Since Galiber et al.  applied PVP in the treatment of 1 case of C2 vertebral invasive hemangioma in 1984, PVP has gradually become one of the effective methods for the treatment of vertebral tumor and OVCFs due to its advantages of simple operation and definite efficacy [12, 16, 17]. KD is a rare special type in OVCFs. After minor trauma, vertebral collapse and kyphosis gradually appear, which are related to vertebral ischemia and necrosis and the formation of vertebral pseudarthrosis. The occurrence of KD is usually manifested as intractable lumbago and back pain, and in severe cases, nerve damage may occur [18, 19]. X-ray and CT examination showed IVC in the vertebral body or MRI suggested limited fluid filling in the vertebral body [1, 2, 20]. IVC is mainly located in the thoracolumbar region, and most of the fractures are wedge-shaped fractures, with fractures occurring near the upper and lower endplates of the vertebral body [11, 21].
Due to the presence of IVC and pseudarthrosis, the injured vertebra of KD can flex during spinal flexion activities, which can widen the cracks in the vertebral body . The partial correction of collapsed vertebral height and kyphosis can be achieved when the spine is extended. In PVP, the bone cement is often confined to the diffusion in the crack, which has the effect of maintaining the extension and correcting kyphosis, without the need of PKP balloon expansion and reduction. Spontaneous reduction can occur in patients with KD in posterior extension without further balloon expansion reduction [23, 24]. Heo et al.  reported that excessive reduction is likely to accelerate the process of vertebral ischemia and necrosis, leading to severe recollapse, so excessive reduction of injured vertebra during surgery should be avoided. Therefore, we selected PVP combined with postural reduction to treat KD.
A small dose of bone cement can restore the mechanical properties of the fractured vertebral body, and the amount of bone cement injected has no obvious correlation with the analgesic effect [26, 27]. Even 1.5 ml of bone cement injected into each vertebral body can obtain satisfactory analgesic effect [25, 26]. Biomechanical studies have confirmed that vertebral strength can be restored by injecting about 2 ml bone cement, and vertebral stiffness can be restored by injecting about 4 ml bone cement . In this study, both groups have reached the requirements of restoring vertebral strength and stiffness. However, the amount of bone cement injection is a one-sided index to reflect the benefit of bone cement and cannot reflect the distribution of bone cement in vertebral body. The diffusion volume of bone cement can more reasonably reflect the distribution of bone cement in vertebral body. This study suggests that different distribution patterns of bone cement are closely related to the degree of osteoporosis, the degree of compression, the position and internal morphology of IVC, and the degree of peripheral sclerosis of IVC.
In this study, VAS and ODI postoperatively were significantly reduced in both groups and were maintained to the last follow-up. Li et al.  found that VAS and ODI were significantly improved at the last follow-up compared with those before treatment in patients with KD after intensive vertebral treatment. In this study, spongy group is superior to the blocky group in terms of pain relief and functional recovery. Due to the obstruction of the ossification band and fibrous membrane at the edge of IVC, bone cement of the blocky group was filled in the IVC region in the form of solid mass, resulting in the failure of effectively filling of bone cement in other gaps and osteoporosis areas around the crack. The limited bone cement mass cannot be connected with the upper and lower adjacent endplates and cannot strengthen cancellous bone of vertebrae, which is more prone to stress shielding leading to recollapse , and cannot support the normal physiological stress from the body resulting in the continued existence of pain symptoms caused by osteoporosis . In the spongy group, the bone cement was intercalated trabeculae with diffuse distribution and had high degree of vertebral strengthening. In this study, the average injection amount of bone cement per vertebra in the blocky group was (4.7 ± 1.55) ml vs. (4.0 ± 1.16) ml that in the sponge group. Although the injection amount was more in the blocky group than in the spongy group, the spongy group showed better recovery of pain relief and function after treatment. Many scholars also found that the injection amount of bone cement was not proportional to the analgesic effect [25, 26]. Therefore, it is suggested that the fractured cracks should be filled and the cement should be better distributed in the vertebrae. We found that the postoperative correction degrees of vertebral compression and kyphosis in the two groups gradually decreased over time, suggesting that the postoperative vertebral height and Cobb angle gradually lost, which was consistent with previous findings [17, 30]. The loss of the blocky group was more obvious than that of the spongy group, which was related to the distribution pattern of the bone cement itself.
