Polymethylmethacrylate-augmented Cannulated Pedicle Screw Fixation for High-energy Vertebral Compression Fracture With Osteoporosis

Yao-yao Liu Third Military Medical University Daping Hospital and Research Institute of Surgery Jun Xiao PLA Rocket Force University of Engineering Xiang Yin Third Military Medical University Daping Hospital and Research Institute of Surgery Ming-Yong Liu Third Military Medical University Daping Hospital and Research Institute of Surgery Zhong Wang Third Military Medical University Daping Hospital and Research Institute of Surgery Lei He Third Military Medical University Southwest Hospital Peng Liu Third Military Medical University Daping Hospital and Research Institute of Surgery Fei Dai (  david-feifei@163.com ) Army Medical University


Introduction
The incidence of osteoporotic fractures is increasing signi cantly with a rapidly ageing population [1], especially among postmenopausal women, the data is up to 50% [2]. Vertebral compression fractures (VCFs) are the most common condition [3,4] and a major socio-economic burden for elderly populations, which widely performed in the treatment of percutaneous vertebroplasty (PVP) and kyphoplasty (PKP).
Up to now, these two minimally invasive surgical methods have been proved to have an excellent clinical effect [5,6].
However, vertebral compression fractures due to high-energy trauma (HVCFs), which de ned as VCF combined with severe clinical manifestations such as spinal nerve compression, spinal stenosis, instability, delayed kyphosis and neurological impairment (Kümmell's disease),need to restore spinal stability by internal xation and decompress spinal canal to relieve the neurological symptoms. The challenge to surgeons is that osteoporosis tends to cause screw loosening and trabecular fracture [7]. Loss of vertebral height, pseudoarthrosis and bone graft non-fusion, followed by inadequate holding power of internal xation, ultimately may lead to chronic lower back pain, kyphosis and dysneuria.
Studies have shown that cannulated pedicle screw augmented by polymethylmethacrylate (PMMA) can signi cantly improve the screw holding power [8][9][10] .However, few studies involve the clinical outcome of cannulated pedicle screw to augment xation in VCF with osteoporosis due to high-energy trauma.
Authors improved the design and created a new type of PMMA-augmented pedicle screw named bone cement-injectable cannulated pedicle screw (CICPS). To evaluate the clinical results and feasibility of this technique in HVCFs with osteoporosis, we retrospectively reviewed the data of eighteen consecutive patients treated with CICPS, and summarized the incidence of complications and the clinical e cacy of PMMA augmentation.

Study design
The study was approved by the Daping Hospital ethics committee (IRB, 2019149); and informed consent was obtained from each participant. This study was a retrospective investigation of osteoporotic patients who underwent surgery at our clinic for HVCFs. Considering that using conventional pedicle screws in cases with a severely osteoporotic vertebral body may lead to screw loosening or pulling out, it would be unethical to include a control group for comparison. Instead, we evaluated the clinical outcomes of this technology by comparing the baseline with post-treatment measurement parameters.

Patients
From October 2012 to May 2017, a total of 30 cases with HVCFs were collected. Inclusion criteria were as follows: (1) Diagnosed as osteoporosis by dual-energy X-ray absorptiometry (T-value < -2.5) [10]; (2) HVCFs with acute or chronic impairment of spinal cord function, spinal stenosis or spinal structural instability; (3) Used CICPS for internal xation; (4) consented the treatment requirement and schedule. Exclusion criteria were as follows: (1) Simple fragility fracture; (2) Old fracture; (3) Not used CICPS for internal xation; (4) Allergy to any implant or having surgical contraindications. Figure 1 shows the ow chart of patient recruitment and study procedures.

Materials
The preliminary clinical application of CICPS had achieved satisfactory results in the previous studies performed by authors [8][9][10]. In references 8 to 10, we had described the design features of CICPS in detail. It is designed as a hollow rod with three side holes at the screw tip so that PMMA can out ow into the centrum through this route. Fixation ability is augmented because of anchoring the screw into cancellous bone with PMMA. In this study, the diameter and length of CICPS were determined according to the measurement results during actual intraoperative conditions. PMMA was a high-viscosity bone cement supplied by Heraeus Holding, Germany.

