Surgical management of cubital tunnel syndrome: A systematic review and meta-analysis of randomised trials

Abstract

Background

Cubital tunnel syndrome (CUTS) is a common upper limb compression neuropathy with significant consequences when left untreated. Surgical decompression remains gold-standard treatment for moderate to severe disease, however the optimal operative technique remains unclear. This network meta-analysis (NMA) of Level I and II randomised prospective studies aims to discern superiority between open in-situ, endoscopic and anterior transposition (subcutaneous or submuscular techniques) with respect to the primary outcome of response-to-treatment and secondary outcomes which include complications, post-operative chronic pain VAS scale, return to work and re-operation.

Methods

This NMA adhered to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. PubMed, Web of Science, Cochrane Central, Science direct and Embase were searched. The MESH database was further searched with the terms ‘cubital tunnel’ to improve sensitivity of the search. Data pertaining to the primary and secondary outcomes were pooled for NMA.

Results

Following abstract and full-text screening, 10 randomised prospective trials were included. There was no statistical difference in the response-to-treatment between the four studied techniques. Endoscopic decompression conferred a significantly higher complication rate compared to open decompression (Odds Ratio [OR], 4.21; 95% CI, 1.22–14.59). Endoscopic decompression had a statistically significant lower risk of post-operative chronic pain compared to open in-situ decompression (OR, 0.03, 95% CI, 0.00–0.32). There were no differences between techniques with respect to return to work or re-operation rates.

Conclusion

Response-to-treatment was similar between the four operative techniques for CUTS. Endoscopic decompression was found to be more hazardous when compared to open-in situ decompression but conferred significantly less post-operative chronic pain. There was significant heterogeneity in reported outcomes between the included articles. The authors suggest conducting more high-quality research with standardised outcome reporting to facilitate comparison.

Level of evidence: ii

Systematic Review and Meta-analysis of Randomised Prospective Trials- Therapeutic study.

1. Introduction

Cubital tunnel syndrome (CUTS) is one of the common compression neuropathies affecting ∼25 cases per 100 000 population per year.¹ Advanced disease may result in the compromise of hand function, quality of life and earnings.² Surgical decompression is the most effective treatment in moderate to severe cases3, 4, 5 with ∼15 000 patients across the UK⁶ and US⁷ undergoing surgical decompression annually.

Given the clinical importance of CUTS, the condition has been identified as a high priority elbow pathology with insufficient evidence in literature for treatment and rehabilitation.⁸ Various surgical techniques exist with open in-situ decompression remaining the commonly performed operation.⁹ Nerve transposition (either submuscular or subcutaneous) was historically a rising trend while endoscopic release is increasing in popularity given the smaller incisions.¹⁰ Currently, there is a lack of consensus regarding the optimal surgical technique with differences in reported patient outcomes and complications. No strong evidence suggests that one technique is better than others. The technique of choice seems to depend on the preference and experience of the operating clinician.¹¹

The main aim of this network meta-analysis is to gather reliable evidence from randomised prospective trials (level I and II evidence) to determine the best surgical technique in achieving optimal results whilst maintaining low incidence of post-operative complications following cubital tunnel decompression.

2. Methods

The authors adhered to PRISMA guidelines throughout the preparation of this network metanalysis (NMA).

2.1. Strategy of literature search

A literature search of PubMed, Cochrane Central, Web of Science, EMBASE and Science direct was performed. The MESH database was interrogated with the query ‘cubital tunnel’ to increase the search sensitivity. Two authors independently screened the retrieved citations. Screening was conducted in two steps: (1) abstract screening and retrieval of full-text articles if deemed relevant to the research question; (2) full-article screening and inclusion if deemed to fulfil inclusion and exclusion criteria.

2.2. Inclusion criteria

Studies meeting the subsequent standards were included in the review.

