Therapeutic Efficacy of Third-Generation Percutaneous Vertebral Augmentation System (PVAS) in Osteoporotic Vertebral Compression Fractures (OVCFs): A Systematic Review and Meta-analysis

Purpose This study aimed to assess whether the third-generation PVAS was superior to percutaneous vertebroplasty (PVP) or percutaneous kyphoplasty (PKP) in treating patients with OVCFs. Methods Databases, including Pubmed, Embase, and Cochrane library, were searched to identify relevant interventional and observational articles in vivo or in vitro comparing the third-generation PVAS to PVP/PKP in OVCFs patients. A meta-analysis was performed under the guidelines of the Cochrane Reviewer's Handbook. Results 11 in vivo articles involving 1035 patients with 1320 segments of diseased vertebral bodies and 8 in vitro studies enrolling 40 specimens with 202 vertebral bodies were identified. The vivo studies indicated no significant differences were found in visual analog scale (VAS), Oswestry Disability Index (ODI), operation time, or injected cement volume (P > 0.05). The third-generation PVAS was associated with significant improvement in vertebral height and Cobb angle (P < 0.05) and also with a significantly lower risk of cement leakages and new fractures (P < 0.05). The vitro studies suggest that the third-generation PVAS was associated with better anterior vertebral height (AVH) and kyphotic angle (KA) after deflation and cement. No significant differences were found in stiffness or failure load after cement between the two groups (P > 0.05). Conclusion Based on current evidence, although providing similar improvement in VAS and ODI, the third-generation PVAS may be superior to PVP/PKP in local kyphosis correction, vertebral height maintenance, and adverse events reduction. Further high-quality randomized studies are required to confirm these results.


Introduction
Over the past few decades, as a minimally invasive procedure, the PVAS has been considered the optimal management for symptomatic OVCFs [1][2][3]. PVP, the first-generation PVAS, can provide effective and rapid pain relief and spinal stabilization via direct injecting polymethylmethacrylate (PMMA) into the inter-trabecular marrow space of a fractured vertebra [4]. However, this procedure is challenging to restore vertebral height and with up to 54.7% of cement leakage [5]. The matters lead to the evolution of the second-generation PVAS, PVP, which can correct kyphosis through inflation of a balloon inside the collapsed vertebral body [6]. The balloon could also create a cavity, allowing more viscous cement to be injected with lower pressure, thereby significantly reducing leakage risk [7]. Notwithstanding, PKP has been proved to be associated with a higher rate of refracture on cemented vertebrae than PVP [8], especially with an intravertebral cleft (IVC) [9][10][11][12]. Moreover, secondary loss of the initial reduction may occur after balloon deflation [13].
These concerns promote the emergence of the thirdgeneration PVAS. This novel expandable scaffolding device is permanently implanted into the vertebral body to restore reduction mechanically before injecting bone cement. Up to now, several systems are available: SpineJack® [14][15][16][17], Vertebral Body Stenting® (VBS) [18][19][20], OsseoFix® System [21,22], and Kiva® System [23][24][25]. In theory, the thirdgeneration PVAS is superior to PVP/PKP in height restoration and height maintenance. However, inconsistent results were obtained from different trials comparing clinical symptoms recovery, vertebral height restoration, and adverse events of the third-generation PVAS versus PVP/PKP in patients with OVCFs [20,24,26]. In order to provide more evidence for clinical decision-making, we conducted a systematic review and meta-analysis to integrate existing evidence from relevant in vivo or in vitro trials to evaluate the superiority of third-generation PVAS over PVP/PKP in the treatment of patients with OVCFs.

Search Strategy.
This systematic review and metaanalysis was conducted based on the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement [27]. The current systematic review protocol was registered on INPLASY.COM (ID: INPLASY202110015) and available in full https://inplasy.com/inplasy-2021-1-0015/. A systematic computer-based retrieval for all relevant published articles in vivo or in vitro was performed in medical databases including Pubmed, Embase, and Cochrane Library from inception to December 31, 2020. The search terms for the study object: "Spinal Fractures [Mesh]" OR " Spinal Fracture * " OR " thoracic fracture * " OR " lumbar fracture * " OR " vertebral fracture * : " The intervention's search terms are as follow: "KIVA" OR "SpineJack" Studies included in quantitative synthesis (meta-analysis) (

in vivo and 8 in vitro)
Records excluded (n=162) Figure 1: Summary of study selection and inclusion process.

