Percutaneous vertebroplasty versus percutaneous kyphoplasty for the treatment of delayed post-traumatic vertebral body collapse (Kümmell’s disease) in Chinese patients: a systematic review and meta-analysis
Article Category: Original article
Published Online: Dec 31, 2019
Page range: 273 - 283
Received: Jan 25, 2019
Accepted: Mar 06, 2018
DOI: https://doi.org/10.2478/FON-2019-0045
Keywords
© 2019 Peng-Fei Han et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
Kümmell’s disease, otherwise known as delayed post-traumatic vertebral collapse and avascular necrosis of the vertebral body, after a mild traumatic experience in an asymptomatic period, was first described by the German surgeon Kümmell in 1895.1, 2 Due to the rare and complicated conditions, Kümmell’s disease has so far been reported few times. Kümmell’s disease causes only mild symptoms at an early stage, and the duration of the disease varies, which can delay diagnosis and treatment.3 In the late stage, vertebral body collapse and kyphotic deformity are caused, which often leads to severe intractable back pain.4 Some patients even have symptoms of spinal cord compression, which severely impacts the quality of life. Kümmell’s disease treatment includes conservative and surgical methods, which includes PVP, PKP minimally invasive surgery, and open surgery. As the vertebral body of Kümmell’s patients is destroyed and does not heal itself, conservative treatment is often ineffective for Kümmell’s disease and requires surgical intervention.5 Surgical treatment of patients with Kümmell’s disease results in better clinical outcomes such as good pain relief, functional improvement, and kyphosis correction. Both the PVP and PKP are minimally invasive surgery, which percutaneously injected cement into the vertebral body under local anesthesia and X-ray to enhance the strength and stability of the vertebral body.6 Percutaneous vertebroplasty (PVP) and kyphoplasty can effectively relieve fracture pain and even partially restore the height of vertebral body. At the same time, the thermal effect generated by polymerization of bone cement can further alleviate the pain of nerve degeneration and necrosis in and around vertebral body. Therefore, the operation has the advantages of simple operation, exact curative effect, small trauma, and early activity of the patient.7, 8
Currently, it is generally believed that PVP surgery is more economical than PKP surgery because there are no consumables such as balloons. Clinical application has confirmed that PVP is a safe and effective surgical procedure;9, 10 PKP is a better choice for patients with significant vertebral height loss and old fractures with pseudo articular formation. PKP can better restore vertebral height and correct kyphosis.11, 12 On the other hand, due to the injection of bone cement into the compressed vertebral body under a large pressure, both types of minimally invasive surgery may cause complications such as leakage of bone cement. Reportedly, both procedures can be used for the treatment of Kümmell’s disease, but the optimal choice between these two procedures remains controversial.13, 14 This meta-analysis intends to systematically compare the efficacy between the two procedures in order to provide some theoretical guidance for clinical practice.
The retrieval was conducted in the online databases include Embase, Pubmed, Central, Cinahl, PQDT, CNKI, CQVIP, Wanfang Data, Cochrane Library, and CBM. We also manually searched some journals’ catalog and references, and strive to find gray literature, such as unpublished academic papers and chapters in monographs. Searching all relevant papers without restricting the language and translating if necessary. Keywords searched for both Chinese and English were: Percutaneous Vertebroplasty, PVP, Percutaneous Kyphoplasty, PKP, and Kümmell’s disease. Search strategy was: Kümmell’s disease AND “Percutaneous Vertebroplasty OR PVP” AND “Percutaneous Kyphoplasty OR PKP.”
Inclusion criteria for the analysis were (1) Chinese adults with phase I and II Kümmell’s disease; (2) randomized controlled trials (RCTs), prospective studies, retrospective studies, and cohort studies; (3) Patients received minimally invasive surgery for the treatment of Kümmell’s disease; (4) study compared results of PVP and PKP, (5) The outcome being measured by Cobb angle, anterior vertebral body height, Oswestry Dysfunction Index (ODI), Visual Analogue Scale (VAS) score, cement volume, cement leakage, and operation time.
The exclusion criteria were: (1) letters, comments, editorials, case reports, proceedings, personal communications, or reviews; (2) study objective or intervention measures failed to meet the inclusion criteria; (3) the original documents of experimental design being not precise; and (4) studied with incomplete data.
