Cubital Tunnel Syndrome

Cubital tunnel syndrome is compression of the ulnar nerve at the elbow, where it passes between the medial epicondyle and the olecranon. It is the second most common peripheral nerve compression, after carpal tunnel syndrome. Paraesthesia in the small and ulnar ring finger comes first; intrinsic weakness follows as it progresses.

Who. Men predominate 2:1, the peak age is 30-50 years and 30% of cases are bilateral. Repetitive elbow flexion and leaning on the elbows are the occupational risks. The largest population study (Osei, US claims data) is more modest, finding incidence only slightly higher in men overall and rising with age in both sexes, so quote the 2:1 ratio and the mid-life peak with that in mind.

Causes. Most cases are idiopathic. Screen every patient for an underlying cause:

• Repetitive flexion, occupational or from sleeping with the elbow bent
• Direct pressure from leaning on the elbows
• Cubitus valgus, producing tardy ulnar palsy after a fracture
• Osteophytes from elbow arthritis

The tunnel. A fibro-osseous tunnel between the medial epicondyle and the olecranon. Its floor is the MCL and elbow capsule; its roof is Osborne's ligament, the arcuate ligament joining the two heads of FCU. Here the ulnar nerve becomes superficial and vulnerable.

The course. Proximally the nerve passes the medial intermuscular septum in the medial arm and the arcade of Struthers, 5-10cm above the epicondyle. It then runs through the retrocondylar groove between the medial epicondyle and olecranon, and leaves between the two heads of FCU under Osborne's ligament. Each of these is a potential compression site.

The roof varies. It ranges from absent or thin fascial bands to thick tendinous or muscular structures. The variation changes flexion-related compression and explains why operative findings do not always match a single textbook form.

View larger

Flexion narrows the tunnel. Elbow flexion stretches the nerve 4-8mm and narrows the tunnel by 55%, and the pressure within it increases sixfold from extension to flexion. The nerve's excursion with full flexion is 10mm.

Blood supply. Multiple coiled vessels enter the nerve segmentally along its length. Excessive circumferential dissection during transposition can devascularise it, so preserve the feeding vessels and a vascularised tissue bed whenever possible.

View larger

What the nerve supplies in the hand. Sensation to the small finger and the ulnar half of the ring finger, palmar and dorsal. The motor supply:

• Hypothenar muscles: ADM, ODM and FDM
• Adductor pollicis, thumb adduction
• Third and fourth lumbricals
• Palmar interossei, finger adduction
• All the dorsal interossei, finger abduction, including the first dorsal interosseous, which abducts the index
• The deep head of FPB

This is the one cubital tunnel presentation in which simple in-situ decompression is not the default answer, and the classic exception to decompression first.

What "tardy" means. A delayed ulnar neuropathy that appears months, years or even decades after an elbow injury in childhood. The classic cause, and the commonest, is a lateral condyle fracture nonunion or malunion, or else a physeal injury. Both arrest lateral-column growth and produce a progressive cubitus valgus.

How the nerve fails. As the carrying angle increases, the nerve is chronically stretched and angulated as it courses behind the medial epicondyle. Repeated traction over years, often compounded by friction and subluxation, eventually produces the neuropathy long after the original fracture has been forgotten.

Why transposition, not in-situ release. The problem is length and tension, not just a tight roof. In-situ decompression leaves the nerve on the stretch behind a valgus elbow. Anterior transposition moves it in front of the axis of elbow flexion, which shortens its path and relieves the traction.

The deformity. Correcting a correctable bony deformity (e.g. a supracondylar/varus-producing osteotomy for the valgus, or addressing a lateral condyle nonunion) is considered alongside the nerve procedure, particularly in younger patients. Deformity correction alone does not reliably reverse an established neuropathy.

Reading the McGowan grade. The grade guides treatment. Grade I is intermittent numbness with a normal examination, and gets a conservative trial of 3 months before surgery. Weakness on examination or a positive Froment's sign is Grade II: motor involvement is a surgical indication. Intrinsic atrophy and a claw hand are Grade III, which calls for urgent decompression and a warning that motor recovery may be incomplete.

History. Numbness in the small and ulnar ring finger, worse at night and with elbow flexion. Weak grip shows as dropping objects, and numbness while holding a phone to the ear is the phone sign. Ask about occupation and sleeping position; rapid progression and severe weakness are red flags.

Examination. Always compare with the other side. Tinel's sign over the cubital tunnel and two-point discrimination, abnormal at over 6mm, sit alongside the provocative and motor signs below.

The motor signs. Froment's sign, Wartenberg's sign and clawing are the three motor signs of ulnar palsy, in order of severity. In Froment's sign the weak adductor pollicis is replaced by FPL, supplied by the median nerve, so the thumb IP joint flexes as the patient pinches. In Wartenberg's sign a weak third palmar interosseous cannot adduct the small finger, which rests abducted and catches on the pocket.