BioMed Research International
OR "vertebral body stent * " OR "Stentoplasty" OR "VBS" OR "OsseoFix." We also checked the reference lists of all including articles to avoid any initially omitted studies. There was no publication language and population limitation during the systematic review. A detailed list of search strategies could be found in Supplemental Appendix 1. (2) Non-original articles (case reports, reviews, letters, meta-analyses, conference abstract, and editorials).

Selection
Criteria. D. CK. and Z. YT. independently screened eligible studies based on the criteria mentioned above. Firstly, the titles and abstracts were reviewed to exclude articles that obviously did not meet the inclusion criteria. Then, a full-text review was conducted to ensure met all the inclusion criteria. All disagreements were resolved by reaching a consensus among the researchers.      BioMed Research International assessed independently by D. CK. and W. HY. using the Cochrane Collaboration's Risk of Bias Tool [29] and Newcastle-Ottawa scale (NOS) [30], respectively. Any discrepancies of data extraction and quality assessment were settled by discussing a third independent author (Z.J.).

Data
Analysis. This meta-analysis was conducted with Review Manager 5.3 software (Cochrane Collaboration, Oxford, UK). Continuous data were calculated through the mean difference (MD) or standardized mean difference (SMD) with 95% CI. We calculated risk ratio (RR) with 95% CI to evaluate the cement leakage and adjacent level fractures. Heterogeneity across studies was assessed using Cochran's Q and I 2 statistics, and P < 0:1 and I 2 > 50% were considered statistical heterogeneity [31]. A fixed-effects model was conducted when I 2 ≤ 50%; otherwise, a randomeffects model was performed. Publication bias was assessed statistically by Stata 12.0 (Begg and Egger tests). Sensitivity analysis was also introduced to detect the result's stability. P < 0:05 was considered statistically significant.
The sensitivity analysis was performed by omitting one study in each round to examine the impact on the overall result. The operation time in the third-generation PVAS group was not significantly different from that in the PKP group when omitting any of the studies except Schützenberger et al. [19]. In addition, the sensitivity analysis suggested no significant variation in bone cement injected attributable to heterogeneity.
Seven studies offered relevant data on new fractures between the two groups. The comprehensive meta-analysis estimated a borderline statistically significant RR of 0.52 (95% CI 0.39 to 0.72; P < 0:0001), suggesting a lower risk of new fractures with the third-generation PVAS (Figure 7(b)). The I 2 value attributed 32% variation to heterogeneity; therefore, a fixed-effects model was used.
The sensitivity analysis of failure load indicated no significant impact on the final pooled result following when omitting any of the eligible studies. The results of the Begg and Egger tests in stiffness and failure load also proved the absence of significant publication bias in terms of the 8 articles in vitro (P > 0:05).

