$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
The protocol was approved by the institutional ethics committee. The trial was registered with the Chinese Clinical Trial Registry (ChiCTR2200056512). All patients provided written informed consent. During the study, patients’ personal information was de-identified and encrypted, and was accessible only to authorized members of the research team.
Study design
This was a single-center, prospective, randomized, parallel-controlled clinical study comparing the incidence and severity of cement leakage between high-viscosity and low-viscosity medical polymethyl methacrylate (PMMA) bone cement in percutaneous vertebroplasty (PVP) for elderly patients with single-level acute osteoporotic vertebral compression fractures (OVCF), and evaluating analgesic and functional improvement, radiographic changes, and perioperative parameters in the two groups. The study was conducted in accordance with the CONSORT guidelines, patients were consecutively enrolled, and baseline data were collected before randomization. The study period was from June 1, 2022, to May 31, 2024; all patients were followed according to the prespecified protocol for 12 months postoperatively. Postoperatively, all patients uniformly received an anti-osteoporotic treatment regimen (elemental calcium 600 mg/day, vitamin D3 800 IU/day, alendronate sodium 70 mg/week, continued for 12 months)10, to reduce confounding arising from treatment differences.
Study participants
Patients were recruited from the spinal surgery outpatient and inpatient departments of our hospital. Inclusion criteria: (1) age ≥65 years; (2) single-level acute OVCF, time from onset to presentation ≤6 weeks, with bone marrow edema visible on MRI STIR sequence of the affected vertebra; (3) bone mineral density T-score ≤−2.5 (DXA within 6 months of surgery); (4) involvement of one of the levels from T5 to L5; (5) VAS ≥5 at rest and unresponsive to standardized analgesic therapy; (6) ASA anesthetic evaluation I–III; (7) able to understand the study and sign written informed consent. Exclusion criteria: (1) traumatic burst fracture or interruption of posterior wall cortical continuity with >25% compromise of the sagittal diameter of the spinal canal (assessed on preoperative CT); (2) pathologic fracture due to tumor, infection, or vascular malformation, or secondary osteoporosis; (3) prior spine surgery or vertebral augmentation at the target level; (4) progressive neurological deficit requiring decompression surgery; (5) coagulopathy (platelets <100×10^9/L or INR >1.5) or inability to discontinue anticoagulation therapy; (6) active infection; (7) allergy to PMMA or local anesthetics; (8) severe cardiopulmonary insufficiency (ASA IV); (9) cognitive impairment or poor adherence precluding completion of follow-up.
Randomization and sample size
A computer-generated random number table was used to allocate patients in a 1:1 ratio to the high-viscosity and low-viscosity groups, with stratification by fracture level (thoracolumbar junction T11–L2 vs non-thoracolumbar junction). The randomization sequence was generated by an external statistician, and allocation concealment was implemented using sequentially numbered, opaque, sealed envelopes. Surgeons could not be blinded due to material characteristics, while outcome assessors and statistical analysts were blinded. The primary endpoint was the overall incidence of cement leakage within 48 h postoperatively. Based on preliminary estimates from our center’s prior case database, a leakage rate of 40% in the low-viscosity group and 20% in the high-viscosity group was set as a clinically meaningful difference; with two-sided α=0.05 and power (1−β)=0.80, the normal approximation method for comparing two independent proportions was used to calculate a required sample size of 82 per group. Allowing for up to 12% loss to follow-up or missing data, we planned to enroll 92 patients per group, for a total sample size of 184.
Surgical methods
Preoperative preparation and anesthesia
Routine preoperative evaluations included a complete blood count, coagulation profile, electrocardiogram, and chest imaging. Prophylactic antibiotics were administered: intravenous cefazolin 1.0 g 30 minutes before surgery, or clindamycin 600 mg for patients allergic to penicillin. Aspirin was discontinued 5 days before surgery with assessment of platelet function. The procedure was performed under C-arm fluoroscopy with the patient in the prone position, with the abdomen suspended, and chest and abdomen padded to reduce intra-abdominal pressure. Anesthesia was local infiltration combined with intravenous sedation and analgesia: 1% lidocaine was infiltrated layer by layer along the puncture tract; propofol was given as a loading dose of 0.5–1.0 mg/kg followed by maintenance via micro-infusion; fentanyl was administered as an intravenous bolus at 0.5–1.0 µg/kg. Blood pressure, heart rate, respiration, and oxygen saturation were monitored throughout.
Percutaneous vertebroplasty operative steps
A unilateral pedicle approach was adopted in principle, with conversion to a bilateral approach when the pedicle was small or correction was required. Under anteroposterior and lateral fluoroscopic guidance, an 11-G working cannula was advanced through the pedicle into the vertebral body via the pedicle safe triangle; on the lateral view, the needle tip was positioned at the junction of the anterior one-third and middle one-third of the vertebral body, and on the anteroposterior view, it was located near the vertebral midline. After reaching the predetermined position, bone marrow was aspirated to confirm the absence of backflow, and then bone cement was injected. Injection was carried out under continuous lateral fluoroscopic monitoring, with an injection rate controlled at 0.1–0.3 mL/s. Injection was stopped immediately upon any termination criterion: the leading edge of the cement reached the posterior one-quarter of the vertebral body, new extravertebral spread appeared on imaging, or the planned upper limit of volume was reached. The planned upper limit of volume was 3.0 mL for the thoracic spine and 5.0 mL for the lumbar spine11. At the end of the procedure, the puncture tract was compressed for hemostasis, and a pressure dressing was applied. Operative time was calculated from the start of skin infiltration anesthesia to the completion of dressing fixation.
