Carpal Tunnel Syndrome: Current ConceptsRobert R. Slater, Jr., MD, Hand and Upper Extremity Surgery Service, Department of Orthopaedic Surgery, University of California, Davis.
[J South Orthop Assoc 8(3), 1999.
© 1999 Southern Orthopaedic Association]
AbstractCarpal tunnel syndrome (CTS) is a clinical syndrome manifested by characteristic signs and symptoms resulting from an entrapment neuropathy of the median nerve at the wrist. It is the most common compression neuropathy in the upper extremity. In this paper, the etiology and pathophysiology of CTS are reviewed, as well as the clinical examination and other tests that may be useful in establishing its diagnosis. A variety of surgical techniques have been espoused for treating CTS refractory to non-operative treatment and the proposed advantages and disadvantages of the new procedures are discussed. If the diagnosis is correct, then surgical results are reliably good.
IntroductionCarpal tunnel syndrome (CTS) is a constellation of signs and symptoms resulting from compression of the median nerve in the carpal tunnel. Paget first described its clinical manifestations in 1854 in a patient who sustained a fracture of the distal radius. In 1913, Marie and Foix described the pathologic changes noted in the median nerve after long-term compression in an 80-year-old patient with thenar atrophy. Moersch provided the first description of spontaneous nerve compression in 1938 and is credited with coining the term "carpal tunnel syndrome." Learmonth described the first surgical release of the median nerve in 1933 in a patient whose nerve was compressed by posttraumatic osteoarthritic osteophytes. Nonetheless, it is Phalen[5-8] who deserves most credit for popularizing this disorder and calling it to the attention of the medical community with a series of publications beginning in 1950. The purpose of this article is to review and summarize the etiology and diagnosis of this disorder and discuss options for its treatment.
AnatomyThe carpal tunnel is clearly defined anatomically. The palmar radiocarpal ligament and the palmar ligament complex between the carpal bones form the floor of the canal. The roof is formed by three continuous segments of flexor retinaculum: the thin proximal segment is the deep investing fascia of the forearm; the transverse carpal ligament proper inserts on the scaphoid tuberosity and part of the trapezium radially, and on the pisiform and the hook of the hamate ulnarly; the distal portion is the aponeurosis between the thenar and hypothenar muscles. The tunnel contains nine tendons -- the flexor pollicus longus tendon, four flexor digitorum superficialis tendons, and four flexor digitorum profundus tendons -- along with the median nerve which lies most superficially, immediately under the ligament.
Pathophysiology and EtiologyThe basic pathophysiology of nerve compression common to all the specific causes has been well described.[10,11] The initial insult is a reduction in epineural blood flow, which occurs with 20 to 30 mm Hg compression. Intracarpal canal pressures in patients with CTS routinely measure at least 33 mm Hg and often up to 110 mm Hg with wrist extension. Continued or increased pressure eventually causes edema in the epineurium and endoneurium. If applied for 2 hours, pressure of 50 mm Hg will cause epineural edema, and if applied for 8 hours, it will increase endoneural fluid pressure fourfold and block axonal transport. As further injury occurs to the capillary endothelium, more protein leaks out into the tissues, which become more edematous, and a vicious cycle ensues. The effects are most pronounced within the endoneurium, since more exudate and edema accumulate there, unable to diffuse across the perineurium. The perineurium resists pressure changes because of its higher tensile strength and acts as a diffusion barrier creating in effect a "compartment syndrome" within the nerve.
Carpal tunnel syndrome occurs more commonly in women than men[14,15] and is most common between the ages of 30 and 60 years. Anything that compromises the space available for the median nerve in the carpal tunnel can cause CTS. Local structural changes and masses at the wrist are known causes, including distal radius fractures, blunt trauma with associated hemorrhage and swelling, and tumors such as lipomas and ganglion cysts. A wide variety of systemic illnesses, metabolic diseases, overuse syndromes, and aberrant anatomic structures also have been described as causes of CTS (Table).
Whether occupation and job-related
hand or wrist overuse are risk factors for developing CTS is highly controversial.
