TKA Neurovascular Injury

Neurapraxia is the commonest after TKA. A conduction block from compression or ischaemia leaves axon and connective tissue intact, so there is no Wallerian degeneration, and sensation is often spared more than motor function. Recovery is expected over days to weeks (up to about 12 weeks); it is the basis of most traction and compression palsies that recover with dressing release and knee flexion.

Axonotmesis disrupts the axon but leaves the endoneurial tubes intact, and Wallerian degeneration occurs distally. It becomes detectable on EMG/NCS once degeneration is established, at 3-4 weeks. Recovery depends on axonal regrowth at roughly 1mm/day, so it takes months and is often incomplete.

Neurotmesis is complete transection or dense intraneural scar, with no spontaneous recovery. It is rare in TKA, and a transection recognised intra-operatively is an indication for repair. Otherwise it is distinguished from axonotmesis only by failure of recovery or by operative findings.

Most post-TKA palsies are Sunderland I-II and recover. Failure to improve clinically and electrically by 3 months signals a higher grade and prompts consideration of exploration or neurolysis.

A practical, exam-friendly way to classify the palsy is by timing of recognition, which carries prognostic weight.

• Immediate, recognised in recovery: suggests intra-operative traction or transection, or tourniquet ischaemia - generally a worse prognostic group.
• Delayed, after the block wears off or the next day: classically attributed to a constrictive dressing, haematoma, or limb positioning under epidural analgesia - more amenable to dressing release and repositioning, often with a better prognosis.

Idusuyi and Morrey highlighted the high frequency of delayed presentation and its association with epidural analgesia and impaired proprioception.

Preoperative valgus is the best-supported risk factor, but resist quantifying it. The largest series (10,361 TKAs) found the association highly significant (p less than 0.0001), with 10 of its 30 palsy patients having valgus of 12 degrees or more, yet it never published a relative risk, and no study cited on this page supports a numerical multiplier such as "3-4 times" or "8-10 times". Say the risk rises steeply with the magnitude of valgus and that a large correction should change your plan; do not attach a number the literature does not contain.

And it is contested. A smaller series (1,476 TKAs, 19 events) failed to reproduce the association at all. That is a study underpowered to detect it rather than one disproving it, but it is why "valgus increases risk" is the defensible claim and any figure attached to it is not.

Why correction matters. A lateral soft-tissue contracture requires extensive release, and correction "lengthens" the lateral structures, the peroneal nerve among them. Valgus correction greater than 10-15° significantly increases nerve palsy risk through lateral soft-tissue stretching; staged correction or a constrained implant are the alternatives to consider.

Flexion contracture. Greater than 20° of fixed flexion does have a real number behind it: an odds ratio of 3.9 (p = 0.002) in 1,166 knees, from the tourniquet study cited below, where the valgus multipliers have none. The posterior capsule release it requires is close to the popliteal vessels, and restoring extension elongates the neurovascular structures.

Fixed valgus with a flexion contracture is the highest-risk combination for peroneal palsy: the lateral release lengthens the nerve.

Varus. The nerve palsy risk is lower than with valgus and the vascular injury risk similar; medial release does not tension the peroneal nerve.

Rheumatoid arthritis carries a 3-5× increased peroneal nerve palsy rate. The soft tissues are fragile, pre-existing neuropathy is common, synovitis may obscure the anatomy, and it is often combined with valgus deformity.

Peripheral vascular disease. Atherosclerotic vessels are less tolerant of manipulation, calcification alters the surgical planes, and the tourniquet carries a thrombosis risk. Pre-operative vascular assessment is recommended; consider a no-tourniquet technique.

Diabetes mellitus brings pre-existing neuropathy, which may mask symptoms, impaired nerve recovery and microvascular disease, with a higher threshold for concern post-operatively.

Peripheral neuropathy of any kind increases vulnerability: a lower threshold for injury and poorer recovery potential.

Previous knee surgery leaves altered anatomy and scarring, and adhesions may tether the neurovascular structures. It carries a higher revision surgery risk; consider pre-operative imaging (MR angiography).