Bone cement preparation and key points of viscosity control
Both groups used medical PMMA bone cement registered with the national drug regulatory authority and prepared with a vacuum mixer, with the ambient temperature maintained at 21–23 °C. The high-viscosity group used manufacturer-labeled high-viscosity bone cement and, per the instructions, entered the recommended injection window 6–9 minutes after mixing; the low-viscosity group used manufacturer-labeled low-viscosity bone cement and entered the injection window 2–4 minutes after mixing. Both groups used a screw-driven injection device to achieve stable, low-speed, continuous injection. There was no crossover between groups in materials or injection windows. If intraoperative imaging monitoring suggested signs of leakage, the injection was paused immediately; after the viscosity further increased, a small supplementary injection was given at a very low speed, or the injection was terminated.
Outcome measures
Cement leakage–related measures
The primary endpoint was the overall incidence of cement leakage within 48 h postoperatively. All patients underwent thin-slice CT re-examination within 48 h postoperatively, which was independently interpreted by two radiologists in a blinded manner. Before review, CT datasets were de-identified, and the readers had no access to randomization assignment, operative records, or information on cement type, preparation, or injection window; image assessment was performed on anonymized scans using predefined leakage criteria only. Leakage types were recorded according to the Yeom classification: Type B (leakage via intravertebral veins into the posterior vertebral venous plexus), Type S (leakage via segmental veins outside the vertebral body), Type C (leakage through a cortical defect into paravertebral soft tissues or the intervertebral disc)12. Severe leakage was defined as imaging-confirmed entry into the spinal canal or into the segmental vein trunk, or the occurrence of clinical events such as new neurological deficits, dyspnea, or hypoxemia. When respiratory symptoms occurred, CT pulmonary angiography (CTPA) was performed immediately to rule out bone cement pulmonary embolism. In addition to the overall incidence, the incidence of each type and the incidence of severe leakage were recorded.
Pain and functional outcome measures
Pain outcomes, as key secondary efficacy endpoints, were assessed using the 0–10 visual analogue scale (VAS)13 preoperatively and at 24 h, 1 month, 3 months, 6 months, and 12 months postoperatively. Functional outcomes, as secondary efficacy endpoints, were assessed using the Chinese version of the Oswestry Disability Index (ODI, 0–100%)14 preoperatively and at 3, 6, and 12 months postoperatively. Differences from baseline at each time point and between-group differences were calculated. Analgesic medication use was recorded and standardized to morphine equivalent dose.
Radiographic and perioperative parameters
Radiographic outcomes, as auxiliary endpoints, included the anterior height ratio (AHR) of the affected vertebral body and the local Cobb angle. AHR was measured as the anterior height of the affected vertebral body divided by the average of the posterior heights of the adjacent superior and inferior vertebral bodies, multiplied by 100%; the Cobb angle was defined as the angle between the upper endplate of the superior vertebra and the lower endplate of the inferior vertebra. Measurements were obtained on standardized lateral X-rays preoperatively, at 48 h postoperatively, and at 3 months, and were independently interpreted by two physicians; agreement was calculated. Perioperative parameters included operative time, number of fluoroscopy exposures, fluoroscopy dose-area product, injected volume of bone cement, and bed rest time within 24 h postoperatively.
Follow-up methods and time points
During hospitalization, vital signs, wound condition, and neurological signs were observed, and CT and the first outcome assessment were completed at 48 h postoperatively. After discharge, outpatient or telephone follow-up was conducted at 1, 3, 6, and 12 months; questionnaires were uniformly distributed and collected by study nurses, who also verified medication use records. If neurological symptoms occurred at any follow-up point or persistent pain markedly worsened, additional imaging re-examination was arranged and recorded as an adverse event. Follow-up data were entered into an electronic data capture system and double-checked.
Statistical methods
The primary endpoint was analyzed according to the intention-to-treat principle. Categorical variables were expressed as counts and percentages, and continuous variables as mean ± standard deviation or median (interquartile range). Baseline comparability between groups was assessed using the t-test or Mann–Whitney U test, and the chi-square test. The primary endpoint was compared using the chi-square test, with relative risk and 95% confidence interval calculated; a multivariable logistic regression model was constructed to adjust for baseline covariates such as age, sex, fracture level, and T-score. VAS and ODI were analyzed as repeated measures using linear mixed-effects models, with individuals as random intercepts and fixed effects including group, time, and their interaction, and between-group marginal mean differences with 95% confidence intervals were reported for each time point; if the residual distribution deviated from normality, robust standard errors were applied. Single-time-point comparisons of radiographic and operative parameters were performed using the t-test or the Mann–Whitney U test; changes from preoperative to follow-up were analyzed using mixed models. Inter-reader agreement for leakage type was assessed with the weighted κ coefficient, and agreement for continuous radiographic measurements was assessed with a two-way random, absolute-agreement ICC. When the proportion of missing data exceeded 5%, multiple imputation (m=20) was used for sensitivity analyses of continuous outcomes; if the 48 h CT for the primary endpoint was missing, two sensitivity analyses were performed: in the worst-case scenario, missing values were treated as 'leakage present', and in the best-case scenario, missing values were treated as 'no leakage'. All tests were two-sided, with P < 0.05 considered statistically significant. Multiplicity was controlled using the Holm method, prioritizing the two key secondary endpoints of VAS at 24 h and 3 months postoperatively; other secondary outcomes were treated as exploratory, with effect sizes and confidence intervals reported.