Stentoplasty with Resorbable Calcium Salt Bone Void Fillers for the Treatment of Vertebral Compression Fracture: Evaluation After 3 Years

Introduction

Vertebroplasty (VP) was first described by Galibert et al. in vertebral angiomas.¹ In the last three decades, VP has been used extensively to treat osteoporotic and malignant vertebral compression fractures. Injection of polymethylmethacrylate (PMMA) bone cements aims to achieve the restoration of vertebral body height and eliminate the pain caused by fractures or tumors. However, bone cement leakage is the most common complication which may result in radicular pain, paralysis and even death.^2–4 Balloon kyphoplasty (BKP) was designed to reduce the pressure of infusion and the risk of bone cement leakage.^5,6 In consideration of the loss of vertebral body height after balloon withdrawal in BKP, stentoplasty was developed. Vertebral body stenting (VBS), using an inflatable balloon for expansion, has been proved to achieve a satisfactory reduction and maintain the restored vertebral height after balloon withdrawal.^7,8 In recent years, a new minimally invasive technique, called Stent-Screw Assisted Internal Fixation (SAIF), was developed based on the VBS system and can be applied in severe neoplastic and osteoporotic fractures.^9,10 It has good clinical efficacy and biomechanical support.^11,12 The VBS/PMMA can effectively reconstruct the load-bearing capacity of the anterior column. The screw anchoring to the vertebral pedicle bridged with VBS/PMMA complex avoids bone cement displacement and reduces the stress distribution of the surrounding structure.¹¹ However, the application of VBS and PMMA may increase the operative vertebral stiffness.

PMMA can provide main stability due to its micro-interlocking with the surrounding bone trabecular. However, PMMA has disadvantages such as high polymerization temperature,¹³ potential cytotoxicity,¹⁴ poor biocompatibility and no biodegradation,¹⁵ which could cause complications such as cement leakage, osteonecrosis in the vertebral body, and fractures in the adjacent vertebral body. In histopathology, we can observe fibrous membrane formation between PMMA and bone trabecular without any remodeling process.¹⁶

Calcium phosphate cements (CPCs) can be considered as an alternative bioactive bone filler to PMMA. The application of CPCs as filler in vertebral compression fractures is still controversial. Some investigators reported favorable clinical results for CPCs used in vertebra fractures.^17,18 However, several literatures emphasized that standalone CPCs are prone to vertebral body refractures.^19,20 Undoubtedly, the mechanical properties of CPCs are inferior to PMMA.²¹ Schutzenberger et al. performed stentoplasty with CPCs to treat vertebral fracture which showed excellent clinical results, but height loss still existed.²² The major disadvantage of CPCs is longer osteogenesis time and resulting weaker stiffness. In order to improve mechanical strength, two different methods are proposed: one is to improve primary stabilization through adding fiber reinforcements; and the other is to build secondary stabilization with callous tissue of new bone rapid formation.²¹

In recent years, there are several new calcium salt bone void fillers that have been adopted in orthopedics and craniomaxillofacial surgery, such as calcium sulfate/calcium phosphate (CSCP) composite and hydroxyapatite/collagen (HAP/COL) composite.^23,24 Both have demonstrated good bioactive properties, which have been confirmed in fast osteogenesis.^25,26 However, studies of these fillers used in stentoplasty are rarely reported.

Therefore, we designed a prospective study to investigate stentoplasty with resorbable bone fillers to treat vertebral compression fractures. Our aim was to clarify the clinical and radiological outcomes of vertebral compression fractures treated by stentoplasty with resorbable calcium salt bone fillers compared with traditional BKP.

Methods

This study was approved by the life ethics committee of Beijing Friendship Hospital and was registered in China clinical research and trial center (ChiCTR-OIC-16008922). We confirm that this study was conducted in accordance with the Declaration of Helsinki. All queries regarding related data sets collected and analyzed in the present study can be addressed to the corresponding author.