Clawing and the ulnar paradox. The ring and small fingers hyperextend at the MCP joints and flex at the IP joints. The claw is worse with a low lesion. In a high lesion FDP is paralysed as well, so there is less IP flexion; in a low lesion FDP works and the clawing is more marked.

The unstable nerve. With elbow flexion the nerve can move from the retrocondylar groove to the front of the epicondyle. Dynamic palpation and ultrasound distinguish symptomatic instability from a static enlarged nerve.

View larger

A second snap may follow the first. The first is the ulnar nerve crossing the epicondyle; deeper flexion brings the medial head of triceps over it as well, and missing the triceps component can leave persistent snapping after an isolated nerve transposition.

View larger

A clinical diagnosis. A typical presentation with a positive Tinel's sign, elbow flexion test and motor signs is diagnosed clinically.

Nerve conduction studies are the gold standard. They show slowing of motor conduction velocity across the elbow, under 50 m/s, compared with the segments above and below.

EMG grades severity. It shows denervation potentials in the intrinsics (FDI, ADM), and fibrillations indicate axonal loss.

Imaging. Radiographs show cubitus valgus, osteophytes or a prior fracture; MRI shows subluxation, a mass lesion or nerve changes. On short-axis ultrasound the compressed nerve is enlarged and hypoechoic, with an increased cross-sectional area and loss of its normal fascicular pattern at the compression site. Compare the proximal and distal segments and interpret the scan alongside the examination and electrodiagnostic findings.

View larger

View larger

Conservative treatment is appropriate for McGowan Grade I with intermittent symptoms only.

• Activity modification: avoid prolonged elbow flexion and leaning on the elbows, with an ergonomic workplace assessment
• A night extension splint, towel wrap or commercial, to keep the elbow extended and prevent nocturnal flexion
• An elbow pad to protect the nerve from direct pressure, worn during the day if there is occupational exposure

When to operate. The absolute indications are intrinsic muscle atrophy (McGowan III), failure of an adequate conservative trial of 3 months, and progressive weakness. McGowan Grade II with motor involvement, recurrent subluxation of the nerve and occupational requirements are relative indications. Do not delay surgery if motor involvement is present.

In-situ decompression is technically simpler, with a lower complication rate.

1. Position supine with the arm on a table, the elbow flexed 20-30 degrees; a tourniquet is optional.
2. Make a curved 6-8cm incision between the medial epicondyle and olecranon, over the retrocondylar groove, protecting the medial antebrachial cutaneous nerve.
3. Identify and protect the ulnar nerve. Release Osborne's ligament (the FCU aponeurosis) and continue distally between the FCU heads, leaving no distal fascial edge to kink the nerve.
4. Release the arcade of Struthers if tight, and the medial intermuscular septum.
5. Check nerve excursion and stability through flexion, then close in layers; no drain is required.

Preserve the epineurium and the vascular supply throughout.

View larger

View larger

View larger

Anterior transposition adds complexity but addresses subluxation. The nerve can be placed in one of three beds:

• Subcutaneous, superficial to the flexor-pronator mass
• Intramuscular, within the flexor-pronator mass
• Submuscular, deep to the elevated flexor-pronator origin

It is selected for instability, deformity, a scarred tunnel or revision rather than used routinely in every case:

• A subluxating nerve
• Failed in-situ decompression
• Cubitus valgus (tardy ulnar palsy)
• A scarred tunnel
• Revision surgery

Subluxation as an indication is contested by the trial evidence; see Controversies.

View larger

Pivot points. The nerve is mobilised from the retrocondylar groove while the branch points and the median nerve are identified. Moving it anteriorly creates new bends at the limits of release. Sufficient proximal and distal neurolysis is required to prevent traction and compression at these pivot points.

View larger

View larger

Submuscular transposition with Z-lengthening. Planned Z flaps lengthen the flexor-pronator fascia rather than closing it tightly over the transposed nerve, which creates a broad, nonconstricting muscular bed. A proximal fascial flap and a distal FCU musculofascial flap are mobilised while the nerve is protected, then reapproximated without tension over the anterior nerve bed. Flex and extend the elbow before closure to exclude kinking, tethering or excessive fascial pressure.

View larger

View larger

View larger

Structures at risk. The anterior and posterior branches of the medial antebrachial cutaneous (MABC) nerve cross near the operative field beside the ulnar nerve, and identifying their course reduces painful neuroma and persistent medial-elbow dysaesthesia after decompression.

View larger

Motor branches. Sequential branches leave the nerve soon distal to the epicondyle. Over-long release, aggressive traction or deep dissection risks denervating FCU and intrinsic pathways during transposition or epicondylectomy.

