RESUMO
OBJECTIVE: To probe the relationship among cement volume/fraction, imaging features of cement distribution, and pain relief and then to evaluate the optimal volume during percutaneous vertebroplasty. METHODS: From January 2014 to January 2017, a total of 130 patients eligible for inclusion criteria were enrolled in this prospective cohort study. According to the different degrees of pain relief, cement leakage, and cement distribution, all patients were allocated to 2 groups. Clinical and radiologic characteristics were assessed to identify independent factors influencing pain relief, cement leakage, and cement distribution, including age, sex, fracture age, bone mineral density, operation time, fracture level, fracture type, modified semiquantitative severity grade, intravertebral cleft, cortical disruption in the vertebral wall, endplate disruption, type of nutrient foramen, fractured vertebral body volume, intravertebral cement volume, and volume fraction. A receiver operating characteristic curve was used to analyze the diagnostic value of the cement volume/fraction and then to obtain the optional cut-off value. RESULTS: The preoperative visual analog scale scores in the responders versus nonresponders patient groups were 7.37 ± 0.61 versus 7.87 ± 0.92 and the postoperative VAS scores in the responders versus nonresponders were 2.04 ± 0.61 versus 4.33 ± 0.49 at 1 week. There were no independent factors influencing pain relief. There were 95 (73.08%) patients who experienced cement leakage, and cortical disruption in the vertebral wall and cement fraction percentage were identified as independent risk factors by binary logistic regression analysis (adjusted odds ratio [OR] 2.935, 95% confidence interval [95% CI] 1.214-7.092, P = 0.017); (adjusted OR 1.134, 95% CI 1.026-1.254, P = 0.014). The area under the receiver-operating characteristic curve of volume fraction (VF%) was 0.658 (95% CI 0.549-0.768, P = 0.006 < 0.05). The cut-off value of VF% for cement leakage was 21.545%, with a sensitivity of 69.50% and a specificity of 60.00%. The incidence of favorable cement distribution was 74.62% (97/130), and VF% were identified as independent protective factors (adjusted OR 1.185, 95% CI 1.067-1.317, P = 0.002) The area under the receiver-operating characteristic curve of VF% was 0.686 (95% CI 0.571-0.802, P = 0.001 < 0.05). The cut-off value of VF% to reach a favorable cement distribution was 19.78%, with a sensitivity of 86.60% and a specificity of 51.50%. CONCLUSIONS: In osteoporotic vertebral compression fracture with mild/moderate fracture severity at the single thoracolumbar level, the intravertebral cement volume of 4-6 mL could relieve pain rapidly. The optimal VF% was 19.78%, which could achieve satisfactory cement distribution. With the increase of VF%, the incidence of cement leakage would also increase.