Surgical techniques and postoperative treatment
The screws were manually placed according to the anatomical landmarks. PMMA powder and water agent were mixed, and then, injected by a special device during a dough-like mass viscosity. No more than 2ml of PMMA mixture was determined by intraoperative monitoring of intravertebral PMMA dispersion. The following measures were taken to ensure the augmentation effect and further to reduce the PMMA leakage risk: 1. CICPS suggested to keep parallel to the upper and lower vertebral endplates; 2. The probe was used to directly detect the required screw length so that the out ow side hole of CICPS was located in the anterior 2/5th of the vertebral body; 3. CICPS was not allowed to penetrate the upper, lower, inner and outer sidewalls of the pedicle and the lateral cortex of the vertebral body; and 4. The entry angle of CICPS was slightly larger than that of a conventional pedicle screw [9]. The quantity of CICPSs was determined by the bone quality, the stability requirement, and manual feeling during the tunnel preparation. In principle, a minimum quantity of CICPSs was recommended due to the di culty in removing and repairing. The remaining screws adopted conventional ones. Decompression, xation and fusion was performed as usual, if necessary.
Antibiotics were routinely used to prevent surgical site infection in the rst 48 hours, and the wound drainage tube was removed after 48-72 hours after the amount of drainage uid was less than 50ml. Three days after the surgery, patients were encouraged to perform rehabilitation exercises by wearing a thoracolumbar brace. Anti-osteoporosis treatment should be supplemented as soon as possible after operation.

Safety and effectiveness evaluation
Operation time, blood loss, and hospitalization time were recorded to evaluate the basic condition of the surgery. A review was carried out at 3 months, 6 months, 1 year, and every half-year after surgery. Moreover, lumbar X-ray lms were obtained to evaluate bony fusion, loss of vertebral height, loss of the Cobb angle, and screw loosening or pull-out. Visual Analog Scale (VAS) and Oswestry Disability Index (ODI) scoring systems were used to evaluate pain and functional recovery in lower limbs, respectively. Complications such as wound infection, cerebrospinal uid leakage, and PMMA leakage were recorded.
The method of measuring the loss of vertebral height: The vertebral body height measured in sagittal Xlm was used to evaluate fracture reduction. Distances were measured between the following points: the most dorsal superior and inferior endplate margins, the most anterior superior and inferior margins, and the height of the centrums from the midpoint of the two lines. The average values were calculated, and vertebral height was determined after taking into consideration the change in vertebral height between the postoperative review and the last follow-up that was regarded as the value of vertebral height loss [10].

The method of measuring the Cobb angle
The Cobb angle was used to assess sagittal deformity on a plain radiograph. The upper and lower vertebral bodies near the fractured part with the largest inclination angle were considered, and the angle between the laminae was measured as the kyphosis Cobb angle. The angle change between the postoperative review and the last follow-up was regarded as the value of Cobb angle loss [10].
Evaluation criteria for spinal fusion on X-ray lms [10,11]: (1) Passage of trabecular bone through the bone graft area; (2) Vertebral movement between exion and extension X-ray lm < 3 mm or change in the intervertebral space angle < 5 degrees; and (3) Bone growth through the intervertebral disc space.
Evaluation criteria for screw loosening on X-ray lms (1) Screw displacement on a lateral X-ray greater than 1 mm [12,13]; (2) Double-circle sign around the screw on an anterior-posterior X-ray lm [14].

Statistical analysis
The experimental data were calculated as mean ± standard deviation (x ± s) and analyzed by using the SPSS 13.0 statistical software package. The paired test was conducted to compare variables, including VAS and ODI scoring and imaging indicators at the pre-operation, post-operation, and at the last followup. The change was considered statistically signi cant if p was less than 0.05.

Results
Twenty cases were enrolled and the follow-up rate at each time point was 90% because 2 cases were lost at three months after operation. Finally, 18 cases were completely recorded for an average follow-up time of 18.5 ± 8.7 months (range, 6-54 months). These cases comprised 6 males and 12 females with an average age of 72.3 ± 6.5 years ranging from 56 to 80 years. The average duration from injury to admission was 20 ± 13.5 days (range, 3 hours to 3 months).
The operation time was 165 ± 23 min (range, 95-462 min), the bleeding amount was 206 ± 75 ml (range, 50-400 ml) and the length of the hospitalization time was 12.5 ± 3.5 d (range, 9-22 d). There was no death, surgical related neurological dysfunction and paraplegia in all cases at each follow-up time point.
One case of super cial surgical site infection was noted, which was under control after an intravenous drip of antibiotics. Cerebrospinal uid leakage occurred in one case, which healed completely after the drainage tube removed following two week of bed rest. With respect to PMMA leakage, there was one case leakage into the paravertebral vein, and another case into the intervertebral disc space during the operation, but no serious complications such as nerve or blood vessel injury and pulmonary embolism were observed (Table 1).  Note: All indexes at the post-operation and last follow-up showed signi cant differences from those at the pre-operation, p < 0.05. a, p < 0.05, a signi cant difference was observed in Cobb angle between the post-operation and the last follow-up. b, p > 0.05, no signi cant difference was observed in the vertebral height, VAS score, and ODI score between the post-operation and the last follow-up.

