6 Management of Complications of Nerve Decompression Surgery
6.1 Part A: Complications of Carpal Tunnel Release
Daniel J. Nagle
Abstract
Keywords: carpal tunnel surgery, carpal tunnel release, complications
6.1.1 Introduction
Carpal tunnel syndrome is the most common nerve compression in the upper extremity. The prevalence of carpal tunnel syndrome in the United States general population has been estimated to be 3.72%,1 and in 2006 approximately 577,000 carpal tunnel releases were performed.2 Carpal tunnel release surgery provides patients with predictably good outcomes; however, complications do occur infrequently. This chapter will review the diagnosis, treatment, and outcomes of intraoperative and postoperative complications associated with carpal tunnel release.
6.1.2 Intraoperative Complications
Nerve Injury
Nerve injury can be the most devastating of all the intraoperative complications that can occur during a carpal tunnel release. The median nerve trunk, the palmer cutaneous branch and motor branch of the median nerve, the ulnar nerve, and the digital nerves are all at risk. Benson et al3 report that a Medline review of the medical literature from 1966 through 2001 for reports of structural damage to nerves, arteries, or tendons during carpal tunnel release revealed an incidence of transient neurapraxias of 1.45% in endoscopic carpal tunnel release (ECTR) cases and 0.25% of open carpal tunnel release (OCTR) cases. Major nerve injuries (median or ulnar nerve) were reported in 0.13% of ECTRs and 0.10% of OCTRs. A review of 54 publications by Boeckstyns and Sorensen4 found the frequency of irreversible nerve damage to be 0.3% for ECTRs and 0.2% for OCTRs.
Ruijs et al5 and others have demonstrated that the shorter the time from nerve injury to repair, the better the outcome. Also, there is abundant evidence that primary repair of nerve injuries produces superior results when compared to delayed repair using nerve grafts or conduits. Immediate repair allows the surgeon to more accurately identify the fascicular pattern of the nerve and thus aid in the preservation of the correct fascicular orientation during the repair. It is therefore critical that the diagnosis of an intraoperative nerve injury be made at the time of the incident in order to afford the patient the best chance of a good outcome. Ideally, any nerve laceration that occurs during a carpal tunnel release would be identified at the time of the surgery and repaired using standard microsurgical techniques. A delay in the diagnosis renders the repair more complex due to the retraction of the nerve ends, the need for extensive neurolysis, and the need for nerve grafts to bridge the inevitable nerve gap. Given what is stated above, if during surgery there is any doubt whatsoever that there may have been a nerve injury, the surgeon should explore the area to rule out any injury.
The communicating branch of Berrettini carries sensory fibers between the median and ulnar common digital nerves. This communicating branch is present in up to 94% of patients,6 and can run very close to the distal edge of the transverse carpal ligament (TCL), placing it at risk during carpal tunnel release. Injury of the communicating branch of Berrettini can lead to decreased digital sensation usually in the long and ring fingers7 as well as an annoying neuroma in the palm. As with any nerve injury, early repair of a lacerated communicating branch of Berrettini offers the best chance of a good recovery.
Vascular Injury
Vascular injuries are very rarely encountered during carpal tunnel release. Benson et al3 reported the incidence of injuries to the superficial palmar arch to be 0.02% in ECTRs and 0.00% in OCTRs. Zhang et al8 report one hematoma among 1144 mini-OCTRs. The vascular supply to the hand is assured by both the radial and ulnar arteries and their anastomoses in the superficial and deep palmar arches.9 This anatomy is protective of the blood supply to the hand in case of a laceration of the superficial palmar arch. The treatment of an injury to the superficial palmar arch is a function of the arterial anatomy of the patient’s hand. In most cases, simple ligation of the lacerated artery is reasonable. However, in the rare situation in which the laceration leads to vascular compromise of a portion of the hand, primary microsurgical repair of the superficial palmer arch would be indicated.
As many hand surgeons perform carpal tunnel surgery under tourniquet control and do not release the tourniquet until the postoperative compressive dressing is applied, the possibility of an undiagnosed arterial injury exists. Such an injury could lead to a postoperative hematoma which could compress the median nerve and cause significant scarring within the carpal tunnel. Early detection and evacuation of the hematoma is the treatment of choice. However, even after appropriate treatment, Kaltenborn et al10 report that patients who suffered from bleeding complications after carpal tunnel release had significantly worse functional results.
