Pre-operative Halo-gravity Traction in the Treatment of Complex Spinal Deformities

What Do We Know So Far?

A Systematic Review

Abstract

Spinal deformity surgeries are complex procedures associated with a high risk of complications. Halo-gravity traction (HGT) is a useful option that provides a gradual traction force to aid in deformity correction. Though its benefits are well-acknowledged, there still exist major ambiguities regarding its role in the management of complex spinal deformities. We performed a systematic review of the electronic databases including EMBASE, MEDLINE, PubMed, and Cochrane on November 12, 2021 to identify relevant articles on HGT; to analyze the existing literature on pre-operative HGT; and to compare the existing protocols for HGT in spinal deformity patients, its varied effects on the radiological parameters and general health status of the patients, and its associated complications. Among the 284 articles available in the literature, 34 articles were finally included and a total of 1151 patients [mean age, 14.6 years] were analyzed. Mean pre-traction coronal Cobb angle of 107° (72°–140.7°) was reduced by 24.8% [to a mean of 80.5° (42°–120.2°)] following HGT. Mean pre- and post-traction sagittal Cobb angles were 88° (56°–134.7°) and 65.4° (36°–113°; a reduction by 25.7%), respectively. Following HGT, mean body weight and body mass index (BMI) of patients improved by 7.2% and 9.1%, respectively. Mean improvement in forced vital capacity and forced expiratory volume has been reported to be 14.5% and 13.9%, respectively. Pre-operative HGT is a useful option in the treatment of spinal deformities. It aids in reducing curve magnitude and provides optimal time for improving general condition (pulmonary and nutritional status) of patients pre-operatively. It is a safe procedure with 2.1% neurological and 11.6% non-neurological complication rates.

INTRODUCTION

Spine deformity surgeries are challenging scenarios for spine surgeons worldwide.[¹] Stiff and rigid curves of large magnitudes need diligent pre-operative evaluation including planning the extent of deformity correction, strategic placement of implants, corrective maneuvers, osteotomies, and the need for staged procedures.[^2,3] These major deformity surgeries are always associated with a high risk of neurological deterioration.[⁴] A pre-operative, slow, sustained correction using halo traction followed by surgical manipulation has been an approach used over past decades to enable better correction of curve magnitudes, with the added advantage of ameliorating the risk of catastrophic neurological injuries.[⁵] Among the options available, halo-gravity traction (HGT) has been widely practiced, as traction forces can be applied while the patient remains ambulant.[⁶] Studies have shown HGT to be a relatively safe, well-tolerated procedure providing significant corrective forces.[^7,8] Moreover, pre-operative optimization of patients with severe spinal deformities is also critical; this can also be performed while the patient remains in HGT.[⁹]

Most studies in the literature have supported the role of HGT in deformity correction, although a few have shown less evidence, contributing to a lack of clear consensus. This article provides a comprehensive review about our current understanding on the actual role of HGT in the management of complex spinal deformities, protocols of application, potential benefits, and complication rates. We also discuss the pitfalls in the existing literature on this subject and elaborate upon the prospects for future research.

MATERIALS AND METHODS

A review of the literature was performed on November 12, 2021 using EMBASE, MEDLINE, PubMed, and Cochrane Database. An elaborate search was made using the keywords “halo-gravity traction,” “spinal deformity,” “scoliosis,” “kyphosis”, “deformity correction surgery.”

Inclusion criteria for article selection

Only articles which were written in English language and had included patients with pre-operative HGT applied for at least 2 weeks were reviewed. Case series, observational cohort, case–control studies, and randomized controlled trials were given preference.

Exclusion criteria for article selection

Articles published before the year 2000, those without relevant outcome measures, articles with effects of only halo-pelvic traction (HPT) or halo-femoral traction (HFT), and review articles were excluded.