• 1)Randomised prospective trials with adult patients (>18 years old) undergoing surgical decompression (open in-situ decompression, endoscopic decompression, subcutaneous or sub-muscular nerve transposition) for CUTS
• 2)Clinically evident CUTS

2.3. Exclusion criteria

• 1)Non-randomised trials, retrospective studies, case series or reports, review articles and conference abstracts
• 2)Studies including post-traumatic ulnar nerve compression (distal humerus fracture and/or elbow dislocation), polyneuropathy or tumors in or around ulnar nerve
• 3)Studies on revision cubital tunnel decompression
• 4)Articles that were not available in English
• 5)Studies where data extraction and analysis were unreliable

2.4. Quality assessment

Article quality was assessed as per the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions (5.1.0, part 2, Chapter 8.5, updated March 2011). The Cochrane risk of bias assessment tool was utilised to appraise the studies. Selection bias (sequence generation, allocation concealment), performance bias (participant and personnel blinding), detection bias (blinding of outcome assessment), attrition bias (incomplete outcome data), reporting bias (selective outcome reporting) and other potential biases were assessed. Judgments were classified as ‘Low, ‘High or ‘Unclear’ risk of bias.

2.5. Data synthesis

2.5.1. Outcomes

The primary outcome for this NMA will be response-to-treatment at one year after surgery. Secondary outcomes include complications, re-operation, post-operative chronic pain and return to work.

Outcome ranks, return to work rate, complication rate, re-operation rate and chronic pain rates were pooled and analysed in a meta-analysis module. Frequentist NMA was performed using MetaInsight V4 tool using the concept of repeated sampling.¹²

Due to heterogenicity in reporting outcomes which included Bishop score, modified Bishop score and rate of patient satisfaction, outcome was divided into two pools: ‘satisfactory’ which included excellent and good outcomes and ‘unsatisfactory’ which included fair and poor outcomes.

Inconsistency analysis presents all pair-wise comparisons of the treatments and evaluates the consistency and agreement between estimates from the direct and indirect effects.

2.6. Assessment of heterogeneity

Visual inspection of the forest plots and the inconsistency model were used to asses heterogeneity. A random-effects model was used when there was significant heterogeneity (P value < 0.1).

3. Results

3.1. Search results

1657 unique articles were identified following the search. 127 titles were selected for full-text review. Ten studies were included in this NMA (PRISMA flow diagram, Fig. 1). The included studies are summarised in Table 1^13-22.

Fig. 1.

PRISMA flow diagram.

Table 1.

Basic characteristics of included study population, outcome assessment and conclusions.

