Brent Graham

Case Presentation

A 29-year-old woman originally presented with a primary complaint of a sensory disturbance affecting the right hand beginning 1 year previously. This symptom was confined to the distribution of the median nerve, affected the digits primarily, and was particularly prominent at night. Nocturnal splinting of the wrist into slight extension was somewhat helpful in reducing this symptom. Eventually a carpal tunnel release was performed. There was an immediate and dramatic reduction in the symptoms for a period of about 3 months. At this point recurrent symptoms of a burning dysesthesia were experienced in the middle and index fingers. Nocturnal symptoms were inconsistently present. The physical examination showed a well-healed wound in the customary location for a carpal tunnel release. There was tenderness in the area of the wound. Passive flexion and extension of the wrist exacerbated the sensory symptoms. The neurological examination was normal. Electro-diagnostic tests showed some minor abnormalities of conduction across the carpal tunnel, somewhat improved over a previous examination. An exploration and neurolysis of the median nerve was performed and the palmaris brevis muscle was interposed between the nerve and the surgical wound. The wrist was immobilized in slight extension for 2 weeks. The symptoms resolved immediately and did not recur after 1 year of follow-up.

Diagnosis

Recurrent carpal tunnel symptoms, with adhesive neuritis of the median nerve

Characteristic Clinical Presentation

The rate of complication or failure of surgical treatment for carpal tunnel syndrome (CTS) varies in the literature but has been estimated to be as high as 12%. Most of the studies reporting on this topic have major methodological flaws so the validity of these figures is difficult to verify. Our local experience, in a multidisciplinary hand surgery program located at a large tertiary center, is that the failure rate after primary surgical treatment is no higher than 1%. This figure should be considered the acceptable benchmark for the operative management of this condition, regardless of the setting or conditions. Successful primary surgical treatment of CTS depends on an accurate diagnosis of the presenting symptoms. The same is true in the evaluation of the patient presenting with a failure of surgical release.

Failures of surgical treatment can be broadly classified within one of three categories: (1) a persistence of the pre-operative symptoms, (2) perioperative development of new symptoms that are associated with either resolution or persistence of the original symptoms, and (3) recurrence of the same or similar symptoms within a period of weeks or months following initially successful surgical treatment.

Persistence of Preoperative Symptoms

The most common explanation for unresolved symptoms after carpal tunnel release is an initially inaccurate diagnosis of CTS. This issue has not been adequately studied and there are several factors related to the evaluation of CTS that indicate the topic is unlikely to be well studied in the foreseeable future.

The major obstacle is that there remains a substantial lack of consensus, even among experts, on the diagnostic criteria for CTS. The role of electrodiagnostic studies (EDS) in this context is particularly controversial. The use of EDS as confirmatory evidence, or even as the sole criterion for CTS, is a common practice throughout North America. It is frequently assumed that EDS represent the gold standard for the diagnosis of CTS; however, most of the available evidence does not support this concept. In fact, the reliability of EDS is largely unknown, and the validity of this investigation is also difficult to evaluate without a widely accepted gold standard with which to compare. Conditions for performing EDS and thresholds for identifying CTS are known to vary widely among institutions and sometimes even within the same laboratory.

In the assessment of the patient presenting with symptoms unresolved by a carpal tunnel release (CTR), the most important task is to determine the nature of the symptoms prior to surgical treatment and to establish retrospectively the likelihood that these actually represented CTS. When the symptoms appear clearly due to CTS and these symptoms have remained entirely unchanged after CTR, the only reasonable explanations are either that there is an unresolved neuropathy of the median nerve, like Wallerian degeneration, or that there is ongoing compression of the median nerve.