Tendon Injuries
The flexor tendons are in harm’s way during a carpal tunnel release. Fortunately, flexor tendon injuries are quite rare. Benson et al3 reported an incidence of flexor tendon injury in 0.008% of ECTRs and 0.00% of OCTRs. Just as with nerve injuries, a flexor tendon injury should, in general, be identified as quickly as possible and repaired. The surgeon has some latitude however, in regard to the need for tendon repair. If, for example, the superficial flexor (FDS) is lacerated and the flexor digitorum profundus (FDP) to that finger is intact, one could, given that fingers can function quite well without the FDS, opt to not repair the lacerated flexor tendon, thus sparing the patient the rigors and complications associated with a flexor tendon repair. Clearly, clinical judgment is required in every case. Flexor tendon repair in the face of a carpal tunnel release presents unique challenges. The loss of the flexor retinaculum precludes placing the wrist in flexion during the postoperative rehabilitation since placing the wrist in flexion would lead to bowstringing of the flexor tendons and median nerve and compromise the surgical outcome. To overcome this, the flexor retinaculum can be reconstructed at the time of the flexor tendon repair using a rotation flap of the residual flexor retinaculum or with a palmaris longus tendon graft.11,12 Of course, reconstruction carries with it the possibility of persistence or recurrence of the carpal tunnel syndrome.
6.1.3 Postoperative Complications
Persistent Symptoms
Unidentified Nerve Injury
Postoperative paresthesias can pose a vexing diagnostic dilemma. Are the paresthesias due to the severity of the presenting carpal tunnel syndrome? Was there an injury to one of the nerves? If there was a nerve injury, was it simply a neurapraxia or was there a laceration? Was the carpal tunnel release incomplete? Given the fact that the outcome of a nerve repair is directly related to the delay between the injury and repair, it is critical to differentiate between these various causes of sensory or motor dysfunction in the immediate postoperative period. Assuming that the surgeon did not identify any nerve injury at the time of wound closure, the first opportunity to make the postoperative diagnosis of a nerve injury would be in the recovery area. Unless faced with a significant deterioration of the nerve function compared to preoperative testing (in which case, serious considerations would have to be given to immediate exploration of the surgical wound), most surgeons would attribute the abnormal postop exam to the effects of residual anesthesia or neurapraxia, and assume the nerve function will improve by the first postoperative visit. A lack of improvement at the first postoperative visit should trigger some concern that an occult nerve injury has occurred. Clinical signs are, however, not always reliable. Median nerve neurapraxia can lead to significant motor and sensory impairment that can completely resolve. The severity of the carpal tunnel syndrome can render the postoperative diagnosis of a nerve injury difficult. A patient presenting with preoperative decreased sensation and weakness of the abductor pollicis brevis is unlikely to miraculously have resolution of those symptoms at the first office visit. If the surgeon is certain the surgery was well performed and can explain the altered nerve status (advanced carpal tunnel syndrome, neurapraxia), he/she might elect to follow the patient expectantly. Otherwise, advanced imaging including high-resolution ultrasound,13 magnetic resonance imaging (MRI), and/or MR neurogram should be performed as soon as possible to rule out nerve laceration. The ultrasound of the median nerve will typically continue to demonstrate findings consistent with median nerve compression14 in addition to any laceration.15 A lacerated nerve should be repaired as soon as possible to avoid nerve retraction and the need for nerve grafting.
Undetected nerve injury can lead to chronic hand pain, motor and sensory dysfunction, as well as complex regional pain syndrome (CRPS). The symptoms associated with partial injuries to the median nerve and its branches are often dismissed as stemming from a neurapraxia. However, nerve dysfunction and pain that does not improve with time should alert the surgeon to the possibility of an underlying healing nerve laceration. The diagnostic steps listed above including ultrasound and MR imaging should be pursued to clearly identify and characterize the nerve injury. Nerve studies can also be performed if the injury occurred more than 30 days earlier. The delayed treatment of a completely or partially lacerated nerve will typically require neurolysis and nerve grafting.