Study ID	Group	Sample	Mean age/years (range)	Gender M:F	Affected side	Assessment	Conclusion
Krejčí 201813	Endoscopic	22	52.4 (22–72)	11 M:11F	Left: 20 (44.4%) Right: 25 (55.6%)	Pain - VAS Bishop score Scar aesthetics Complications Working or activity status	Follow-up: 3 months and 1 year Pain: No difference in post-operative or chronic pain between groups at 1 year Bishop score: Both techniques produce equally satisfactory outcomes. Endoscopy has better scar aesthetics and earlier return to work at the expense of longer operative duration. No complications reported.
	Open	23	56.9 (44–74)	12 M:11F
Gervasio 200414	Endoscopic	35	53.1 (34–75)	25 M:10F	Left: 21 (60%) Right: 14 (40%)	Bishop score Electrophysiological Outcome Complications	Follow-up: 3 weeks, 6 months and a last follow-up. Mean follow-up was ∼47 (7–72) for both groups. Bishop score and Electrophysiological outcomes were comparable between the two techniques. No major complications noted at one year mark.
	Submuscular transposition	35	52.2 (32–74)	23 M:12F	Left: 20 (57%) Right: 15 (43%)
Schmidt 201515	Endoscopic	29	50.3 ± 10.7	17 M:12F	Left: 31 patients (57.4%) Right: 21 patients (38.9%) Bilateral: 2 (3.7%)	Modified Bishop Score Electrophysiological outcome Pain Complications Re-operation rate	Follow-up: 1 week, 3 months, 6 months, 1 year and 2 years Modified Bishops Score: Both techniques yielded good to excellent results in more than 80% of patients with no difference between groups. Electrophysiological outcome: No difference between groups. Pain: No significant differences between groups No disadvantages in the open technique while using a small incision. No injuries to the medial antebrachial nerve observed in the open group.
	Open	27	47.9 ± 12.9	16 M:11F
Keiner 200916	Submuscular transposition	16	46 (14–75)	8 M:8F	Left: 5 Right: 11	Pain Sensory and Motor Symptoms Atrophy Patient satisfaction	Follow-up: 2 months, Long-term telephone interview: median 63.1 (41–73) months Outcome: Both techniques had a good outcome at long-term follow-up. Simple decompression was recommended by authors.
	Open	17	52 (29–74)	12 M:5F	Left: 8 Right: 9
Bartels 200517	Open	75	47.2 ± 12.9 (20.5–76.4)	46 M:29F	Left: 48 Right 27	Short Form 36 McGill Pain Questionnaire Sensory and Motor Exam Atrophy Complications Re-operation rate	Follow-up: 6 weeks, 18 weeks, 1 year Outcome: Both techniques were equally effective. On average, open decompression was 17.6 min quicker. Complications: Significantly fewer complications occurred following open decompression (7/30 versus 23/30). Simple decompression was favoured for its simplicity.
	Subcutaneous transposition	77	47.1 ± 12.1 (20.3–76.6)	48 M:29F	Left: 59 Right: 18
Biggs 200418	Open	23	56.7 (27–77)	16 M:7F	Not available	Pain levels Sensory and Motor function Functionality McGowen Louisiana State University Medical Centre grading systems Complications	Follow up: 6 weeks, 6 months and 1 year Outcome: Both simple decompression and sub-muscular transposition were effective in improving clinical outcome with no significant difference. Transposition is associated with greater complexity and risk.
	Submuscular transposition	21	61.1 (29–83)	17 M:4F
Heikenfeld 201319	Open	15	Not available	Not available	Not available	Modified Bishop score Patient satisfaction Electrophysiology Complications	Follow-up: 10 days, 3 months and 1 year Endoscopic cubital tunnel decompression leads to comparable results to open in situ decompression after 1 year.
	Endoscopic	15
Zarezadeh 201220	Subcutaneous transposition	24	47.58 ± 12.1	13 M:11F	Not available	Pain - VAS Sensation and Motor exam Muscle atrophy	Follow-up: 2–3 weeks, 1 year Pain: Submuscular transposition demonstrated significant pain reduction. No other differences were identified between groups.
	Submuscular transposition	24	47.41 ± 12.2	14 M:10F
Jaddue 200921	Subcutaneous transposition	13	34 (range not available)	26 M:8F	Not available	Pain - VAS Bishop score Incision length Operative time Complications	Follow-up: 2 weeks, 6 months and 1 year Pain and Bishops score: Subcutaneous transposition was significantly superior. The authors recommend the subcutaneous technique but recognise study limitations.
	Submuscular transposition	13
Schwarm 202222	Endoscopic	25	63 (45.0–73.0)	5 M:20F	12 (54.5%) dominant arm	Modified Bishop score Incapacity to work Duration of postoperative pain Hypoesthesia Muscle atrophy and weakness	Follow-up: 3 months and 1 year. No differences was identified for outcome measures between treatment groups
	Open	15	50.0 (43.0–62.5)	11 M:4F	8 (32.0%) dominant arm

3.2. Quality of included studies

Overall, there was no serious risk of bias. However, the included studies exhibited variability in their risk of bias assessments which may be attributed to differences in study designs, methodologies, and participant characteristics. A quality assessment overview of the included studies is presented in Fig. 2, Fig. 3.

Fig. 2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Fig. 3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

3.3. Results analysis

3.3.1. Response to treatment

Four interventions (open in-situ decompression, endoscopic decompression, subcutaneous and sub-muscular nerve transposition) in 10 studies were pooled and a NMA performed, including a total of 544 patients. Both direct and indirect comparison demonstrated no statistical difference in response-to-treatment between the studied techniques as demonstrated by all 95% CI crossing 1.00 (Fig. 4). P value for inconsistency analysis ranged from 0.44 to 0.73, indicating no significant evidence of inconsistencies were detected.

Fig. 4.

Response-to-treatment pooled results.

