骨质疏松性椎体压缩骨折经皮椎体成形术中骨水泥椎间盘渗漏对手术疗效的影响

Effects of intradiscal bone cement leakage on the clinical outcome of percutaneous vertebroplasty for treatment of osteoporotic vertebral compression fracture

目的:观察骨质疏松性椎体压缩骨折(osteoporotic vertebral compression fracture,OVCF)经皮椎体成形术(percutaneous vertebroplasty,PVP)中骨水泥椎间盘渗漏对手术疗效的影响。方法:回顾性分析2016年1月至2019年1月采用PVP治疗的162例OVCF患者的病例资料,其中30例出现骨水泥椎间盘渗漏(渗漏组)、132例未出现骨水泥渗漏(无渗漏组)。比较2组患者的伤椎疼痛视觉模拟量表(visual analogue scale,VAS)评分、Oswestry功能障碍指数(Oswestry disability index,ODI)、Cobb角、椎间盘高度及邻近椎体骨折发生率。结果:渗漏组随访时间19~32个月,中位数26个月。无渗漏组随访时间17~33个月,中位数25个月。①伤椎疼痛VAS评分。时间因素与分组因素存在交互效应(F=3.571,P=0.039)。2组患者伤椎疼痛VAS评分总体比较,组间差异无统计学意义,即不存在分组效应(F=2.875,P=0.092)。手术前后不同时间点伤椎疼痛VAS评分的差异有统计学意义,即存在时间效应(F=792.352,P=0.000)。2组患者的伤椎疼痛VAS评分随时间变化均呈下降趋势,但2组的下降趋势不完全一致。术前、术后1d,2组患者伤椎疼痛VAS评分的组间差异均无统计学意义[(7.21±1.32)分,(7.15±1.41)分,t=-0.227,P=0.821;(2.81±0.59)分,(2.89±0.83)分,t=0.508,P=0.612];末次随访时,渗漏组伤椎疼痛VAS评分低于无渗漏组[(1.58±0.64)分,(2.23±0.84)分,t=4.734,P=0.000]。②ODI。时间因素与分组因素不存在交互效应(F=1.283,P=0.279)。2组患者ODI总体比较,组间差异无统计学意义,即不存在分组效应(F=0.007,P=0.935)。手术前后不同时间点ODI的差异有统计学意义,即存在时间效应(F=662.717,P=0.000)。2组患者的ODI随时间变化均呈下降趋势,且2组的下降趋势完全一致[渗漏组:(41.29±5.74)%,(22.75±5.97)%,(11.46±5.24)%,F=192.697,P=0.000;无渗漏组:(42.51±5.19)%,(21.27±6.37)%,(11.56±5.71)%,F=941.928,P=0.000]。③伤椎Cobb角。时间因素与分组因素不存在交互效应(F=0.157,P=0.773)。2组患者伤椎Cobb角总体比较,组间差异无统计学意义,即不存在分组效应(F=0.890,P=0.347)。手术前后不同时间点伤椎Cobb角的差异有统计学意义,即存在时间效应(F=633.877,P=0.000)。2组患者的伤椎Cobb角随时间变化均呈减小趋势,且2组的减小趋势完全一致(24.07°±4.15°,13.64°±1.48°,13.54°±1.49°,F=184.863,P=0.000;24.70°±4.36°,13.91°±1.94°,13.86°±1.86°,F=886.434,P=0.000)。④伤椎椎间盘高度。时间因素与分组因素存在交互效应(F=94.251,P=0.000)。2组患者伤椎椎间盘高度总体比较,组间差异无统计学意义,即不存在分组效应(F=1.189,P=0.277)。手术前后不同时间点伤椎椎间盘高度的差异有统计学意义,即存在时间效应(F=786.319,P=0.000)。2组患者的伤椎椎间盘高度随时间变化均呈下降趋势,但2组的下降趋势不完全一致。术前、术后1d,2组患者伤椎椎间盘高度的组间差异均无统计学意义[(7.01±1.49)mm,(6.92±1.30)mm,t=-0.320,P=0.750;(6.91±1.44)mm,(6.96±1.29)mm,t=0.172,P=0.864];末次随访时,渗漏组伤椎椎间盘高度低于无渗漏组[(5.17±1.46)mm,(6.06±1.25)mm,t=3.450,P=0.001]。⑤邻近椎体骨折发生率。末次随访时,渗漏组8例出现邻近椎体骨折,其中胸段2例、胸腰段4例、腰段2例;无渗漏组25例出现邻近椎体骨折,其中胸段6例、胸腰段14例、腰段5例。2组患者邻近椎体骨折发生率比较,差异无统计学意义(χ^(2)=0.900,P=0.343)。结论:OVCF患者PVP治疗后出现骨水泥椎间盘渗漏,不影响伤椎的Cobb角及活动功能恢复、不会增加邻近椎体骨折的风险,反而可以在短期内缓解疼痛症状,但远期可能加速椎间盘退变。

