Is anterior decompression and fusion superior to laminoplasty for cervical myelopathy due to ossification of posterior longitudinal ligament? A systematic review and meta-analysis

Abstract

Context: Considerable controversy exists over surgical procedures for ossification of the posterior longitudinal ligament (OPLL).

Objective: The purpose of the meta-analysis was to compare the clinical outcome of anterior decompression and fusion (ADF) with laminoplasty (LAMP) in treatment of cervical myelopathy due to OPLL.

Methods: PubMed, EMBASE and the Cochrane Register of Controlled Trials database were searched to identify potential clinical studies compared ADF with LAMP for cervical myelopathy owing to OPLL. We also manually searched the reference lists of articles and reviews for possible relevant studies. Thirteen studies with 1120 patients were included in our analysis. Subgroup analyses were performed by the canal occupying ratio of OPLL.

Results: Overall, the mean preoperative Japanese Orthopaedic Association (JOA) score was similar between two groups. Compared with LAMP group, ADF group was higher at the mean postoperative JOA scores and mean recovery rate, reoperation rate, and longer at mean operation time. There was not significantly different in mean blood loss and complication rate between two groups. In subgroup analysis, ADF had a higher mean postoperative JOA score and recovery rate than LAMP in cases of OPLL with occupying ratios ≥ 50%, while those difference were not found in cases of OPLL with occupying ratios < 50%.

Conclusion: ADF achieves better neurological improvement compared with LAMP in treatment of cervical myelopathy due to OPLL, especially in cases of OPLL with occupying ratios ≥ 50%. Complication rate is similar between two groups, but ADF can increase the risk of reoperation

Introduction

Ossification of posterior longitudinal ligament (OPLL) is characterized by ectopic bone formation in spinal ligaments.¹ The prevalence of OPLL in Asian countries is reported to be 0.4–4.3%, while the incidence is estimated to be 0.1–1.7% among North Americans and Europeans.^2,3 OPLL mostly occurs in the cervical spine and is a common cause of cervical myelopathy in Asian population, especially Japanese.⁴ Conservative treatment could be chosen for asymptomatic or mild symptomatic myelopathy, but surgical decompression should be reserved for moderate or severe symptomatic myelopathy due to cervical OPLL.^5,6

Both anterior decompression and fusion (ADF) and posterior laminoplasty (LAMP) are frequently used for cervical myelopathy caused by OPLL.^7,8 ADF allows definitive resection or mobilization of the ossified lesion, enables direct decompression of the spinal cord and maintains suitable alignment of the cervical spine. LAMP increases the available space of the spinal canal and achieves indirect decompression by shifting the spinal cord posteriorly.

Recent studies have compared the surgical outcomes of both ADF and LAMP for cervical myelopathy owing to OPLL.^9–15 Some studies have reported that neurological improvement after ADF was superior to that after LAMP.^10,13,14 However, other studies have showed that ADF was similar or inferior to LAMP in the surgical outcomes.^9,11,12,15 There is therefore considerable controversy over which surgical technique, ADF or LAMP, is better for cervical myelopathy due to OPLL. The purpose of the study was to perform a systematic review and meta-analysis to investigate whether ADF achieved better neurological outcome than LAMP in the treatment of cervical myelopathy due to OPLL. Other clinical parameters, such as operation time, blood loss, rates of complications and reoperation, were also compared between the two groups.

Materials and methods

Eligibility criteria

Studies were eligible for inclusion if they met the following criteria: (1) clinical studies comparing ADF with LAMP for cervical myelopathy due to OPLL, (2) study population 18 years and older, (3) minimum 12-months follow-up, (4) minimum 10 of sample size per group, (5) English literature. Studies were excluded if they met the following criteria: animal studies, biomechanical studies, duplicate publications of one trial, case report, letter, revision, technology note, thoracic OPLL, commentaries, reviews and meta-analysis.

Search strategy

A systematic and comprehensive literature search was conducted in PubMed, EMBASE and the Cochrane Register of Controlled Trials database up to 16 September 2018, the search strategy included use of Medical Subject Headings (MeSH) terms and keywords. Terms for OPLL included: Ossification of Posterior Longitudinal Ligament [Mesh] OR Calcification of Posterior Longitudinal Ligament OR Posterior Longitudinal Ligament Calcification OR Posterior Longitudinal Ligament Ossification. They were combined with terms specifying surgery ((Laminoplasty [Mesh] OR Laminoplasties OR Laminaplasty OR Laminaplasties) OR (anterior decompression and fusion OR anterior cervical corpectomy OR anterior cervical discectomy OR cervical decompression and fusion)). Reference lists of all relevant retrieved articles and reviews were manually searched to identify additional studies that might have been missed. If necessary, we also contacted the corresponding author to clarify some ambiguous data. However, it was difficult to identify unpublished studies which were not included in this meta-analysis.