Inclusion decisions were made independently by two reviewers participated in according to the pre-stated eligiblity criteria. Disagreement between the two reviewers was resolved by discussion or consulting to a third reviewer when necessary. The risk-of-bias assessment tool outlined in Cochrane Handbook was used to measure the methodological quality of case-controlled trials (CCTs). Six domains are evaluated: random sequence generation, allocation concealment, blinding of patients and personal, blinding of outcome assessment, incomplete outcome data, and selective reporting risk. The Newcastle–Ottawa Scale (NOS) was used to assess the quality of cohort studies, the full score is 9 points. Trials with a score of more than 6 points are considered high-quality study. Relevant data were recorded in this analysis, including: first author’s name, published year, sample size of PVP and PKP in the treatment of Kümmell’s disease, and so on.
Data were independently entered into RevMan 5.3 software by two reviewers. Dichotomous outcomes were expressed in terms of Odds ratio (OR) and the weighted mean difference (WMD) or the standard mean difference (SMD) was used for continuous outcomes, both with 95% confidence intervals (95% CI). Heterogeneity was tested using both the chi-square test and
Based on the above mentioned search strategy, a total of 132 related articles were retrieved. By reading the titles and abstracts, we excluded 55 noncontrolled studies, repeated publications, and 40 articles that are not related to the research purpose. Thirty-seven related studies were screened out preliminarily; further reading the full texts and the screening criteria were strictly followed according to the inclusion and exclusion criteria. Finally, one English article and seven Chinese articles were included. Patient’s characteristics and conditions in included study were compared such as age, gender, and so on. The differences were not significant (
Figure 1
Flow diagram of search strategy.
To compare preoperative and postoperative Cobb angle changes in PVP and PKP groups of Chinese Kümmell patients, five clinical studies were included. These studies were divided into three subgroups based on preoperative, postoperative 1–2 days and last follow-up. Random effect model was employed in meta-analysis because the heterogeneity between the studies and subgroups was significant (
Figure 2
Forest plot to assess Cobb angle between two procedures.
According to the preoperative, 1–2 days after operation, and the last follow-up, data were divided into three subgroups. A total of six trials were included, and preoperative and postoperative vertebral body height of Chinese Kümmell’s patients between PVP and PKP groups were compared. Random effect model was employed in meta-analysis because the heterogeneity between the studies and subgroups was significant (
Figure 3
Forest plot to assess anterior vertebral body height between two procedures.
According to the preoperative, 1–2 days after operation, and the last follow-up, data were divided into three subgroups. A total of six trials were included, and preoperative and postoperative ODI of Chinese Kümmell’s patients in PVP and PKP groups were compared. Random effect model was employed in meta-analysis because the heterogeneity between the studies and subgroups was significant (
Figure 4
Forest plot to assess ODI between the two procedures.
According to the preoperative, 1–2 days after operation, and the last follow-up, data were divided into three subgroups. A total of eight trials were included, and preoperative and postoperative VAS score of Chinese Kümmell’s patients between PVP and PKP groups was compared. Random effect model was employed in meta-analysis because the heterogeneity between the studies and subgroups was significant (
The results showed that in PVP group and PKP group, preoperative VAS score (95% CI [−0.09, 0.45],
Figure 5
Forest plot to assess VAS score between the two procedures.
For 6 studies included we compared the amounts of cement injected between PVP and PKP procedures. Random effect model was employed in meta-analysis because the heterogeneity between the studies was significant (
Figure 6
Forest plot to assess cement volume between two procedures.
Seven studies included have compared the cases of cement leakage of PVP and PKP procedures. Fixed effect model was employed in meta-analysis with the absence of heterogeneity (
Figure 7
Forest plot to assess cases of cement leakage between the two procedures.
Five trials included have compared the operation time between PVP and PKP procedures. Random effect model was employed in meta-analysis because the heterogeneity between the studies was significant (
General characteristics of included studies.