Patients

Patients with fresh thoracolumbar vertebral compression fractures referred to our institute from January 2014 to January 2016 were included in this prospective study. Inclusion criteria are: (i) older than 50 years old; (ii) refractory back pain caused by low-trauma damage after conservative treatment for at least 6 weeks; (iii) definite diagnosis of fresh thoracolumbar vertebral compression fracture through MRI or radionuclide bone scan; (iv) single vertebra fracture. Exclusion criteria are: (i) unstable vertebral compression fracture; (ii) malignant tumor; (iii) neurologic symptoms due to compression and stimulation to spinal marrow or nerve root; (iv) primary disease unable to tolerate the operation. It included patients who met the requirements in the study based on the voluntary principle. All included patients needed to sign informed consent forms. All included patients were randomly divided into three groups (group A, group B and group C).

Surgical Technique

The operation was performed through unilateral pedicle approach under local anesthesia. All the operative processes were directed under X-ray fluoroscopy. The patients in group A and B were operated on following the procedures outlined below (Figure 1). After a working cannula was established in the vertebral body, a VBS system (Weigao Group Co Ltd, CHN) was propped up through the balloon to restore the vertebral body height. Then, bone filler was injected into the vertebral body after the balloon was removed. The patients in group A were injected with CSCP fillers (Gene-X, Biocomposites Ltd, UK), while patients in group B with HAP/COL composite filler (Bongold, Beijing Allgens Medical Science & Technology Co Ltd, CHN). Then, the working cannula was removed and skin was sutured. The patients in group C had traditional BKP filled with PMMA. All patients were inquired about the sensation and movement of the lower extremities immediately after surgery.

Figure 1 Intraoperative radiological images. (A) Build a working cannula through unilateral pedicle approach. (B) Inflate balloon and push open VBS. (C and D) Adjust the balloon placement and completely push open VBS. (E and F) Remove balloon and inject filler material, then remove working cannula. Abbreviation: VBS, vertebral body stenting.

Clinical and Radiological Evaluation

Clinical evaluation was implemented through visual analogue pain scale (VAS) and Oswestry disability score (ODI). VAS was used to assess back pain (0–10, 0 = no pain, 10 = worst pain). Meanwhile, ODI was investigated to evaluate disability and the quality of life (ODI: the actual score/highest possible score × 100%, 0 = normal, the closer to 100%, the more serious the dysfunction is). The operation time, blood loss, bone filler volume and complications were recorded and evaluated. Radiological evaluation was measured by the vertebral height and Cobb angle of the operative vertebra on the lateral radiograph. Cobb angle was measured between superior and inferior endplate of operative vertebrae. Vertebral height compression rate was calculated by a ratio of anterior to posterior height. The osteogenesis was evaluated from transverse section computed tomography (CT) images. Three circular areas (2 mm²) inside the VBS were randomly selected to calculate average CT value (Hounsfield unit, Hu). Three circular areas (2 mm²) randomly selected outside the VBS in the same vertebra were considered as the standard control. The CT values comparison between inside and outside of the VBS and the observation of bone trabecula were applied as the osteogenesis evaluation method. All the above data were collected and recorded preoperatively, 1 day after surgery, 1 year after surgery (short term) and 3 years (long term) after surgery.

Statistical Analysis

All statistical analysis was conducted in SPSS 20.0. The age, body height, body weight, BMI, BMD, VAS, ODI, AH, preoperative Cobb angle, preoperative compression rate, intraoperative blood loss, operation time and bone fillers volume were shown as average ± standard deviation. Paired t-test was used for comparison within the group. One-way ANOVA test was used for comparison between the three groups. Chi-square test was used for comparison of the sexual distribution of the three groups. P < 0.05 was considered as statistically significant.

Results

The 42 patients included in the study were randomly divided into three groups (14 patients per group). All patients went through surgeries successfully. Mean follow-up time was 3 years. There was no significant difference in age, gender, body height, weight, BMI, BMD, preoperative Cobb angle, preoperative compression rate, intraoperative blood loss, operation time and bone filler volume in the three groups (Table 1).

Table 1 Clinical Data Comparison in Three Groups

The preoperative VAS showed a mean score of 7.71±0.61 in group A, 7.71±0.67 in group B and 7.39±0.81 in group C. The postoperative VAS evidently decreased to 2.39 ±0.45 in group A, 1.86±0.63 in group B and 2.07±0.51 in group C (P< 0.05). At 1-year follow-up, VAS was maintained to 1.82 ±0.37 in group A, 1.75±0.58 in group B and 1.79±0.58 in group C. The 3-year VAS was 2.14 ±0.53 in group A, 1.93±0.65 in group B and 1.68±0.61 in group C. There was no significant difference in VAS among the three groups preoperatively and at each postoperative time point (Figure 2).