Objective:To observe the effects of intradiscal bone cement leakage on the clinical outcome of percutaneous vertebroplasty (PVP) in treatment of osteoporotic vertebral compression fracture(OVCF). Methods: The medical records of 162 patients who underwent PVP for OVCF from January 2016 to January 2019 were analyzed retrospectively. The intradiscal bone cement leakage was found in 30 patients(leakage group) and unfound in 132 patients(non-leakage group). The injured vertebrae pain visual analogue scale(VAS) score,Oswestry disability index(ODI),Cobb angle,intervertebral disc(IVD) height and incidence rate of adjacent vertebral fracture(AVF) were compared between the 2 groups. Results: The patients in leakage group were followed up for 19-32 months with a median of 26 months,and the ones in non-leakage group for 17-33 months with a median of 25 months.(1)There was interaction between time factor and group factor in injured vertebrae pain VAS scores(F = 3. 571,P = 0. 039). There was no statistical difference in injured vertebrae pain VAS scores between the 2 groups in general,in other words,there was no group effect(F = 2. 875,P = 0. 092). There was statistical difference in injured vertebrae pain VAS scores between different timepoints before and after the PVP,in other words,there was time effect(F = 792. 352,P =0. 000). The injured vertebrae pain VAS scores presented a time-dependent decreasing trend in the 2 groups,while the 2 groups were inconsistent with each other in the variation tendency. There was no statistical difference in injured vertebrae pain VAS scores between the 2 groups before the PVP and at 1 day after the PVP(7. 21 ± 1. 32 vs 7. 15 ± 1. 41 points,t =-0. 227,P = 0. 821;2. 81 ± 0. 59 vs 2. 89 ±0. 83 points,t = 0. 508,P = 0. 612). The injured vertebrae pain VAS scores were lower in leakage group compared to non-leakage group at the last follow-up(1. 58 ± 0. 64 vs 2. 23 ± 0. 84 points,t = 4. 734,P = 0. 000).(2)There was no interaction between time factor and group factor in injured vertebrae ODI(F = 1. 283,P = 0. 279). There was no statistical difference in injured vertebrae ODI between the 2 groups in general,in other words,there was no group effect(F = 0. 007,P = 0. 935). There was statistical difference in injured vertebrae ODI between different timepoints before and after the PVP,in other words,there was time effect(F = 662. 717,P = 0. 000). The injured vertebrae ODI presented a time-dependent decreasing trend in the 2 groups,and the 2 groups were completely consistent with each other in the variation tendency(leakage group: 41. 29 ± 5. 74,22. 75 ± 5. 97,11. 46 ± 5. 24 %,F = 192. 697,P = 0. 000;non-leakage group: 42. 51 ± 5. 19,21. 27 ± 6. 37,11. 56 ± 5. 71 %,F = 941. 928,P = 0. 000).(3)There was no interaction between time factor and group factor in injured vertebrae Cobb angle(F = 0. 157,P = 0. 773). There was no statistical difference in injured vertebrae Cobb angle between the 2 groups in general,in other words,there was no group effect(F = 0. 890,P = 0. 347). There was statistical difference in injured vertebrae Cobb angle between different timepoints before and after the PVP,in other words,there was time effect(F = 633. 877,P = 0. 000). The injured vertebrae Cobb angle presented a time-dependent decreasing trend in the 2 groups,and the 2 groups were completely consistent with each other in the variation tendency(24. 07 ± 4. 15,13. 64 ± 1. 48,13. 54 ± 1. 49 degrees,F = 184. 863,P = 0. 000;24. 70 ± 4. 36,13. 91 ± 1. 94,13. 86 ±1. 86 degrees,F = 886. 434,P = 0. 000).(4)There was interaction between time factor and group factor in injured vertebrae IVD height(F =94. 251,P = 0. 000). There was no statistical difference in injured vertebrae IVD height between the 2 groups in general,in other words,there was no group effect(F = 1. 189,P = 0. 277). There was statistical difference in injured vertebrae IVD height between different timepoints before and after the PVP,in other words,there was time effect(F = 786. 319,P = 0. 000). The injured vertebrae IVD height presented a time-dependent decreasing trend in the 2 groups,while the 2 groups were inconsistent with each other in the variation tendency. There was no statistical difference in injured vertebrae IVD height between the 2 groups before the PVP and at 1 day after the PVP(7. 01 ± 1. 49 vs6. 92 ± 1. 30 mm,t =-0. 320,P = 0. 750;6. 91 ± 1. 44 vs 6. 96 ± 1. 29 mm,t = 0. 172,P = 0. 864). The injured vertebrae IVD heights were lower in leakage group compared to non-leakage group at the last follow-up(5. 17 ± 1. 46 vs 6. 06 ± 1. 25 mm,t = 3. 450,P = 0. 001).(5)At the last follow-up,the AVF was found in thoracic segments in 2 cases,thoracolumbar segments in 4 cases and lumbar segments in 2 cases in leakage group;while the AVF was found in thoracic segments in 6 cases,thoracolumbar segments in 14 cases and lumbar segments in 5 cases in non-leakage group. There was no statistical difference in incidence rate of AVF between the 2 groups(χ^(2)= 0. 900,P = 0. 343).Conclusion: The intradiscal bone cement leakage doesn’t affect the injured vertebra Cobb angle and recovery of activity function,and it cann’t increase the risk of AVF in OVCF patients who underwent PVP. Instead,it can relieve pain symptoms in the short term,whereas it may accelerate IVD degeneration in the long term.