Several authors have listed occupation and heavy manual labor as causal
factors for the disorder.[16-19] The opposite
view has been argued for as long, and Phalen himself stated that CTS was
not an occupational disease.[5,7,8,20]
Nathan et al reported that results
of nerve conduction studies of large numbers of industrial employees showed
no consistent association between the prevalence of CTS (detected by decreased
sensory nerve conduction) and the type and level of occupational hand activity,
length of employment, or bilateral versus unilateral activity. It has been
shown that patients covered by workers' compensation respond differently
to standard treatment of CTS.
DiagnosisThe diagnosis of CTS is based primarily on symptoms and findings on physical examination. The most common symptom is paresthesias in the distribution of the median nerve. Other symptoms include feelings of clumsiness and weakness in the affected hand, often made worse with activity. Proximal radiation of pain or paresthesias to the elbow or even the shoulder is not uncommon, and even shoulder pain has been described as the presenting symptom of CTS.[14,23] Thenar atrophy is a sign of advanced CTS of long-standing duration.
Night pain is a consistent symptom,
though its cause has been debated. Phalen
postulated that when the hand and wrist are at rest during sleep, engorgement
and relative venostasis occur in the small vessels within the flexor synovialis,
producing swelling, more nerve compression within the tunnel, and worse
pain. Active motion of the fingers and wrist decreases venous engorgement
and relieves pain, a phenomenon consistent with many patients' histories
and their need to shake or exercise their hands to relieve night pain.
It is an interesting theory that emphasizes the vascular etiology of the
disorder, but there is no published evidence to prove the hypothesis. Alternatively,
it may be that patients hold their wrists flexed while sleeping, thus increasing
the pressure on the median nerve and causing pain. Therefore, preventing
wrist flexion would be expected to decrease symptoms, and may be why many
patients find it beneficial to wear neutral-position wrist splints at night.
Provocative TestsA variety of provocative tests have been described to help confirm the diagnosis clinically. Phalen's test was described by him in 1957. Flexion of the wrist causes compression of the nerve between the transverse carpal ligament and the flexor tendons in the tunnel, causing paresthesias in the median nerve distribution and reproducing the patients' symptoms. The testing technique is important because even normal subjects will have paresthesias if the wrist is flexed passively with enough force or held flexed long enough. As described by Phalen, the test is done by having patients rest their elbows on the examination table with their forearms perpendicular to the floor and let their wrists drop into flexion with gravity assistance. The test is considered positive when paresthesias develop in less than 60 seconds. Patients with advanced CTS often notice paresthesias in less than 20 seconds. Tinel's sign is another provocative test commonly used to evaluate CTS. Interestingly, there is often discrepancy in how this test is done, interpreted, and even described by various examiners and authors. Tinel's original description was published in 1915. He noted that tingling sensations occurred when an injured nerve was percussed over its proximal stump, and he speculated this was a sign of axonal regeneration.
Testing technique is important when eliciting Tinel's sign, and subtle differences in test performance probably account for some of the discrepancies in reported prevalence. Phalen, noted the presence of Tinel's sign in all three patients presented in 1950 and described its presence as one of the two most reliable diagnostic findings, but he did not describe the technique for eliciting the sign. Later publications by Phalen and others have variously described "soft tapping," "light percussion,"[7,8] and "gentle tapping" of the median nerve at the transverse carpal ligament to elicit paresthesias in the median nerve distribution and produce a positive test result. It is difficult to quantify precisely how much pressure to use to elicit the sign and interpret it accurately. Too much force or a sharp blow over a normal median nerve will produce finger tingling, and as Phalen cautioned this must not be interpreted as the presence of Tinel's sign.
