. In contrast, HPT establishes a closed mechanical loop between the skull and pelvis, providing an active, 24-hour continuous, and powerful distraction force that functions independently of patient positioning. A recent matched cohort study comparing these 2 modalities found that HPT achieved significantly superior radiographic correction in shorter durations for severe rigid curves compared to HGT.¹⁶ Crucially, the study demonstrated that HPT patients were more likely to be down-staged to lower-grade osteotomies, whereas HGT patients with comparable baseline severity often still underwent 3-CO. We acknowledge that HPT is associated with higher invasiveness, risks and patient discomfort compared to HGT; however, we believe that the superior corrective efficacy and the potential to avoid high-risk 3-CO justify the use of HPT, provided that the manageable traction-related risks are meticulously monitored.

The current study found that both the ORT and the EBL were significantly reduced in the HPT+PSF group. A systematic review by Yang et al suggested that the mean ORT for PVCR was 430 min and the EBL was 2639 mL; However, study by Zhang et al reported that the ORT and EBL could be respectively shortened to 283 min and 842.31 mL by using multiple-level asymmetrical Ponte osteotomies for patients requiring VCR.¹⁷ We considered that the reductions in ORT and EBL could be attributed to several factors. First, preoperative HPT decreases the severity of spinal deformity, facilitating easier exposure of the upper thoracic and lumbar region for pedicle screw placement during surgery, improving intraoperative efficiency. Second, HPT markedly reduces the degree of angular deformity of the main curve, transforming it from an angular shape into a smoother, arc-shaped configuration. Following HPT, the DAR was significantly reduced from 21.2 ± 6.4 to 10.5 ± 4.8 (P < 0.001). This morphological change alleviates allows for satisfactory correction with lower-grade osteotomies, thus avoiding the need for complex and high-risk 3CO. For patients who achieved satisfactory correction through HPT, osteotomy could be avoided, and only in situ spinal instrumentation and fusion were performed. If the placement of pedicle screws in the apical vertebra was challenging, this step could be omitted, and additional satellite rods spanning the apical region could be used for reinforcement instead, which would significantly reduce the ORT. Third, the significantly higher incidence of IONM alerts in the one-stage PSF group inherently prolonged the overall operative time. The process of addressing the IONM alerts introduces substantial intraoperative delays that are largely avoided in the HPT cohort. Collectively, these findings indicate that preoperative HPT not only simplifies the surgical procedure but also enhances its safety and efficiency in managing severe rigid spine deformities.

Neurological deficit represents one of the most devastating complications of spinal deformity surgery. The reported incidences of transient neurologic deficits ranging from 2.9% to 29.3% and permanent deficits occurring in 2.0-6.3% of cases.^15,18-20 In neurological deficits, the spinal cord injuries were primarily attributed to 2 mechanisms: acute mechanical injury and ischemic injury. The permanent deficits observed in the one-stage PSF group were directly tied to the highly demanding technical steps of the 3-CO procedures. Upon review of the operative records for the 2 patients who sustained permanent deficits, the loss of IONM signals occurred during specific critical maneuvers. In one case, the alert was triggered during the closure of the osteotomy site. Despite immediate reversal of the maneuver, the signals did not fully recover, suggesting a direct mechanical compression related to dural buckling during apical shortening. In the second case, signal loss occurred during the posterior vertebral wall resection and ventral dural sac decompression. Severe adhesions between the ventral dura and the vertebral body complicated the dissection, leading to an acute massive hemorrhage (approximately 1000 mL) from the epidural venous plexus and a sudden drop in mean arterial pressure (MAP) to 70 mmHg. Despite immediate blood transfusion and rapid pharmacological elevation of MAP, the MEPs were permanently lost. These actual case events underscore that while 3-CO provides essential apical shortening, the requisite manipulations carry profound mechanical and acute physiological risks to the spinal cord. Previous studies have demonstrated that preoperative HPT can markedly reduce the incidence and severity of neurological complications—a finding consistent with our results.^8,10 In the present study, the incidence of IONM alerts was significantly lower in the HPT+PSF group (3.3%) compared with the PSF group (20.0%); also, the incidence of neurological deficit was significantly lower in the HPT+PSF group, indicating enhanced neurological safety. Several factors may explain this safety advantage. First, HPT utilizes the viscoelastic properties of the spinal cord to mitigate mechanical and ischemic risks. The slow, continuous distraction allows neural tissue to lengthen gradually via biological creep, preventing the spike in intramedullary pressure associated with acute correction. Simultaneously, this gradual stretching promotes neovascularization and allows spinal cord blood flow to autoregulate to the new length over weeks, avoiding the ischemic shock of rapid lengthening. Second, clinically, HPT facilitates correction and decreases the need for high-risk 3-CO, thereby minimizing direct mechanical disturbance to the spinal cord. Third, the significantly reduced intraoperative blood loss in the HPT group helps maintain adequate mean arterial pressure, ensuring optimal spinal cord perfusion. In contrast to the abrupt alignment alterations that occur during one-stage 3-CO, HPT offers the spinal cord sufficient time to adapt to both pathological and physiological changes. For patients with IONM alerts during surgery, reducing the traction distance or removing the traction devices could be tried to protect the spinal cord from distraction injury. Collectively, these factors contribute to a safer correction process and a lower risk of neurological injury.