Study selection

In the initial search, 422 potentially relevant publications were identified. After removing 62 duplications, the titles and abstracts of 360 publications were screened. 341 publications were excluded for the following reasons: not involved ADF versus LAMP (n = 150), review (n = 51), case report (n = 31), meta-analysis (n = 9), letter (n = 5), technology note (n = 6), commentaries (n = 7), biomechanical (n = 5), animal study (n = 6), thoracic OPLL (n = 19), revision surgery (n = 3), no English (n = 10) and no related (n = 39). Based on the remaining 19 publications, a comprehensive review of the full-text was conducted. Studies that did not fulfill the inclusion criteria in any manner were excluded. Two reviewers independently selected eligible publications and any disagreement was settled by discussion with a third reviewer. Finally, thirteen studies were identified to be included in this meta-analysis.^10,13–24 The process was shown in Figure 1.

Figure 1.

The flowchart shows the process of publication selection. ADF, anterior decompression and fusion; LAMP, laminoplasty; OPLL, ossification of posterior longitudinal ligament.

Data extraction

Study characteristics and outcomes in the included studies were extracted independently by two reviewers, with discrepancies being solved through consensus with a third reviewer. The primary outcomes of interest were neurological functional outcomes including the preoperative and postoperative Japanese Orthopaedic Association (JOA) scores and recovery rate. Secondary outcomes included operation time, blood loss, rate of complications and rate of reoperation.

Study characteristics

Of the 13 included studies, twelve were retrospective studies^13–24 and one was prospective study.¹⁰ All included studies were conducted in Asia, eight studies were done in Japan,^{10,13,14,16,18,20,22,24} three were from china^15,17,19 and two was from korea.^21,23 The years of publication ranged from 1995 to 2017, and the lengths of follow-up ranged from 1 to 24.8 years. Sample sizes ranged from 26 to 252, and had a total of 1120 patients (560 in the ADF group and 560 in the LAMP group). Baseline characteristics and postoperative data of the included studies were shown in Tables 1 and 2, respectively.

Table 1. Baseline characteristics of included studies.

Study ID	Study design	Country	Surgery approach	Number of patients	Patients Age mean ± SD or(range)	Follow-up time(m)	Surgical Segments mean ± SD or(range)or levels	Preoperative JOAs mean score ± SD	Occupied rate (%)
Goto S et al. 1995	Retrospective	Japan	ADF LAMP	34 65	54.5 (NA) 58.2 (NA)	104.4 (13.2-297.6) 84 (20.4-180)	NA NA	7.8 ± 3.2 7.6 ± 2.9	≥50 ≥50
Baba et al. 1995	Retrospective	Japan	ADF LAMP	88 47	47 (NA) 56 (NA)	204 175.2	<3 ≥3	9.4(NA) 8.4 ± 1.1	≥30 ≥30
Tani T et al. 2002	Retrospective	Japan	ADF LAMP	14 12	62 ± 11 66 ± 6	49 ± 34 50 ± 43	3.5 ± 1 4 ± 1.2	9.4 ± 2.5 8.8 ± 2.8	≥50 ≥50
Iwasaki et al. 2007	Retrospective	Japan	ADF LAMP	27 66	58(41-74) 57(41-75)	72 (24-120) 122.4(60-240)	≥2 ≥2	9.5 ± 2.4 9.2 ± 3.8	≥40 ≥40
Masaki et al. 2007	Retrospective	Japan	ADF LAMP	19 40	52.8 ± 6.6 62.6 ± 10.3	≥12 ≥12	2.9 ± 0.9* 4.6 ± 0.5	8.3 ± 2.9 8.6 ± 2.4	≥50 ≥50
Lee et al. 2008	Retrospective	Korea	ADF LAMP	20 27	56.8 (42-72) 54.7 (30-70)	21.8 (6-61) 29.1 (11-64)	≥3 ≥3	NA NA	NA NA
Chen et al. 2011	Retrospective	China	ADF LAMP	22 25	57.2(41-73) 54.2(32-66)	≥48 ≥48	≥3 4-5	9.3 ± 1.8 8.5 ± 0.7	<50 <50
Sakai et al. 2012	Prospective	Japan	ADF LAMP	20 22	59.5 ± 9.3 58.4 ± 9.6	≥60 ≥60	3.1(1-5)* 4.5(4-5)	11.4 ± 2.8 10.9 ± 2.3	<50 <50
Liu et al. 2013	Retrospective	China	ADF LAMP	68 59	54.4 ± 12.8 57.9 ± 9.5	≥60 ≥60	NA NA	7.6 ± 2.5 7.7 ± 2.7	≥40 ≥40
Fujimori et al. 2014	Retrospective	Japan	ADF LAMP	12 15	55.6 ± 7.8 58.7 ± 9.1	118.8 ± 49.2 122.4 ± 68.4	3.3 ± 0.9* 5.4 ± 1.2	9.5 ± 2.2 9.1 ± 2.6	≥50 ≥50
Kim et al. 2015	Retrospective	Korea	ADF LAMP	71 64	57.3 (35-76) 56.4 (35-76)	12-68 24-64	NA NA	12 ± 4 12 ± 3	≥50 ≥50
Koda et al. 2016	Retrospective	Japan	ADF LAMP	15 16	57.7(49-69) 60.3(36-82)	13-134 12-131	≥3 NA	9.8 ± 2.6 9.5 ± 2.6	≥50 ≥50
Hou et al. 2017	Retrospective	China	ADF LAMP	150 102	47.8 (29-79) 45.9 (30-77)	35.4 (29-38) 36.1 (28-40)	≤3 >3	9.5 ± 2.1 9.8 ± 1.9	≥50 ≥50