To help answer the question of how much pressure to apply, Durkan described a new variation of Tinel's test in 1991. An atomizer bulb connected to a pressure manometer allows the examiner to apply 150 mm Hg directly over the carpal tunnel. Thirty-one patients (46 hands) in whom the presence of CTS had been proved clinically and electrodiagnostically and most of whom had subsequent surgical release of the transverse carpal ligament were tested preoperatively by the direct compression method. Eighty-seven percent of the patients had onset of pain or paresthesias in the median nerve distribution within 30 seconds. The same results were achieved when the examiner applied direct pressure to the area with his thumbs. In the same study, a 10% false-positive rate was reported among a group of 50 control subjects. Similar results reported by Williams et al in 1992 verified the usefulness of the test. Subsequently, Durkan developed and described the use of a calibrated piston to achieve the same result.
The interpretation of the presence or absence of either of these two classic signs presents an interesting dilemma when evaluating suspected CTS. Kuschner et al surveyed the literature and reported a summary of the frequency of Tinel's and Phalen's signs in CTS. Tinel's test has been reported present in anywhere from 8% 19 to 100% of patients with CTS, while Phalen's test has been reported in 10% to 88% of patients. Kuschner et al also studied the presence of Tinel's and Phalen's signs in 200 normal wrists, and found they were present in 45% and 20%, respectively. Other reports in the literature have indicated similar findings and the presence of either or both signs in normal subjects.
Given that neither Phalen's nor Tinel's test is completely reliable in making the diagnosis of CTS, other provocative tests have been described and investigated. In general, these can be considered as either innervation-density tests or threshold tests. Innervation density tests evaluate multiple overlapping peripheral receptive fields and the density of innervation in the area being tested. The test depends on the cortical integration of peripheral impulses and can be interpreted as normal in a patient with nerve compression as long as a few fibers remain to conduct impulses to their correct end points in the cerebral cortex. Input to the central nervous system is changed drastically after nerve repair or regeneration, since impulses no longer reach their same (correct) end points in the cortex. Correct interpretation of an innervation-density test indicates that normal peripheral-cortical connections have been achieved or maintained. Therefore, innervation-density tests are better tests for nerve regeneration, such as after nerve repair, than for nerve compression.
The most commonly used innervation-density test is the static two-point discrimination test, performed by applying a force through two dull points placed a known distance apart (such as 5 mm) in the longitudinal axis of a digit without blanching the skin.[32,33] One study found two-point discrimination was abnormal in only 22% of patients with confirmed CTS, underscoring the point that such changes only occur late in the disease process. A variation of the test has been described using a wheel anesthesiometer, which is better able to detect subtle disturbances of discriminatory sensation than standard two-point discrimination, but this is not widely used.
Threshold tests evaluate a single nerve fiber innervating a receptor or group of receptor cells. In compression neuropathy, the sensory relay system remains uninterrupted, but threshold testing can demonstrate a gradual, progressive change in value as a greater proportion of nerve fibers is lost while others maintain their proper central connections.
Examples of threshold tests include Semmes-Weinstein monofilament pressure testing, the tourniquet test, measurements of vibratory sensibility, and electrical current perception tests. The Semmes-Weinstein pressure test is done by applying a monofilament perpendicularly to the palmar surface of a digit until it bends. Each given monofilament requires a certain known amount of applied force to bend. The subject is asked to localize verbally, without looking, which digit is being touched. The tourniquet test is done by applying a pneumatic blood pressure cuff proximal to the elbow and inflating it to a pressure higher than the subject's systolic blood pressure. The test is positive if numbness and tingling in the median nerve distribution develop or are exaggerated within 60 seconds.
Vibratory sensibility can be tested with a fixed-frequency, variable amplitude vibrometer and noting the threshold voltage needed to deliver a stimulus detectable by the patient. Alternatively, it can be tested by applying a vibrating 256 Hz tuning fork to a fingertip on the affected hand; the test is positive when sensation is altered between the subject's normal and affected hands. In one detailed report of sensibility testing of 23 hands in 20 patients with CTS, testing with a vibrometer was the most sensitive and reliable test and was positive in 87% of hands. The monofilament test was next best at 83%. Vibratory testing with the 256 Hz tuning fork, Phalen's test and the tourniquet test were all about equivalent at 70%. Tinel's test was even less sensitive at only 61%.
