A second puncture and injection technique for treating osteoporotic vertebral compression fractures

Percutaneous vertebroplasty (PVP) has been widely used for treating osteoporotic vertebral compression fractures (OVCFs). PVP can achieve pain relief, reduce bedrest duration, and improve the quality of life in elderly patients. However, with the development of the PVP technique, some related complications have followed, such as recompression of the cemented vertebral body, which can cause a series of problems, such as back pain, limited mobility, kyphosis, neurological compression, and even revision surgery [1‐3]. Inadequate cement filling in the vertebral body, especially in the area where the cement is not strengthened between the upper and lower endplates, easily leads to the recompression of the cemented vertebral body [4]. Therefore, how to improve the dispersion of bone cement in the vertebral body during PVP, especially in the unreinforced area between the upper and lower endplates, has become our research direction.

We performed a second puncture and injection technique to improve the dispersion of cement (Mendec Spine, Tecres Medical, Verona, Italy) in these areas that were not strengthened between the upper and lower endplates during PVP (Vertebroplasty System, Guanlong Medical, China). We retrospectively analysed the clinical outcome of patients who underwent this PVP procedure for treating OVCFs and compared the results with the traditional PVP technique (single puncture and injection).

A total of 193 patients with PVP were reviewed, of whom 86 underwent second puncture and injection, and 107 underwent single puncture and injection. We defined these 86 patients and 107 patients as groups A and B, respectively. The final follow-up duration for all patients ranged from 12 to 20 months, with an average of 15.64 months. The cemented vertebral body levels were T₁₀ to L₄ in group A and were T₆ to L₄ in group B (Fig. 2). There were no significant differences in sex, age, weight, height, BMD, follow-up time, hospital days, or intraoperative blood loss between the two groups (Tables 1 and 2). The injected cement volume in group A was 2.5–5.9 ml, with an average of 4.10 ml, and that in group B was 1.8–4.3 ml, with an average of 3.24 ml. The operation time in group A (23–55 min; average, 35.49 min) was longer than that in group B (18–61 min; average, 31.30 min) (Table 2).

AVH and MVH after surgery were significantly restored compared with the pre-operative values in the two groups. Regarding the pre-operative AVH, MVH, RAVH, and RMVH, there were no differences between the two groups (P > 0.05). At the final follow-up, the LAVH and LMVH in group A were 0.01 ± 0.03 and 0.02 ± 0.36, respectively, while those in group B were 0.14 ± 0.17 and 0.14 ± 0.13, and the differences were statistically significant (Tables 3 and 4).

Pain was significantly relieved after surgery compared with pre-operative pain in both groups. There were no significant differences in the VAS score or ODI at 3 days post-operation between the two groups (P > 0.05). However, the VAS score and ODI in group B were significantly higher than those in group A at the final follow-up; the VAS score and ODI in group B were 1.65 ± 0.70 and 14.50 ± 4.16, respectively, and those in group A were 1.00 ± 0.74 and 12.81 ± 4.02, respectively (Table 5). Three patients in group A and two in group B experienced adjacent vertebral fractures. Regarding mild, moderate, and severe cement leakage, there were 25 (29%), 5 (5%), and 0 cases, respectively, in group A, and 28 (26%), 3 (2.8%), and 1 (0.009%) case, respectively, in group B (P > 0.05).

PVP, as a minimally invasive technique, is an important choice for treating OVCFs. With the development of this technique, however, a series of complications have followed, especially the recompression of cemented vertebral bodies, and it is an important cause of long-term low back pain and kyphosis after PVP [1]. Heo et al. [7] reported that the incidence of re-collapse in 343 patients with OVCFs was 3.21% after PVP. Chen et al. [8] also reported that this incidence was 9.7%, accompanied by post-operative refractory low back pain, limited spinal activity, and other symptoms. What is the cause of vertebral body recompression? Lin et al. [9] retrospectively analysed 137 patients with single-segment PVP and concluded that if the injured vertebral body is not completely augmented with cement, the hard bone cement will destroy the trabecular bone in the unfilled area when an external force acts on the vertebral body, causing these areas to collapse again. Kim et al. [4] revealed a similar result that there is a region that is not cement-augmented between the upper and lower endplates because the bone cement is not sufficiently dispersed, which is more likely to cause recompression of the cemented vertebral body. Liang et al. [10] found through three-dimensional finite element analysis that uneven distribution of bone cement increases the maximum von Mises stress of cancellous bone around bone cement, suggesting that uneven distribution of the bone cement in the vertebral body causes the destruction of the cancellous bone in the unfilled area, leading to recompression of the cemented vertebral body. Zhang et al. [11] also reviewed 177 patients with PVP and found that bone cement distributed around both the upper and lower endplates resulted in a significantly lower incidence of recompression.