NA, not available; ADF, anterior decompression and fusion; LAMP, laminoplasty; *significant difference; JOAs, Japanese Orthopaedic Association score; SD, standard deviation.

Table 2. Postoperative Data after anterior decompression and fusion or laminoplasty.

Study ID	Study design	Country	Surgery approach	Number of Patients	Postoperative JOAs	Recovery rate %	Complication rate	Reoperation rate	Blood loss (ml)	Operation time (min)
					mean score ±SD	mean ±SD			mean score ±SD	mean ±SD
Goto S et al. 1995	Retrospective	Japan	ADF LAMP	34 65	NA NA	66.2±22.8 52.8±29	NA NA	NA NA	NA NA	NA NA
Baba et al. 1995	Retrospective	Japan	ADF LAMP	88 47	14.1(NA) 13.1±1.5	65.7 54.6	0 6%	0 0	NA NA	NA NA
Tani T et al. 2002	Retrospective	Japan	ADF LAMP	14 12	13.9±2.3 10.1±3.4	58±24 13±39	57.1% 66.7%	7.1% 16.7%	NA NA	NA NA
Iwasaki et al. 2007	Retrospective	Japan	ADF LAMP	27 66	13.2±1.9 12.2±3	51±23 55±30.3	29.6% 13.6%	26.0% 1.5%	513±415 464±437.5	302±75.8 177±76.3
Masaki et al. 2007e	Retrospective	Japan	ADF LAMP	19 40	14.2±2.3 13±2.6	68.4±27.3 52.5±30	NA NA	NA NA	NA NA	NA NA
Lee et al. 2008	Retrospective	Korea	ADF LAMP	20 27	NA NA	NA NA	25.0% 18.5%	10.0% 0	NA NA	NA NA
Chen et al. 2011	Retrospective	China	ADF LAMP	22 25	14.2±1.3 10.9±0.4	63.2±15.2 25.1±8.5	22.7% 32.0%	4.5% 0	NA NA	NA NA
Sakai et al. 2012	Prospective	Japan	ADF LAMP	20 22	15.1±2.2 14±2.2	71.4±26 55.3±29.6	25.0% 0.0%	15.0% 0	292.8±192.8 289.6±215.8	300.3±78.6 183.2±41.1
Liu et al. 2013	Retrospective	China	ADF LAMP	68 59	15.2±2.8 12.8±2.7	80.6±9.7 55.7±13.9	35.3% 18.6%	NA NA	460±240 330±110	190±42.5 160±22.5
Fujimori et al. 2014	Retrospective	Japan	ADF LAMP	12 15	13.3±1.9 11.7±2.7	52.5±18.5 30.1±30.5	25.0% 20.0%	33.3% 20.0%	600±459 240±153	314±89.9 128±60
Kim et.al. 2015	Retrospective	Korea	ADF LAMP	71 64	15.6±2.8 15±4	72.6±38.2 51.9±66.3	16.9% NA	NA NA	NA NA	NA NA
Koda et al. 2016	Retrospective	Japan	ADF LAMP	15 16	14.6±2.3 10.3±3	72.5±58.65 14.4±58.65	93.3% 6.3%	6.7% 0	218±277.5 251±111.3	302±106 152±63.8
Hou et al. 2017	Retrospective	China	ADF LAMP	150 102	14±2.1 15.2±2.8	NA NA	41.3% 51.0%	NA NA	142.6±48.5 186.7±64	152±37.2 128.5±25.6

