Comparison of Anterior Cervical Discectomy and Fusion versus Posterior Cervical Foraminotomy in the Treatment of Cervical Radiculopathy: A Systematic Review

Abstract

Controversy remains over whether anterior cervical discectomy and fusion (ACDF) or posterior cervical foraminotomy (PCF) is superior for the treatment of cervical radiculopathy. We therefore performed a systematic review including three prospective randomized controlled trails (RCT) and seven retrospective comparative studies (RCoS) by searching PubMed and EMBASE. These studies were assessed on risk of bias according to the Cochrane Handbook for Systematic Reviews of Interventions, and the quality of evidence and level of recommendation were evaluated according to the GRADE approach. Clinical outcomes, complications, reoperation rates, radiological parameters, and cost/cost‐utility were evaluated. The mean complication rate was 7% in the ACDF group and 4% in the PCF group, and the mean reoperation rate was 4% in the ACDF group and 6% in the PCF group within 2 years of the initial surgery. There was a strong level of recommendation that no difference existed in clinical outcome, complication rate and reoperation rate between the ACDF and the PCF group. There was conflicting evidence that the ACDF group had better clinical outcomes than the PCF group (one study with weak level of recommendation). PCF could preserve the range of motion (ROM) of the operated segment but did not increase the ROM of the adjacent segment (weak level of recommendation). Meanwhile, the average cost or cost‐utility of the PCF group was significantly lower than that of the ACDF group (weak level of recommendation). In conclusion, the PCF was just as safe and effective as the ACDF in the treatment of cervical radiculopathy. Meanwhile, PCF might have lower medical cost than ACDF and decrease the incidence of adjacent segment disease. Based on the available evidence, PCF appears to be another good surgical approach in the treatment of cervical radiculopathy.

Introduction

Cervical radiculopathy is defined as a pain and/or sensorimotor deficit syndrome caused by compression of a cervical nerve root1. It is common among middle‐aged to elderly individuals, and often occurs secondary to disc herniation or foraminal stenosis due to cervical degeneration. The symptoms of cervical radiculopathy range from pain, numbness, or tingling in the upper extremity to electrical type pains or even weakness. Conservative treatment is initially recommended for degenerative cervical radiculopathy, and surgery is indicated for patients that are refractory to non‐operative treatment2.

Anterior cervical discectomy and fusion (ACDF) and posterior cervical foraminotomy (PCF) are safe and effective surgical approaches for the treatment of cervical radiculopathy3, 4. The two surgical approaches were initially reported in the 1940s and1950s and have been modified over time3. Compared with PCF, ACDF has the advantages of easier and wider exposure of the intervertebral space and less patient discomfort, and has been considered the standard operation for cervical radiculopathy over the past two decades. However, there are also many disadvantages with the ACDF, including symptomatic adjacent disc disease, pseudoarthrosis, instrumentation failure and, ventral approach‐related complications5, 6.

Use of PCF can avoid ventral approach‐related complications, such as postoperative dysphagia, hematoma, and recurrent laryngeal nerve palsy3, 6. In addition, it does not require fusion, which avoids fusion‐related complications, such as pseudoarthrosis and instrumentation failure, and might preserve the mobility of operated segments, which might decrease the incidence of adjacent level degeneration. Another advantage of PCF is that surgeons can visualize the compressed nerve root directly and confirm the decompression5. However, higher rates of neck pain, muscle spasm, and more blood loss were reported associated with PCF for the damage of dorsal cervical musculature7, 8, 9. Furthermore, as PCF does not require reconstruction or fusion the intervertebral space, it may be associated with a higher incidence of revision surgery than ACDF3. Meanwhile, the indication for PCF is limited to patients with lateral disc herniation and foraminal stenosis without cord compression and large central disc herniation, because the cervical myelon must not be mobilized toward medial7, 10. PCF can be performed following an open dorsal approach or a minimally‐invasive (MIS) approach using a tubular retractor; the MIS approach is more popular nowadays. However, a meta‐analysis indicated that there was no significant difference in clinical outcomes when patients with symptomatic cervical radiculopathy from foraminal stenosis were treated with traditional open or an MIS foraminotomy11.

