Contents

• Part I - General Principles of Nerve Compression
  • Compression versus entrapment
  • Pathophysiology of chronic compression
  • Grading and the double-crush phenomenon
  • Clinical evaluation
  • Electrodiagnostics
• Part II - Median Nerve Compression
  • Carpal tunnel syndrome
  • Pronator syndrome (median compression in the proximal forearm)
  • Anterior interosseous nerve (Kiloh-Nevin) syndrome
• Part III - Ulnar Nerve Compression
  • Cubital tunnel syndrome (ulnar nerve at the elbow)
  • Ulnar tunnel syndrome (Guyon’s canal)
• Part IV - Radial Nerve Compression
  • Posterior interosseous nerve (PIN) syndrome
  • Radial tunnel syndrome
  • Wartenberg syndrome (superficial radial nerve)
• Part V - Thoracic Outlet Syndrome
• Part VI - Lower-Limb Entrapment Syndromes
  • Common peroneal (fibular) nerve at the fibular neck
  • Tarsal tunnel syndrome
  • Meralgia paraesthetica
  • Morton’s metatarsalgia (interdigital neuroma)
  • Other lower-limb entrapments
• References

Part I - General Principles of Nerve Compression

A peripheral nerve can fail in many ways, but the commonest by far is slow strangulation where it passes through a tight anatomical corridor. These compression and entrapment neuropathies are among the most frequent problems in hand and upper-limb surgery. Knowing why a nerve suffers in a tunnel, how that suffering shows itself, and when to intervene applies just as well to the median nerve at the wrist as to the ulnar nerve at the elbow or the common peroneal nerve at the knee.

Compression versus entrapment

The two words are often used interchangeably, but Birch draws a useful distinction. Compression is a force applied to a nerve from outside, pressing it against a rigid underlying structure (the classic example being the “Saturday-night palsy” of the radial nerve). Entrapment is reserved for a nerve passing through an opening too small for it: the corridor tethers the nerve so that it also becomes subject to traction with movement of the adjacent joint.[1] In practice the two coexist, and pure compression is rare; traction, shear and vibration usually compound it.[2]

Pathophysiology of chronic compression

Sustained compression sets off a well-described histological cascade. The earliest change is in the small vessels of the endoneurium and perineurium, with breakdown of the blood-nerve barrier and subperineurial oedema; Renaut corpuscles aggregate at the site; perineurial and epineurial fibrosis follow; segmental demyelination develops, first locally and then diffusely; and, at the end of the process, the unmyelinated fibres and axons themselves are lost.[3] The clinical picture parallels this sequence: intermittent paraesthesia (from episodic ischaemia) precedes constant paraesthesia and weakness (continued ischaemia), which precede fixed numbness and muscle atrophy (axonal degeneration).[4]

Where a nerve is also tethered, joint motion loads it in traction. Function is impaired once a nerve is elongated by about 12%; venous flow is obstructed at around 8% stretch, and frank ischaemia appears at about 15%, the point at which the spiral bands of Fontana (the normal wave-like alignment of fibres) are erased.[5] An entrapped common peroneal nerve is therefore under more tension with the knee straight and less with it flexed, and the cubital tunnel tightens in elbow flexion.

Grading and the double-crush phenomenon

The Seddon and Sunderland gradings of nerve injury apply to compression: a brief insult produces a neurapraxic conduction block (recovery certain), more prolonged compression produces axonotmesis, and severe or neglected compression produces axonal loss with slow, often incomplete recovery.[6]

The double-crush phenomenon of Upton and McComas (1973) holds that a proximal site of compression along an axon makes the nerve more vulnerable to a second, distal compression, the two summating through impaired axoplasmic flow; their original observation was of patients with a cervical root lesion plus a distal entrapment, and a reverse double crush (distal lesion affecting a proximal site) is also described.[7] The concept is not universally accepted. Birch, following Wilbourn and Gilliatt, considers it over-cited, because most quoted examples fail the basic requirement of anatomical continuity between the two sites: Morgan and Wilbourn’s review of 12,736 limbs found an anatomically appropriate coexisting radiculopathy in only 0.8%.[8]

Clinical evaluation

Because compression damages the nerve in a predictable order, the examination is built around what fails first. Threshold sensory tests (vibration with a tuning fork, Semmes-Weinstein monofilaments) become abnormal before innervation-density tests (static and moving two-point discrimination); a two-point discrimination worse than about 8 mm is regarded as non-functional.[9] In the earliest stage all sensory testing may be normal and provocative manoeuvres are the only positive finding.[10] The tools are the Tinel sign (percussion over the nerve producing electric paraesthesiae in its distribution), site-specific positional and compression tests held for about a minute, and the scratch-collapse test (renamed the sensory-collapse test). For the last of these, with the patient resisting isometric shoulder external rotation, a light scratch over the compression site causes a transient loss of strength, and an ethyl-chloride “freeze-out” of the most irritable site can then unmask a second one.[11]