NA, not available; ADF, anterior decompression and fusion; SD, standard deviation; LAMP, laminoplasty; *significant difference; JOAs, japanese orthopaedic association score; LAMF, laminectomy and Fusion.

Quality assessment

The quality assessment of the included studies was performed by the Newcastle-Ottawa Quality Assessment Scale (NOS), as recommended by the Cochrane Non-Randomized Studies.²⁵ This scale allocated a maximum of nine points for risk of bias in three domains: (1) selection of study groups (four points); (2) comparability of groups (two points); and (3) ascertainment of exposure and outcomes (three points) for case–control and cohort studies, respectively. Study that scored 6 or more was eligible for data-pooling and study that scored 7 or more was considered high quality.²⁶ The evaluation process was conducted by two reviewers independently. All argument were solved by discussion with a third reviewer.

Of the 13 included studies, one obtained 6 points of NOS,²¹ seven studies obtained 7 points of NOS^{14,15,17,18,22–24} and other five received more than 7 points of NOS.^{10,13,16,19,20} The quality assessment of included studies was summarized in Table 3.

Table 3. Quality assessment of included studies.

Study	Selection	Comparability	Outcome	Total scores
Baba et al. 1995	3	1	3	7
Goto S et al. 1995	4	1	3	8
Tani T et al. 2002	3	1	3	7
Masaki et al. 2007	3	2	2	7
Iwasaki et al. 2007	3	1	3	7
Chen et al. 2011	3	1	3	7
Sakai et al. 2012	4	2	3	9
Liu et al. 2013	4	1	3	8
Lee et al. 2013	3	1	3	7
Fujimori et al. 2014	3	2	3	8
Kim et al 2015	3	2	1	6
Koda et al. 2016	3	2	3	8
Hou et al. 2017	3	1	3	7

Subgroup analysis

Subgroup analysis was performed basing on mean preoperative canal occupying ratio of OPLL. Subgroup A included the patients of OPLL with canal occupying ratio ≥ 50%, while subgroup B included the patients of OPLL with canal occupying ratio < 50%.

Statistical analysis

All statistical analyses were conducted using Stata 14.0. Treatment effects were calculated as odds ratio (OR) for dichotomous outcomes and standardized mean difference (SMD) for continuous outcomes with a 95% confidence interval (CI). The heterogeneity among studies was examined by the I² statistic and considered significant if P value < 0.05 or I² > 50%. If no evident heterogeneity existed, the fixed effects model was selected to pool results. If present, a random effects model was utilized, and a Galbraith plot was performed to look for outliers in effect sizes. The expectation is that 95% of the studies is within the area defined by two CI lines. Then, a sensitivity analysis was performed by eliminating one or more study which was not within or far away from the area defined by two CI lines, until heterogeneity was not presented and results compared.²⁷ Publication bias was formally assessed by funnel plot and Egger test (P > 0.05 suggest no significant bias).

Results

Preoperative and postoperative JOA scores

Preoperative JOA score was available in eleven studies.^10,13–22 There was not a significant difference in mean preoperative JOA score between ADF and LAMP groups (P = 0.807, SMD = 0.02, 95% CI −0.12 to 0.15, Figure 2), and heterogeneity across studies was not statistically significant (P = 0.745, I²= 0%). Postoperative JOA score was available in ten studies.^10,14–22 Mean Postoperative JOA score was significant higher in ADF group than in LAMP group (P = 0.003, SMD = 0.82, 95% CI 0.28–1.37, Figure 3), and significant heterogeneity was identified (P < 0.001, I²= 91.9%). A sensitive analysis was performed by removing Chen et al.,¹⁷ Koda et al.¹⁶ and Hou et al.¹⁵ to decrease heterogeneity (P = 0.065, I² = 49.4%). Statistical analysis using fixed effect model showed the stability of our results (P < 0.001, SMD = 0.51, 95% CI 0.33–0.69) (Appendix 1).