Revision surgery carries a 2-3× increased nerve palsy rate. Scar tissue distorts the planes, component extraction applies force, and bone loss may require augments (increased height).

Tourniquet. Direct compression injury is possible, and there is an ischaemic component to the nerve injury.

Technique. Aggressive retractor placement, excessive tissue tension, inadequate haemostasis leading to haematoma, and tight dressings or splints all contribute.

Motor. Foot drop, the inability to dorsiflex the ankle, with weakness of the toe extensors (EHL, EDL) and of eversion (peroneus longus and brevis). The gait shows foot slap and a high-stepping steppage pattern to clear the foot.

Sensory. Numbness over the dorsum of the foot, in the first web space (deep peroneal territory) and over the lateral leg (superficial peroneal territory). The lateral foot (sural nerve) may be spared.

When it appears. Regional anaesthesia may mask it initially, and it is typically recognised when the block wears off, at 12-24 hours. An outpatient must be examined before discharge, and ankle dorsiflexion power documented systematically.

The examination grades and maps the deficit:

• Ankle dorsiflexion (L4-5), graded 0-5
• Great toe extension (L5), graded 0-5
• Ankle eversion (L5-S1), graded 0-5
• Sensation in the first web space (deep peroneal)
• Sensation over the dorsum of the foot (superficial peroneal)

Motor. Weakness of ankle plantarflexion (gastrocnemius, soleus), toe flexion (FHL, FDL) and the foot intrinsics, with difficulty at push-off during gait.

Sensory. Numbness of the sole and altered sensation over the plantar surface of the toes, sometimes with burning or dysaesthesia.

Context. Less common than peroneal nerve injury, it is often associated with haematoma or direct trauma, and may have a vascular component given the nerve's proximity to the popliteal vessels.

Features. Combined peroneal and tibial deficits: total below-knee motor loss, foot drop with loss of plantarflexion, and sensory loss over the entire foot except the medial side (saphenous).

Causes. A proximal injury is rare in TKA. A sciatic palsy after TKA should prompt consideration of hip pathology and of stretching of the nerve at the sciatic notch, and an epidural haematoma or abscess must be excluded.

1. Recognise and document: motor power (MRC grade 0-5) and the sensory examination, compared with pre-operative status if available, and a photograph of any skin changes.
2. Remove compression: loosen all dressings immediately, remove circumferential bandages, release any splint, and make sure nothing presses on the fibular head.
3. Position: flex the knee to 30-45° to reduce nerve tension, avoid external rotation of the leg, pad the fibular head, and elevate the leg slightly.
4. Rule out other causes: epidural haematoma or abscess (back pain, bladder, bilateral signs), compartment syndrome (compartment tension) and vascular injury (pulses, perfusion).
5. Document and communicate: inform the patient and family, record it in the notes, refer early to physiotherapy, and consider early surgical exploration if a compressive haematoma is a concern.

Conservative management is the standard. Continue knee flexion positioning, fit an ankle-foot orthosis early for safe mobilisation and to prevent equinus contracture, and examine weekly. Physiotherapy maintains ankle range of motion to prevent contracture, strengthens the available muscles and trains gait with the AFO.

Investigate with NCS/EMG at 3-4 weeks as a baseline, MRI if a haematoma is suspected, and repeated neurological examination.

Surgical exploration is indicated for:

• A large haematoma on imaging causing compression
• A complete palsy with no recovery by 2-3 weeks
• Nerve transection recognised intra-operatively
• An associated vascular injury requiring surgery

Continue conservative care: regular physiotherapy, the AFO for walking, and serial clinical and EMG assessment.

Watch for recovery. The first signs are twitching in tibialis anterior, and sensation returns before motor function. Document the improvement in MRC grade; EMG at 3 months gives the prognosis.

No recovery. When EMG shows no reinnervation at 3-6 months and there is no clinical improvement, consider the surgical options: neurolysis, nerve repair or tendon transfer, described under Surgical Technique.

Indication. No clinical or electrical recovery by about 3 months, or a documented compressive lesion.