The most common situation in which an adequate CTR has failed to resolve symptoms clearly attributable to CTS is Wallerian degeneration of the median nerve. This is typically seen in elderly patients in whom the extent and duration of median nerve compression have been extreme. Normally there are clinical findings of median nerve denervation in the hand, such as a loss of two-point discrimination or atrophy of the thenar eminence. When these findings are present preoperatively, the goal of CTR should be limited to resolving the reversible component of the sensory loss. It is unlikely that there will be any significant reinnervation of the hand, except in those rare instances where Wallerian degeneration of this nature is encountered in a young person with greater regenerative capacity than that normally expected in the elderly patient.

The middle-aged or younger individual who continues to experience symptoms of CTS after a CTR is almost always affected by ongoing median nerve compression. By far the most common reason for this is simply an inadequate release of the transverse carpal ligament at the time of the original procedure. Examination of the healed operative wound will usually disclose that the incision either was made in an inappropriate location or is of an inadequate length. A frequent finding is a short transverse wound that has been made proximal to the transverse carpal ligament. This approach is often used for a blind subcutaneous release of the carpal tunnel without the aid of endoscopic visualization. The previous operative report should be reviewed, although this seldom provides any helpful detail.

Rarely, there has been an adequate release of the transverse carpal ligament but the median nerve continues to experience significant compression due to an aberrant structure. Some examples of these include proximally located lumbricals within the carpal tunnel, a reversed palmaris longus muscle, and other anomalous subcutaneous muscles located proximal to the transverse carpal ligament. A persistent, large median artery may also lead to continuing compression of the median nerve. The prevalence of each of these abnormalities in the population with CTS is not known, but clinical experience has shown that, even taken together, anomalous anatomical structures are an infrequent cause of median nerve compression.

When an evaluation of the symptoms leading to the failed surgery suggests that CTS is unlikely, attention should be focused on establishing the actual diagnosis. This requires a careful history, taking special care to fully understand the nature and pattern of the symptoms. It is particularly important to draw a clear distinction between the symptom of pain and the complaint of a sensory disturbance like numbness or tingling.

The most common conditions misdiagnosed as CTS are other compressive neuropathies of the major peripheral nerves in the upper extremity or a cervical radiculopathy. These include compression of the proximal median nerve in the forearm, ulnar nerve compression at the cubital tunnel or in the Guyon canal, and compression of the radial sensory nerve in the distal forearm, or C6 or C7 radiculopathy. Most of the time the pattern of sensory symptoms and reflex asymmetry is sufficient to allow these conditions to be distinguished from one another, even in retrospect. The distinction between median nerve compression in the forearm and at the carpal tunnel may be subtle because the only anatomical difference is the involvement by the sensory disturbance of the palm when the compression occurs proximal to the formation of the palmar cutaneous branch. The sensory symptoms associated with proximal median nerve compression tend to have a burning or dysesthetic nature more frequently than do those associated with compression of the nerve at the carpal tunnel.

In addition to nerve compression outside the carpal tunnel, there are other conditions that may be misdiagnosed and treated as CTS. A partial list of these seen in our clinic includes de Quervain tenosynovitis, osteoarthritis of the peritrapezial joints, and chronic nonunion of the scaphoid. In general, these diagnoses are associated with a primary symptom of pain rather than by a sensory disturbance. Distinguishing any of these from CTS would appear to be relatively easy; however, these cases continue to be misdiagnosed as median nerve compression. This is particularly likely to occur when there is an inappropriate emphasis on the results of EDS that may have been obtained during the evaluation of these cases.

New Onset of Symptoms in the Immediate Preoperative Period

Recurrence of Carpal Tunnel Syndrome Symptoms after Carpal Tunnel Release

Late recurrence of CTS symptoms of median nerve compression weeks or months after an apparently successful carpal tunnel release usually indicates a complication of surgical wound healing. The nerve may become adherent to the healing wound in the transverse carpal ligament or even to the skin incision. As these adhesions mature, traction may be placed on the nerve during movements of the wrist into flexion and extension. This may lead to CTS-like symptoms. The superficial location of the nerve within the carpal canal places it at risk for this complication. Although there is no evidence to support a brief period of immobilization after CTR, splinting the wrist in moderate extension minimizes the propensity of the nerve to displace anteriorly into the surgical field if the wrist is flexed during the immediate postoperative period. Leaving the fingers free to move allows the flexor tendons to glide through the carpal tunnel, even if the nerve does not move. This should be sufficient to impede the formation of any adhesions between the median nerve and these tendons.