However, there are conflicting outcomes on reoperation and success rates for ACDF and PCF. Some studies report that PCF might be associated with a higher incidence of revision surgery and that it has a lower less success rate than ACDF12, 13. Other studies conclude that reoperations and success rates are similar for the two approaches3, 7. No systematic review has been published to compare ACDF and PCF in the treatment of cervical radiculopathy. Therefore, we performed this systematic review to summarize the evidence in the literature to evaluate the clinical outcomes, radiological outcomes, complications, reoperation rates, and cost/cost effectiveness of ACDF and PCF in the treatment of cervical radiculopathy.

Methods

We developed a protocol prior to this systematic review that was registered in PROSPERO. The registration number is CRD42016032776.

Search Strategy

The primary sources of the studies reviewed in this systematic review were PubMed and EMBASE. The search included literature exclusively in English and published up to 28 December 2015. The following terms were used in our search: cervical foraminotomy OR cervical laminoforaminotomy OR cervical keyhole OR posterior cervical microdiscectomy, OR posterior cervical discectomy. Reference lists of all included studies were scanned to identify potentially relevant studies. Two reviewers independently screened the titles and abstracts of studies identified from the search. Full text copies of all potentially relevant studies were obtained.

Inclusion and Exclusion Criteria

Studies were included if they met the following criteria: (i) study design: prospective or retrospective comparative study; (ii) patients with cervical radiculopathy due to lateral disc herniation or foraminal stenosis; (iii) clinical outcomes, radiological outcomes, complications, reoperation rates, and cost/cost effectiveness differences were compared between ACDF and PCF; and (iv) published in English.

Studies on tumors, trauma, infection, previous surgeries, revision surgeries, combined anterior and posterior surgeries, and other posterior approaches were excluded. Non‐English studies were excluded.

Data Extraction and Analysis

We extracted data from each study, including study design, study location, number of cases, age, follow‐up time, surgical levels, clinical outcomes, radiological outcomes, complications, reoperations, and cost/cost effectiveness. Statistical analysis was performed using Review Manager software (RevMan Version 5.2; The Nordic Cochrane Center, The Cochrane Collaboration, Copenhagen, Denmark) and SPSS version 18.0 (LEAD Technologies, NC, USA). Heterogeneity was tested using the χ²‐test and quantified by calculating the I ² statistic, in which P < 0.05 and I ² > 50% was considered statistically significant. For the pooled effects, the weighted mean difference (WMD) was calculated for continuous variables, and the odds ratio (OR) was calculated for dichotomous variables. Continuous variables are presented as mean differences and 95% confidence intervals (CIs), whereas dichotomous variables are presented as ORs and 95% CI. Random‐effects or fixed‐effects models were used depending on the heterogeneity of the studies included. Publication bias was tested using a funnel plot.

Quality Assessment

The quality of evidence and strength of recommendation was assessed according to the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach14. Details regarding treatment strategy and outcome parameters were registered by two independent reviewers. Differences in risk of bias, quality of evidence, strength of recommendation, and outcome parameters were discussed in a consensus meeting.

Results

Characteristics of Studies

A total of 769 articles were first identified by screening the titles and abstracts. Of these, 757 articles were excluded for the following reasons: duplication, irrelevant studies, case reports, revision surgeries, non‐English written studies, and reviews. The remaining 12 articles underwent a detailed and comprehensive evaluation (Fig. 1); 1 trial protocol article and 1 expert opinion article were excluded. The remaining 10 studies were finally included, comprising 3 prospective randomized controlled trails (RCT) and 7 retrospective comparative studies (RCoS)3, 4, 7, 13, 15, 16, 17, 18, 19, 20. Details of these studies are listed in Table 1.

