Iatrogenic common peroneal nerve injury during harvesting of semitendinosus tendon for anterior cruciate ligament reconstruction

Background

Autologous hamstring grafts are commonly used for anterior cruciate ligament (ACL) reconstructions. While injury to the infrapatellar branch of the saphenous nerve is a well-documented complication of harvesting the hamstring graft, injury to the common peroneal nerve (CPN) has rarely been described.1 2 The CPN originates from the bifurcation of the sciatic nerve, which occurs at a variable level, from as proximal as before exiting the pelvis, to as distal as 7 cm above the epicondylar line.3 The CPN courses through the posterolateral compartment of the thigh and typically runs along the medial side of the long head of the biceps femoris. In around 20% of cases, the CPN is found deep to the long head and along the medial border of the short head of the biceps femoris.4 The exact anatomical relation and average distance between the CPN and the semitendinosus tendon have, to the best of our knowledge, not been reported in the literature.

Case presentation

A 26-year-old man sustained an isolated ACL rupture of the left knee following a soccer injury. Two months after the injury, he was planned for an ACL reconstruction using an ipsilateral autologous semitendinosus tendon. Surgery was performed by one of the consulting orthopaedic surgeons (AvH) at our institution, at the time of his Knee and Sports Medicine fellowship. The procedure was performed under spinal anaesthesia with an adductor canal block. The tendon was harvested using a mini-incision technique5 (figure 1) (the technique described by Prodromos was later altered to a single posterior mini-incision, as described in the current paper). With this technique, the tendon is palpated at the posteromedial side of the popliteal fossa and is luxated through the skin via a small (±2–3 cm) incision at that level. An open-ended tendon stripper (figure 2) is advanced proximally over the tendon to strip it off of the muscle. A closed-ended stripper is then used to strip the tendon from its insertion on the anteromedial side of the proximal tibia.

Figure 1

Diagram showing the anatomy of the posterior thigh, including the common peroneal nerve and the location of the posterior mini-incision. Reprinted from Chadwick C, Prodromos MD, et al. Posterior mini-incision technique for hamstring anterior cruciate ligament reconstruction graft harvest. Arthroscopy 2005;21:130–137 by the Arthroscopy Association of North America, with permission from Elsevier. a., artery; n., nerve; v., vein.

Figure 2

Detail of the tip of an open-ended tendon stripper as used for the proximal part of the semitendinosus tendon.

In this particular case, on retraction of the tendon stripper that was advanced proximally, a twitch of the ipsilateral foot was noted. Because the graft of the semitendinosus tendon was too short (18.5 cm), a semitendinosus allograft was used instead. The surgical procedure was completed without further complications.

After the patient had returned to the ward, a drop foot was noted on the operated leg. This was explained at that time due to residual effects of the spinal anaesthesia.

Investigations

Neurological examination the next morning showed paralysis of the tibialis anterior and peroneus muscles, with normal strength in the quadriceps, hamstring and triceps surae muscles. Sensation of touch was absent over the medial aspect of the distal thigh, the anterolateral aspect of the leg and the dorsum of the foot. No signs of wound complications, thrombosis or large haematoma were present. An electromyographic (EMG) study was not performed since at our institution we usually only request one, when a consulted neurologist or neurosurgeon advises to do so. Depending on the diagnostic purpose, EMG is usually performed weeks or months after the injury is sustained.6 In the first week, an EMG is primarily useful to localise the site of injury. Distinguishing complete from incomplete nerve injury is most sensitive between 1 week and 2 weeks after the injury is sustained.6 Since the location of the injury was roughly know given the mechanism of injury, an EMG would not have been of added value in the first week.

Differential diagnosis

A CPN palsy proximal to the knee was diagnosed, caused by either an iatrogenic transection, compression by haematoma, neuropraxia of the CPN caused by the tendon stripper or neuropraxia caused by the tourniquet or the leg holder. Both sharp nerve transection and compression by haematoma are indications for surgical exploration within days.6

Treatment

An ankle-foot orthosis was fitted, and after immediate consultation with a neurosurgeon, the patient was transferred to a tertiary referral centre for neurosurgery the same day. Complete loss of function of the CPN was confirmed. On the second day after ACL reconstruction, surgical exploration was performed. The peroneal and the tibial nerves were identified below the level of the bifurcation of the sciatic nerve and tested with electrical stimuli, which showed a complete loss of function of the CPN. No discontinuity of the CPN was seen. An intraneural haematoma, which was causing axonotmesis (damage to the peripheral axons with an intact nerve sheet), was observed roughly at the level of the myotendinous junction of the semitendinosus. An extended neurolysis was performed to decompress the fascicular bundles of the CPN.

Outcome and follow-up

One month after the ACL reconstruction and surgical exploration, the wounds had healed well, knee range of motion was 125°–0°–0°, and the anterior drawer and Lachman test were symmetrical for the left and right knee. Motor function of the peroneal nerve had not improved. There was dysesthesia over the anterolateral side of the lower leg and the dorsum of the foot, for which amitriptyline and pregabalin were prescribed. Return of motor function was first observed 6 months after the ACL reconstruction, with Medical Research Council (MRC) grade-4 muscle power.7 At 1-year follow-up, the result of the ACL reconstruction was satisfactory, with the knee stable in all planes, a full range of motion and no signs of muscle atrophy compared with the contralateral thigh and leg. Some hypoaesthesia remained in the distal anterolateral portion of the leg and the lateral aspect of the foot. Motor function of the tibialis anterior and the peroneus muscles had recovered, showing MRC grade-5 muscle power.