RESULTS

An initial search using the keywords “halo-gravity traction,” “spinal deformity,” “scoliosis,” “kyphosis,” and “deformity correction surgery” in EMBASE, MEDLINE, PubMed, and Cochrane Database yielded 284 articles. The initial screening involved exclusion of duplicate articles, articles unrelated to HGT, and articles in non-English literature based on abstracts or the titles of articles. Entire manuscripts were obtained for all these selected articles and thoroughly reviewed during the second stage of article selection on the basis of aforementioned inclusion and exclusion criteria. Finally, 34 articles (having a total of 1151 patients) were analyzed and included in this review. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were applied for strategically including the articles in the study and are shown in Figure 1. Year of publication, patient profile, deformity characteristics, neurological status of the population, radiological parameters, details of traction characteristics, surgeries performed, reported complications, and final clinical and radiological outcomes were analyzed.

Figure 1: Flowchart showing the selection process of articles in the systematic review

DISCUSSION

HGT was initially popularized by Stagnara[¹⁰] as a convenient and safe modality to provide a gradual traction force in patients with severe, rigid spinal deformities. Unlike HFT and HPT tractions, in which mobility of the patient is limited, HGT can be applied while the patient remains ambulant.[^6,11] The countertraction force in HGT is provided by the patient’s own body weight.[¹²]

Despite HGT being a commonly used pre-operative strategy in deformity correction surgeries, there is a significant paucity of data in the existing literature on various issues related to it. There are only two systematic reviews or meta-analysis published on this topic hitherto.[^13,14] Both these reviews focussed on only specific parameters and did not discuss various other factors evaluated in these individual articles. These reviews were also highly selective in the inclusion of studies [only a total of 25 studies (633 patients) were included in both these reviews combined]. As a result, the overall sample population included in these studies was relatively small. The individual studies on this subject were also heterogeneous and included patients with multiple underlying etiologies. We therefore planned the current study to comprehensively review the current literature and discuss the demographic details, patterns of deformities and etiologies, traction protocols employed, clinical and radiological outcome measures, effect of HGT on neurological status, as well as the spectra of complications reported with HGT.

BIOMECHANICAL PRINCIPLES OF HGT

Biomechanically, spinal column is constituted by a long and slender bone pillar (which alternates with intervertebral disc) in a laminated pattern, adequately supported by adjoining muscles, ligaments, and other soft tissues. The biomechanical properties of spine are thus the resultant of combined viscoelastic properties of each individual component.[¹⁵] In a patient with spinal deformity, this viscoelastic column tends to progressively collapse over time, owing to the vector secondary to gravitational forces acting in a vertical direction in upright postures. Therefore, in accordance with the age-old adage “deformity begets deformity,” the curves tend to undergo a vicious cycle of progression.[¹⁶]

HGT is based on the principle of “tissue creep,” which is explained as the property of fascia or other adjoining tissues to lengthen (by virtue of visco-elastic nature), when subjected to sustained tension load resulting in reduced resistance to application of a second load.[¹⁷] Therefore, when the traction is applied for a specific time length, progressive deformation ensues. Prolonged traction thus elongates and relaxes soft tissues, thereby enabling gradual correction of deformities in a controlled manner. This has been reported to offer potential benefits including alleviation of neurological symptoms, evaluation of curve flexibility, reduction of force required during intra-operative correction, mitigation of spinal cord injury risk, and enhancement of curve correctability.[¹⁸]

DETAILS OF ARTICLES AND PATIENT PROFILE

A total of 284 studies were identified using our search criteria. Of these, 168 were duplicates and were removed. Eighty-four studies did not meet the inclusion criteria. Overall, 34 studies were reviewed. All the articles included in the study were published during the period of 2000–21 (five articles were published between 2000 and 2010, 10 articles between 2011 and 2015, and 19 were published in the last 6 years) [Table 1]. Eleven articles were published in European journals and the remaining 23 were from North American journals. Nineteen articles were published in orthopedic spine journals, 3 in neurosurgical, 3 in general orthopedic, 4 in pediatric orthopedic, and 5 were published in medical journals. Thirteen studies (488 patients) emanated from North America, 12 (290 patients) from Asia, 5 (212 patients) from Europe, 3 (146 patients) from Africa, and one study (15 patients) was conducted in South America. Among them, nine were comparative studies; of which six studies compared HGT patients with those who did not undergo any form of peri-operative traction. Studies by Chen et al.[⁴] and Shi et al.[¹⁹] compared HGT with HPT and HFT, respectively. The study by Davies et al.[²⁰] compared the outcomes of HGT applied in the in-patient and out-patient settings.