Figure 1.

Flow chart of studies selection.

Table 1.

Characteristics of included studies

Study	Design	Country	Number of cases	Surgical levels	Follow‐up time*	Age (years)* *
Ruetten et al. 7	RCT	Germany	ACDF: 86PCF: 89	Single‐level monoradiculopathy	ACDF: 2 years† PCF: 2 years†	43† , ‡
Wirth et al. 16	RCT	USA	ACDF: 25PCF: 22	Single‐level monoradiculopathy	ACDF: 60 monthsPCF: 60 months	ACDF: 41.7† PCF: 43.8†
Herkowitz et al. 15	RCT	USA	ACDF: 17PCF: 16	Single‐level monoradiculopathy	4.2 years† , §	ACDF: 43† PCF: 39†
Selvanatha et al. 4	RCoS	UK	ACDF: 150PCF: 51	N/A	ACDF: 24.0 ± 1.4 monthsPCF: 25.0 ± 1.2 months	ACDF: 48† PCF: 50†
Lubelski et al. 3	RCoS	USA	ACDF: 188PCF: 140	ACDF: 2.6 ± 0.7 PCF: 2.7 ± 0.8	ACDF: 2 years† PCF: 2 years†	ACDF: 53.3 ± 11.6PCF: 52.9 ± 12.2
Cho et al. 19	RCoS	Korea	ACDF: 30PCF: 31	Single‐level monoradiculopathy	ACDF: 61.22 ± 20.22 monthsPCF: 62.61 ± 19.90 months	ACDF: 49.73 ± 10.41PCF: 52.45 ± 9.65
Korinth et al. 13	RCoS	Germany	ACDF: 124PCF: 168	Single‐level monoradiculopathy	72.1 ± 25.9 months§	ACDF: 45.9 ± 8.2PCF: 46.9 ± 10.4
Mansfield et al. 20	RCoS	USA	ACDF: 79PCF: 22	Single‐level monoradiculopathy	3 years† , §	ACDF: 49† PCF: 49†
Alvin et al. 18	RCoS	USA	ACDF: 45PCF: 25	Single‐level radiculopathy (unilateral or bilateral)	1 year† , §	ACDF: 49.3 ± 9.6PCF: 46.5 ± 11.2
Tumialán et al. 17	RCoS	USA	ACDF: 19PCF: 19	ACDF: 1.1 ± 0.3PCF: 1.1 ± 0.3	ACDF: 18.1 monthsPCF: 11.2 months	ACDF: 41.5 ± 8.4PCF: 49.3 ± 8.1

^* Data are represented as mean ± standard deviation unless otherwise indicated.

^† Only mean values are reported.

^‡ Mean age of all cases.

^§ Mean follow‐up time of all cases.

ACDF, anterior cervical discectomy and fusion; N/A, not applicable; PCF, posterior cervical foraminotomy; RCoS, retrospective comparative study; RCT, prospective randomized controlled trail.

Quality Assessment

According to the GRADE approach, the quality of evidence was high in 1 study, moderate in 3 studies, and very low in 6 studies (Table 2). The level of recommendation was strong for 4 studies and weak for the 6 other studies.

Table 2.

Quality of included studies according to the GRADE approach

Study	Design	Selection bias	Performance bias	Attrition bias	Detection bias	Quality of evidence	Recommendation
Ruetten et al. 7	RCT	No	No	Yes	No	High	Strong
Wirth et al. 16	RCT	No	No	No	Yes	Moderate	Strong
Herkowitz et al. 15	RCT	Yes	No	No	Yes	Moderate	Strong
Selvanathan et al. 4	RCoS	Yes	Yes	Yes	No	Very low	Weak
Lubelski et al. 3	RCoS	No	Yes	No	No	Moderate	Strong
Cho et al. 19	RCoS	Yes	Yes	No	No	Very low	Weak
Korinth et al. 13	RCoS	Yes	Yes	No	No	Very low	Weak
Mansfield et al. 20	RCoS	Yes	Yes	No	Yes	Very low	Weak
Alvin et al. 18	RCoS	Yes	Yes	No	Yes	Very low	Weak
Tumialán et al. 17	RCoS	Yes	Yes	No	Yes	Very low	Weak

RCoS, retrospective comparative study; RCT, prospective randomized controlled trail.