A critical consideration in adopting the HPT strategy is the duration of treatment. Our cohort underwent a mean of 20 weeks of traction, which indeed places a burden on patients and their families, representing a major downside compared to the one-stage surgery. While prolonged, this duration is strictly underpinned by both biomechanical safety and physiological necessity. Biomechanically, slow distraction (0.5-1 mm/day) is essential to avoid neurological ischemia and release the soft-tissue. Physiologically, a recent study utilizing 3D lung reconstruction demonstrated that the optimal improvement in lung volume is typically achieved after 4 to 6 months of HPT.²¹ Notably, in the current study, standard PFT parameters did not show statistically significant improvement immediately post-traction. We attribute this to the physiological lag between anatomical expansion and functional recovery. Although HPT successfully expands the constricted thoracic cage and increases lung volume, the recruitment of atelectatic alveoli for effective gas exchange is a delayed process. The restoration of functional ventilation often requires a longer remodeling period, typically up to one year, to align with the structural expansion. Ultimately, surgeons might argue that foregoing the morbidity associated with prolonged traction is preferable if neurological deficit are merely transient or mild. However, the critical question is whether this burden could be justified by the safety benefits. In the current study, the risks of the one-stage procedure were not limited to transient issues; 2 patients in the PSF group (4.0%) sustained permanent spinal cord injuries. In contrast, the HPT+PSF group incurred no permanent neurological deficits. Thus, the clinical rationale for HPT is to exchange the temporary and manageable morbidity of traction for minimizing the risk of permanent and catastrophic disability. For young patients with such extreme spine deformities, we believe this investment in safety is justified.

One patient in the HPT+PSF group developed severe adding-on, necessitating revision surgery. The principal cause of this mechanical complication was the inappropriate selection of the lower instrumented vertebra (LIV). Although HPT can substantially reduce the magnitude of spinal deformity, the determination of the LIV should be guided by preoperative radiographic evaluation rather than post-traction alignment.

Limitation

Several limitations should be acknowledged in this study. First, the retrospective design and single-center nature may introduce selection and information bias, limiting the generalizability of the results. Second, although all surgeries were performed by an experienced surgical team following standardized protocols, individual variations in surgical decision-making—such as the selection of osteotomy grade or the precise duration of traction—might have influenced the outcomes. Third, the follow-up period, while exceeding 2 years for all patients, may still be insufficient to fully assess long-term complications such as implant failure, loss of correction, or junctional problems. Fourth, the relatively small sample size, particularly in subgroups undergoing 3-column osteotomies, restricts the statistical power for detecting differences in rare complications. Finally, we could not definitively claim the superiority of HPT because of the absence of patient-reported outcome measures (PROMs) such as SRS-22 or SF-36, which limits the evaluation of functional and quality-of-life improvements following different treatment strategies. Therefore, prospective multicenter studies with larger sample sizes, longer follow-up, and inclusion of PROMs are warranted to validate and expand upon our findings.

Conclusion

Staged HPT followed by PSF provides a safe and effective strategy for the management of severe rigid spine deformities. In terms of radiographic correction and perioperative safety profile, this staged approach demonstrated significant advantages over one-stage 3-CO, notably reducing operative time, blood loss, and neurological risks. Preoperative traction effectively decreases deformity stiffness and magnitude, allowing for safer and less aggressive corrective procedures. However, careful patient selection, standardized traction protocols, and precise surgical planning remain essential to minimize traction-related and mechanical complications.

Footnotes

ORCID iDs

Honghao Yang

Jixuan Huang

Yihan Yang

Yong Hai

Ethical Considerations

Ethical approval was obtained from the institutional review board at Beijing Chao-Yang Hospital (Approval No. 2025-HU-505).

Consent to Participate

All participants provided informed consent prior to enrollment.

Author Contributions

YH and LJZ were responsible for conceptualizing and designing the study. HHY, JXH, YHY, and HSG collected and analyzed the data. HHY and JXH drafted the original manuscript. YSW, YQZ, and YPZ provided substantial supervision and critical revisions. All authors have read and approved the final manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. If anyone needs the data and information from this study, please contact the correspondence authors.*

References

Camargo

Abode-Iyamah

Shah

Bechtle

Freeman

. Comprehensive perioperative approach to complex spine deformity management. Clin Spine Surg. 2022;35(7):310-318. doi:10.1097/bsd.0000000000001240.

Zhou

Yang

Wang

Zhang

Wang

Hai

. The effectiveness and safety of staged halo-pelvic traction combined with posterior spinal fusion in the treatment of severe rigid spine deformity: a prospective cohort study. J Orthop Surg Res. 2025;20(1):341. doi:10.1186/s13018-025-05739-7.

Passias

Krol

Passfall

, et al. Three-column osteotomy in adult spinal deformity: an analysis of temporal trends in usage and outcomes. J Bone Joint Surg Am. 2022;104(21):1895-1904. doi:10.2106/jbjs.21.01172.