Table 1: Publication data regarding various articles included in the systematic review

From these 34 articles, a total of 1151 patients were finally included in the systematic review. The mean age of the study population was 14.6 years and 47% (540 patients) were males.

DEFORMITY CHARACTERISTICS

A majority of the studies (635 patients) included a mixed cohort of patients with kyphosis, kyphoscoliosis, and scoliosis.[^21,22,23] Thirteen studies included only scoliosis patients (283 patients), five studies included only patients with kyphoscoliosis (214 patients), and another study by Li et al.[²⁴] included only patients with kyphotic deformities (19 patients). The underlying etiologies of deformities included congenital (28 studies), idiopathic (27 studies), neuromuscular (13 studies), neurofibromatosis (12 studies), syndromic (10 studies), and post-infectious (6 studies) causes [Table 2]. Hui et al.[²⁵] and Sun et al.[²⁶] studied patients with spinal deformities associated with intra-spinal anomalies. The effect of HGT on spinal deformity patients with underlying osteogenesis imperfecta was studied by Janus et al.[²⁷]

Table 2: Demographic data of included patients

TRACTION STRATEGIES EMPLOYED

In the included studies, a minimum of 4 and a maximum of 10 pins were used for the application of halo.[^3,27,28] HGT was started with a lower weight initially (range: 1.5–5 kg or 20% body weight) and increased gradually over a few weeks to about an average of 50% (range: 33–60%) of their total body weight (TBW).[^{29,30,31,32,33,34}] Apart from the study by Han et al.,[³⁵] in which a maximum traction weight equivalent to 60% of TBW was used, all the other studies used a maximum weight of 50% of TBW [Table 3]. In a delphi-based study by Benjamin et al., the authors suggested that the maximum traction weight needs to be reached by 2 weeks and must be continued throughout the traction duration.[³⁶] A 24-h traction protocol was preferred universally across the studies, and a minimum of at least 12 h per day traction application was followed. The weight was reduced during sleep, at times of personal hygiene or if the patient developed neurological symptoms such as weakness or paresthesia. The mean duration of traction was 60.58 (14–138) days. Neurologically intact patients were ambulant throughout the entire duration of HGT application. A pre-traction/pre-operative surgical release was performed in 14 studies. Although the studies have agreed upon strict compliance of HGT for ameliorating the outcomes in deformity patients, the exact data related to patient compliance are deficient in the current literature.

Table 3: Data regarding the HGT protocol and complications associated with HGT

SURGICAL DETAILS

Instrumented fusion as the definitive procedure following traction was performed in 22 studies (823 patients). In five studies (105 patients), growth-preserving techniques such as growth rod/vertical expandable prosthetic titanium ribs/guided growth were performed as the patients were too young for a fusion surgery.[^37,38] Four studies included patients with both fusion and growth-preserving modalities.[^23,28,30,39] The details of the final surgical procedure were not mentioned in three studies. Since the procedures undergone were highly heterogeneous, the reporting of overall intra-operative blood loss (IOBL) and duration of surgery was difficult to compare among the included studies. Among the five studies involving fusion-only procedures, the mean reported IOBL was 1673.70 (767.8–2900) mL. The mean duration of surgery was 380.6 (282–474) min in these patients.