These results indicate that the full distribution of bone cement in the vertebral body, especially between the upper and lower endplates, is the key factor in preventing recompression of the cemented body. In our previous study, we applied a technique of accurate puncture and ultra-early injection of low-viscosity cement to improve cement diffusion in the vertebral body [5]. However, some patients still experienced recompression of the cemented vertebral body. We observed that the cement in these re-collapsed vertebral bodies was not sufficiently distributed in the vertebral body, especially below the upper endplate or upon the lower endplate. How to fill these areas with cement during a PVP procedure has become our research focus.

In this study, if an unfilled area, especially upon the lower endplate or below the upper endplate, was found by C-arm fluoroscopy during PVP, we used the second puncture and injection technique. This technique allowed the cement to diffuse into these areas, and we confirmed that the AVH and MVH were not significantly different from those in the control group, and VAS score and ODI were lower than those in the control group at the last follow-up. Most importantly, the incidence of leakage and adjacent vertebral fractures was not increased.

We observed that the post-operative AVH and MVH had different degrees of restoration. We considered that this restoration might be related to the patient’s hyperextended position caused by the patient’s back being elevated with soft pillows before operation, the chest and pelvis being elevated with soft pillows in the prone position during operation, or the pressure from the forward advancement of puncture needles during surgery, rather than related to the PVP procedure itself [12]. This type of restoration will inevitably cause the vertebral body to stretch. If the cement is insufficiently filled in these areas, it is more likely to cause the vertebral body to collapse again [9, 13]. In our study, once these unfilled areas were found during operation, puncture and cement injection were performed again for the unfilled area so that these areas were fully filled with cement, thereby reducing the incidence of recompression. The quality of life of patients was improved, and this was confirmed by the results of the VAS score and ODI. We observed that the VAS score and ODI at the final follow-up in group A were significantly lower than those in group B, indicating that the pain and dysfunction of the patients in group B were more obvious than were those in group A. This result reaffirmed the importance of adequate cement filling to prevent vertebral recompression and improve the quality of life of patients.

This technique of second puncture and injection did not increase cement leakage. We consider this result to be related to the gradual withdrawal of the needles during surgery so that there is space for the cement filling in the front. In addition, the cement is relatively viscous when it is injected again, which may be one of the reasons why it does not easily leak. This type of injection method may cause an increase in the cement volume. Some researchers have reported that an increase in the volume of cement will lead to the occurrence of adjacent vertebral fractures [14]. However, our results are inconsistent with the previously reported result. Lee et al. [15] followed up 351 patients with OVCFs who underwent PVP and found that the average BMD of patients with adjacent vertebral fractures was − 3.1 ± 1.5, while the average BMD of patients without adjacent vertebral fractures was − 2.7 ± 1.5. Therefore, he considered that adjacent vertebral fractures might be mainly due to the natural progression of osteoporosis. Ning et al. [16] also reached similar conclusions by reviewing 365 cases. Consequently, we considered that adjacent vertebral fractures are mainly related to the development of osteoporosis. In addition, this increase in cement volume may play a role in preventing re-collapse. Chen et al. [17] revealed that less bone cement perfusion in the injured vertebral body was an important factor in the loss of vertebral height after surgery. Clark et al. [18] also suggested that adequate perfusion of the bone cement was required for PVP in patients with OVCFs.

This technique of second puncture and injection cannot be performed in all vertebral bodies. In the upper and middle thoracic vertebrae, this technique may present a high risk of neurological or spinal cord injury because the pedicles are relatively narrow. The nearest vertebral body we treated with this technique was T₁₀. In addition, it is necessary to carefully measure the size of the pedicles and the angle of the puncture on X-ray and CT images before PVP. This technique is not recommended if the pedicle is too small or the angle of adjustment is too narrow.

We acknowledge that our study has several limitations. First, this is a single-centre study. Second, the sample size is relatively small. Third, there may be some confounding factors or bias because this is a retrospective study. In the next stage, multi-centre, large-sample, and prospective study may be needed for further validation.

In summary, the second puncture and injection technique can be applied during PVP (below T₁₀), especially if the cement is not sufficiently dispersed below the upper endplate or above the lower endplate. This method may effectively increase the dispersion of cement in these areas, thus preventing recompression of the cemented vertebral body; additionally, this technique does not increase the risk of cement leakage or adjacent vertebral fracture.