The correction rate of postoperative kyphosis and vertebral compression was more obvious in both groups, but these correction rate of blocky group was more obvious than that of spongy group. Course of the disease, age, preoperative Cobb angle and vertebral compression rate of the blocky group were significantly higher than those of the spongy group, while the anterior vertebral height and BMD of the blocky group were significantly lower than those of the spongy group. It is suggested that the condition of KD in the blocky group is more serious than that of the spongy group, the formation time of IVC is longer, the volume of IVC is larger, and the pseudarthrosis is obvious. Therefore, when the body position is reduced, the vertebral body has a large reduction space and presents the "open-mouth state". The injected bone cement presents a mass distribution, which effectively restore the height of vertebral body and correct kyphosis.
The bone cement leakage rate of the blocky group was 16.67% lower than that of the sponge group (21.21%), but there was no statistical difference between the two groups. Krauss et al.  reported that the bone cement leakage rate was 18.2% in the PVP treatment of KD. Wang et al.  reported that the bone cement leakage rate was 7.4% in the PKP treatment of KD. In the blocky group, the anterior vertebral edge leakage was the main part, while in the spongy group, the intervertebral disc leakage and the paravertebral vein leakage were the main part. However, there were no related clinical symptoms in both groups. In the blocky group, due to the obstruction of the sclerosis band and fibrous membrane around the IVC area, the injection of bone cement showed a lumped distribution. The bone cortex in the anterior edge of the vertebral body was often incomplete, which could leak out along the anterior edge to the front of the vertebral body. Therefore, the incidence of leakage in the anterior edge of the vertebral body of the blocky group was the highest. However, the spongy group of bone cement was widely inserted into the trabecular space and distributed in a diffuse manner, which was easy to be squeezed into the intervertebral disc and the paravertebral vein plexus. In order to prevent the occurrence of bone cement leakage, the operator should carefully analyze the imaging data before treatment, fully understand the IVC position, size and fracture site, and grasp the position of pedicle injection.
The blocky group had an incidence of adjacent vertebral fracture of 16.67% vs. the spongy group of 9.09%. But there was no statistical difference between the two groups. It indicates that the blocky distribution of bone cement is more likely to lead to the occurrence of adjacent vertebral fractures. Yang et al.  found that 14.1%-39.1% OVCFs patients in the PVP treatment experienced an adjacent vertebral fractures during the first year after surgery. In the blocky group, the vertebral body was seriously compressed and kyphosis was obvious. The biomechanics of adjacent segments changed greatly. The mechanical transfer between the vertebral body and the intervertebral disc was changed, resulting in the increase of the stress of adjacent vertebrae and the higher probability of new fracture. In addition, the bone cement was filled in the IVC area in the form of solid mass, which is more prone to stress shielding, increases the stress load of adjacent intervertebral discs and vertebral bodies, and increases the risk of recurrent fractures . The bone cement in the spongy group was evenly distributed, which made the stress of the whole vertebral body relatively evenly distributed, and reduced the concentrated stress on the adjacent vertebral body and intervertebral disc. Therefore, the complication rate of new adjacent vertebral fracture is low.
This study has some limitations. Firstly, KD is relatively rare, and easy to be complicated with OVCFs and multiple fractures, so there are few cases that meet the inclusion criteria. Therefore, the number of cases in this study was small, and the sample size needs to be expanded for further verification and analysis. Preoperative ICV morphology and size were not measured in three dimensions, and biomechanical studies of cement distribution in the vertebrae on adjacent vertebral bodies and intervertebral discs were lacking to support the results. The researchers did not completely blind in the collection of data, which may lead to bias in the data. Therefore, the current findings provide a clinical reference, but need to be validated in further multi-center, randomized, double-blind clinical trials.