Pain relief and functional recovery
The VAS score improved from 8.3 ± 2.6 before operation to 2.2 ± 0.8 after operation, and to 1.8 ± 0.3 at last follow-up. The ODI score improved from 75.5 ± 12.2% to 35.2 ± 7.2%, and to 30.5 ± 13.1% at these three time points. These indexes were signi cantly improved after operation (P < 0.05), but no signi cant difference was observed comparing post-operation with last follow-up (P > 0.05) ( Table-2).

Typical cases
Typical case 1 (Fig. 2): A 68-year-old female patient (T-value = -3.0), who had suffered from traumatic fracture of the T12 vertebral body 2 months ago, was diagnosed as HVCF and kyphosis, and she did not receive any bed rest and brace treatment. Imaging examinations showed wedge-shaped deformation of the T12 vertebral body and thoracolumbar kyphosis ( Fig. 2A-D). Thus, CICPS was used to strengthen the internal xation, and the kyphosis was corrected after distraction and reduction (Figs. 2E and F). Pouring PMMA with PVP device was performed to maintain vertebral height. 6 and 12 months after operation, bony fusion was achieved with maintenance of the spinal curvature and vertebral height, and there was no PMMA displacement or screw loosening (Fig. 2G-J).
Typical case 2 (Fig. 3): A 73-year-old female patient (T-value = -3.5) suffered from L1 HVCF, kyphosis, and cauda equina injuries due to a fall one month ago (Fig. 3A-D). She was treated with spinal canal decompression, posterolateral arthrodesis, and CICPS-augmented internal xation of upper and lower vertebral bodies (Fig. 3E and F). The symptoms were then relieved after operation and CICPS loosening was not detected at the 18-month follow up. Although the vertebral height was reduced, the kyphosis angle was maintained and bony fusion was successfully achieved (Fig. 3G and H). PMMA was leaked into the anterior venous plexus at L1 without any symptom.