The diagnosis of postoperative neurapraxia is one of exclusion. Only if the surgeon is confident no intraoperative nerve laceration occurred can a “wait-and-see” posture be adopted. If there is any question regarding the presence of a nerve injury, advanced imaging should be pursued as discussed above. If, however, a neurapraxia is the correct diagnosis, one can be relatively confident that with the passage of time the patient will improve. However, there are a few caveats. Clearly a neurapraxia of a median nerve in a patient with an advanced carpal tunnel syndrome will recover more slowly than a neurapraxia in a patient with a mild carpal tunnel syndrome. Advanced age and diabetes can negatively impact the recovery after any nerve injury including neurapraxia. Recovery after a severe neurapraxia can require months and even years.
The discussion of injuries to the palmer cutaneous branch of the median nerve was not undertaken earlier in this chapter as it is typically not identified until well into the postoperative period. Patients who suffer this injury complain of numbness in the proximal thenar eminence. They also complain of shooting pain at the volar wrist. The clinical exam can include decreased sensation in the distribution of the nerve and also a Tinel’s sign over the resultant neuroma. Injuries to this nerve are relatively rare and are associated with incisions that are placed radial to the palmaris longus near the flexor carpi radialis at the wrist flexion crease. The diagnosis is typically based on the clinical examination, though occasionally high-resolution ultrasound can demonstrate the neuroma.16 A diagnostic local anesthetic block performed proximal to the area of the suspected neuroma can be helpful. The treatment of this problem is surgical. The neuroma of the palmar cutaneous nerve must be identified, and the palmar cutaneous branch followed back to its origin off the median nerve. The neuroma is then buried in soft tissue or bone depending upon the surgeon’s preference. While repair of the palmar cutaneous branch of the median can be performed, due to the typical delay in diagnosis and the distal nature of the nerve injury in the region of its arborization, primary repair is rarely possible.
Incomplete Release
Incomplete release of the TCL can occur with both OCTR and ECTR. The possibility of an incomplete release should be considered when confronted with a patient who has undergone a carpal tunnel release whose symptoms worsen or do not improve after what is thought to be a successful carpal tunnel release. It is rare that a patient who has undergone a successful carpal tunnel release does not notice some improvement in their symptoms within the first couple of days or weeks of surgery. Of course, if the patient presents with a very advanced carpal tunnel syndrome the improvement will be more nuanced and may take longer. A patient that undergoes a carpal tunnel release and is no better or worse after the surgery, assuming there has been no nerve injury (see above), must be considered to have residual compression of the median nerve in spite of what appeared to be an adequate release of the TCL. In this situation, it is reasonable to follow the patient’s progress over a few weeks. If, however, the patient shows no improvement, an ultrasound and possibly an MRI of the carpal tunnel, distal forearm, and palm should be carried out in order to rule out residual compression of the median nerve. While one could perform postoperative nerve studies to confirm the presence of continued or worsened nerve compression, the fact that even after a successful nerve decompression the nerve studies do not improve for several months makes the interpretation of early postoperative nerve studies difficult. X-rays of the wrist should be used to rule out any undiagnosed advanced arthrosis (scapholunate advanced collapse deformity [SLAC] and scaphoid nonunion advance collapse deformity [SNAC]), gout, or pseudogout, all of which can produce abundant synovitis which can erode through the volar wrist capsule and cause continued median nerve compression. Should the above-mentioned diagnostic tests fail to clearly identify a site of residual compression, a diagnostic steroid injection can be helpful. If the patient’s symptoms temporarily improve after injection, exploration of the carpal tunnel should be undertaken. If the patient experiences absolutely no improvement after the steroid injection, another source of the patient’s symptoms should be considered (e.g., cervical radiculopathy, brachial plexopathy, thoracic outlet syndrome, pronator syndrome, diffuse peripheral neuropathy, multiple sclerosis, Waldenstrom’s macroglobulinemia17).
The exploration of the median nerve for any sign of residual compression should start in the distal forearm taking care to open the distal forearm fascia. The nerve should be dissected through the carpal tunnel all the way into the palm visualizing the median nerve and all its branches. Postoperative splinting with the wrist in mild extension should be used to avoid the development of bowstringing of the flexor tendons and median nerve.18
Occasionally, the median nerve will be severely scarred and simple decompression will not be sufficient and more complex treatment may be required. Such treatment can include wrapping the nerve with a vein graft,19 the use of local rotation flaps such as a hypothenar fat pad flap,20 and local muscle flaps utilizing the abductor digiti minimi,21 the pronator quadratus,22 or the palmaris brevis.23 Synovial flaps have been used as well.24 Pedicled flaps such as the reverse radial forearm adipofascial flap,25 and even free omental transfer26 have been used with some success.