Diagnosis of this type of adhesive neuritis should rely primarily on the history of the recurrence of a sensory disturbance in the median nerve distribution, within several months of an otherwise successful carpal tunnel release. The sensory symptom may be characterized by a burning or dysesthetic component. Movement of the wrist or digits may worsen the symptoms. In some instances this may lead to the inappropriate conclusion that the condition is activity related. In fact, hand activity may manifest this problem, but the symptoms actually have a physical basis at the site of the previous surgery. This possibility should be considered where the symptoms occur in work environments.

Characteristic findings of CTS, like nocturnal numbness, may not be present. The physical examination should look for evidence of significant contracture at the previous surgical wound, although the wound itself may appear largely unremarkable. There is usually tenderness in the area. Holding the wrist in either flexion or extension may precipitate the symptoms.

The role of EDS in the diagnosis of CTS has already been discussed. The place of this investigation in the evaluation of the patient with unresolved or recurrent symptoms of CTS is important to consider carefully. Because even expert clinical evaluation may not clearly indicate the diagnosis in these complicated patients, EDS adds important additional information in some circumstances. The result of this investigation must be interpreted with care, taking into consideration the previous surgery in the area. Comparison with previous studies obtained preoperatively is only meaningful if the studies are performed in the same laboratory under similar conditions. Even under these circumstances it may be difficult to draw firm conclusions based heavily on the outcome of electrical studies because there may be an effect of surgery itself on the median nerve. Furthermore, the capacity of surgical decompression of the carpal tunnel to induce measurable changes in median nerve function is unknown. It would be expected that relieving pressure on the nerve would lead to more normal electrical function, but factors like the extent of preoperative median nerve compromise and patient age may affect this relationship in a significant way. Studies of this question have been uniformly flawed by the absence of electrical studies in patients who are functioning well after CTR. As usual, results that are inconsistent with the clinical evaluation should be treated with care. The usual bias is to give excessive weight to the results of the electrodiagnostic evaluation and to minimize the clinical findings.

Sometimes advanced imaging techniques, such as magnetic resonance imaging (MRI), may also be helpful, although the yield of new information in this situation is frequently limited. The presence of extensive scarring in the area should be expected following open CTR, even in cases where there are no complications. The MRI appearance of clinically important perineural fibrosis after CTR remains incompletely defined.

Outcome and Prognosis

The patient who is thought to be manifesting symptoms of adhesive neuritis of the median nerve requires an approach unique to this condition. Surgery should be considered when the symptoms are persistent and sufficiently troubling to the patient that intervention is warranted. The long-term outcome of this type of median nerve compression is not known. In particular, it is not known whether median nerve function deteriorates over time as might be expected from significant, longstanding compression in primary CTS. As a result, the indications for treatment in this condition must be dictated by the extent of the patient’s symptoms.

There is no evidence guiding the treatment to minimize recurrent scarring between the median nerve and the healing surgical wound. A 2-week period of immobilization with the wrist in 30 degrees of extension may allow the median nerve to remain within the carpal canal and discourage anterior displacement into the wound. This movement is probably more likely if the wrist is held in flexion, as it is likely to be after surgery, on the volar aspect of the wrist and hand. Interposition of a pedicled muscle flap, such as the palmaris brevis, between the nerve and wound may also diminish the extent of wound scar adhesion to the median nerve. Other local muscle flaps, such as the pronator quadratus and abductor digiti quinti, have also been used for this purpose. These muscle transfers are technically challenging, but they have been reported to be useful in several small clinical series in the literature.