Electrodiagnostics

Nerve-conduction studies and electromyography localise the lesion, distinguish a demyelinating (compressive) from an axonal injury, and help stage severity and prognosis. A compressive/demyelinating lesion shows focal slowing of conduction velocity, prolonged distal latency, and conduction block (a drop in compound muscle action potential amplitude across the lesion), localised by short-segment “inching” studies; an axonal lesion shows reduced sensory and motor amplitudes, with fibrillations and positive sharp waves on needle EMG.[12] Two caveats matter clinically: studies assess only large myelinated fibres, so pain and early compression (which affect small fibres first) can occur with normal studies; and EMG is falsely negative for denervation within about three weeks of an acute injury.[13]

Part II - Median Nerve Compression

Carpal tunnel syndrome

Carpal tunnel syndrome is the most commonly diagnosed compression neuropathy of the upper limb, with an estimated cumulative incidence around 8% and an annual carpal-tunnel-release rate of up to 1.5 per 1000 population; it is commoner in women and usually presents in the fourth or fifth decade.[14]

Anatomy. The carpal tunnel is roofed by the flexor retinaculum (transverse carpal ligament), which spans the hamate and triquetrum on the ulnar side to the scaphoid and trapezium on the radial side; its floor is the carpus. It contains nine flexor tendons (four FDS, four FDP, one FPL) and the median nerve, the nerve lying just deep to the retinaculum in the volar-radial quadrant; the narrowest point is about 2 cm distal to the leading edge, where the nerve shows its morphological change.[15] Two anatomical facts are clinically decisive. The recurrent (thenar) motor branch arises just distal to the retinaculum to supply abductor pollicis brevis, opponens pollicis and the superficial head of flexor pollicis brevis; in the Lanz classification of variations it is extraligamentous in most people but subligamentous (~31%) or transligamentous (~23%) in a substantial minority, and these variations must be respected at surgery.[16] The palmar cutaneous branch arises about 5 cm proximal to the wrist crease and runs superficial to the retinaculum, so it does not pass through the tunnel. That is why sensation over the thenar eminence and central palm is spared in carpal tunnel syndrome, the cardinal discriminator from a proximal (pronator) lesion.[17]

Figure 1. Transverse section of the carpal tunnel: the median nerve lies superficial to the nine flexor tendons, beneath the transverse carpal ligament. Source: DoPhotoShop, via Wikimedia Commons, CC BY-SA 3.0.

Figure 2. Cutaneous territory of the median nerve (palmar and dorsal): thumb, index, middle and the radial half of the ring finger. Source: Dbult, via Wikimedia Commons, CC BY-SA 4.0.

Risk factors and pathophysiology. Diabetes, hypothyroidism, rheumatoid arthritis, obesity, pregnancy and acromegaly are recognised associations, along with forceful, repetitive and vibratory hand use (keyboarding is a weak or doubtful association).[18] Bilateral disease in an older patient should prompt thought of amyloidosis, of which carpal tunnel syndrome may be the presenting sign.[19] The mechanism is raised canal pressure (highest in wrist flexion and extension, lowest in neutral) causing ischaemia and then demyelination.[20]

Clinical features. Patients describe nocturnal paraesthesiae of the radial three-and-a-half digits (Mackinnon calls these “nearly pathognomonic”) and paraesthesiae during fixed-wrist activities such as driving or holding a book, with relief on shaking the hand (the “flick sign”).[21] Advanced disease brings thenar wasting and weakness of thumb abduction. The diagnosis is clinical, confirmed and staged by electrodiagnostics.

Provocative tests. The Tinel sign at the wrist has low sensitivity; the Phalen test (wrist held in 90° flexion for up to a minute) and the Durkan carpal-compression test (direct thumb pressure over the nerve) are more useful, especially combined, though reported sensitivities and specificities vary widely.[22] The sensory-collapse test has high specificity.[23]

Figure 3. Phalen’s manoeuvre: sustained wrist flexion reproduces median paraesthesiae in carpal tunnel syndrome. Source: LittleT889, via Wikimedia Commons, CC BY-SA 4.0 (cropped).

Electrodiagnostics. A distal median sensory latency over 3.5 ms or motor latency over 4.5 ms is taken as positive; sensory amplitudes fall before motor, and EMG denervation of the thenar muscles is a late sign. About 11% of clinically definite cases have normal studies, so a normal study does not exclude the diagnosis.[24]

Conservative treatment. A neutral-position night splint (not the 20-30° extension of an off-the-shelf cock-up splint, which raises canal pressure) is first-line, and a corticosteroid injection gives temporary relief; a good response to injection is a favourable predictor of surgical success.[25]

Surgical treatment. Division of the transverse carpal ligament can be done open or endoscopically. Open release is performed through a short palmar incision placed about 6 mm ulnar to the thenar crease, dividing the ligament on its ulnar border to keep scar off the nerve and to protect the palmar cutaneous and the recurrent motor branches; a crossing cutaneous branch is found in 15-20% of distal incisions.[26] Endoscopic release (single-portal Agee or two-portal Chow) offers slightly less incisional pain and earlier return to work but a higher rate of transient nerve injury; randomised trials show no difference in long-term symptom relief, and both authors personally prefer the open technique.[27] Internal neurolysis, epineurotomy and routine tenosynovectomy confer no benefit.[28]

Figure 4. Open carpal tunnel release: the palmar incision exposes the transverse carpal ligament for division. Source: Dr. Harry Gouvas, via Wikimedia Commons, public domain.

