The current case series presents the short-term results of a very early series of the first 100 cases treated with the VBS system. In the meantime four other studies have presented clinical and radiographic outcomes of the VBS system, but all patient samples were considerably smaller (~ 50–20 cases) [10–13]. Overall, good restoration of vertebral body height and alignment were obvious, but a direct translation of realignment into a significantly reduced rate of adjacent vertebral fractures, as suggested in the biomechanical literature, could not be shown in our series , and much larger case numbers will be needed for a conclusive assessment of this aspect.
VBS was developed for overcoming the weakness of the BKP system in incompletely maintaining the restored vertebral body realignment and height that is achieved with a fully inflated ballon [15, 16]. This weakness of BKP leaves a considerable one third of treated cases without an appreciable height restoration . The percentage and the overall extent of realignment shall be improved with VBS based on the principles of balloon kyphoplasty and vascular stenting. Using VBS, the stent remains within the newly created vertebral cavity so the balloon can be removed after deflation while preventing the vertebral body from collapsing, so that, in an ideal scenario, a virtually physiological vertebral body height and shape can be restored and preserved. In their recent RCT, Werner et al. did not find significant differences in vertebral kyphosis correction between BKP and VBS. Both systems achieved around 4.6° of mean reduction . Despite a sufficiently powered analysis and presumably well balanced fracture characteristics, preoperative kyphotic deformity angles of the two groups were not reported, which, according to our findings, have an influence on the reduction potential of a fracture. Moreover, stratified results by fracture type were not reported, which may have revealed differences between the two therapies. Thaler et al. could only achieve an average 3.5° reduction of vertebral kyphosis in 27 patients, but the mean preoperative Beck Index of 0.87 implies a smaller reduction potential than that of our group (preop 0.73), which may explain the better reduction of 4.2° in the current study .
Using a different unit for describing re-alignment but kyphotic angles, Maestretti et al. reported Beck Index improvements of 0.14 in their traumatic BKP case series treated with calciumphosphate cement  and Krüger et al. of about 0.07 in their series of osteoporotic incomplete burst fractures . However, in the latter series the preoperative BI values were around 0.8, whilst those in the Maestretti series were 0.7 and hence better comparable with the extent of VB deformation in our series. The overall BI improvement we found was 0.2, and even 0.26 in an idealized scenario ignoring the non-moving and poor-moving fractures. These values do also make clear that they can only be achieved in fractures with preoperative BI values of about 0.7 or lower.
The indication for applying a vertebral augmentation method from the meanwhile available spectrum of “simple” vertebroplasty with no intrinsic mechanical method for height restoration but patient positioning, to balloon kyphoplasty and vertebral body stenting should not only be based on fracture type, patient characteristics and extrusion risks but also on the extent of vertebral body deformation. An only mildly deformed fracture has a generally small realignment potential, and fractures with Beck Indices around 0.8 and local kyphotic angles of 8.5°  are probably more suitable for BKP or even VP than fractures with Beck Indices < = 0.7 and local kyphotic angles of 17°  where VBS can develop its full realignment potential. As shown in Figure 5, about a third and up to half of the fractures had mean BI improvements of 0.3-0.5, but such extent of restoration can only be achieved with a corresponding preoperative deformity. If not given, the realignment potential of therapies such as VBS and other implant based augmentation technologies is limited by a “ceiling effect”, i.e. restoration of BI to values greater 1.0 is largely impossible and the resulting overcorrection clinically not meaningful. In addition, we could show that reports of Beck Indices and local kyphotic angles are influenced by fracture morphology and should hence be described separately. In our case series there were about 20% biconcave fractures and the relation between vertebral body middle height and posterior wall height, herein introduced as alternative Beck Index is the most appropriate way of describing these fractures. The local kyphotic angle or the original Beck index reflecting the relation between anterior and posterior vertebral height are less suited for a description of biconcave fractures. The anterior-posterior wall relations deal with the important local and the resulting segmental kyphotic deformity which is, based on biomechanical considerations, responsible for increased risks of new adjacent and distant vertebral fractures and patients’ postural decompensation in the kyphotic plane .
Klezl pointed out that kyphosis correction with VBS was better in the traumatic group where even reduction of the fractured endplate with the stent could be achieved with possible implications on future performance of injured discs in young patients . Research suggests that a vertebral trauma and especially a fractured endplate can cause disc cell apoptosis and disc degeneration [19, 20]. The anterior spinal column, especially the fragmented superior endplate could be well reconstructed by the stent provided that it was inserted accurately. Our results confirm these observations where the alternative BI was improved from 0.53 to 0.75 in biconcave but also from 0.64 to 0.85 in crush fractures while it only increased from 0.68-0.79 in wedge shaped fractures. Endplate and mid vertebral height restoration are new aspects in minimally invasive fracture treatment whereas improved realignment and decreased cement leakage was the original goal of the balloon kyphoplasty principle.
Looking at all moving wedge shaped fractures, the local kyphotic angle was even improved by 10°, from preoperative 17.7° to 7.6°. Diel et al. report average improvements of 4° with VP  and Hulmes reports an average of 6.6° improvement for BKP and VP in his systematic review . In contrast, Papanastassiou et al. reported 4.8° kyphotic angle change from baseline for BKP and only 1.7° for VP in their systematic review of randomized and non-randomized controlled studies. Such comparisons highlight the potential of VBS. However, a translation into significantly reduced adjacent fracture rates could not be deduced from our data yet. A 9% rate of new fractures is lower than the 10.4% after BKP but higher than the 8.4% after VP as reported by Papanastassiou . Considering our analysis of cases with new fractures after surgery, the sex distribution in these groups should always be reported since female gender seems to represent a risk factor.
The overall 29% cement extrusion rate we observed is comparable with BKP rates if assessed in an independent fashion. The FREE study reported a 27% extrusion rate whereas the systematic analysis of Hulme et al. calculated an only 9% rate based on published literature [2, 22]. In many reports, however, authors have assessed their own extrusions which probably lead to a gross underreporting. Thaler et al. reported a 25.5% extrusion rate in their VBS series and Werner et al. had a 20% minor and 10% major leakage rate. In both studies, a viscometer was used [12, 13]. The final rate of symptomatic extrusions in our series was 1% which is lower than the 2.6% in VP and comparable with the 1.3% in BKP reported by Hulme . Klezl had 2 asymptomatic leakages in 20 treated vertebral bodies and Muto had none [10, 11]. Klezl pointed out the more difficult situation in osteoporotic bone where a slight and careful overfilling of cement should be aimed at for achieving a good interdigitation of cement with bone . Considering the 50% osteoporotic fractures in our series and its 23.8% extrusion rate, the cementation challenge in this group seems to have been well met, this despite most surgeons in our series probably not having passed their VBS learning curve yet. New cementing techniques like radiofrequency kyphoplasty may help to further reduce leakage rates. Kurth et al. reported a respectably low rate of only 15.5% in a multicenter study with 186 treated vertebral bodies . High viscosity cement is another promising option for leakage prevention, which has mostly been applied and assessed in vertebroplasty. Georgy found only 8% of moderate or severe leakage in a chart/xray review of 66 treated levels , but this figure is not easy to compare with other reports that have not graded leakage into none, mild, and more severe types. In the meantime, a higher viscosity cement is also available for VBS (Vertecem “V+”) and future studies will have to show if it can help reduce leakage rates to levels comparable with the above cited percentages.