Discussion
Known as a microinvasive technology, PVP and PKP, is considered as the "gold standard" for the treatment of osteoporotic VCFs [5,6,15]. However, PVP and PKP may no longer be applicable when VCFs accompanied with severe high-energy trauma. Some special reasons are as follows: Internal xation must be performed to restore spinal stability based on the three-column fracture due to high-energy trauma. 3. Delayed neurological symptoms, causing by local kyphosis and wedging of the injured vertebral, require kyphosis correction by spinal xation [18], and even require support to the anterior and middle columns [19]. However, when internal xation is performed, osteoporosis seriously reduces the stability in HVCFs patients. Studies have documented that xation stability in osteoporotic spine reduced signi cantly with a loosening rate up to 12% [20].
Up to now, PMMA-augmented cannulated pedicle screw is considered an optional technique which has broad application prospects [21,22]. Some researchers recommended this type of pedicle screw for lumbar spondylolisthesis with osteoporosis [23,24] In current study, the VAS and ODI score signi cantly improved after operation immediately and at the last follow-up compared with those before operation (P < 0.05), which shows that CICPS is a safe and effective treatment for patients with HVCFs. Additionally, there are no signi cant differences between VAS and ODI scores immediately after operation and at the nal follow-up more than 6 months (P > 0.05). The result shows that satisfactory mid-term clinical outcomes can be achieved in this group. Singh V et al. [29] did a systematic analysis for PMMA-augmented cannulated pedicle screw. Their results summarized from published studies indicated that the average VAS score before operation was 8.4 (range 8-9.2) compared to 2.3 (range 1.42-4.8) at the last follow up. The average improvement ODI for assessment of functional recovery was 42.1. In this study, VAS score before operation was 8.3 compared with 2.2 at last follow-up and ODI improved from 30.5 to 75.5. These results show that CICPS has a great improvement in pain relief and function recovery similar to previous series.
PMMA leakage is a focus point to surgeons because neurological or spinal cord damage may cause by the situation [12,18,30]. Traditionally, PMMA is injected into the preset nail tunnel using a PVP device during laddering period, and then the screw is inserted [25]. The challenges of this procedure lie in uncontrollable diffusion of PMMA during the screw insertion process. Even if PMMA is removed immediately when the spinal canal leakage occurs, these surgical procedures will lead to the time window of PMMA injection being exceeded. Therefore, we chose to further research CICPS instead of traditional injection method. The advantage of this method is that screws are left in situ and monitored by intraoperative X-ray when PMMA injection, which have less possibility of PMMA leakage due to pressurization. Moreover, the design of side holes located at distal third of screw also play an important role to avoid PMMA leakage [8][9][10].
Authors used CICPSs in 18 patients with HVCFs and osteoporosis. 3 screws in 2 patients (3/40, 7.5%) were observed PMMA leakage, but without any neurological symptoms. Wuisman et al. [31] reported four out of 49 screws (8.2%) had PMMA leakage into spinal canal using traditional technique and removed PMMA by laminectomy and dura manipulation. Janssen et al. [32] reported that the incidence of asymptomatic PMMA leakage was 66.7% and that of symptoms was 5.5%, in 1.2% of which needed to remove the PMMA and screws. Martı´n-Ferna´ndez et al. [33] observed a 62.3% incidence of spinal leakage in 313 patients. 1.55% of the cases had symptoms such as radicular pain of lower limbs and transient sensorimotor de cit. In brief, PMMA leakage presented no symptoms or only mild symptoms in most studies and the leakage rate showed a big difference, which reason might be related to the small sample size or different screw design in different studies.
In our series, the loss of kyphosis cobb angle increased signi cantly (P < 0.05) comparing the date of last follow-up with that of post-operation, while the vertebral height was no signi cant change (P > 0.05) and the screw loosening was not observed. The reason for this phenomenon can be explained by physiological progress. Studies have shown that kyphosis progressed an average 7 degree over 15 years in healthy women, especially between age of 50 and 59 [34,35]. In thoracolumbar burst fracture following internal xation without PMMA augmentation, progressive kyphosis was also observed [36]. Therefore, trauma-related disc apoptosis or osteoporosis process [37][38][39] may play important roles in kyphosis progression spite of PMMA augmentation or not.
HVCFs usually occurs in elderly patients over 50 years. These patients may have had many comorbidities and severe osteoporosis, which resulted in complex surgical procedures might not be tolerated. In one study of Huang et al [30], they observed the clinical outcomes of traditional PMMA augmentation technology in 28 cases of Kummell's disease. The operation time, blood loss and hospitalization time were 115.9 ± 10.0 minutes, 214.3 ± 17.5 ml and 7.6 ± 1.3 days respectively. While in this study, these data were 165 ± 23 minutes, 206 ± 75 ml and 12.5 ± 3.5 days. Furthermore, 3 patients had asymptomatic PMMA leakage, one developed a urinary tract infection, one contracted pneumonia, and one experienced delayed wound healing. There was no loosening at the nal follow-up. We cannot draw any better conclusions by compared these data with our ndings. However, the absolute values of the data between CICPS and the traditional PMMA augmentation technology only have small differences, which may be related to the technical pro ciency of surgeons and postoperative management procedures.
This study has several limitations. Firstly, this study is not a randomized controlled study, which contributed to a low level of evidence. Secondly, the sample size is small and the follow-up time is short, thus a study with more patients and longer follow-up duration is required to further con rm our results. Thirdly, since the safety and effectiveness of CICPS have been biomechanically evaluated before [8], here only clinical evaluation was performed. Finally, standard application protocol of CICPS should be developed to avoid the catastrophic consequences of excessive application, which will be our future research focus.

Conclusion
In conclusion, CICPS using PMMA augmentation maintains their functional outcomes after internal xation at latest follow-up in spite of kyphosis progression. The incidence of asymptomatic PMMA leakage, screw loosening and complications is not higher than that of previous reports. CICPS may be a clinically safe and effective technology which is used to augment pedicle screws in HVCFs with osteoporosis. This study could provide evidence-based basis for developing the guidelines of the cannulated pedicle screw application, especially in HVCFs with osteoporosis. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
The study was approved by the Daping Hospital ethics committee (IRB, 2019149).

Consent for publication
Informed consent for publication was obtained from all participants.   A-D, the preoperative X-ray, MRI and CT indicated T12 HVCF with kyphosis. E-F, the postoperative X-ray indicated that the height of T12 was restored and the kyphosis was corrected. G-J, there was no signi cant loss of vertebral height and kyphosis angle at 6 and 12-month follow-up.