Outcomes and complications. More than 80% of well-selected patients are satisfied, though up to half report some residual symptoms and thenar recovery may take a year; reoperation for persistent or recurrent symptoms runs below 5%.[29] Complications include injury to the median or palmar cutaneous branch (the third-web-space fibres are most at risk), incomplete release, and “pillar pain”, a poorly understood thenar/hypothenar pain after either technique that can take months to settle.[30]

Pronator syndrome (median compression in the proximal forearm)

Far less common than carpal tunnel syndrome, this is compression of the median nerve in the proximal forearm at one of five sites, from proximal to distal: a supracondylar process and the ligament of Struthers (present in under 5% of people), the lacertus fibrosus, the two heads of pronator teres, and the fibrous arch of FDS.[31] The presentation is volar forearm pain worsened by resisted pronation, and the decisive difference from carpal tunnel syndrome is that the palm is not spared, because the compression lies proximal to the origin of the palmar cutaneous branch.[32] Electrodiagnostics are usually normal. Most cases settle with stretching and activity modification; surgical decompression releases each tight structure in turn.[33]

Anterior interosseous nerve (Kiloh-Nevin) syndrome

The anterior interosseous nerve is a purely motor branch supplying flexor pollicis longus, the FDP to the index (and sometimes middle) finger, and pronator quadratus. Its palsy is therefore motor only, painless, with no sensory loss, producing a characteristic inability to make a normal tip-to-tip pinch (the abnormal “OK sign”).[34] It must be distinguished from rupture of the FPL or FDP tendons (an intact tenodesis effect indicates a palsy). Most cases are not a mechanical compression at all but a neuritis (frequently part of Parsonage-Turner syndrome, or neuralgic amyotrophy, heralded by severe periscapular pain) and recover spontaneously; modern imaging shows intrinsic fascicular constrictions in many.[35] Management is observation for several months, with decompression or nerve transfer reserved for those without recovery.[36]

Part III - Ulnar Nerve Compression

Cubital tunnel syndrome (ulnar nerve at the elbow)

Compression of the ulnar nerve at the elbow is the second most common compression neuropathy of the upper limb.[37] The nerve, the terminal branch of the medial cord (C8-T1), is vulnerable at five potential sites: the arcade of Struthers (a fascial band roughly 8-10 cm proximal to the medial epicondyle), the medial intermuscular septum, the condylar groove at the medial epicondyle, Osborne’s ligament (the cubital-tunnel retinaculum), and the deep flexor-pronator (FCU) aponeurosis.[38]

Figure 5. The ulnar nerve passing through the cubital tunnel at the medial elbow, beneath the overlying fascia (Osborne’s band). Source: InjuryMap, via Wikimedia Commons, CC BY-SA 4.0.

Elbow flexion narrows the tunnel and raises intraneural pressure (worsened by wrist extension and shoulder abduction), so sustained flexion, especially during sleep, is a key driver.[39] Causes include old elbow fracture with cubitus valgus (tardy ulnar palsy), osteoarthritis, ganglion, a subluxating nerve and an anconeus epitrochlearis muscle.[40]

Clinical features. There is numbness of the little and ulnar-half ring fingers; crucially, the dorsal ulnar hand is involved in an elbow lesion but spared in a wrist (Guyon) lesion, because the dorsal cutaneous branch leaves the nerve about 9 cm above the wrist.[41] Motor signs include intrinsic wasting (first dorsal interosseous), the claw hand (Duchenne sign), Wartenberg’s sign (abducted little finger), and Froment’s sign (thumb IP flexion by FPL substitution for a weak adductor pollicis on key pinch); the “ulnar paradox” notes that clawing is worse in a distal lesion, because in a proximal lesion the ulnar FDP is also weak.[42] The Jeanne sign (compensatory thumb MCP hyperextension) is a further sign of adductor weakness.[43] The elbow-flexion test with digital pressure and the scratch-collapse test (which localises to Osborne’s band) are the most useful provocations.[44]

Figure 6. Froment’s sign: a weak adductor pollicis is compensated by flexor pollicis longus, flexing the thumb IP joint during key pinch. Source: Athikhun.suw, via Wikimedia Commons, CC BY-SA 4.0.

