目的  本研究课题对传统的关节突融合技术进行了改良，提出了一种较实用的融合方法。本研究拟达到以下几个目的：（1）利用有限元分析方法研究分析腰椎改良关节突融合术、后路椎间融合术（posterior lumbar interbody fusion，PLIF）和横突间融合术（posteriorlateral lumbar fusion，PLF）术后腰椎稳定性的差异。（2）总结改良关节突融合术治疗腰椎退行性疾病的临床疗效和融合率及并发症情况。（3）比较改良关节突融合术和PLIF后路椎间融合术在退行性腰椎滑脱患者中的融合率和临床效果的差异。（4）评估改良关节突融合技术中自体松质骨和同种异体松质骨在融合率方面有无差异。

方法  （1）利用有限元方法建立L4-5单节段退变的L3～S1节段完整三维模型，以及改良关节突融合术、横突间融合术、PLIF椎间融合术的术后成功融合的腰椎模型，并施加正常生理载荷，记录分析各模型在前屈、后伸、左右侧屈、左右旋转等6个工况下运动范围的变化，刚度及应力分布情况。（2）回顾性分析行腰椎改良关节突融合术并具有1年完整随访资料的491名患者。于术后6个月、12个月随访时分别行腰椎CT平扫+矢状重建检查，评估其融合率。于术前，术后1、3、6及12个月分别使用视觉模拟疼痛评分（Visual Analogue Scale Score，VAS）、日本骨科学会评分（Japanese Orthopaedic Association Score，JOA）、功能障碍指数评分（Oswestry Disability Index，ODI）评估患者的症状改善情况。末次随访时，使用MacNab标准评估患者症状改善情况的整体优良率。（3）回顾性分析行改良关节突融合术（modified facet joint fusion，MFF）和PLIF后路椎间融合术并具有1年完整随访资料的Ⅰ度或Ⅱ度退行性腰椎滑脱患者。首要观察指标为融合率，次要观察指标为临床症状情况、手术时间、手术出血、切口长度、住院时间、住院总费用及并发症情况。于术后6个月、12个月随访时分别行腰椎CT平扫+矢状重建检查，评估其融合率。于术前，术后1、3、6及12个月随访时分别使用VAS疼痛评分、JOA评分、ODI评分评估所有患者的症状改善情况。末次随访时，使用MacNab标准评估其症状改善情况的整体优良率。（4）按照入排标准随机纳入行腰椎改良关节突融合术的腰椎退行性疾病患者，以进行融合的关节突关节为研究对象。根据使用植骨材料种类的不同分为自体骨组和异体骨组。术中随机确定融合节段左侧关节突植骨床植入何种植骨材料，右侧则相应使用另一种植骨材料。两组分别以进行融合的关节突数量为样本量。所有患者分别在术后6个月、12个月随访时行腰椎CT平扫+矢状重建检查，评估其融合率，并记录随访期内患者并发症及二次手术发生率。

结果  （1）PLF横突间融合术模型、PLIF椎间融合术模型和改良关节突融合术模型三种融合模型在前屈、后伸、侧屈和旋转的活动范围均明显小于正常完整腰椎模型（P ＜ 0.05），但三种融合模型的活动范围无显著差异（P ＞ 0.05）。三种融合模型的刚度均明显大于完整模型。（2）术后6个月、12个月随访时，融合率分别为56.8%和96.1%。术后12个月随访时，患者临床症状改善情况的整体优良率为93.6%。随访期内没有与改良关节突融合技术相关的并发症发生。（3）本研究共纳入136例患者（MFF组，71例；PLIF组，65例）。术后6个月、12个月随访时，MFF组和PLIF组在融合率方面均无统计学差异（P ＞ 0.05）。两组在VAS评分、JOA评分和ODI评分等方面均无显著差异（P ＞ 0.05）。但MFF组患者中，手术时间和手术出血明显少于PLIF组（P ＜ 0.001），且MFF组患者的住院总费用明显低于MFF组患者（P ＜ 0.0001）。（4）本研究共纳入患者65例，其中有61例患者顺利完成随访，随访率93.8%。研究中进行融合的关节突数量共226例，其中自体骨组和异体骨组分别为113例。术后6个月随访时，自体骨组融合率明显高于异体骨组融合率（P = 0.016）。术后12个月随访时，自体骨组融合率与异体骨组融合率无显著性差异（P ＞ 0.05）。

结论  改良关节突融合技术是一种融合率高、操作简单、安全性好，且性价比高的融合技术，值得在临床上，尤其是经济不发达及医保制度欠完善的国家或地区推广。

Purpose   (1) To explore the difference in range of motion among MFF, posteriorlateral fusion (PLF) and PLIF techniques on degenerative single segmental lumbal spinal by a finite element method. (2) To evaluate the efficacy and safety of modified facet joint fusion technique (MFF), a retrospective clinical studies was were conducted. (3) To compare fusion rate and clinical outcomes between MFF technique and posterior lumbar interbody fusion (PLIF) for degenerative spondylolisthesis. (4) To compare fusion rate between autologous local bone and allograft bone in the same patient with degenerative lumbar spinal disease underwent instrumented MFF technique.