ASSESSMENT OF DEFORMITY IN THE INCLUDED STUDIES

Most of the studies compared the magnitude of sagittal and coronal Cobb correction using HGT.[^20,40,41] A study by Shi et al.,[⁴²] in addition to the sagittal and coronal plane deformities, included the impact of HGT on rotational subluxation. Improvement in spinal height (T1-S1) and thoracic height (T1-T12) was studied in eight and two studies, respectively [Table 4].

Table 4: Radiographic data of the deformity in the included studies

The mean pre-traction coronal Cobb’s angle was 107° (72°–140.67°). The mean post-traction and final follow-up coronal Cobb’s angles were 80.5° (42°–120.2°; corresponding to an overall improvement of 24.8%) and 57° (33.6°–80.6°; overall 46.7% improvement), respectively. The mean pre- and post-traction sagittal Cobb’s angles were 88° (56°–134.7°) and 65.4° (36°–113°) (corresponding to an overall improvement of 25.7%), respectively. At the final follow-up, the mean sagittal Cobb’s angle was 52° (35.4°–73.7°; overall 41% improvement). Among the reviewed studies, 4 studies included patients with a mean age of ≤ 10 years (early onset), whereas the remaining 30 included patients with a mean age of >10 years (adolescent or late-onset). While comparing the outcome between the early onset and late-onset scoliosis, there was no significant difference between the two groups in the mean correction of coronal Cobb’s angle following HGT [mean difference between post- and pre-traction Cobb’s angle of 29.9º (in the age group ≤10 years) vs. 29.6º (in the age group >10 years)]. However, better correction in the Cobb angles was achieved at the final follow-up time point, in comparison with the pre-traction status in the older age group patients [mean difference between final follow-up and pre-traction Cobb’s angle of 40.3º (in the age group ≤10 years) vs. 49.4º (in the age group >10 years)]. This may be explained by the more unfavorable natural history of deformity progression in the younger, early onset scoliosis patients.

The mean T1-S1 length pre-traction and post-traction and at final follow-up were 26.6 and 31.2 cm (an overall improvement of 17.3%) and 34.5 cm (an overall improvement of 29.7%), respectively. The pre-traction and final follow-up T1-T12 length were 14.9 and 19 cm (an overall improvement of 27.5%), respectively. Thus, in coherence with our expectation, we could observe that a majority of the studies reported the use of HGT in patients with severe deformities only. An overall improvement of curve-related parameters following the application of HGT alone ranged approximately between 17% (for T1-S1 length) and 26% (for sagittal Cobb). The data regarding other crucial parameters in deformity surgeries like coronal (i.e., CSVL) and sagittal (i.e., C7 plumbline) truncal balance, apical vertebral rotation, and pelvic morphological parameters (including pelvic incidence, pelvic tilt, and sacral slope) are deficient in the existing literature. Based on our review, we would therefore recommend the need for further large-scale prospective studies to evaluate the influence of HGT on these additional morphometric characteristics of the deformities.

NEUROLOGICAL STATUS OF PATIENTS AND THE INFLUENCE OF HGT ON ITS OUTCOME

A majority of studies (25/34 studies) published on HGT hitherto had only included patients with normal neurological status. In the remaining studies, a total of 76 patients with pre-existing baseline neuro-deficits were reported. The classification systems broadly used to evaluate the severity of neuro-deficit among these studies were ASIA (two studies), Frankel (one study), Ranawat (one study), and MRC grading (one study). In two studies, neurological status was only graded as progressive or non-progressive.

Among the 1075 patients with normal baseline neurological status, post-traction neurological deterioration in the form of motor weakness or myelopathy was noted in 7 (0.6%) patients. Seven (0.6%) patients had post-traction cranial nerve involvement and two (0.2%) had isolated sensory disturbances. One patient developed brachial plexus palsy[¹⁹] and two developed transient, post-traction bladder disturbance. The neurological disturbances in all these reported cases were transient; and once the weight of traction was reduced, all patients improved completely (the final status of the patient who developed brachial plexus injury is not mentioned in the article). Sink et al.[⁴³] suggested that instead of weights used in most of the studies, spring scale used for applying traction could reduce the incidences of neurological complications. Overall, six patients developed features of raised intra-cranial tension (i.e., nausea, vomiting, dizziness, abnormal pupillary reactivity) during the course of HGT.[^23,26,28,44] Thus, the overall neurological complication rate in HGT was 2.1% (25 events).