Although the diagnosis of CTS is made primarily by these observations and clinical tests, a variety of other more elaborate and expensive tests have been used to confirm the diagnosis. Electrodiagnostic studies are commonly used, but there is a wide range of opinion among medical and surgical specialists about the necessity of such tests to diagnose CTS, and data are insufficient to conclude that nerve conduction studies should serve as the definitive standard for the diagnosis. Recently a group of 12 medical researchers with experience conducting epidemiologic studies on CTS agreed there is no perfect standard for establishing the diagnosis, and they could not reach a consensus about what to do with the patient with classic symptoms but a negative electrodiagnostic study.
A recent study examined the utility of clinical tests used individually and in various combinations with each other for the purpose of diagnosing CTS. The authors used multivariate analysis techniques to investigate whether a combination of tests might be more powerful than a single test to establish the diagnosis. It was shown that the combination of a patient-completed hand diagram, the presence of night pain, abnormal sensibility on monofilament testing, and a positive Durkan's test will provide a diagnostic tool with high sensitivity and specificity. Via a mathematical regression model, it was shown that if results of all four of those tests were abnormal, then the probability that CTS would be correctly diagnosed was 86%. If all four tests were negative, then the probability that the given patient had CTS would be 0.68%. The addition of electrodiagnostic tests did not increase the diagnostic power of the combination of the four clinical tests.
Electrodiagnostic studies are most useful when trying to distinguish CTS from other conditions such as thoracic outlet syndrome or cervical radiculopathy. They also provide objective data on nerve condition that may be useful when patients' secondary gain is suspected. The official clinical policy statement of the American Academy of Orthopaedic Surgeons states that electrodiagnostic studies can be negative in some patients with the disorder and the tests offer supporting evidence but are not required for diagnosis.
Other diagnostic studies are rarely
necessary. Computed tomography shows bony anatomy well but does not define
the soft tissues clearly enough. Ultrasonography can show flexor tendon
movements within the canal but does not define nerve compression and soft
tissue planes well. Magnetic resonance imaging (MRI) shows the soft tissues
with high contrast, and Healy et al
showed good correlation between scan images and surgical pathology, but
the time and cost involved in procuring an MRI scan remains prohibitive.
Gelberman et al reported that pressure
in the carpal tunnel is elevated in CTS; the average canal pressure with
the wrist flexed was 94 mm Hg in 15 patients with CTS and only 31 mm Hg
in 12 control patients. Most of these tests remain investigational tools
only and are not ready or intended for routine use.
Differential DiagnosisIt is important to distinguish true CTS from other conditions that can produce similar signs and symptoms. Nerve compression caused by a cervical disk herniation, thoracic outlet structures, and median nerve compression proximally either in the forearm or at the elbow can masquerade as CTS. Thenar atrophy from other causes (disuse, other neuropathies) and pain due to osteoarthritis of the first carpal-metacarpal joint are sometimes confused with CTS. Trigger fingers and deQuervain's stenosing tenosynovitis are commonly associated with CTS, and Phalen's first report in 1950 described 4 patients with CTS, one of whom had an associated "snapping finger" (trigger finger). These conditions should be ruled out in any patient considered to have CTS.
Nonoperative TreatmentOnce the diagnosis is established, treatment begins with nonoperative management. Splinting the wrist in a neutral position will help reduce and may even completely relieve CTS symptoms. An initial trial of full-time splinting for 3 to 4 weeks followed by part-time night splinting is recommended.[14,41] Nonsteroidal anti-inflammatory drugs (NSAIDs) are frequently prescribed and are beneficial in many cases, even though there is no acute inflammatory process in most patients with CTS. An injection of corticosteroids into the carpal canal can be useful therapeutically as well as diagnostically but must be placed accurately to avoid complications. Response to injection is important; in a large series of patients, Green noted that a good response to injection with immediate pain relief from the local anesthetic correlated with an excellent response to subsequent surgery. However, poor relief from injection does not necessarily mean that the patient is a poor candidate for surgery. Between 65% and 90% of patients can be expected to have recurrence of symptoms after a successful steroid injection, usually after 2 to 4 months, but 11% of patients in Green's series had no recurrence for up to 45 months. Other authors are less optimistic about the lasting efforts of steroid injections, though most agree short-term benefit is to be expected. Eversmann believes steroid injections should be limited to cases in which nerve entrapment is expected to be temporary, such as in pregnancy or when sufficient activity modifications can be made promptly to diminish the contributing stresses at the wrist.