Figure 2.

The forest plot shows the comparison of preoperative JOA score between ADF group and LAMP group. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. SMD, standardized mean difference; ADF, anterior decompression and fusion; LAMP, laminoplasty; JOA, Japanese Orthopaedic Association.

Figure 3.

The forest plot shows the comparison of postoperative JOA score between ADF group and LAMP group. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. SMD, standardized mean difference; ADF, anterior decompression and fusion; LAMP, laminoplasty; JOA, Japanese Orthopaedic Association.

Neurological recovery rate

Recovery rate was presented in ten studies.^{10,13,14,16–22} The mean recovery rate in the ADF group was higher than that in the LAMP group (P < 0.001, SMD = 1.00, 95% CI 0.45–1.55, Figure 4), with high heterogeneity across studies (P < 0.001, I²= 90.1%). Sensitive analysis with fixed effect model, by removing Chen et al.,¹⁷ Liu et al.¹⁹ and Iwasaki et al.,¹⁸ showed the stability of our results (P < 0.001, SMD = 0.58, 95% CI 0.38–0.78) without heterogeneity (P = 0.360, I²= 9.0%) (Appendix 2).

Figure 4.

The forest plot shows the comparison of recovery rate between ADF group and LAMP group. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. SMD, standardized mean difference; ADF, anterior decompression and fusion; LAMP, laminoplasty.

Complication rate

Data of complication rate was available in ten studies.^{10,14–20,23,24} No significant difference was found in complication rate between ADF and LAMP groups (P = 0.246, OR = 1.59, 95% CI 0.73–3.45, Figure 5), with high heterogeneity across studies (P < 0.001, I²= 70.4%). A sensitive analysis was performed by removing Koda et al.,¹⁶ Sakai et al.¹⁰ and Baba et al.²⁴ to decrease the heterogeneity (P = 0.108, I²= 42.5%), and statistical analysis using fixed effect model indicated the stability of our results (P = 0.697, OR = 1.07, 95% CI 0.75–1.53) (Appendix 3). The most common complications were CSF leakage and axial pain in ADF population, whereas axial pain and C5 paralysis were most common complications in LAMP population. The distribution of complications was shown in Table 4.

Figure 5.

The forest plot shows the comparison of complication rate between ADF group and LAMP group. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. OR, odds ratio; ADF, anterior decompression and fusion; LAMP, laminoplasty.

Table 4. Distribution of complications in ADF and LAMP Population.

Complication	ADF, No, (%) LAMP, No, (%)
CSF Leakage	20(15.74) 5(5.2O5)
Hematoma	5(3.94) 3(3.13)
Axial pain	45(35.43) 63(65.63)
C5 paralysis	14(11) 20(20.83)
Pseudarthorosis	5(3.94) 0
Neurological deterioration	11(8.7) 5(5.205)
Graft extrusion	12(9.45) 0
Hoarseness	14(11) 0
Dysphagia	1(0.8) 0
Total	127 96

CSF, cerebrospinal fluid; LAMP, laminoplasty; ADF, anterior decompression and fusion.

Reoperation rate

Data of reoperation rate was available in seven studies.^{10,14,16–18,20,23} ADF group suffered from a higher rate of reoperation than LAMP group (P = 0.005, OR = 3.92, 95% CI 1.51–10.16, Figure 6), and heterogeneity across studies was not statistically significant (P = 0.328, I²= 13.3%).

Figure 6.

The forest plot shows the comparison of reoperation rate between ADF group and LAMP group. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. OR, odds ratio; ADF, anterior decompression and fusion; LAMP, laminoplasty.

Operation time

Data of operation time was available in six studies.^{10,15,16,18–20} The mean operation time was longer in ADF group compared with LAMP group (P < 0.001, SMD = 1.44, 95% CI 0.94–1.94, Figure 7), with significant heterogeneity across studies (P < 0.001, I²= 82.1%). Sensitive analysis with fixed effect model, by removing Liu et al.¹⁹ and Hou et al.,¹⁵ showed the stability of our results (P < 0.001, SMD = 1.81, 95% CI 1.46–2.16) without heterogeneity (P = 0.531, I²= 0%) (Appendix 4).