View larger

Medial epicondylectomy occupies the middle ground between in-situ decompression and transposition, and it is a recurring viva option.

The principle. Rather than moving the nerve, remove the bony prominence it is tethered against. After the tunnel is decompressed, the medial epicondyle is osteotomised, partially or completely, so the nerve can glide anteriorly on its own during flexion without the formal mobilisation, dissection and potential devascularisation of a transposition. It preserves the nerve's intrinsic blood supply better than transposition while still removing the retrocondylar tether. It avoids a new anterior bed, but shares the same need to preserve the branches and the blood supply.

What the evidence shows. Outcomes are broadly equivalent to the other operations but stage-dependent. A randomised trial within the Cochrane review found no difference between medial epicondylectomy and anterior transposition in clinical or neurophysiological outcome (Cochrane, PMID 27845501). The staged meta-analysis (Mowlavi, PMID 10946931) found that in minimal or mild disease total symptom relief was achieved most often after medial epicondylectomy, but in severe disease it produced the poorest operative result. Those rankings compare separate case series rather than randomised arms, so hold them loosely.

The practical message. Epicondylectomy is a reasonable alternative in milder disease and for the subluxating nerve. It is not a rescue for the atrophic hand, and the MCL must be respected to avoid trading neuropathy for instability.

MABC nerve injury causing a painful neuroma is the most common complication, and careful handling and protection during dissection are essential.

Return to work. Desk work at 1-2 weeks and manual work at 6-8 weeks.

Success rates. In-situ decompression and anterior transposition both give 80-90% good or excellent results, with no significant difference between them.

Prognosis. Motor recovery is often incomplete if intrinsic atrophy is present pre-operatively. Intrinsic recovery depends on the duration of symptoms, and sensation recovers better than motor function.

Global Epidemiology

• Second most common upper-limb entrapment neuropathy after carpal tunnel syndrome (Cochrane, PMID 27845501).
• Population incidence approximately 30 per 100,000 person-years (US claims database, PMID 28362959); rises with age, slightly higher in men.
• Around 40% of diagnosed patients eventually undergo surgery; surgical conversion increases with age.

Side-by-Side Guidance

Across societies the convergent position—supported by Level I trials (PMID 15730578) and meta-analyses (PMID 18056489, 27845501)—is that simple in-situ decompression is the default operation, with transposition reserved for documented subluxation, revision, or specific deformity (e.g. cubitus valgus tardy ulnar palsy).

Registry & Outcome Notes

• No dedicated international cubital tunnel registry exists; evidence rests on RCTs and administrative databases (e.g. PMID 28362959).
• Reported good/excellent outcomes are 80-90% across techniques, with revision rates of roughly 4-5% (PMID 29481399).

High- vs Limited-Resource Practice

• High-resource: routine NCS/EMG, ultrasound nerve cross-sectional area, day-case surgery, endoscopic options.
• Limited-resource: clinical diagnosis and the elbow flexion/Tinel tests guide management; open in-situ decompression under local or regional anaesthesia is effective, low-cost, and avoids the need for advanced neurophysiology.

Documentation & Consent (universal)

• Record motor examination (Froment's, Wartenberg's) and any intrinsic wasting at each visit.
• Consent must include risk of incomplete motor recovery when atrophy is present, MABC nerve injury, wound complications, and recurrence.
• Capture occupational exposure where work-related causation is plausible.

Decompression or transposition in severe disease. Older dogma reserved transposition for severe (McGowan III) cases. Gervasio's RCT (PMID 15617592) found equivalence even in Dellon Grade 3 disease and the Cochrane review (PMID 27845501) reaches the same conclusion, but the two are not independent, because Gervasio is one of the trials Cochrane pools. The severe-disease evidence is therefore 35 patients per arm, which excludes only a large difference, and the optimal operation for the atrophic hand remains genuinely debated.

The subluxating nerve. Subluxation is the classic indication for transposition, yet Bartels Part 1 (PMID 15730578) found outcomes unaffected by subluxation after simple decompression. Whether subluxation alone justifies transposition is unsettled.

Endoscopic or open release. Endoscopic in-situ release gives equivalent clinical outcomes (Buchanan, PMID 29481399) with less scar pain but more haematoma. Cost, the learning curve and limited long-term data keep its role contested.

Ultrasound and nerve CSA. High-resolution ultrasound, measuring an enlarged nerve cross-sectional area, is increasingly used, but diagnostic thresholds are not standardised and NCS/EMG remains the reference standard. The place of imaging in equivocal cases is evolving.

Other open questions. When, if ever, to operate in purely sensory McGowan I disease that fails conservative care. The value of supplementary nerve-gliding exercises, which added no benefit over advice in the one RCT analysed by Cochrane. And the lack of a unified outcome metric across trials, which complicates pooling.