In the 6 studies (including 44 patients) which reported the influence of HGT on the overall neurological outcome of patients with pre-existing neuro-deficits, 31 (70.5%) patients improved neurologically following HGT. Only in one case, gross further deterioration of neurological status was reported during traction. This isolated case of major neuro-deficit was a 14-year-old female patient with neuromuscular kyphoscoliosis and pre-existing neuro-deficit.[⁴¹] Her neurological status deteriorated and she developed spastic paraplegia while in traction, which further worsened following the final surgery (anterior release-posterior spinal fusion).

Although the overall literature seems to be optimistic regarding the influence of HGT on the neurological status of an individual, the reporting of neurological outcome per se in these studies is fairly limited (not all studies mention these details elaborately). This is especially important as neurological safety is, in fact, purported as one of the major benefits of HGT. In this context, we would like to reiterate upon the need for future trials foregrounding the influence of HGT on the neurological outcome of patients.

HGT AND OTHER PATIENT-RELATED SYSTEMIC PARAMETERS

The effect of HGT on pulmonary function was evaluated in ten studies, among which the study by Bao et al.[⁴⁵] only included patients with severe respiratory dysfunction. Ten studies discussed the change in forced vital capacity (FVC), whereas eight analyzed the improvement in forced expiratory volume (FEV1).[^46,47] The mean pre- and post-traction FVC has been reported as 50.4% and 57.7%, respectively. This corresponds to an overall improvement of 14.5%. The average pre- and post-traction FEV1 has been reported to be 43.8% and 49.9%, respectively (an overall improvement of 13.9%) [Table 5]. Wang et al.,[¹³] in their systematic review, also noted a significant improvement in pulmonary function following HGT and suggested that HGT would be beneficial in reducing the anesthetic complications in the perioperative period.

Table 5: Effect of HGT on pulmonary function

Comparison of improvement in nutrition was performed in six studies [Table 6]. The effect of HGT on nutrition was assessed in terms of weight gain (four studies) and change in body mass index (BMI) (two studies).[^47,48] The mean pre-traction weight in the study population was 37.3 kg. Following HGT application, the mean weight of patients improved to 40 kg (corresponding to an overall improvement of 7.2%). Among the studies which reported BMI, the overall mean improvement in BMI was 9.1% (pre-traction BMI of 16.4 kg/m² to post-traction BMI of 17.9 kg/m²). It may thus be observed that HGT broadly improves different systemic clinical parameters (especially pulmonary function and nutrition) during the pre-operative period in patients undergoing spinal deformity surgeries. This has been reported as one of the major benefits by the proponents of this strategy.

Table 6: Effect of HGT on nutritional status of the patients

Rationally, any long-term procedure, which relatively restricts the mobility of patients over a period of time (even though HGT still maintains patients’ ambulation when compared with HPT or HFT), is expected to result in certain major systemic complications such as osteoporosis or deep venous thrombosis (DVT). The literature evidence regarding both these complications is limited. While Han et al.[³⁵] published the only study which reported the effect of HGT on bone mineral density (BMD), none of the existing studies has reported any evidence on DVT. In their series of 20 patients with severe kyphoscoliosis, Han et al. noted a significantly lower BMD in their cohort following HGT (Z-score changed from -2.2 to -2.7). The mean traction duration in their study was 80 days (which was longer than a majority of other studies), which could have at least partly influenced their results. They also observed a strong correlation between the duration of HGT and fall in BMD. Surprisingly, no other studies have assessed these parameters heretofore. Thus, this is one of the areas where the current literature on HGT is profoundly deficient and offers substantial prospects for future research.