Approach. Supine or lateral, with a tourniquet optional and used judiciously. A lazy-S or curvilinear incision over the fibular neck, behind the biceps femoris tendon, finds the common peroneal nerve posterior to biceps femoris, and it is traced distally around the fibular neck.

Key steps at the fibular neck:

• Release the fibrous arch at the origin of peroneus longus, the classic constriction point
• External neurolysis of epineurial fibrosis, decompressing the superficial and deep branches
• Intraneural neurolysis only if a discrete fascicular constriction is found
• Preserve the nerve's vascular supply and avoid devascularising mobilisation

Rationale. Mont and colleagues found 30/31 patients (97%) improved and discontinued the AFO after decompression, against 3/9 (33%) managed non-operatively, supporting decompression when non-operative measures fail.

Pitfalls. Operating too early (most Sunderland I-II palsies recover spontaneously), missing a more proximal sciatic lesion, and iatrogenic injury during dissection at the fibular neck.

Principle. Restore perfusion within the ischaemic window; simple thrombectomy alone is frequently insufficient. Endovascular options (stent graft, covered stent, coil embolisation of a pseudoaneurysm) are increasingly first-line at experienced centres.

Open options depend on the lesion:

• Intimal flap or thrombosis: thrombectomy with or without a patch; often requires bypass
• Transection or laceration: end-to-end anastomosis, interposition or reversed vein graft (popliteal/below-knee bypass)
• Pseudoaneurysm or AVF: open repair or endovascular exclusion

Adjuncts. Four-compartment fasciotomy where reperfusion follows prolonged ischaemia.

Evidence. The Pennsylvania Hospital series reported a shift towards endovascular treatment over time with equivalent limb salvage, and emphasised that thrombectomy alone rarely sufficed for ischaemic complications.

Indication. A permanent, non-recovering foot drop, typically after 12 months without recovery and with no further regeneration on EMG.

Posterior tibial tendon transfer (Bridle procedure). Tibialis posterior is harvested and routed anteriorly (interosseous or circumtibial) to the dorsum of the foot, then fixed to the lateral cuneiform or midfoot with the foot in neutral dorsiflexion. It restores active dorsiflexion and removes AFO dependence in selected patients.

Prerequisites. A supple ankle without fixed equinus contracture (which may need adjunctive lengthening), good tibialis posterior power, and a compliant, motivated patient.

Alternatives. An AFO indefinitely gives acceptable function; arthrodesis is for fixed deformity.

Retractors. A lateral retractor on the fibular head compresses the peroneal nerve directly, so place it on the proximal tibia, and do so with the knee flexed. Retract gently and intermittently, avoiding prolonged pressure, and consider visualising the nerve in high-risk cases.

Correcting the deformity. Excessive valgus correction is a traction injury to the peroneal nerve. Limit valgus correction to 10-15° in a single stage; for severe valgus consider a constrained implant to reduce the release, stage the correction for extreme deformity, and accept slight under-correction rather than a nerve palsy.

Tourniquet. Prolonged tourniquet time causes ischaemic nerve injury. A wider cuff (10cm minimum) distributes the pressure, the inflation pressure should be lower (limb occlusion pressure + 100mmHg), duration is limited to less than 90-120 minutes, and tourniquet-free TKA is worth considering in high-risk patients.

Closure and dressings. Tight closure and dressings cause compression and compartment syndrome. Avoid excessive tension on closure and circumferential tight bandages, use loose, well-padded dressings, and monitor post-operatively.

Posterior work. An aggressive posterior release injures the popliteal vessels directly. Keep the knee flexed during posterior work, which moves the vessels away, use bent Hohmann retractors behind the tibia and a subperiosteal dissection technique, remove posterior osteophytes carefully under direct vision, and avoid blind instruments posteriorly.

The tibial cut. Saw over-penetration lacerates the posterior vessels. Control the depth of the cut with an oscillating saw fitted with a guard or depth gauge, and warn the assistant to protect posteriorly. Extramedullary guides may be safer than intramedullary.

Haemostasis. A haematoma compresses neural structures. Release the tourniquet before closure to identify bleeding, achieve meticulous haemostasis, address genicular vessel bleeding carefully, and consider a drain in complex cases.