Clinical Outcomes

Clinical outcomes were reported in 6 studies, including 3 RCT and 3 RCoS (Table 3)4, 7, 13, 15, 16, 18. Ruetten et al. 7 evaluated visual analogue scale (VAS), German version North American Spine Society Instrument, and Hilibrand criteria in an RCT, and no significant difference was observed between ACDF and PCF groups at preoperative and 2‐year follow‐up time points; the satisfaction rate was 91% in the ACDF group and 96% in the PCF group7. Wirth et al. report that the incidence of pain relief was 100% in the PCF group and 96% in the ACDF group in an RCT; and no significant difference was observed16. Relief of pain and weakness was evaluated by Herkowitz et al. in an RCT; a 75% excellent/good relief rate in the PCF group and a 94% excellent/good relief rate in the ACDF group was observed, respectively; however, no significant difference was observed between the two groups15.

Table 3.

Clinical outcomes of included studies

Study	Design	Number of cases	Outcome criteria	Clinical outcome (ACDF vs PCF)
Ruetten et al. 7	RCT	ACDF: 86PCF: 89	VAS,German version NASS Instrument, Hilibrand Criteria	P > 0.05
Wirth et al. 16	RCT	ACDF: 25PCF: 22	Incidence of pain relief	96% vs 100% (P > 0.05)
Herkowitz et al. 15	RCT	ACDF: 17PCF: 16	Relief of pain and weakness (excellent/good rate)	94% vs 75% (P > 0.05)
Selvanathan et al. 4	RCoS	ACDF: 150PCF: 51	NDI, VAS‐neck, VAS‐arm	P > 0.05
Korinth et al. 13	RCoS	ACDF: 124PCF: 168	Success rate (Odom I + II)	93.6% vs 85.1% (P < 0.05)
Alvin et al. 18	RCoS	ACDF: 45PCF: 25	VAS, PDQ, PHQ‐9, EQ‐5D	P > 0.05

ACDF, anterior cervical discectomy and fusion; EQ‐5D, euroQol‐5 dimensions; NASS, North American Spine Society; NDI, neck disability index; PCF, posterior cervical foraminotomy; PDQ, pain disability questionnaire; PHQ, patient health questionnaire; RCoS, retrospective comparative study; RCT, prospective randomized controlled trail; VAS, visual analogue scale.

Three RCoS also compared clinical outcomes of both groups. Selvanathan et al. report that the improvement of neck disability index (NDI), VAS‐neck and VAS‐arm were similar between both groups with a trend favoring the PCF group4. Korinth et al. report that the success rate (Odom I + II) was significantly higher in the ACDF group (93.6%) than the PCF group (85.1%)13. Alvin et al. evaluated VAS, the Pain Disability Questionnaire (PDQ), the Patient Health Questionnaire (PHQ‐9), and the EuroQol‐5 Dimensions (EQ‐5D); and no significant difference was observed between ACDF and PCF groups for any of these measures at preoperative and 1‐year postoperative time points18.

Radiological Outcomes

Only 1 RCoS reported on postoperative range of motion (ROM) of cervical spine19. The ROM of the operated segment decreased in the ACDF group but was preserved in the PCF group; it was no motion in the ACDF group and 8.82° ± 6.65° in the PCF group, respectively. Meanwhile, the ROM of the adjacent segment increased in the ACDF group but did not increase in the PCF group; the postoperative ROM of the caudal adjacent segment was 11.33° ± 5.07° in the ACDF group and 8.73° ± 5.92° in the PCF group, respectively19.