3.3.2. Complication rate

Four interventions in seven studies were pooled and analysed regarding complications rate including infection, keloid formation, delayed hematoma, wound dehiscence, and sensibility loss around scar. These studies included a total of 437 patients. A NMA demonstrated that open in-situ decompression had statistically significant lower complication rate compared to endoscopic decompression (OR 4.21, 95% CI 1.22–14.59) but not to subcutaneous (OR 2.78, 95% CI 0.71–10.92) or submuscular (OR 2.03, 95% CI 0.66–6.18) transposition (Fig. 5). P value for inconsistency analysis ranged from 0.31 to 0.73, indicating no significant evidence of inconsistencies were detected.

Fig. 5.

Complication rate pooled results.

3.3.3. Post-operative chronic pain

Four interventions in six studies were pooled and analysed, including a total of 288 patients. A NMA demonstrated that endoscopic decompression had a statistically significant lower risk of post-operative chronic pain (OR 0.03, 95% CI: 0.00–0.32) (Fig. 6). Conversely, there were no statistically significant differences observed in comparison to open in-situ decompression for both subcutaneous transposition (OR 1.35, 95% CI: 0.12–14.71) and submuscular transposition (OR 0.11, 95% CI: 0.01–1.36). P value for inconsistency analysis measured 0.63, indicating no significant evidence of inconsistencies were detected.

Fig. 6.

Post-operative chronic pain pooled results.

3.3.4. Return to work rate

Three interventions in three studies were pooled and analysed, including a total of 155 patients. No study including the anterior subcutaneous transposition technique assessed return to work as an outcome. A NMA demonstrated no statistically significant superiority in earlier return to previous employment or hobbies. Endoscopic decompression trended towards earlier return to work compared to open in-situ decompression (OR 5.33, 95% CI 0.99–28.84) (Fig. 7). In contrast, there were no statistically significant differences observed when comparing submuscular transposition (OR 0.81, 95% CI: 0.22–2.93) to open in situ decompression.

Fig. 7.

Return to work pooled results.

3.3.5. Re-operation rate

Three interventions in four studies were pooled and analysed, including a total of 293 patients. No study including anterior submuscular transposition assessed re-operation rate as an outcome. A NMA demonstrated no statistical difference between the studied techniques (Fig. 8).

Fig. 8.

Re-operation rate pooled results.

4. Discussion

CUTS is the second most common compressive neuropathy in the upper limb which can result in significant pain, paraesthesia and hand weakness. If left untreated, patients may develop irreversible muscle atrophy resulting in the compromise of hand function and quality of life.² Cubital tunnel syndrome was identified as a research priority due to the lack of consensus on optimal surgical management.⁸ Multiple decompression techniques exist and this is the first NMA of level I and II randomised prospective trials evaluating superiority. At a minimum follow-up of one year, all four interventions (open in-situ decompression, endoscopic release, anterior submuscular and anterior subcutaneous transposition) are comparable with respect to response-to-treatment, return to work and re-operation rate. Open in-situ decompression is associated with a lower complication rate compared to endoscopic decompression. Patients undergoing endoscopic decompression reported significantly less chronic pain when compared to open in-situ decompression.

A large systematic review and meta-analysis of 30 articles was undertaken by Wade et al. (2020) and included both experimental and observational studies, with a total of 2894 limbs undergoing eight different operative techniques.²³ The authors concluded that open in-situ decompression was more effective than nerve transposition with superior patient outcomes. Wade et al.’s inclusion criteria was less stringent with data pooling from lower-tiered studies, whereas we exclusively incorporated randomised control trials to reduce bias and improve methodological consistency. Despite this, the finding that endoscopic decompression was the more hazardous operation is concordant with our NMA of level I and II studies.²³

Another meta-analyses compared open in-situ decompression to endoscopic release for CUTS with the inclusion of two randomised control trials and five observational studies.²⁴ A significant difference in surgical outcome, patient satisfaction or complication rate was not demonstrated.²⁴ Byvaltsev et al. (2020) also compared open in-situ decompression to endoscopic release.⁵ Three randomised controlled trials and five retrospective studies were included. A significant difference was not demonstrated in Bishop score, VAS score reduction, post-operative satisfaction, hematoma rate and re-operation rate between the two groups. Patients reported a statistically significant improvement in scar tenderness and elbow pain following endoscopic release,²⁵ which align with our study findings and is likely due to the smaller surgical incision and less invasive nature of the endoscopic procedure.