Figure 7. The ulnar “claw hand”: MCP hyperextension and IP flexion of the ring and little fingers. Source: Mcstrother, via Wikimedia Commons, CC BY 3.0.

Grading and electrodiagnostics. Severity is graded by the motor-based McGowan classification (I, no weakness; II, weakness without atrophy; III, atrophy), and the Dellon classification adds sensory criteria for mild, moderate and severe disease.[45] Electrodiagnostics show conduction slowing across the elbow (a velocity below about 50 m/s, or a fall of more than 10 m/s across the segment), with conduction block defined as an elbow CMAP at least 50% below the wrist value.[46]

Treatment. Mild and moderate disease is managed conservatively (avoiding sustained flexion, with a night extension splint or padding), and about half improve.[47] When surgery is needed the options are in situ decompression (open or endoscopic), medial epicondylectomy, and anterior transposition (subcutaneous, intramuscular, submuscular or, the authors’ preference, transmuscular).[48] The important evidence is that for mild-to-moderate disease without prior elbow injury simple decompression equals transposition, with meta-analyses showing no significant difference; transposition (especially submuscular) is favoured for revision, subluxation and throwing athletes.[49] “Distal kinking of the nerve is the greatest error in transposition surgery,” and injury to the medial antebrachial cutaneous nerve is the commonest technical complication.[50] Outcomes are less reliable than for carpal tunnel release, and failure rates as high as 25% are reported.[51] In severe cases with intrinsic wasting, a supercharged end-to-side (SETS) transfer of the anterior interosseous nerve to the ulnar motor branch can accelerate recovery.[52]

Ulnar tunnel syndrome (Guyon’s canal)

Compression at the wrist occurs in Guyon’s canal, bounded by the volar carpal ligament (roof) and transverse carpal ligament (floor), with the pisiform and the hook of the hamate as landmarks. The Gross-Gelberman zones define the deficit: a zone 1 lesion (proximal to the bifurcation) is mixed motor and sensory, zone 2 (the deep motor branch) is purely motor, and zone 3 (the superficial branch) is purely sensory.[53] The commonest cause is a ganglion, followed by a hook-of-hamate fracture, ulnar artery thrombosis (hypothenar hammer syndrome) and repetitive pressure (“handlebar palsy” in cyclists); the dorsal ulnar sensory territory is spared, localising the lesion to the wrist.[54] Because of the propensity for space-occupying lesions, imaging is important, especially with an isolated motor palsy; treatment is surgical release of the canal, with care to identify the deep motor branch as it curves around the hook.[55]

Figure 8. Guyon’s canal: the distal ulnar nerve and its branches around the pisiform and hook of hamate, with the three compression zones labelled. Source: Jmarchn, via Wikimedia Commons, CC BY-SA 3.0.

Part IV - Radial Nerve Compression

Posterior interosseous nerve (PIN) syndrome

The PIN is the motor continuation of the radial nerve through the radial tunnel, where it can be compressed at four points, from proximal to distal: fibrous bands at the radiocapitellar joint (Capener’s bands), the radial recurrent vessels (the leash of Henry), the fibrous medial edge of ECRB, and, the principal site, the arcade of Fröhse, the tendinous proximal edge of the supinator.[56] A complete PIN palsy produces inability to extend the fingers at the MCP joints and to extend the thumb, with no sensory loss; wrist extension is preserved but deviates radially, because ECRL (innervated proximal to the PIN) is spared while ECU is not.[57]

Figure 9. Wrist drop of a complete (high) radial nerve palsy. In PIN syndrome, by contrast, wrist extension is preserved (deviating radially) because the radial wrist extensors are spared. Source: Wellcome Collection, via Wikimedia Commons, CC BY 4.0.

It must be distinguished from extensor-tendon rupture (the tenodesis effect is preserved in a palsy) and from C7 radiculopathy. Most closed cases are observed for about three months with a splint; persistent palsy is decompressed, releasing each structure to the distal edge of the supinator.[58]

Figure 10. The supinator and the deep branch of the radial nerve (posterior interosseous nerve) at the arcade of Fröhse. From Gray’s Anatomy (1918), public domain, via Wikimedia Commons.

Radial tunnel syndrome

Radial tunnel syndrome shares the same anatomy but presents as pain without motor loss, typically a “resistant tennis elbow,” with tenderness over the mobile wad and supinator rather than the lateral epicondyle (Lister’s localising rule), pain on the resisted middle-finger extension test and on resisted supination.[59] It is a genuinely controversial diagnosis: electrodiagnostic studies are usually negative and it is largely subjective, so reproduction of pain on deep palpation and exclusion of other causes are key.[60] Tennis-elbow straps are avoided because they press on the nerve; persistent cases are decompressed as for PIN syndrome.