Figure 7.

The forest plot shows the comparison of operation time between ADF group and LAMP group. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. SMD, standardized mean difference; ADF, anterior decompression and fusion; LAMP, laminoplasty.

Blood loss

Data of blood loss was available in six studies.^{10,15,16,18–20} There was no significant difference in mean blood loss between two groups (P = 0.679, SMD = 0.13, 95% CI –0.49 to 0.76, Figure 8), with significant heterogeneity across studies (P < 0.001, I²= 90.9%). Sensitive analysis with fixed effect model, by removing Hou et al.,¹⁵ Fujimori et al.²⁰ and Liu et al.,¹⁹ showed the stability of our results (P = 0.855, SMD = 0.03, 95% CI −0.29 to 0.35) without heterogeneity (P = 0.815, I²= 0%) (Appendix 5).

Figure 8.

The forest plot shows the comparison of blood loss between ADF group and LAMP group. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. SMD, standardized mean difference; ADF, anterior decompression and fusion; LAMP, laminoplasty.

Subgroup analysis

In subgroup A, postoperative JOA score was available in eight studies.^{14–16,18–22} The higher mean postoperative JOA score was found in ADF group than in LAMP group (P = 0.019, SMD = 0.66, 95% CI 0.11–1.21, Figure 9), with heterogeneity (I²= 87.1%, P < 0.001). Sensitive analysis with fixed effect model, by removing Koda et al.¹⁶ and Hou et al.,¹⁵ showed the stability of our results (P < 0.001, SMD = 0.49, 95% CI 0.25–0.73) without heterogeneity (P = 0.076, I²= 49.9%, Appendix 6). Recovery rate was available in eight studies.^{13,14,16,18–22} The mean recovery rate was higher in ADF group than in LAMP group (P < 0.001, SMD = 1.03, 95% CI 0.56–1.50, Figure 10), with heterogeneity (I²= 76.4, P < 0.001). Sensitive analysis with fixed effect model, by removing Liu et al.,¹⁹ showed the stability of our results (P < 0.001, SMD = 0.61, 95% CI 0.4–0.82) without heterogeneity (I²= 32.7%, P = 0.178, Appendix 7).

Figure 9.

The forest plot shows the comparison of postoperative JOA score between ADF group and LAMP group in the patients with canal occupying ratio ≥ 50%. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. SMD, standardized mean difference.

Figure 10.

The forest plot shows the comparison of recovery rate between ADF group and LAMP group in the patients with canal occupying ratio ≥ 50%. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. SMD, standardized mean difference; ADF, anterior decompression and fusion; LAMP, laminoplasty; JOA, Japanese Orthopaedic Association.

In subgroup B, postoperative JOA score and recovery rate were available in four studies.^10,17–19 There was no significant difference in mean postoperative JOA score (P = 0.117, SMD = 0.99, 95% CI −0.25 to 2.22, Figure 11) and recovery rate (P = 0.065, SMD = 1.35, 95% CI −0.085 to 2.79, Figure 12) between ADF and LAMP groups, with heterogeneity (I²= 94.3, P < 0,001 and I²= 95.5%, P < 0.001, respectively). Owing to the limited number of included studies, sensitive analysis over those results could not be performed. Therefore, careful interpretation should be urged for those results.

Figure 11.

The forest plot shows the comparison of postoperative JOA score between ADF group and LAMP group in the patients with canal occupying ratio < 50%. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. SMD, standardized mean difference.

Figure 12.

The forest plot shows the comparison of recovery rate between ADF group and LAMP group in the patients with canal occupying ratio < 50%. The size of the squares reflects the weight of the trial in pooled analysis. The horizontal bars represent the 95% CI. SMD, standardized mean difference. ADF, anterior decompression and fusion; LAMP, laminoplasty; JOA, Japanese Orthopaedic Association.

Publication bias

The shape of funnel plot was symmetric, indicating no existence of publication bias in recovery rate (P_egger= 0.236, Figure 13) and complication rate (P_egger= 0.243, Appendix 8) identified studies reported. However, the shape of funnel plot presented asymmetry in preoperative (P_egger= 0.017, Appendix 9) and postoperative JOA scores (P_egger= 0.006, Appendix 10), which might be attributed to the bias of measurement and follow-up. There were fewer than 10 references to other secondary outcome index, publication bias related to those outcomes was therefore not properly assessed by a funnel plot asymmetry or more advanced regression-based tests.²⁸

Figure 13.