Typically, the identification and decompression of the residual median nerve compression leads to good recovery. Stang et al27 reported that 84% of patients who underwent surgery for incomplete release of the carpal tunnel were satisfied with the outcome of the surgery. However, 78% of patients complained of persistent neurological symptoms.
Flexor Tendon Complications
Postoperative complications related to the flexor tendons are rare. There are two major complications that can be encountered. One is the subluxation of the ulnar flexor tendons over the hook of the hamate. The other is bowstringing of the flexor tendons.
Subluxation of the flexors over the hook of the hamate may be the consequence of the TCL incision being carried out through or very close to its bony attachment on the hook of the hamate without leaving a remnant of the ulnar edge of the TCL to inhibit flexor tendon subluxation. Symptoms consist of a popping sensation along the volar aspect of the wrist when the wrist is flexed and ulnarly deviated while the patient makes a fist. If this is noticed early in the postoperative period, splinting the wrist in extension for several weeks can be curative. Should nonoperative treatment not succeed, excision of the hook of the hamate should be considered.28
Bowstringing can occur after an extended carpal tunnel release in which the distal forearm fascia and TCL are completely released combined with mobilization of the median nerve out of its bed. The palmar displacement of the flexor tendons pushes the median nerve out of the carpal tunnel into the subcutaneous tissue. The subcutaneous position of the median nerve renders it susceptible to trauma. Even light pressure applied to the displaced median nerve will cause paresthesias and pain. Just as with subluxation of the flexor tendons over the hook of the hamate, if the bowstringing is noted early during the postoperative period, splinting with the wrist in extension can be helpful. However, if this conservative approach does not solve the problem, reconstruction of the TCL can be performed using either residual TCL or a free palmaris longus graft.11,12
Rarely, postoperative scarring of the flexor tendons can lead to “triggering” of the enlarged scarred tendons as they pass beneath the distal edge of brachial fascia. This entity should not be confused with the slightly more common “trigger wrist” which is associated with pathology affecting the flexor tendons in patients with an intact TCL, the treatment of which includes carpal tunnel release.29 The treatment of a post-carpal tunnel release “trigger” wrist consists of the surgical release of the distal aspect of the antebrachial fascia as well as release of the neo-TCL that will have formed after the prior carpal tunnel release. Postoperative splinting should be used to prevent bowstringing of the flexor tendons and median nerve.
Trigger fingers are often noticed after carpal tunnel release but are not, strictly speaking, a complication of carpal tunnel release surgery. King et al30 reported the incidence of trigger finger in 1185 patients during the first 6 months after carpal tunnel release to be 6.6% on the operated side as opposed to 3.5% in the unoperated contralateral hand. The thumb was the most frequently involved digit. The incidence of triggering in the index, long, ring, and small fingers was similar in the operated and nonoperated hands. Lin et al31 noted the increased incidence of trigger finger after carpal tunnel release primarily during the first 6 postoperative months. The treatment for a trigger finger is well established and includes steroid injections and, if needed, release of the A1 pulley.
Infection
The likelihood of developing an infection after a carpal tunnel release has been reported to be as low as 0.32%.32 Werner et al32 noted several independent positive risk factors for infection following an open carpal release, including younger age, male sex, obesity, tobacco use, alcohol use, diabetes, inflammatory arthritis, peripheral vascular disease, chronic liver disease, chronic kidney disease, chronic lung disease, and depression. Werner et al in another study33 noted an increase in surgical site infection after OCTR in patients with poorly controlled diabetes and a perioperative (within 3 months of surgery) HbA1c of greater than 8 mg/dL. While multiple studies34,35 suggest that prophylactic antibiotics are of no benefit when used for procedures such as carpal tunnel release, some authors have suggested prophylactic antibiotics should be considered when performing a carpal tunnel release in diabetic patients.36
The surgeon must be ever vigilant regarding patient complaints of increasing pain and erythema at the surgical site. Even a simple stitch abscess can rapidly devolve into a deep infection. Early treatment including stitch removal and antibiotics will often lead to the resolution of a superficial infection. In the unfortunate case in which a deep infection develops, the treatment includes incision and drainage and appropriate IV antibiotic therapy. Unfortunately, in such situations the patient is likely to develop extensive scarring of the flexor tendons and median nerve that could possibly require tenolysis, neurolysis, and vascularized soft tissue coverage.