DOMICILIARYVS. IN-HOSPITAL HGT

In all articles except one, patients were kept under observation in the hospital throughout their course of HGT. In the study by Davies et al.,[²⁰] a comparison was drawn between the application of HGT under domiciliary care and on an in-patient basis. At the end of HGT, there was no statistically significant difference in the coronal (P = 0.5) and sagittal deformity (P = 0.8) correction. There was no statistically significant difference in the incidence of complications such as pin loosening, pin-site infections, or neurological complications between the two groups. However, it was observed that the mean cost of treatment per patient was 2.8-fold higher than that for the in-patient group when compared with the domiciliary population. They concluded that pre-operative HGT in an out-patient setting is an option to be considered in patients with a low risk of spinal deformity-related neurologic complications and those not requiring close supervision from the nursing staff.

NON-NEUROLOGICAL HGT-RELATED COMPLICATIONS

Overall, in a total of 159 (out of 1151 patients; 13.8%) direct complications of HGT have been reported in the literature. The most commonly reported non-neurological complication is pin-site loosening or infection. This complication alone accounts for 83% (132 events) of reported complications. Among them, on only one occasion, the patient went on to develop serious, deep-seated infection and cranial abscess,[⁴⁷] which necessitated craniotomy, revision of halo, and antibiotics. HGT was continued in this patient for a total of 202 days and she finally underwent growth rod placement. All other patients reported in the literature thus far only presented with superficial, local infection, which settled with pin-site care (with or without pin removal or revision halo application) and antibiotics. It has also been postulated that increasing the number of pins reduces the torque distribution at individual pins.[⁴⁹] This in turn reduces the chances of loosening and infection. Cervical or trapezial soreness is also reported as a common complication related to the use of HGT in many studies,[^31,50] although the exact number of patients has not been documented. On a majority of occasions, the soreness has been reported to improve with conservative management including traction weight reduction, massage therapy, or muscle relaxants. Gastrointestinal discomfort has also been reported in two patients.[²⁸]

COMPLICATION RATES AND DURATION OF HGT

Of the 34 studies included, HGT was applied for ≤6 weeks in 11 studies (262 patients). A total of 32 (12.2%) complications were observed in these patients. Among them, 22 (8.4%) complications were pin-related (pin-site infection: 12, pin loosening: 10). One (0.4%) patient developed cranial abscess, two (0.8%) had gastrointestinal discomfort, and seven (2.7%) patients had transient neurological complications (motor/sensory/bladder disturbances); all of which improved with reduction of traction weight.

HGT >6 weeks was applied in 23 studies (889 patients). A total of 127 (14.3%) complications were noted. About 110 (12.4%) complications were pin-related (pin-site infection: 89, pin loosening: 21). Eleven patients had transient neurological complications, whereas six had features of raised intracranial tension.

LIMITATIONS

Though the systematic review has analyzed various studies existing in the literature, it has several limitations. Most of the studies are quite heterogeneous and include patients with deformities (kyphosis/scoliosis/kyphoscoliosis) of varied etiologies (congenital/neuromuscular/infective/syndromic). This poses difficulty in making standard recommendations regarding the use of HGT for specific underlying pathologies. Although a formal quality assessment of the included studies was not performed in our review, a majority of studies thus far published on this subject are of low quality.

CONCLUSION

Overall, from all the available studies, we find that pre-operative HGT improves the clinical and radiological outcomes in patients with spinal deformities such as kyphosis, scoliosis, or kyphoscoliosis. It aids in reducing the curve magnitude and provides optimal time for improving the general condition (pulmonary function and nutritional status) of the patients pre-operatively. It is a relatively safe procedure with 2.1% neurological and 11.6% non-neurological complication rates. There is still a substantial deficiency in existing literature on the true role of HGT in the management of complex spinal deformities. In this context, the need for large-scale, prospective studies on this subject cannot be understated.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.