Complications

Complications were reported in 6 studies, including 2 RCT and 4 RCoS4, 7, 13, 15, 17, 18. A total of 809 cases were involved, including 441 cases of ACDF and 368 cases of PCF. The mean complication rate was 7% in the ACDF group and 4% in the PCF group. There was no significant difference in the complication rate between the ACDF group and the PCF group (P > 0.05, OR, 1.53 [0.84, 2.79]; Fig. 2). Moderate heterogeneity existed between these studies (I ² = 41%, P = 0.16).

Figure 2.

Forest plot illustrating the complication rate in comparison between anterior cervical discectomy and fusion (ACDF) and posterior cervical foraminotomy (PCF) groups.

Reoperations

Reoperations within 2 years of the initial surgery were reported in 3 studies, including 1 RCT and 2 RCoS3, 4, 7. 704 cases were involved including 424 cases of ACDF and 280 cases of PCF. The mean reoperation rate was 4% in ACDF group and 6% in PCF group. There was no significant difference in reoperation rate between ACDF group and PCF group (P > 0.05, OR, 0.74 [0.36, 1.52]; Fig. 3). The χ ² test indicated no statistical evidence of heterogeneity (I ² = 0%, P = 0.41).

Figure 3.

Forest plot illustrating reoperations within 2 years of the initial surgery in comparison between anterior cervical discectomy and fusion (ACDF) and posterior cervical foraminotomy (PCF) groups.

Longer follow‐up time was reported in 2 studies. Wirth et al. reported that reoperation rate was 28% in ACDF group and 27% in PCF group respectively with 60 months follow‐up time16. Korinth et al. reported that reoperation rate at the index level was 2.4% in ACDF group and 7.1% in PCF group with about 72.1 months follow‐up time13. No significant difference in reoperation rate was observed in both these studies.

Cost/Cost‐Effectiveness

Three RCoS reported cost or cost‐effectiveness of both groups in the treatment of single‐level cervical radiculopathy17, 18, 20. Mansfield et al. reported that the average cost of an ACDF was 89% more than a PCF with 3 years follow‐up time20. Alvin et al. reported that the cost‐utility ratio of PCF group was 39% lower than that of ACDF group with 1 year follow‐up time18. Tumialán et al. reported that the direct cost of PCF group was 65% lower than that of ACDF group, and the indirect cost of PCF group was 76% lower than that of ACDF group17.

Publication Bias

The Stata 12.0 software (StataCorp LP, College Station, TX, USA) was used to examine the publication bias of the main results. Both funnel plots were largely symmetrical including complication rate and reoperation rate. It indicated that publication bias did not play a vital role in the observed effects and the conclusions were reliable.

Discussion

ACDF is more popular than PCF in the treatment of cervical radiculopathy. However, conflicting evidence still exists for the 2 approaches. We therefore performed this systematic review to summarize the evidence in the literature to evaluate the clinical outcomes, radiological outcomes, complications, reoperation rates and cost/cost effectiveness between ACDF and PCF in the treatment of cervical radiculopathy.

Ten studies were included in this systematic review including 3 RCTs and seven RCoS. Quality of evidence and strength of recommendation was evaluated by GRADE approach. We followed weak level of recommendation only when strong level of recommendation was lack.

Clinical outcomes were reported in 3 RCTs and 3 RCoS4, 7, 13, 15, 16, 18. Although different outcome criteria were measured, satisfactory clinical outcomes of both approaches were reported in all these studies. Success rate or incidence of pain relief was 93.6%–96% in ACDF group and 75%–100% in PCF group respectively13, 15, 16; no significant difference was observed in all 3 RCT and 2 RCoS4, 7, 15, 16, 18. Only one RCoS reported that ACDF group had better clinical outcome than PCF group, however, the recommendation level of this study was weak13. There was no significant difference in clinical outcomes between ACDF and PCF group basing on the strong level of recommendation of 3 RCTs. It indicated that both surgical approaches were very effective in the treatment of cervical radiculopathy.