Discussion
As the aging process accelerates, OVCFs contribute to a major health problem worldwide due to the loss of healthrelated quality of life and high healthcare costs [41]. Treatment options usually contain conservative management (analgesics, bracing, bed rest, and physical therapy) and minimally invasive surgery (PVP and PKP). Although most studies suggested PVP and PKP appear to be associated with longer post-discharge survival rates and a cost-effective alternative to nonoperative management [2,[41][42][43], two high-quality RCTs [44,45] indicated that patients could not benefit from vertebral augmentation in resolving pain and disability. Furthermore, refractures and new fractures, the most severe complications, were not avoided after PVP or PKP. Currently, various forms of third-generation PVAS involving SpineJack, KIVA, VBS, and OsseoFix have been evaluated by cadaver and clinical studies [15,20,25,37,39]. However, it is still questionable whether the third-generation PVAS is superior to PKP or PVP. As far as we know, our study is the first systematic review and metaanalysis to comprehensively compare the efficiency of the third-generation PVAS versus PVP/PKP for OVCFs in vitro and in vivo.
The application of the third-generation PVAS in OVCFs has expanded enormously during the last decade [46][47][48][49][50]. Although the facilities of each system are different, all of them are characterized by implanting permanent expandable devices to hydraulically or mechanically control reduction of the vertebral fracture and the sagittal balance of the spine [48,49]. Due to the lack of mechanical reduction ability of PVP, the recovery of vertebral height depends on intraoperative posture or the use of stents to induce scoliosis. PKP can restore vertebral height utilizing a balloon dilatation; nevertheless, it is difficult to maintain height after balloon deflation, even in a lordotic position, where an approximately 110 N compression is still imposed on the vertebrae, resulting in the collapse of the created cavity [36].
Concerning in vitro experiments, our meta-analysis indicated that sagittal height restoration and kyphosis correction were significantly better when using the third-generation PVAS than PKP. The correction loss of PKP may attribute to the deflation effect: The created cavity may collapse after balloon deflation, before cement augmentation, due to the existing constant preload exerted on the vertebra even in a lordotic position [35]. The third-generation PVAS, as specific mechanical properties permanent implant devices, can provide the immediate intraoperative load-bearing capability to  9  4  4  35  6  16  12  30  15   131   Total  42  133  15  74  26  44  55  177  50   616   Total  32  122  25  78  26  13  61  178  50   585 Weight  offset the deflation effect before bone cement injection [36]. Wang et al. [40] found that the VBS could withstand a compressive load of 226 N, exceeding the existing preload of 110 N which exerted on the vertebrae. Despite the lack of accurate values for other devices in included studies, we believe that the effects are similar. No significant differences were found in failure load and stiffness after cement augmentation between the two groups in our meta-analysis, which demonstrated that the implanted permanent expandable devices did not affect the biomechanical behavior of the treated vertebral body. Without considering the influence of surrounding soft tissue, in vitro studies simulate the conditions of immediate post-operation and offer a promising result for the thirdgeneration PVAS, whereas the effects of healing and gradual restoration of activity cannot be evaluated. Thus, we also compared the radiological and clinical results of the thirdgeneration PVAS and PVP/PKP in clinical studies in our systems review. Similar to cadaver studies, experimental results in clinical indicate that the third-generation PVAS was more effective in restoring vertebral body height and correcting kyphosis angle than PVP/PKP at all time points.
On the contrary, no significant differences were observed in terms of short-, mid-, and long-term VAS and ODI, except the mid-term VAS. Previous meta-analysis [7,51] had found that painful and functional improvement were positive correlated with vertebral height recovery and kyphosis correction after PKP/PVP for OVCFs, which is not consistent with our research. To date, it could not yet be established with certainty that height gain and improved outcomes in pain relief and quality of life are clinically relevant. Crucially, the common denominator for pain relief after the third-generation PVAS or PKP/PVP is the internal cement splint [52]. In addition, the surrounding ligaments, muscles and osteoporosis could also affect the outcomes.
Treatment-associated complications, such as cement leakage and new fractures, have caused widespread concern among surgeons. Our results illustrate that the thirdgeneration PVAS could lower the risk of cement leakage and refracture compared to PKP/PVP. It is generally accepted that the high-pressure injection of low-viscosity bone cement would lead to a higher risk of cement leakage [7,53]      14 BioMed Research International and the supported surrounding trabeculae for low-pressure injection. In contrast to PKP, the new system maintains the cavity under the expandable intravertebral implant support, which, in theory, further reduces the possibility of cement extravasation [54]. Moreover, retaining the implant can reduce the use of bone cement, thus theoretically decreasing the occurrence of cement leakage. However, our meta-analysis indicated no differences in the amount of bone cement injected between the 2 groups, which may be attributed to the good maintenance of the cavity supported by expandable implants. Thoracic fractures are often referred to as kyphotic fractures for being associated kyphotic spinal angulation, which lead to the center of gravity being shifted more anteriorly, increasing the lever arm of the forces and the forward bending moments on the already fragile spine. These mechanical changes often result in a further compression of the fractured vertebral but also put adjacent vertebrae at a higher risk of developing new fractures [48]. Although it is still controversial whether these adjacent fractures are due to the surgical procedure or natural evolution, obtaining and maintaining a more adequate reduction via these expandable implants is essential to prevent the domino effect, which is the consecutive occurrence of OVCFs in adjacent vertebrae due to excessive anterior overload after the first uncorrected wedge-shaped vertebral body [54]. Our results also prove that expandable intravertebral implants could provide adequate stability without increasing vertebral stiffness to decrease the risk of adjacent fractures. The limitations of our meta-analysis were as follows: First, the lack of random allocation, allocation concealment, and blinding in the no-RCTs might result selective and performance bias; second, the methods using for evaluating vertebral height change and other outcomes, surgical technologies, and instruments varied among studies, all of which increased the risk of heterogeneity; and third, given the limited number of the included studies in the analysis, the findings should be confirmed in future research with more relevant RCTs to obtain more reliable and conclusive data.

Conclusions
Based on our current evidence, third-generation PVAS provided a similar effect on pain relief and functional improvement compared with PVP/PKP at each follow-up period. However, the third-generation PVAS was more effective for local kyphosis correction, vertebral height maintenance, and with a significantly reduced risk of incidence of cement leakage and new fractures. Further high-quality RCTs are required to confirm these results.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.