Wartenberg syndrome (superficial radial nerve)

Wartenberg’s syndrome, or cheiralgia paraesthetica (Wartenberg, 1932), is compression of the superficial sensory branch of the radial nerve where it emerges between the brachioradialis and ECRL tendons in the distal forearm; the tendons scissor across the nerve as the forearm pronates.[61] Causes are external (tight wristbands, watch straps and handcuffs) or iatrogenic, from injury during de Quervain or thumb-base surgery. There is pain and numbness over the dorsoradial hand with a positive Tinel sign; the Finkelstein test can be positive (it overlaps with de Quervain disease, which it must be distinguished from). Most cases respond to removing the cause, splinting in supination and injection; release is occasionally needed.[62]

Part V - Thoracic Outlet Syndrome

Thoracic outlet syndrome (TOS) is compression of the neurovascular structures (the brachial plexus, subclavian artery or subclavian vein) as they cross the upper thorax, and it is the most contested diagnosis in peripheral-nerve surgery: in one survey nearly 20% of hand surgeons doubted it exists, and the average patient had seen almost five doctors before diagnosis.[63]

Anatomy. Three spaces are described: the interscalene (scalene) triangle (anterior scalene, middle scalene and first rib; the commonest site, transmitting the artery and plexus, while the vein passes anterior to the anterior scalene), the costoclavicular space (between clavicle/subclavius and first rib, narrowed by arm abduction and poor posture), and the subcoracoid/pectoralis-minor space.[64] Predisposing anatomy includes a cervical rib (present in 0.5-1% of people, often bilateral, commoner in women), an anomalous first rib, fibrous bands (Roos described nine) and a scalenus minimus muscle, together with the muscle imbalance of a head-forward, slumped posture.[65]

Figure 11. Neurovascular anatomy of the thoracic outlet: the brachial plexus and subclavian artery pass between the anterior and middle scalenes over the first rib. Source: Nicholas Zaorsky MD, via Wikimedia Commons, CC BY-SA 3.0.

Classification. TOS is vascular (arterial or venous) or neurogenic, the latter subdivided into “true” and “disputed”. The proportions are striking: arterial TOS is only 1-2% and venous (Paget-Schroetter) 2-3%, “true” neurogenic TOS (with objective lower-trunk wasting) is rare, on the order of 1 in a million, and the “disputed” or electrically-negative neurogenic form accounts for well over 95% of cases.[66] True neurogenic TOS produces the Gilliatt-Sumner hand, a thenar-predominant intrinsic wasting from C8-T1 compression, usually associated with a cervical rib or band.[67]

Figure 12. A right-sided C7 cervical rib (arrow), the bony anomaly that predisposes to arterial and true neurogenic thoracic outlet syndrome, often missed on MRI. Source: James Heilman MD, via Wikimedia Commons, CC BY-SA 4.0.

Clinical features and tests. Neurogenic TOS causes shoulder-girdle and neck pain with paraesthesiae (present in up to 95%), usually in the medial arm and ulnar digits, brought on after overhead activity. Arterial TOS threatens the limb with claudication, coolness and emboli and almost always reflects a bony anomaly. Venous TOS is the effort thrombosis of a young athlete, with a swollen, cyanotic arm and prominent collateral veins.[68] The provocative tests are notoriously non-specific and frequently positive in normal people, so reproduction of symptoms matters more than pulse change: Adson (head rotation with inspiration, watching the radial pulse), Roos/EAST (arms abducted and externally rotated, opening and closing the hands for three minutes), Wright (hyperabduction) and the costoclavicular (military-posture) test.[69]

Investigation. A cervical-spine and chest radiograph (the apical lordotic view best shows a cervical rib, which MRI often misses) is part of the work-up; electrodiagnostics are usually normal in neurogenic TOS and are obtained mainly to exclude distal entrapment, though a reduced medial antebrachial cutaneous sensory amplitude is a useful early lower-trunk sign; vascular studies and angiography/venography are reserved for the vascular forms.[70]

Treatment. Neurogenic TOS is managed conservatively first, with a structured physiotherapy programme to restore posture and muscle balance (stretching the tight scalenes/pectoralis, strengthening the scapular stabilisers, while avoiding scalene strengthening), succeeding in 50-80% or more; a scalene block or botulinum toxin can help and predicts surgical response.[71] Surgery (after at least three months) typically combines scalenectomy, first-rib resection where it compresses the plexus, excision of any cervical rib or band, and plexus neurolysis, by a transaxillary or supraclavicular approach; scalenotomy alone has a high recurrence rate (up to 65%), and the combination of scalenectomy with first-rib resection gives the best results (Sanders: 99% versus 57% satisfied).[72] Arterial TOS needs urgent decompression and arterial reconstruction; venous (Paget-Schroetter) TOS is treated by thrombolysis followed by first-rib decompression, stents faring poorly.[73] Surgery succeeds in about 80%, with recurrence of 5-25%, usually within months and usually from incomplete resection.[74]