Pillar Pain
“Pillar” pain is characterized by discomfort over the “pillars” of the hand centered over the hook of the hamate and the unciform process of the trapezium. Pillar pain is typically not noted immediately after surgery but becomes apparent as the patient applies pressure to the palm while pushing off to rise from a chair or when doing a pushup. Larsen et al37 noted that 73% of 90 patients undergoing standard OCTR, mini-OCTR, and ECTR (30 patients in each group) developed pillar pain. The operative technique had no impact on the incidence of pillar pain. The etiology of pillar pain has been thought to possibly be related to changes in the carpal anatomy though the exact etiology is still unknown.38 The patient presenting pillar pain should be counseled to avoid aggravating activities though this can delay the return to work. The use of the antioxidant alpha-lipoic acid (ALA) for 40 days has been reported to reduce pillar pain in a small cohort of patients.39 Extracorporeal shock wave therapy has also been shown in a small controlled study by Haghighat et al40 to be beneficial in the treatment of pillar pain. Fortunately, pillar pain almost always resolves spontaneously during the first few postoperative months and reassuring the patient of this fact is typically the only treatment needed.
Pisiform Pain
Occasionally, patients develop discomfort at the pisotriquetral joint after carpal tunnel release. Patients who develop this problem will complain of discomfort with direct compression of the pisotriquetral joint. Stahl et al41 have opined this is related to latent pisotriquetral arthritis that is unmasked as a result of a change in the biomechanics of the pisotriquetral articulation noted after carpal tunnel release. The transection of the TCL is thought to lead to a loss of the radial restraint of the pisotriquetral joint, which in turn leads to maltracking of the pisiform and aggravation of the underlying arthrosis. Stahl et al41 reviewed the charts and X-rays of 700 patients who had undergone either OCTR or ECTR and identified 14 cases of pisotriquetral dysfunction and arthrosis. Both the diagnosis and treatment can include a steroid injection carried out under fluoroscopic control. If the steroid injection does not lead to long-term relief, excision of the pisiform is typically curative.
Scar Tenderness
The incidence of scar tenderness after carpal tunnel release has been reported to be between 19% and 61%.42,43 ECTR appears to be associated with less scar pain than OCTR.44 The possibility that scar tenderness after OCTR is related to injury of the terminal branches of the palmar cutaneous branches of the median and ulnar nerves has been suggested. Siegmeth et al45 performed a prospective randomized study comparing postoperative scar pain in two cohorts of 42 patients each. One cohort underwent an OCTR during which no attempt was made to preserve the superficial nerve branches. The other cohort underwent a similar OCTR during which the superficial nerve branches were meticulously protected. The authors noted no difference in scar pain between the two methods at 6 weeks, 3 months, and 6 months. The treatment of scar tenderness can include postoperative scar massage (once the wound is healed), and the application of silicone padding and steroid cream. The avoidance of heavy hand use during the early healing phase is also beneficial.
Complex Regional Pain Syndrome
Complex regional pain syndrome (CRPS) is encountered in 2 to 8% of patients who have undergone carpal tunnel release regardless of the surgical technique. Post carpal tunnel release, CRPS is more frequently noted in middle aged (40–64 y) women.46 The hand surgeon must consider CRPS in the differential diagnosis of patients who are noted to have postoperative pain out of proportion to what one would expect. Also, CRPS should be considered in patients who seem to be doing relatively well after surgery but subsequently develop pain and swelling for no apparent reason. The hand surgeon can treat early cases of CRPS with a 2-week course of oral steroids as well as frequent gentle therapy. Should this initial treatment not lead to the resolution of the patient’s symptoms, referral to a pain clinic for more aggressive treatment should be considered.
Prolonged Neurological Symptoms
Elderly patients who present with very advanced carpal tunnel syndrome can occasionally develop postoperative nerve pain that seems to be associated with the gradual recovery of median nerve function. This is a relatively rare phenomenon. The treatment of this can include the use of gabapentin as well as duloxetine, though the mainstay of treatment is constant reassurance that the symptoms will resolve. Serial light touch localization and two-point discrimination testing and nerve studies can provide the patient with “objective” encouraging evidence of nerve improvement. Unfortunately, the recovery can be very slow and require a year or two.