Zlowodzki et al. (2007) conducted a meta-analysis of four randomised control trials, with a total of 261 patients, comparing open in-situ decompression to anterior nerve transposition (either subcutaneous or submuscular).¹¹ The included studies were homogenous and no difference in nerve conduction velocities or clinical outcome scores were demonstrated between treatment groups. The authors concluded that open in-situ decompression was a safe technique.¹¹ Chen et al. (2014) included 13 prospective and retrospective studies in a meta-analysis and concluded that open in-situ decompression had similar results to anterior transposition, although the complication rate was significantly lower in the open in-situ decompression group.²⁶ Our study conclusion aligns with the meta-analysis of both Zlowodzki and Chen et al. with respect to response-to-treatment. However, we were unable to perform a robust network meta-analysis with respect to return-to-work and re-operation rate following anterior transposition.

Liu et al. (2015) performed a meta-analysis of randomised control trials and observational studies comparing anterior subcutaneous versus submuscular transposition of the ulnar nerve.²⁷ In subgroup analysis, the risk of adverse events was significantly lower in the subcutaneous transposition group compared to the submuscular group. The study findings were underpowered with high risk of selection bias and the authors have advocated for future high-quality randomised control trials with standardised clinical outcomes to clarify this topic.²⁷

This NMA of Level I and II studies did not demonstrate a significant difference in response-to-treatment suggesting equivocal efficacy of the four studied techniques. The finding that endoscopic release is likely associated with an increased complication rate has been corroborated by literature.²³ The minimally invasive nature of the endoscopic approach risks inadvertent injury to surrounding structures including the ulnar nerve proper and medial antebrachial nerve.⁹ The limited exposure may preclude inadequate neurolysis and the need for revision surgery. We were unable to discern a difference in complication rates with respect to anterior transposition techniques. The more extensive dissection required to transpose the nerve may explain the increased complication rates reported in literature.²⁶ Reportedly, the anterior transposition of the ulnar nerve is linked to a reduction in regional blood flow to the ulnar nerve, persisting for a minimum of three days post-operation.²⁶ During anterior nerve transposition, it is frequently necessary to decompress the nerve over a minimum distance of 10 cm. The stripping of vasculature results in a significant decline in the regional blood flow to the nerve.²⁸ Moreover Ogata et al. (1991) demonstrated in animal studies that a devascularised nerve is more vulnerable to pressure.²⁹ All transposition procedures potentially risk nerve vascularity.³⁰

Despite our stringent inclusion and exclusion criteria the risk of bias remained as demonstrated by our assessments. Furthermore, the included randomised control trials were heterogenous with respect to the outcomes measured. These ranged from validated scores such as the VAS and Bishop score to unvalidated subjective items such as patient satisfaction. Not all studies reported comparable electrophysiological data, weakness or incapacity to work which the authors deem significant outcome measures. At present, we recommend that clinicians continue with the surgical technique that yields the best response-to-treatment and lowest complication rate in their practice. Further high quality multi-centre randomised control trials with standardised outcomes is necessary to confidently delineate superiority between available techniques.

Disclaimer

None.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Funding statement

The authors can confirm that this research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Patient consent

The authors can confirm that this systematic review did not require patient or guardian consent. No patient was directly involved in this research.

Ethical statement

This is a systematic review and meta-analysis of randomised trials. There were no instances of experimentation on human subjects and hence institutional ethical approval was not required. The principles of Good Clinical Practice including relevant laws and guidelines during the conduct of the review.

CRediT authorship contribution statement

Eslam Abourisha: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Visualization, Roles/, Writing – original draft. Ananth S. Srinivasan: Data curation, Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing – review & editing. Ahmed Barakat: Supervision, Formal analysis, Writing – review & editing. Han Hong Chong: Supervision, Formal analysis, Validation, Visualization, Roles/Writing, Writing – review & editing. Harvinder P. Singh: Conceptualization, Validation, Roles/, Writing – review & editing.