Part VI - Lower-Limb Entrapment Syndromes

Mackinnon’s overarching plea is that surgeons should bring the same rigour to lower-limb entrapments that they bring to carpal tunnel syndrome, rather than adopting a “wait-and-see” attitude to a foot drop; Birch, by contrast, is markedly more cautious, warning that several of these diagnoses are over-made and that a missed proximal cause or tumour can be disastrous.[75]

Common peroneal (fibular) nerve at the fibular neck

This is the commonest lower-limb mononeuropathy. The nerve is vulnerable because it lies subcutaneously over the fibular neck and passes under the tendinous edge of peroneus longus (the posterior crural intermuscular septum, its main entrapment point), and it is under more tension with the knee extended.[76] Causes include habitual leg-crossing, squatting, plaster casts, weight loss, prolonged surgery, knee dislocation and arthroplasty, and a ganglion from the proximal tibiofibular joint.[77] It produces a foot drop (weak dorsiflexion and eversion) with sensory loss over the lateral leg and dorsum of the foot; an L5 radiculopathy is excluded because tibialis posterior is spared in a peroneal lesion.[78] Conduction block across the fibular head carries a good prognosis; decompression alone restores about two-thirds, better when combined with a tendon transfer for established drop.[79]

Figure 13. The sciatic nerve dividing into the tibial and common peroneal (fibular) nerves; the common peroneal is vulnerable as it crosses the fibular neck. Source: InjuryMap, via Wikimedia Commons, CC BY-SA 4.0.

Tarsal tunnel syndrome

Tarsal tunnel syndrome (Keck, 1962) is entrapment of the posterior tibial nerve and its medial/lateral plantar and calcaneal branches behind the medial malleolus, deep to the flexor retinaculum.[80] Causes are ankle trauma with swelling, a pronated/flat foot, arthritis and the metabolic neuropathies; it produces plantar pain, paraesthesiae and numbness, often worse at night, and is commonly misdiagnosed as plantar fasciitis. A Tinel sign behind the malleolus and a positive scratch-collapse test help, since electrodiagnostics are inconsistent.[81] Treatment is release of the flexor retinaculum and the separate plantar tunnels; Birch regards it as the most over-diagnosed lower-limb nerve lesion and insists a proximal cause and a space-occupying tumour be excluded first.[82]

Figure 14. The tarsal tunnel: the posterior tibial nerve passes behind the medial malleolus, deep to the flexor retinaculum, before dividing into the plantar nerves. From Williams et al., via Wikimedia Commons, CC BY 4.0.

Meralgia paraesthetica

Meralgia paraesthetica (Bernhardt-Roth syndrome) is entrapment of the lateral femoral cutaneous nerve (L2-L3) as it passes near the anterior superior iliac spine and inguinal ligament.[83] It is purely sensory, with pain, paraesthesiae and numbness over the anterolateral thigh, worse on standing and relieved by sitting; obesity, pregnancy, tight belts and iliac-crest graft harvest are typical causes.[84] Most cases are managed conservatively (weight loss, removing compression, injection); persistent cases are decompressed, with results of 60-95%.[85]

Figure 15. Meralgia paraesthetica: the sensory territory of the lateral femoral cutaneous nerve over the anterolateral thigh. By Jochen Lengerke (overlay on a public-domain figure), via Wikimedia Commons, public domain.

Morton’s metatarsalgia (interdigital neuroma)

Morton’s metatarsalgia (Morton, 1876) is a lesion of the common plantar digital nerve, classically of the third web space, where the nerve is tethered and repetitively traumatised under the intermetatarsal ligament. The pathology is in fact chronic nerve compression with perineurial thickening rather than a true neuroma.[86] It produces burning pain radiating into the toes on walking, relieved by rest and by removing the shoe, and, unlike tarsal tunnel syndrome, it does not worsen at night; web-space compression reproduces the pain (the classic Mulder click).[87] Because it behaves as a compression, both Mackinnon and Birch favour division of the intermetatarsal ligament over neurectomy, since cutting the nerve inevitably produces a neuroma.[88]

Figure 16. Morton’s (interdigital) neuroma: the classic location in the third intermetatarsal space (circled). Source: N. Z. Cascianini, via Wikimedia Commons, CC BY-SA 4.0.

Other lower-limb entrapments

The superficial peroneal nerve can be entrapped where it pierces the deep fascia about 12 cm above the lateral malleolus (sensory loss on the dorsum of the foot with pain on forced inversion); the deep peroneal nerve under the extensor retinaculum gives the anterior tarsal tunnel syndrome (first-web-space sensory change with EDB wasting); the saphenous nerve in the adductor (Hunter) canal gives medial knee and leg pain (gonyalgia paraesthetica); and the pudendal nerve in Alcock’s canal gives the “perineal paralysis of the cyclist”.[89]

References

1. Birch, Surgical Disorders of the Peripheral Nerves, p. 98, citing Dyck, Dyck & Engelstad (2005). The median nerve beneath the flexor retinaculum is the archetypal entrapment.