Figure 2 The clinical and radiological evaluation indicators trend. Three groups were not significantly different from each other for each evaluation indicator preoperatively and at each postoperative follow-up point (P>0.05). (A) AH. (B) Cobb angle. (C) VAS. (D) ODI. Abbreviations: AH, anterior height; VAS, visual analogue scale; ODI, Oswestry disability score.

The preoperative assessment of ODI showed a mean score of 81.79±5.03 in group A, 81.36±4.36 in group B and 79.93±4.76 in group C. There was no significant difference in ODI among the three groups before surgery. The postoperative ODI scores decreased to 36.79±5.26 in group A, 34.0±4.80 in group B and 32.71±3.81 in group C (P< 0.05). At 1-year follow-up, the scores dropped to 21.48±4.68 in group A, 23.64±4.92 in group B and 26.64±4.53 in group C. At 3-year follow-up, the scores remained 22.21±6.18 in group A, 19.36±7.32 in group B and 23.21±5.89 in group C. There was no significant difference in ODI among the three groups at each postoperative time point (Figure 2).

As shown in Table 2, the AH and Cobb angle significantly improved compared with preoperative in three groups (P< 0.05). Meanwhile, there was no significant difference among the three groups at preoperative and each postoperative time point (Figure 2). However, the AH was remarkably reduced from postoperative (2.46±0.62 cm) to 1-year (2.40±0.62 cm) in group A (P = 0.005) and from postoperative (2.49±0.62 cm) to 1-year (2.40±0.54 cm) in group B (P = 0.029). Both group A and group B presented loss of AH in the comparison within the group at 1-year follow-up. At 3-year follow-up, the loss of AH did not continue compared with 1-year follow-up, and was still higher than preoperative (P< 0.05). A typical case of vertebral height loss during the follow-up is presented in Figure 3.

Table 2 The Radiological Data Preoperatively and at Each Postoperative Time Point

Figure 3 A typical case of vertebral height (L1) loss during the follow-up. (A) Coronal image at 1 day after operation. (B) Coronal image at 1 year after operation. (C) Coronal image at 3 years after operation. (D) Sagittal image at 1 day after operation. (E) Sagittal image at 1 year after operation. (F) Sagittal image at 3 years after operation.

As to the osteogenesis efficacy of fillers, we could observe remarkable bone trabecula formation from the transverse section CT images in group A and group B at 1-year and 3-year follow-up (Figure 4). In group A, the CT value inside the VBS was gradually dropping, but always higher than outside the VBS with the time prolongation. The CT value in the group B was gradually rising and higher than outside the VBS at 1-year follow-up (Table 3). After 1 year, the CT value inside the VBS was higher than the control area in group A and group B (P< 0.05).

Table 3 The CT Value of Inside and Outside VSB at Each Postoperative Time Point

Figure 4 The evaluation of bone trabecula formation and density variation inside and outside the VBS at 1 day, 1 year and 3 years after operation. (A–C) Group A. (D–F). Group B.

Two patients in group C and one in group A experienced asymptomatic cement leakage. One patient suffered adjacent vertebral body fracture in group C (Table 1). No complications such as nerve root compression, spinal cord compression, pulmonary embolism, fat embolism, thrombus or wound infection were observed in the three groups.

Discussion

VP and BKP are the most common minimally invasive treatments for patients with vertebral compression fractures. Both techniques can relieve pain and improve the quality of life. Our clinical data show that stentoplasty with resorbable bone fillers also has an outstanding contribution to relieve back pain and improve quality of life. The postoperative VAS and ODI scores improved evidently compared with their preoperative values. At each follow-up time, the VAS and ODI were maintained within satisfactory levels after operation. Both in the short term and long term, this operation played an important role in pain relief and quality of life improvement. In this study, we reported for the first time the similar clinical efficacy between stentoplasty with resorbable bone fillers and BKP with PMMA. This result can help better understanding of the mechanism of BKP pain reduction. Some authors insisted that sensory nerve ending fibers can be destroyed through polymerization heat reaction and monomer toxicity of PMMA.¹⁸ However, resorbable bone fillers did not produce any heat or toxic substance that influenced sensory nerve ending fibers. Therefore, the stabilization of vertebra is the critical factor for immediate and persistent pain alleviation. Mechanical stabilization can prevent stimulation of algesi receptors in the periosteum, bone marrow and Haversian system.

