The clinical features of Kummell’s disease (KD) include: a history of trauma, asymptomatic period after trauma, delayed vertebral collapse and the existence of the vacuum or fissure signs in the injured vertebral body. It is reported that the pathological mechanism of KD is complicated and may be the result of the interaction of a variety of factors such as avascular necrosis in the vertebral body, changes in the biomechanics of the spine, and incomplete repair of bone microfractures . The majority of researchers believes that the disease starts from osteoporotic vertebral fractures, followed by avascular necrosis of the injured vertebrae and pseudojoint formation, eventually leading to vertebral body collapse. The treatment of KD mainly includes conservative treatment, minimally invasive PVP or PKP and open reduction internal fixation . Most of KD is still treated with surgery at present. PKP and PVP have been widely used in patients with neurologically intact osteoporotic KD, especially those who are intolerant to general anesthesia [16–18]. So far, many studies have reported the good effects of PVP and PKP in the treatment of KD [19–21], but it is still controversial that which surgical method is more advantageous. On the basis of these researches, we initiated a comparative study on patients with type I or II KD treated with PVP and PKP.
Our data showed that the operation time, operation cost, and intraoperative fluoroscopy times in the PVP group were statistically lower than those in the PKP group (P < 0.05). The volume of bone cement injection in the PKP group was higher than that in the PVP group, and the incidence of postoperative bone cement leakage in the PVP was higher than that in the PKP group, however, these differences were not statistically significant. Many previous similar studies have also confirmed that the rate of bone cement leakage in PVP was higher than that of PKP [22–24]. The possible explanation is: in PKP, the surgeon creates a low pressure environment before injecting bone cement into the injured vertebrae, which could reduce the penetration of bone cement to surroundings, compared with the high pressure environment of PVP. Meanwhile, during the balloon expansion of the PKP, the surrounding bone can be compressed more tightly, further sealing the fracture and reducing leakage. Our team has carried out a number of researches on PKP since 1999 and gradually formed a set of bone injection technique that can effectively reduce bone cement leakage, which was called temperature gradient cement injection technique. The key technical points include: (1) Choosing the best initial injection time point for bone cement; (2) Following the principle of injection slowly with low pressure; (3) Temperature gradient injection technique (ie interval injection technology) ; (4) Bone cement layered modulation technique if necessary. Both temperature gradient injection technique and bone cement layered modulation technique are applied in our cases. The insignificant statistical results obtained in this study may be related to the insufficient sample size or the different tendency to choose PVP and PKP. Since the PKP group is mostly patients with poor preoperative postural reduction, we will continue to expand sample size in order to draw more accurate conclusion. The serious surgical complications of PKP and PVP also include pulmonary embolism, nerve injury, and delayed cement displacement . Fortunately, there was no serious complications in our study. We believed that antiosteoporotic treatment and rehabilitative exercise of the muscle strength of the waist and back are helpful to prevent delayed displacement. In addition, regular and timely follow-up after operation is also necessary. Regarding choice of surgical methods for type I or II KD, we prefer the safer PKP to reduce or avoid surgical risk.
Past studies have found that whether it is unilateral or bilateral injection of bone cement, the axial vertebral rigidity and strength can both be restored, and that the injection of bone cement into the pedicle on both sides can achieve good biomechanical performance . Even so, in order to achieve consistency in the treatment of KD patients, we only selected bilateral puncture cases. It was reported that PKP is superior to PVP in treating KD with regard to improvement of vertebral heights and kyphosis deformity [27, 28]. In our study, compared with before operation, two groups of height of the anterior edge of the injured vertebrae and Cobb’s angle at 1 day, 6 months and the final follow-up after operation were significantly corrected (P < 0.05). The improvement of injured vertebral height and kyphotic deformity in the PKP group was significantly better than that of PVP group at each time point. At the last follow-up, there was no significant loss in the vertebral height and the Cobb’s angle of the kyphosis did not change significantly. Similar results were also confirmed by Zhang's study  that PKP can obtain more satisfactory reduction results in the treatment of KD. According to Zhang et al , the correction of Cobb’s angle in the PKP group was slightly better than the PVP group and there were no significant differences between two groups. The reason for the different results may be the different sample size and follow-up time in the two groups. The sample size and follow-up period in our study was larger and more easier to find the difference, and our conclusion was therefore more convincing. Our finds were also verified by a systematic review and meta-analysis that PKP is superior to PVP in terms of vertebral height recovery and correction of kyphosis . The reason that PKP is better than PVP in correcting the height and kyphotic deformity of the injured vertebral body may be: PKP can increase the space for bone cement implantation through the mechanical expansion of the balloon in the injured vertebrae. At the same time, the balloon can directly restore the vertebral height and correct the kyphotic deformity, which is not available in PVP. The comparison of imaging data in our research indicated that the injected bone cement all reached or exceeded the midline of the injured vertebrae, and well dispersed in the middle or anterior position of the vertebral body. In addition, there are cavities and fissures in the vertebral body of KD patients, and the compression and collapse of the vertebral body mainly occur in the anterior or middle vertebral column, which also accorded with the biomechanical characteristics of the spine .
The purpose of surgical treatment of KD is not only to obtain better imaging results, but also to relieve pain and restore function as soon as possible. For some elderly patients with or without severe osteoporosis, it is not even necessary to excessively restore the vertebral height and Cobb’s angle. More than two years of follow-up found that VAS and ODI scores of both PVP and PKP groups were significantly lower at 1 days, 6 months, 1 year, 2 years and the final-up after surgery than before surgery, but no significant difference between PVP and PKP groups was found at all time points postoperatively. This indicated that PVP and PKP have basically the same curative effect in the treatment of type I or II KD and both surgical methods can significantly relieve pain and improve the life quality of KD patients. We believe that the main reason of bone cement injection for pain relief may be that bone cement fills the IVC of the injured vertebrae and plays the role of physical support. Similar conclusions was also made by some studies which believed that both PKP and PVP operation can obviously relieve pains and have little difference in improving the postoperative function of patients [29, 32, 33]. Considering the safety and the better reduction of PKP, we suggest that for patients with no financial concerns, PKP treatment should be given priority on the basis of informed consent to patients and their families.
However, there are some important limitations of this study. The sample size is small and non-RCT articles might induce various types of bias. To confirm our findings, a large multi-center randomized RCT should be conducted. In addition, duration of follow-up period in both groups varied and might bring negative impact on our results.