2. Birch p. 291, quoting Comtet (2005).

3. Birch pp. 98-100; Green’s Operative Hand Surgery, ch. 28, pp. 1299-1300 (Patterson, Novak, Mackinnon).

4. Green’s OHS p. 1299; Birch p. 98 attributes the nocturnal paraesthesiae of carpal tunnel syndrome specifically to episodic ischaemia.

5. Birch pp. 100-101, citing Lundborg & Rydevik (1973); Green’s OHS p. 1300.

6. Birch pp. 97-98; a nerve freed from a fracture or joint within roughly 7-10 days can be expected to recover.

7. Green’s OHS pp. 1299-1300; Mackinnon, Nerve Surgery, ch. 2, p. 45. The clinical lesson is to evaluate and, where appropriate, treat all compression sites along a symptomatic limb.

8. Birch p. 223. A useful example of a genuine inter-source disagreement: Mackinnon endorses and extends double crush; Birch is sceptical.

9. Mackinnon ch. 2, pp. 41-45; Green’s OHS pp. 1306-1307.

10. Mackinnon p. 41.

11. Mackinnon pp. 46-48; Green’s OHS p. 1306, which dates the renaming to Abdo et al. (2019). The scratch-collapse test is especially useful with multilevel disease or normal nerve-conduction studies. Examine distal sites before proximal ones, as any positive provocation tends to make subsequent tests positive too.

12. Birch §6.5-6.6, pp. 205-223; Mackinnon ch. 3.

13. Green’s OHS p. 1307; Birch p. 214. A negative study does not exclude a clinically evident compression - these remain clinical diagnoses confirmed, not defined, by electrodiagnostics.

14. Mackinnon ch. 9, p. 230; Green’s OHS p. 1308.

15. Green’s OHS pp. 1308-1309; Mackinnon pp. 228-229.

16. Green’s OHS p. 1308; Mackinnon p. 229.

17. Mackinnon pp. 228, 232; Green’s OHS p. 1314.

18. Green’s OHS pp. 1299-1300; Mackinnon pp. 235-237. Becker’s case-control odds ratios: obesity 2.9, female sex 3.7, diabetes 1.8.

19. Green’s OHS p. 1311.

20. Mackinnon p. 238; Green’s OHS p. 1309.

21. Green’s OHS p. 1308; Mackinnon pp. 232, 237.

22. Mackinnon pp. 232-233; Green’s OHS p. 1305.

23. Mackinnon p. 234 reports sensitivity 64%, specificity 99%.

24. Mackinnon p. 235; Green’s OHS p. 1308. Studies are obtained mainly to stage severity and predict the rate of recovery.

25. Mackinnon pp. 238-239; Green’s OHS p. 1309. A Cochrane review found no benefit for vitamin B6, diuretics, NSAIDs or magnets.

26. Green’s OHS pp. 1312-1313; Mackinnon pp. 241-242.

27. Mackinnon p. 240; Green’s OHS p. 1309.

28. Green’s OHS p. 1313.

29. Mackinnon p. 243.

30. Mackinnon pp. 247-248; Green’s OHS p. 1313.

31. Mackinnon pp. 226-227, 253; Green’s OHS pp. 1314-1315.

32. Mackinnon p. 263; Green’s OHS p. 1314.

33. Mackinnon pp. 263-264; Green’s OHS pp. 1315-1316.

34. Mackinnon p. 263; Green’s OHS pp. 1314-1315.

35. Green’s OHS pp. 1314-1315; Mackinnon pp. 263-265.

36. Mackinnon p. 265 gives an observation window of roughly 5-9 months; Green’s p. 1315 suggests 7-10 months.

37. Mackinnon ch. 10, p. 269; Green’s OHS p. 1319. Compression at the wrist is only about one-twentieth as common.

38. Mackinnon pp. 269-272; Green’s OHS p. 1320. The arcade-of-Struthers distance differs slightly between sources - Mackinnon ~10 cm, Green’s 8 cm.

39. Green’s OHS p. 1319 cites a 55% reduction in tunnel area with flexion (Apfelberg & Larson); Mackinnon pp. 273-274 gives 30-45%.

40. Mackinnon p. 273; Green’s OHS p. 1319.

41. Mackinnon pp. 270, 276; Green’s OHS p. 1316.

42. Mackinnon pp. 276-278; Green’s OHS pp. 1316, 1319.

43. Standard teaching; not present in the Green’s ch. 28 or Mackinnon ch. 10 extracts, which grade only by McGowan.

44. Green’s OHS pp. 1305, 1319; Mackinnon pp. 278-280.

45. Green’s OHS p. 1319 (McGowan); the Dellon scheme is standard but absent from these extracts, which is flagged in the source mining.