Kaplan et al
tried to define more accurately those patients likely to respond to nonsurgical
treatments by identifying five "risk factors": patient age greater than
50 years, the presence of symptoms for 10 months or more, constant paresthesias,
the presence of associated trigger fingers and/or positive results of Phalen's
test after 30 seconds or less. In a study of 331 subjects with CTS, Kaplan's
patients were treated with splints and anti-inflammatory agents, either
with oral NSAIDs (65%) or an intracanal corticosteroid injection (16%)
or both, and were considered cured if their symptoms resolved for at least
6 months. Approximately 60% of patients were cured without surgery if they
had only one risk factor, but 93% of those with 3 factors and 100% of those
with 4 or more risk factors had unsuccessful nonoperative management.
Operative TreatmentWhen nonoperative management fails, surgical treatment is indicated. Traditionally this has meant division of the transverse carpal ligament under direct vision with an open procedure, and this is still the most commonly performed operation for CTS. It is important for the surgeon to recall the variable anatomy of the palmar cutaneous branch of the median nerve to avoid damaging it and causing a painful neuroma. One must also recall the variations in the motor branch to the thenar muscles. Lanz described four groups of variations in his classic 1977 article: (1) a single thenar motor branch may be derived from the median nerve, but its course under, through, or over the transverse carpal ligament may vary; (2) there may be an accessory motor branch; (3) the thenar motor branch origin may be high, proximal to the transverse carpal ligament; and (4) there may be an accessory branch that originates proximal to the carpal tunnel. The importance of these observations is that they call attention to the variable neuroanatomy, and surgeons must be cognizant of this during dissection. Complications are rare but have been reported, including the devastating complications of complete median nerve transection and massive necrosis of the palm.
The need for neurolysis has been debated in the literature.[47-52] First, it should be distinguished from epineurotomy and epineurectomy. Complete epineurectomy should not be done because it will devascularize a segment of the nerve by disrupting the subepineural plexus of vessels. An epineurotomy is an incision in the epineurium, which may be indicated in some cases where there is an obvious constricting brand. Some surgeons perform this routinely, likening it to a decompressive "fasciotomy."
Early reports in the literature suggested
there was value in performing internal neurolysis routinely.
More recently, authors have questioned the need for neurolysis, especially
given the risk of disrupting the interfascicular plexus and the capillary
network in the nerve by manipulation and dissection, which could lead to
increased fibrosis.[14,49,51,54] Current
consensus, given the abundance of clinical and animal research data, is
that neither an epineurotomy nor an internal neurolysis is needed for adequate
nerve decompression or good results from surgery.
Endoscopic Carpal Tunnel ReleaseThe latest innovation in CTS surgery is the use of endoscopic equipment to perform the release of the transverse carpal ligament. Many methods have been described.[55-59] The equipment required for each is more expensive and takes longer to set up in the operating room than that required for a standard open procedure, thus adding to the total surgical cost. Incomplete releases of the transverse carpal ligament requiring second operations have been reported clinically and occurred in 50% of wrists in a cadaver study. Persistent or recurrent CTS after endoscopic carpal tunnel release (ECTR) may respond well to open carpal tunnel release to complete the release. There is a risk of injuring neurovascular structures in the hand, given their close proximity to the carpal canal with small margins for error and occasional difficulty visualizing them endoscopically. The ulnar and superficial palmar arch arteries and the common digital nerve in the third web space are most at risk.[55,56,60,62]
Postoperative pain is reportedly less severe after endoscopic release than after open procedures, particularly in the "pillars" on either side of the carpal canal. This is related to the limited dissection and the supporting soft tissues that remain after endoscopic surgery. It is possible that the decreased pain alone facilitates an earlier return of grip strength and an earlier return to work or regular activities after surgery. However, results from a recent study support the hypothesis that there may be a biomechanical advantage to leaving the soft tissues intact over the transverse carpal ligament. In a cadaver study, Kline and Moore showed that division of the transverse carpal ligament results in increased tendon prolapse (bowstringing) at the wrist and decreased grip strength.