6.2 Part B: Complications and Failure of Cubital Tunnel Surgery
Godard C.W. de Ruiter
Abstract
In cubital tunnel surgery, various procedures are available to the surgeon to choose from, ranging from simple decompression to different techniques of ulnar nerve transposition. In this chapter, an overview of the different complications for the various procedures is provided. In addition, technical issues associated with these procedures are discussed, as well as possible reasons for failure (persistence or recurrence of symptoms) and different options for revision surgery. Finally, treatment recommendations are presented for the management of the different complications and practical tips are provided for the detection and surgical treatment of potential anatomical variations.
Keywords: ulnar nerve, neuropathy, elbow, decompression, transposition, subcutaneous, submuscular, neuroma
6.2.1 Introduction
Cubital tunnel syndrome (CuTS), also known as ulnar neuropathy at the elbow (UNE), is the second most common neuropathy after carpal tunnel syndrome. The severity of symptoms can be classified as mild, moderate, or severe, according to the classification introduced by Dellon.47 Conservation treatment, consisting of pain medication, the elimination or reduction of the frequency of external compression on the nerve and sometimes a night splint, often is the first treatment especially in mild cases. Surgery is offered to the patient in case conservative treatment fails or in case the patient presents with moderate-to-severe symptoms.47
6.2.2 Surgery for Ulnar Neuropathy: Different Procedures and Potential Complications
There are various surgical procedures available to the surgeon and patient to choose from, ranging from simple and endoscopic decompression to different techniques of ulnar nerve transposition (subcutaneous, submuscular, and intramuscular).48 These procedures all have different pros and cons. In the Cochrane review from 2016 no difference in outcome was found for the various procedures (clinical improvement averaged around 70%), but a higher chance for complications following transposition, especially the risk on superficial and deep wound infection.48 Simple decompression, therefore, nowadays is the first surgical treatment of choice, but in some cases (e.g., in case of luxation or severe neurologic deficit) the surgeon can decide, together with the patient, to directly transpose the ulnar nerve in addition to the decompression. Below, the different surgical treatments for CuTS are discussed and specifically the complications associated with these procedures and potential reasons for failure. In the second part of this chapter (6.2.3) options for revision surgery are discussed. Finally, in the last part (6.2.4) recommendations are provided to prevent and manage complications, and practical tips for the detection and surgical treatment of anatomical variations.
Complications after Decompression of the Ulnar Nerve
The most standard procedure in CuTS is simple decompression of the ulnar nerve. A slightly curved incision is made between the medial epicondyle and the olecranon (Fig. 6‑1) to identify the nerve. Subsequently, potential compression sites are inspected (most often Osborne’s ligament or the aponeurosis of the flexor carpi ulnaris [FCU] muscle). The most frequent side effect of this procedure, which is often probably not even reported as complication, is numbness around the elbow due to injury to the posterior branches of the medial antebrachial cutaneous nerve (MABCN). This might be caused by traction, which most likely will recover slowly in time, but in severe cases (complete transection) sometimes a neuroma might develop which can be quite debilitating to the patient (Fig. 6‑2). The risk for the development of a MABCN neuroma after simple decompression is not exactly known, but in a study by Mackinnon and Novak in 73 out of 100 cases that underwent revision cubital tunnel surgery, a neuroma of the MABCN was found.49 Management of MABCN neuroma will be discussed in the final part of this chapter (6.2.4), but it is important to realize that, even in the absence of a neuroma, pain/numbness around the elbow can substantially reduce patient satisfaction with the outcome, despite the relief of tingling sensation in the hand, and therefore care should be taken during the exposure of the ulnar nerve to prevent MABCN injury.
Fig. 6.1 Anatomical drawing illustrating the course of the medial antebrachial cutaneous nerve (MABCN) that divides into a posterior (PB) and anterior branch (AB). Most commonly the posterior branch crosses the usual site of incision (dashed black line) 2 cm distally to the medial epicondyle (ME), but in about 23% of the cases crossing branches are also found proximal to or at the level of the ME (as in this drawing).50