46. Green’s OHS pp. 1319, 1321; Mackinnon p. 280, who uses a velocity below 40 m/s as a surgical-trial threshold.

47. Mackinnon p. 280 (Padua); Green’s OHS p. 1321.

48. Mackinnon pp. 280-290; Green’s OHS pp. 1321-1326.

49. Mackinnon p. 291; Green’s OHS p. 1321.

50. Mackinnon pp. 286, 299.

51. Mackinnon pp. 299, 304; Green’s OHS p. 1320.

52. Mackinnon pp. 303-304; Green’s OHS p. 1326.

53. Green’s OHS p. 1316; Mackinnon p. 270.

54. Mackinnon p. 274; Green’s OHS p. 1316.

55. Green’s OHS p. 1318; Mackinnon p. 292.

56. Green’s OHS pp. 1330-1331; Mackinnon pp. 309-311.

57. Mackinnon p. 309; Green’s OHS p. 1329.

58. Mackinnon pp. 311-312; Green’s OHS pp. 1331-1332.

59. Mackinnon p. 311; Green’s OHS p. 1329.

60. Green’s OHS p. 1329; only 19% of Roles & Maudsley’s original 36 patients had EMG changes.

61. Green’s OHS p. 1327; Mackinnon p. 308.

62. Green’s OHS pp. 1327-1328. Mackinnon (p. 307) calls the radial sensory nerve “the most unforgiving nerve in the body” for the misery a neuroma here can cause.

63. Mackinnon ch. 12, p. 329; Green’s OHS p. 1341.

64. Green’s OHS pp. 1344-1346; Mackinnon pp. 330-331.

65. Green’s OHS p. 1346; Mackinnon p. 331.

66. Green’s OHS pp. 1341-1342; Mackinnon p. 332 (“upwards of 98%”). Green’s argues for renaming “disputed” as “electrically-negative neurogenic TOS”.

67. Green’s OHS p. 1342; the eponym describes exactly the thenar-predominant wasting these texts attribute to true neurogenic TOS.

68. Green’s OHS pp. 1341-1342, 1347; Mackinnon p. 334.

69. Green’s OHS p. 1348; Mackinnon pp. 333-336.

70. Green’s OHS pp. 1348-1352; Mackinnon pp. 340-345. SSEPs are not of value.

71. Green’s OHS pp. 1352-1353; Mackinnon p. 346.

72. Green’s OHS pp. 1354, 1360; Mackinnon p. 350. Mackinnon takes a more conservative surgical line, holding that first-rib resection is often unnecessary unless there is bony or vascular pathology or a cervical rib - a genuine source disagreement with Green’s.

73. Green’s OHS p. 1354; Mackinnon p. 350.

74. Green’s OHS pp. 1359-1360.

75. Mackinnon ch. 13, p. 390; Birch pp. 303-305.

76. Mackinnon pp. 364, 385; Birch p. 219.

77. Mackinnon p. 364; Birch p. 219.

78. Mackinnon p. 388; Birch p. 219.

79. Mackinnon pp. 389-391.

80. Mackinnon p. 399; Birch p. 304.

81. Mackinnon pp. 382, 401.

82. Birch pp. 304-305; Mackinnon pp. 399-404. Dellon counters that careful surgery is valuable in diabetic neuropathy - a genuine source disagreement.

83. Mackinnon p. 372; Birch p. 240.

84. Mackinnon pp. 372-373.

85. Mackinnon pp. 374-375. Birch (p. 303) stresses it as a potential red flag for serious underlying disease - he records a pelvic chondrosarcoma presenting as meralgia - and prefers simple decompression over neurectomy.

86. Mackinnon p. 405; Birch pp. 220-221.

87. Mackinnon p. 405; the Mulder eponym is standard teaching but is not named in these extracts, which describe the web-space compression sign.

88. Mackinnon pp. 405-406; Birch p. 305, who has seen 14 patients made worse by excision.

89. Mackinnon pp. 392-404; Birch pp. 221, 303-304.

90. Birch p. 98, 291.

91. Birch pp. 98-100; Green’s OHS p. 1299.

92. Green’s OHS pp. 1299-1300; Mackinnon p. 45; Birch p. 223.

93. Green’s OHS pp. 1308, 1314; Mackinnon pp. 228-232.

94. Mackinnon pp. 232-243; Green’s OHS pp. 1305-1313.

95. Mackinnon pp. 269-278; Green’s OHS pp. 1316-1320.

96. Mackinnon p. 270; Green’s OHS p. 1316.

97. Mackinnon p. 291; Green’s OHS p. 1321.

98. Mackinnon p. 311; Green’s OHS pp. 1329-1331.

99. Green’s OHS pp. 1341-1342, 1354; Mackinnon pp. 332, 350.

100. Mackinnon pp. 364, 388; Birch p. 219.

101. Mackinnon pp. 405-406; Birch p. 305.