VBS can effectively restore and maintain the vertebral height, and has been widely used in the treatment of vertebral fractures.^22,27 Restoration of the vertebral body height can reduce the stresses on the adjacent endplates and decrease the risk of refracture.²⁸ In our study, postoperative vertebral height and Cobb angle were obviously improved. However, the AH of the vertebral body appeared to have a slight loss within 1 year, but was still higher than preoperatively. At present, the strength of resorbable bone fillers is far weaker than that of PMMA. VBS with bone fillers cannot provide enough strength to support a routine load before formation of actual bone architecture. After bone structure is reconstituted, the vertebral height is kept relatively stable with a higher strength which prompts secondary stabilization. Therefore, the height did not change from 1-year to 3-year.

In this study, two types of resorbable bone fillers significantly enhanced bone cancellous reconstruction of the vertebral body. In fact, two types of bone fillers prompt osteogenesis with different mechanisms. CSCP is a composite material composed of calcium sulfate and calcium phosphate based on a certain ratio, which has been used in clinics for several years.²⁹ The resorption of CSCP is accompanied with vascular infiltration, osteoid deposition and finally restoration of new mineralized bone trabeculae.³⁰ In addition, it has been shown that CSCP composite can stimulate cell proliferation and enhance angiogenesis, which can be directly bonded to host bone. In a rabbit femur model, CSCP can repair the defect with newly formed bone at 6 months.³¹ In our previous study, we reported prophylactic VP with CSCP in sandwich vertebrae. The 2-year follow-up study result showed prophylactic VP with CSCP decreased the fracture risk of sandwich vertebrae.²³ HAP/COL composite is composed of type I collagen fiber and hydroxyapatite nanocrystals, which is like that of natural bone.^32,33 Moreover, the porous structure provides effective space for bone formation not only around the fillers but also into their pores. The open porous structures both allow cells to migrate into the fillers and blood vessels to grow into material, which provides the oxygen and nutrition for osteogenesis.^34,35 In a rabbit tibial model, bone defect was transplanted with HAP/COL composite. Bone defects were repaired and matured at 2 months and fillers were fully resorbed in the bone marrow cavities at 3 months.³⁶ Feng et al. studied the application of HAP/COL in human extraction sockets and analyzed height of new bone formation. They concluded that mineralized HAP/COL has better osteogenesis compared with ordinary and has broad clinical application prospects.³⁷

In our study, CSCP and HAP/COL have good osteogenesis effect based on CT observation. Actually, CSCP and HAP/COL presented different density change trends in the process of osteogenesis. The density of CSCP area gradually decreased with fillers resorption and new bone formation.³⁸ However, HAP/COL presented low density at the beginning of implantation and increased with the calcium salt deposition. Therefore, the osteogenesis effect cannot be directly compared between CSCP and HAP/COL by CT values.

Most studies reported that the risk of adjacent vertebrae fracture increased because of higher stiffness of PMMA.^39,40 Ottardi et al. reported that effective restoration of vertebral height can reduce adjacent vertebrae fracture incidence.²⁸ The rate of adjacent vertebrae fracture in our study did not show significant differences, which may be associated with consistent vertebral height. So far, it is not clear about the stiffness of VBS compared with PMMA. Cement leakage is the most common complication of vertebral augmentation.⁴¹ In our study, two patients in the PMMA group and one in the CSCP group were observed to have asymptomatic cement leakage through CT scan.

There are a few limitations to this study. First, the sample size is relatively small. Further studies may need to recruit more patients. Second, there is a lack of precise method to compare osteogenesis properties of these two kinds of fillers. Third, the wearing time of a waist brace has not been standardized, which could influence the changes of vertebral height in the early postoperative period.

Conclusion

The stentoplasty with resorbable calcium salt bone void fillers demonstrated similar clinical outcomes to traditional BKP for vertebral compression fractures. Both HAP/COL and CSCP have performed certain osteogenesis. However, stentoplasty with studied fillers exhibited slight loss of the AH within 1 year after surgery.