Endoscopic procedures leave the lesser thenar muscle fibers, palmaris brevis, palmar fascia, subcutaneous tissues, and skin intact. These remaining tissues may then act as a flexor tendon pulley in place of the divided transverse carpal ligament, thereby decreasing flexor tendon bowstringing and improving grip strength. Endoscopic carpal tunnel release has many limitations, and these must be kept in mind to minimize complications. It should not be used in cases with advanced thenar atrophy warranting a neurolysis or release of the thenar motor branch of the median nerve. Endoscopic release should not be used for patients with obvious tenosynovitis, metabolic problems such as amyloidosis, or an obvious mechanical problem, such as a bone fragment in the canal. Previous surgery and scar tissue in the wrist may be a relative contraindication, even for the advocates of endoscopic techniques.
Complications may follow either open or endoscopic carpal tunnel release (CTR). The list of complications reported after open release includes incomplete division of the transverse carpal ligament, damage to the median nerve or its palmar cutaneous or motor branches, tendon adhesions and bowstringing, arterial injury, palmar hematoma, hypertrophic scar, pillar pain, wrist stiffness, and infection. Many of these problems have been reported as single cases and are rare, and most known complications after open CTR are nondisabling.
The cost, safety, and efficacy of
endoscopic carpal tunnel release have been the subject of numerous reports,
and still are being investigated carefully.[60,66-68]
The most common complication is incomplete release of the transverse carpal
ligament which has been found in up to 50% of specimens in some series.
The incidence of iatrogenic neurovascular injury during ECTR is low, but
it has been reported many times[59,67-70]
and may be independent of experience, as Feinstein
reported that a median nerve laceration occurred on the 59th case in a
series of 61 ECTR procedures. Some authors have stated that ECTR is more
complex than open CTR and therefore "carries greater risk,"
but that has been difficult to prove convincingly, and whether ECTR is
inherently more likely to result in a complication remains an unanswered
question. It remains to be seen whether the benefits of slightly less pain
and transient grip strength differences reported by the advocates of ECTR
will in the long run outweigh the risks of decreased visualization of associated
pathologic conditions and potential for iatrogenic injuries to important
Results of TreatmentThe results of surgical treatment of CTS are generally excellent. One recent prospective observational outcomes study examined the results of operative and nonoperative treatment of a group of patients with CTS. Patients treated nonoperatively had no meaningful change in symptoms after 6, 18, and 30 months. In contrast, patients treated operatively showed 23% to 45% improvement in clinical scores (Symptom Severity and Functional Status scales), and the improvements persisted over the 30-month follow-up period. Numerous other studies report similar findings. Grip strength returns to preoperative levels after 3 months in patients treated by open release and perhaps slightly sooner after endoscopic release and improves to 116% preoperative levels after 6 months.
ConclusionIn summary, CTS is a common entrapment neuropathy of the median nerve at the wrist that usually can be diagnosed clinically. In some cases, a variety of provocative tests, nerve conduction velocity tests, or other imaging studies may be necessary to help confirm the diagnosis or rule out other causes of a particular patient's symptoms. Surgical release of the transverse carpal ligament provides reliably good results when nonoperative treatment fails, but a patient-specific management approach is recommended.
Reprint requests to Robert R.
Slater, Jr., MD, Department of Orthopaedic Surgery, 4860 Y St, Suite
3800, Sacramento, CA 95817.
Table. Causes of Carpal Tunnel SyndromeTrauma-related structural changes
Gellman H, Kan D, Gee V, et al: Analysis of pinch and grip strength after carpal tunnel release. J Hand Surg Am 14:863-864, 1989