Methods    (1) A finite element model of L3～S1 segments with a single segmental degeneration at L4-5 level was established. Different models of L4-5 segmental fusion after MFF, PLF and PLIF techniuqes were established, respectively. Physical loads were applied to the models and the changes of range of motion at L4-5 level in different models were recorded during flexion, extension, lateral bending and rotation. (2) 491 consecutive patients with 931 segments who underwent MFF with pedicle screw instrumentation for lumbar degenerative diseases were retrospectively reviewed after 1-year follow-up. Computed tomography with fine-cut axial images and sagittal reconstruction views was used to evaluate the fusion rate of MFF at 6-, 12-month follow-up postoperatively. Clinical outcomes included visual analogue scale pain scores for low back pain (VAS-LBP) and leg pain (VAS-LP), Japanese Orthopedic Association scores (JOA), and Oswestry Disability Index (ODI), all of which were obtained preoperatively and postoperative subsequently at 1-, 3-, 6-, 12-month follow-up after surgery. The clinical outcomes were determined to be excellent, good, fair, or poor according to the MacNab classification at the last follow-up time. (3) A single-center retrospective study including patients with grade Ⅰ or Ⅱ degenerative spondylolisthesis managed by MFF or PLIF with pedicle screw instrumentation was conducted. Computed tomography with fine-cut axial images and sagittal reconstruction views was used to evaluate the fusion rate at 6-, 12-month follow-up postoperatively. Clinical outcomes included VAS-LBP, VAS-LP, JOA and ODI, all of which were obtained preoperatively and postoperative subsequently at 1-, 3-, 6-, 12-month follow-up after surgery. The clinical outcomes were determined to be excellent, good, fair, or poor according to the MacNab classification at the 1-year follow-up time. The blood loss, length of incision, operative time, length of hospital stay and overall hospitalization costs were also compared between the two groups. (4) Patients with degenerative lumbar spinal disease and underwent instrumented MFF were enrolled in our prospective randomized controlled study. Allograft bone was randomized to be used for MFF on one side of the facet joints as the Allograft group, then autologous local bone was used on the other side as the control group in the same patient. Computed tomography with fine-cut axial images and sagittal reconstruction views was performed to evaluate the fusion rate between two groups at 6-, 12-month follow-up postoperatively. Fusion status was evaluated separately by two radiologists blinded to the group affiliation. The complications and reoperations throughout the follow-up period were also compared.

Results   (1) As compared to preoperative model, different models of L4-5 segmental fusion after MFF, PLF and PLIF techniques all obtained reliable stability. No difference in range of motion at L4-5 level was observed among the three fusion techniques. (2) Of the 491 patients, the fusion rates at 6-month follow-up were 56.8% (279/491) and 96.1% (472/491) at the 1-year follow-up. Between baseline and 1-year follow-up time, VAS-LP and VAS-LBP improved from 5.6 ± 0.9 to 0.4 ± 0.5 and 5.1 ± 1.2 to 1.5 ± 0.9, respectively. JOA improved from 9.0 ± 2.0 preoperatively to 27.7 ± 1.0 at the 1-year follow-up and ODI decreased from 64.0 ± 2.0 to 19 ± 1.0. At the final evaluation, 460 (93.6%) patients showed excellent or good results, 16 (3.2%) fair, and 15 (3%) poor. There were no MFF technique-related complications. (3) One hundred and thirty-six patients were included in the study (MFF, n = 71; PLIF, n = 65). Fusion rates at the 6-month and 12-month follow-up time were not statistically significant between the MFF and PLIF groups (54.9% vs 58.5%, P = 0.67 and 94.4% vs 98.5%, P = 0.20). Clinical outcomes including VAS scores, JOA and ODI showed significant improvement compared with preoperative baseline throughout all follow-up time points and no significant difference between two groups. No complications directly related to the fusion techniques and reoperation occurred. (4) A total of 65 patients were enrolled in this trial. Among them, 61 patients with 226 fact joints completed the 1-year follow-up (Allograft group, n = 113; Control group, n = 113). The 1-year follow-up rate was 93.8%. At the 6-month follow-up, fusion rate in control group was higher than Allograft group (51.3% vs 35.4%, P = 0.016). At the 12-month follow-up, fusion rate between two groups has no significant difference (93.8% vs 91.2%, P ＞ 0.05). No complications attributable to allograft bone and reoperation occurred.

Conclusions   (1) MFF technique could achieve comparable spinal stability with PLF and PLIF techniques. (2) MFF with pedicle screw instrumentation achieved satisfactory clinical outcomes and fusion rate. (3) MFF provided fusion rate and clinical outcomes comparable to those of PLIF, which appears to be a promising alternative fusion technique for the treatment of lumbar degenerative diseases. (4) Allograft bone provided comparable fusion rate with autologous local bone as a fusion material for lumbar modified facet joint fusion technique, and could be a notable alternative.