The shape of funnel plot was symmetric, indicating no existence of publication bias in recovery rate identified studies reported. SE, standard error; SMD, standard mean difference.

Discussion

Cervical myelopathy is mainly due to compression of ossified lesion in patients with cervical OPLL, surgical decompression is therefore recommended.^29,30 The optimal surgical option for cervical myelopathy due to OPLL remains a subject of considerable debate, and controversy always exists over the surgical approach: anterior or posterior surgery. The traditional mainstream method in anterior surgery is ADF and typical posterior method is LAMP. Although many studies have recently compared the surgical outcomes between ADF and LAMP for cervical myelopathy due to OPLL, they all were non-randomized controlled studies and offered conflicting results.^9–15 In the current study, our results showed that ADF had a higher recovery rate of neurological function compared with LAMP in treatment of cervical myelopathy owing to OPLL, especially in those patients of OPLL with canal occupying ratio ≥ 50%. The incidence of complications was similar between two groups, but ADF was associated with higher rate of reoperation.

ADF is widely accepted as a standard surgical treatment for moderate or severe symptomatic cervical myelopathy.^7,8 ADF can directly remove the ossified lesion, which theoretically eliminates anterior compression of spinal cord and offers the probability of neurological improvement. Furthermore, spinal fusion can stabilize the cervical spine and relieve dynamic pressure on the injured cord. In this meta-analysis, we demonstrated that, compared with LAMP, ADF offered better neurological outcomes with higher postoperative JOA scores and recovery rate. However, the operation is more technically demanding. In addition, ADF is associated with a series of complications, such as pseudarthrosis, cerebrospinal fluid leakage, dislodgement, hematoma, dysphagia and esophageal perforation.^31–34 Several previous studies demonstrated that ADF resulted in a relatively high rate of complications, and most of those complications were attribute to high demanding of procedure and anatomical feature.^35–37 But, our study found no significant difference in complication rate between ADF and LAMP groups, which therefore should be interpreted with some caution, while more large-size, long follow-up and high quality trials should be performed to confirm this result.

LAMP allows indirect decompression through a posterior shift of the spinal cord. However, inadequate shift of the spinal cord and poor neurological recovery after LAMP often present in patients with cervical kyphosis or thick OPLL (occupying ratio ≥ 50%).^38–41 It was reported that LAMP showed poor surgical outcomes in patients with massive OPLL of occupying ratio ≥ 50%.^20,21 In the present study, we found that LAMP was inferior to ADF in the neurologic improvement, especially in the patients of OPLL with canal occupying ratio ≥ 50%. However, LAMP is a less technically demanding and achieves shorter operation time compared with ADF. We also found that the incidence of reoperation in the ADF group was higher than that in the LAMP group. Therefore, LAMP might be a safe alternative to ADF for cervical myelopathy due to OPLL.

This meta-analysis has some limitations. First, all included studies were non-randomized controlled trials, which limited the evidence level of our study. Despite all of 13 included studies obtained 6 or more points of NOS, the inherent evidence defect of cohort study was not nullified. Second, the detail of surgical treatment differed for each included study and the number of involved segment were available in only nine studies.^{10,14,15,17,19,20,22–24} Reported surgical segments in the LAMP group were 3 or more in these nine studies, but surgical segments in the ADF group were 3 or less in some studies^15,19,22,24 which could induce bias in this meta-analysis. Third, all included studies were performed in Asia. This may induce region bias, which means that the surgical effectiveness in Asians might not be applied to Caucasian or other race.

Conclusions

In summary, our results suggest that ADF achieves better neurological improvement compared with LAMP in treatment of cervical myelopathy due to OPLL, especially in the patients of OPLL with canal occupying ratio ≥ 50%. Complication rate is similar between two groups, but ADF can increase the risk of reoperation. Additional prospective random studies with high-quality, large-scale and long follow-up are required to update this meta-analysis to better investigate which procedure, ADF or LAMP, is better for cervical myelopathy caused by OPLL.

Disclaimer statements

Contributors None.

Funding None.

Conflict of interest The authors declare that they have no conflict of interest.