| Author | Study design | Group | Cases | Age (y) | Gender (M/F) | Outcomes | Newcastle-Ottawa Scale |
|---|---|---|---|---|---|---|---|
| Chen et al. 201215 | Retrospective | PVP | 33 | 69.2 ± 6.3 | 4/29 | D, E, F, G | ⋆⋆⋆⋆⋆ |
| PKP | 30 | 68.7 ± 6.5 | 3/27 | ||||
| Feng and Sun | Retrospective | PVP | 20 | 72.3 ± 5.4 | 12/28 | A, B, C, D, E, F, G | ⋆⋆⋆⋆⋆⋆ |
| 201816 | PKP | 20 | |||||
| Gao et al. 201617 | Retrospective | PVP | 38 | 73 ± 6 | 20/18 | B, C, D, F | ⋆⋆⋆⋆⋆ |
| PKP | 35 | 75 ± 6 | 16/19 | ||||
| Li 201318 | Retrospective | PVP | 7 | 71.8 ± 7.69 | 3/4 | A, C, D, F, | ⋆⋆⋆⋆⋆⋆ |
| PKP | 5 | 2/3 | |||||
| Shi et al. 201719 | Retrospective | PVP | 10 | 75.1 ± 3.6 | 7/16 | A, B, C, D, E, G | ⋆⋆⋆⋆⋆⋆ |
| PKP | 13 | ||||||
| Yu et al. 201620 | Retrospective | PVP | 14 | 74.47 ± 5.79 | 5/9 | A, B, C, D, E, F | ⋆⋆⋆⋆⋆⋆ |
| PKP | 28 | 71.56 ± 8.35 | 9/16 | ||||
| Yu et al. 201621 | Retrospective | PVP | 48 | 74.6 (63–85) | 10/38 | A, B, C, D, E, F, G | ⋆⋆⋆⋆⋆⋆ |
| PKP | 20 | 75.9 (65–87) | 4/16 | ||||
| Zhang et al. 201522 | Retrospective | PVP | 38 | 75.58 ± 4.97 | 10/28 | B, D, E, F, G | ⋆⋆⋆⋆⋆⋆ |
| PKP | 35 | 73.74 ± 4.35 | 9/26 | ||||
Figure 8
Forest plot to assess the operation time between the two procedures.
All the eight studies included in this meta-analysis had gone through a strict quality assessment. All of them were CCTs and the possibility of a bias was low. But the funnel figure showed that there was a small bias, which may be associated with the incomplete collection of relevant literature, insufficient sample size, and the different level of clinical physicians. Sensitivity analysis showed a good overall result (Figures 9 and 10).
Figure 9
Risk of bias graph.
Figure 10
Risk of bias summary.
Each risk of bias item is presented as a percentage across all included studies and indicates the proportional level for each risk of bias item.
Methodological quality of the included studies. This risk of bias tool incorporates assessment of randomization (sequence generation and allocation concealment), blinding (participants, personnel, and outcome assessors), completeness of outcome data, selection of outcomes reported, and other sources of bias. The items were scored with “yes,” “no,” or “unclear.”
Kümmell’s disease is a special type of osteoporosis vertebra compressed fracture (OVCF),23 which accounts for about 10% of OVCF.24 It is reported few times due to the rare incidence and difficult diagnosis. However, as China enters an aging society with the explosion of the elderly population and advancement of imaging diagnostic technology, particularly the popularity of MRI has increased the reports of Kümmell’s disease.25 The main symptoms of Kümmell’s disease are back pain with or without spinal cord injury.26, 27 The disease progresses progressively, and due to the obvious vertebral body necrosis and collapse, it can lead to severe kyphotic deformity in the late stage, which has a higher incidence of nerve injury than common OVCF.28, 29 Li et al. divided Kümmell’s disease into three stages: in Stage I, vertebral body height reduction was < 20% without adjacent disc degeneration; in Stage II, vertebral body height reduction was > 20%, often with adjacent disc degeneration; in Stage III, the posterior cortex of vertebrae ruptured with spinal cord compression.19 For stages I and II, PVP and PKP procedures are often used for vertebral bone cement treatment.30, 31, 32 But the vertebral body in stage III is incomplete due to posterior cortex collapse, and greater risk of cement leakage in vertebral canal exists during surgery, therefore open surgery is recommended.33, 34 At present, PVP and PKP have become one of the most effective methods to treat vertebral tumors and OVCF35, 36 (including Kümmell’s disease).
The purpose of this meta-analysis was to compare the efficacy of PVP and PKP in the treatment of Kümmell’s disease in Chinese patients. We selected seven outcomes including Cobb angle, vertebral body height, ODI, VAS score, amounts of cement injected, cement leakage, and operation time and strive to fully compare the efficacy between the two procedures. By comparison, Cobb angle (95% CI [0.54, 4.42],
This systematic review included seven Chinese articles and one English article, and the methodological quality evaluation results were all high. Most studies were case–control studies. The purpose of this study is to compare the clinical efficacy of two cement augmentation procedures in Kümmell’s disease treatment at I and II stages in Chinese patients. Therefore, the patient’s informed consent, the choice of a specific treatment plan, the medical ethical issues, and the other inevitable bias also impact the reliability of the meta-analysis. In the same outcome measurements system, we include into the maximum of eight articles least of five and the heterogeneity between groups will increase. Therefore, the above conclusions still need to further verify depending on the emergence of RCTs with higher quality and larger sample sizes in the future.