Kümmell's disease (KD) is a form of delayed ischemic osteonecrosis following vertebral trauma, usually secondary to osteoporotic vertebral compression fractures (OVCFs) in the elderly (1). The radiological features are characterized by vertebral body collapse, intravertebral vacuum cleft (IVC), and pseudoarthrosis formation (2). Patients with KD tend to have intractable low back pain and severe kyphotic deformity. As the incidence of OVCF increases, the incidence of KD also tends to increase annually and can be as high as 12.1%–42.4% (3). Given that vertebrae with osteonecrosis fail to heal spontaneously, conservative treatment usually cannot yield satisfactory outcomes (4). In addition, there is a high risk of delayed neurological deficit following the vertebral collapse. Therefore, surgical treatments are highly recommended, especially for patients with intact neurological function, minimally invasive procedures are more popular (5, 6).

Percutaneous vertebroplasty (PVP), percutaneous kyphoplasty (PKP), and bone-filling mesh containers (BFC) are three minimally invasive procedures for the treatment of OVCFs in the elderly (7, 8). Likewise, their application in the treatment of KD has been gradually carried out in clinical practice with satisfactory outcomes (9). However, due to the specificity and complexity of KD, the application of minimally invasive techniques focuses not only on pain relief but also on preventing the deterioration of the deformity and the consequent nerve damage. In general, the concept of the minimally invasive procedure is to achieve treatment by delivering bone cement into the vertebral body (10). Specifically, PVP was performed by injecting bone cement into the compressed vertebral body under high pressure (11), whereas PKP was achieved by dilating the vertebral body with a balloon prior to cement injection (12). BFC was designed to control the dispersion of the bone cement within the vertebral body through a mesh container (13). Evidence from OVCFs showed that cement injection under high pressure can lead to associated complications, with bone cement leakage being the most common one (14). The presence of cleft within the vertebral body and even the formation of pseudo-articulations in KD implies a higher incidence of cement leakage. Therefore, how to reduce the incidence of cement leakage has become an unavoidable issue when choosing a treatment plan for KD.

To date, numerous studies have reported the clinical efficacy of the three procedures for the treatment of KD, but no consensus has been reached. To our knowledge, no study has systematically evaluated the advantages and disadvantages of these three minimally invasive procedures. Here, we collected the best available evidence to determine which approach has the highest effectiveness and fewest complications for KD by using a network meta-analysis (NMA), to provide useful evidence for clinical decision making.

This study was conducted in accordance with the Preferred Reporting Items for systematic reviews and meta-analysis (PRISMA) guidelines (15) and the PRISMA NMA extension statement (16).

Compared to OVCFs, KD manifests as a rare and relatively complex spinal disorder, characterized mainly by greater vertebral instability, progressive deterioration, and more likely combined neurological deficit (41). Since the results of conservative treatment are usually limited, surgical treatment has become an option for more patients. For KD with neurological integrity, minimally invasive treatments are the preferred surgical approaches (42). With advances in disease understanding, the standardized diagnosis and therapeutic management of KD needs to be established with the support of high-level evidence. Numerous clinical studies have compared the clinical efficacy of PVP, PKP, and BFC for the treatment of KD (9, 35). Currently, the consensus is that all three minimally invasive procedures can improve patients’ symptoms, such as the relief of pain, improvement of functional status, and quality of life. However, there is still controversy regarding safety. It is inconclusive whether one procedure is better than the other in the treatment of KD. Therefore, we performed a network meta-analysis of the three minimally invasive procedures, ranking the likelihood of being the best procedure for each clinical outcome and presenting an objective and comprehensive view of their pros and cons.

With regard to the clinical outcomes, even after the addition of BFC, our findings were generally consistent with that of previous conventional meta-analysis which directly compared PVP vs PKP for the treatment of KD (9). That was, no statistical difference was found among the three procedures in terms of VAS and ODI scores, but PKP provided better kyphosis correction than PVP. This was attributed to the balloon expansion effect related to PKP (43). In the case of BFC, although homogeneous diffusion of the bone cement was achieved by the mesh container, it did not show the advantage of correcting the deformity over PVP. However, it demonstrated a definite advantage over the other two procedures in terms of preventing cement leakage. PVP performed the worst outcome in terms of bone cement leakage management. A meta-analysis showed that the incidence of cement leakage was as high as 54.7% and 18.4% for PVP and PKP, respectively (44). There is no doubt that the application of BFC offered a new option for the prevention of bone cement leakage.

In addition, other attempts have been carried out to optimize injection protocols to reduce bone cement leakage. Taking the most economical PVP as an example, surgeons have developed a sequential infusion of bone cement was prepared into a late-phase filiform or early-phase mass shape (similar to high-viscosity bone cement) and injected slowly, followed by the infusion of the bone cement as an early- or midphase filiform shape. By doing so the bone cement leakage rate can be reduced from 41.7% to 14.3% (45). This was similar to the outcome of treating KD with a high-viscosity bone cement product, which reduced the cement leakage rate from 45.2% to 13.6% (46). Moreover, studies have been conducted to modify the composition of bone cements to better match the biological and biomechanical characteristics of the human body. The addition of mineralized collagen to bone cement has been shown to have similar clinical efficacy as traditional bone cement and can reduce the incidence of bone cement leakage (47). The results of these new technologies are promising, and we look forward to more clinical studies based on these new technologies to comprehensively assess their clinical applicability. Meanwhile, regarding the issue of cement leakage, there is no doubt that it has the potential to lead to catastrophic consequences. However, we would like to have more discussion about this. We think more attention should be paid to how many cement leaks are true “adverse events,” i.e., what type of cement leak occurred, whether it led to new clinical symptoms, and what percentage of early or late revisions resulted from it. On this basis, we believe it is more meaningful to compare the pros and cons of the different procedures, but unfortunately, we were not able to obtain enough information from the included literature for a comprehensive analysis, and we expect new studies to cover the details of adverse events.

Undeniably, this study has some limitations. It is well known that the lack of large sample size, multicenter, prospective randomized controlled trials is a common gap in current clinical studies. When RCTs are not sufficient to answer the question of interest, nonrandomized studies can be included for meta-analysis (48). Sixteen of the enrolled studies we included were CCTs, which may have a low-quality grade due to lack of randomization and blinding, as well as the possibility of greater potential bias. When we evaluated the methodological quality of these studies using the ROBINS-I tool, their measurement of outcomes and selection of the reported results were both at high risk, implying a possible overestimation of treatment effects. Therefore, this needs to be taken into account when referring to the results of this study. There is no doubt that more prospective multicenter RCTs with long-term follow-up are needed to evaluate the clinical efficacy of the three procedures for KD. In addition, the study locations included were all in China, or rather, this study tended to reveal the effectiveness of these three minimally invasive procedures for the treatment of KD in the Chinese population. The good point is that we performed NMA, an advanced form of meta-analysis, to integrate and compare both direct and indirect evidence from clinical studies to inform clinical decisions by means of a ranked manner (49). To the best of our knowledge, this is the first NMA evaluating minimally invasive approaches for the treatment of KD.

In conclusion, this NMA performed a hierarchical ranking of the effectiveness and safety of three minimally invasive procedures for patients with KD. The ranking analysis showed that BFC had the highest likelihood of being the best procedure for the treatment of KD based on a combined assessment of effectiveness in improving patient symptoms and safety in the occurrence of adverse events. Our findings present new evidence for the selection of minimally invasive treatments for KD, providing surgeons with informative support in clinical practice, decision making, and guideline designation.