Passias

Williamson

Mir

, et al.

Comparison of multilevel low-grade techniques versus three-column osteotomies in adult spinal deformity surgery: does harmonious correction matter?

J Neurosurg Spine. 2024;40(6):677-683. doi:10.3171/2024.1.Spine23802.

Williamson

Mir

Smith

, et al. Contemporary utilization of three-column osteotomy techniques in a prospective complex spinal deformity multicenter database: implications on full-body alignment and perioperative course. Spine Deform. 2024;12(6):1793-1801. doi:10.1007/s43390-024-00906-y.

Sun

Zhang

Tan

Zhang

. Halo-pelvic traction in the treatment of severe scoliosis: a meta-analysis. Eur Spine J. 2023;32(3):874-882. doi:10.1007/s00586-023-07525-7.

Yang

Huang

Zhao

, et al. Clinical effects of modified Halo-pelvic traction in patients with severe spinal deformities and respiratory disorders. J Orthop Surg Res. 2025;20(1):617. doi:10.1186/s13018-025-06050-1.

Zhou

Liu

Deng

, et al. The contribution of preoperative balanced halo-pelvic traction to severe rigid spinal deformity correction. Eur Spine J. 2023;32(11):4054-4062. doi:10.1007/s00586-023-07916-w.

Zhao

Yang

, et al. Staged short-term modified halo-pelvic traction and posterior spinal fusion versus posterior vertebral column resection for severe rigid kyphoscoliosis: a multicenter comparative study. Spine J. 2025;26:602-611. doi:10.1016/j.spinee.2025.10.024.

10.

Wang

Liu

. Presurgical short-term halo-pelvic traction for severe rigid scoliosis (Cobb angle >120°): a 2-Year Follow-up review of 62 patients. Spine. 2021;46(2):e95-e104. doi:10.1097/brs.0000000000003740.

11.

Wang

Lenke

Thuet

Blanke

Koester

Roth

. Deformity angular ratio describes the severity of spinal deformity and predicts the risk of neurologic deficit in posterior vertebral column resection surgery. Spine. 2016;41(18):1447-1455. doi:10.1097/brs.0000000000001547.

12.

Lewis

Keshen

Lenke

, et al.

The deformity angular ratio: does it correlate with high-risk cases for potential spinal cord monitoring alerts in pediatric 3-Column thoracic spinal deformity corrective surgery?

Spine. 2015;40(15):E879-E885. doi:10.1097/brs.0000000000000984.

13.

Lenke

Hassan

Mohanty

Gupta

Ames

. Three-column osteotomies: past, present, and future. Seminars in Spine Surgery. 2023;35(4):101059. doi:10.1016/j.semss.2023.101059.

14.

Riley

Lenke

Chapman

Jr. Sides

Blanke

Kelly

. Clinical and radiographic outcomes after posterior vertebral column resection for severe spinal deformity with five-year Follow-up. J Bone Joint Surg Am. 2018;100(5):396-405. doi:10.2106/jbjs.17.00597.

15.

Yang

Zheng

Liu

Wang

Kim

Cho

. Posterior vertebral column resection in spinal deformity: a systematic review. Eur Spine J. 2016;25(8):2368-2375. doi:10.1007/s00586-015-3767-7.

16.

Zhou

Wang

Yang

Zhang

Wang

Hai

. Efficacy of Halo Pelvic Traction versus Halo gravity traction in treating severe rigid spinal deformities: a matched retrospective study. Glob Spine J. 2025;16(1):21925682251348670. doi:10.1177/21925682251348670.

17.

Zhang

Tao

Hai

, et al. One-stage posterior multiple-level asymmetrical ponte osteotomies versus single-level posterior vertebral column resection for severe and rigid adult idiopathic scoliosis: a minimum 2-Year Follow-up comparative study. Spine. 2019;44(20):E1196-e1205. doi:10.1097/brs.0000000000003101.

18.

Smith

Shaffrey

Klineberg

, et al. Complication rates associated with 3-column osteotomy in 82 adult spinal deformity patients: retrospective review of a prospectively collected multicenter consecutive series with 2-year follow-up. J Neurosurg Spine. 2017;27(4):444-457. doi:10.3171/2016.10.Spine16849.

19.

Saifi

Laratta

Petridis

Shillingford

Lehman

Lenke

. Vertebral column resection for rigid spinal deformity. Glob Spine J. 2017;7(3):280-290. doi:10.1177/2192568217699203.

20.

Yang

Sui

Deng

Yang

Huang

. One-stage spine-shortening osteotomy treated severe spinal deformity associated with spinal cord malformations. Clin Spine Surg. 2023;36(4):e160-e166. doi:10.1097/bsd.0000000000001397.

21.

Zhou

Yang

, et al. Impact of halo-pelvic traction duration on pulmonary improvement quantified by pulmonary function tests and three-dimensional lung reconstruction in severe rigid kyphoscoliosis: a prospective cohort study. J Orthop Surg Res. 2025;20(1):705. doi:10.1186/s13018-025-06126-y.