Some studies evaluated changes of cervical sagittal alignment after PCF, no significant change was observed in focal and global cervical curvature with at least 30 months follow‐up time21, 22. In this systematic review, only one RCoS reported postoperative ROM of cervical spine19. The ROM of operated segment decreased in ACDF group but was preserved in PCF group; meanwhile, the ROM of the adjacent segment increased in ACDF group but did not increase in PCF group. PCF might impose less stress on the adjacent segments and decrease the incidence of adjacent segment disease19. However, the recommendation level of this study was weak, higher level of evidence was needed.

Complications were reported in 2 RCT and 4 RCoS4, 7, 13, 15, 17, 18. The mean complication rate was 7% in ACDF group and 4% in PCF group. There was no significant difference in complication rate between ACDF group and PCF group. In included studies, the complications were implanted‐related and ventral approach‐related complications including dysphagia, hematoma, hoarse, recurrent laryngeal nerve palsy, pseudarthrosis, adjacent segment disease, wound infection and so on in ACDF group, and dermatoma‐related hypesthesia, wound infection, cerebrospinal fluid fistula and so on in PCF group4, 7, 13, 15, 17, 18. No high rate of neck pain or muscle spasm was reported in PCF group in these studies. Nerve injury or neurological deterioration was occasionally reported in both groups. It indicated that both ACDF and PCF were safe in the treatment of cervical radiculopathy.

Reoperations within 2 years of the initial surgery were reported in 3 studies3, 4, 7. The mean reoperation rate was 4% in ACDF group and 6% in PCF group, and no significant difference was observed. Longer follow‐up time was reported in 2 studies. Wirth et al. reported that reoperation rate was 28% in ACDF group and 27% in PCF group respectively with 60 months follow‐up time16. Korinth et al. reported that reoperation rate at the index level was 2.4% in ACDF group and 7.1% in PCF group with about 72.1 months follow‐up time13. No significant difference in reoperation rate was observed in all these studies. It indicated that PCF did not put patients at higher risk for revision surgery than ACDF.

Three RCoS reported cost or cost‐effectiveness of both groups in the treatment of single‐level cervical radiculopathy17, 18, 20. All these studies reported that the average cost or cost‐utility of PCF group was significantly lower than ACDF group. However, the recommendation levels of these studies were weak; higher level of evidence was needed.

Meanwhile, there are some limitations to this systematic review. First, the included studies are in English; thus, a potential language bias may exist in this systematic review. Second, the sample size may not be large enough to find possible existing evidence, especially the sample size of 2 RCT and 4 RCoS15, 16, 17, 18, 19, 20; larger‐scale studies are further needed to provide more reliable evidence for future evaluation. Third, there was a variable length of time in the follow‐ups between some of the studies and this complicated evaluating and comparing the surgical results. Fourth, different outcome criteria were measured in included studies; this might result in bias outcomes. Fifth, 70% of the included studies had a very low level of evidence, resulting in a weak level of recommendation. Finally, clinical heterogeneity may be caused by the various indications for surgery and for the use of certain surgical technologies at the different treatment centers.

Conclusion

In this systematic review, we summarized the evidence in the literature to compare ACDF with PCF in the treatment of cervical radiculopathy, and no significant difference was found between them in clinical outcomes, complication rates, and reoperation rates. PCF was just as safe and effective as ACDF in the treatment of cervical radiculopathy. Meanwhile, PCF might have lower medical cost than ACDF and did not increase the ROM of the adjacent segment, which might decrease the incidence of adjacent segment disease. Based on the available evidence, PCF seems to be another good surgical approach in the treatment of cervical radiculopathy.