➢ Carpal tunnel syndrome is the most common peripheral nerve compression syndrome.
➢ Treatment options include wrist-neutral bracing, corticosteroid injections, operative release of the transverse carpal ligament, and symptom-relief options.
➢ Endoscopic carpal tunnel release may give patients a faster recovery compared with traditional open release, but there are no ultimate differences in outcome among the various surgical options.
Anatomy and Pathophysiology
The carpal tunnel is bounded by the carpal bones and metacarpals dorsally and by the transverse carpal ligament palmarly. The radial border is the scaphoid, trapezium, and flexor carpi radialis sheath. Ulnarly, the transverse carpal ligament attaches to the pisiform and hamate hook1,2. The distal volar wrist crease is a surface landmark for the carpal tunnel’s proximal limit. Intraoperatively, the distal transverse carpal ligament border can be identified by visualization of the midpalmar fat, which appears approximately 2 mm proximal to the distal edge of the ligament and just proximal to the superficial arch3. There are ten structures within the carpal tunnel: the median nerve, the flexor pollicis longus, and the superficial and deep finger flexors. The median nerve is the most superficial structure in the tunnel4.
The median nerve’s palmar cutaneous branch originates from the radial aspect of the median nerve 5 to 8 cm proximal to the distal volar wrist crease and travels distally, adjacent to the median nerve beneath the forearm fascia, until it passes obliquely radially toward the flexor carpi radialis sheath between 1 and 1.5 cm proximal to the distal volar wrist crease5,6. It then runs with the flexor carpi radialis (and at times within its sheath) at a depth between the palmar aponeurosis and the transverse carpal ligament. About 1.5 cm distal to the scaphoid tubercle, the nerve pierces the palmar aponeurosis to provide sensory innervation to the palm and thenar eminence7. Surgeons should be aware of this nerve branch’s usual course while understanding that variations are possible.
The median nerve’s recurrent motor branch innervates most of the thenar musculature, with the exception of the deep head of the flexor pollicis brevis, which receives dual innervation by the motor branches of the median and ulnar nerves. The recurrent median motor branch most commonly branches from the median nerve distal to the transverse carpal ligament, in an extraligamentous fashion; however, subligamentous (branching deep to the ligament), transligamentous (piercing the ligament), and other variations are possible4,8.
Carpal tunnel syndrome is the result of an increase in pressure on the median nerve within the carpal tunnel. There are myriad etiologies, both intrinsic and extrinsic, that result in median nerve compression. Both wrist flexion and extension result in higher carpal tunnel pressures than baseline, even in healthy individuals. Additionally, power grip, tip pinch, and metacarpophalangeal joint flexion are associated with increased pressure within the tunnel8.
Primary (idiopathic) carpal tunnel syndrome is by far the most common form of carpal tunnel syndrome. Primary carpal tunnel syndrome is idiopathic in terms of a direct cause, although it is associated with systemic conditions such as diabetes, thyroid disease, rheumatoid arthritis, and obesity. The pathophysiologic processes associated with primary carpal tunnel syndrome are notably devoid of inflammation in most9. Abnormalities are found in the subsynovial connective tissue, including thickening and fibrosis of tendon sheaths, as well as vascular proliferation, intimal hyperplasia, and small-vessel thrombosis1. These fibrotic changes are the result of abnormal cellular interaction with collagen, elastin, and proteoglycans10. Examination of these subsynovial tissues has demonstrated deformed, spiraling collagen fibrils that are not present in similar tissues of non-affected individuals10.
Secondary carpal tunnel syndrome causes include sources of compression extrinsic to the carpal canal, such as carpus and distal forearm fractures, or masses and space-occupying lesions within the canal. Prolonged elevated pressure on peripheral nerves causes gradual demyelination and axonal degeneration and, ultimately, loss of nerve conduction8. Regardless of cause, when the median nerve is compromised because of increased pressure either within or external to the canal, this pressure should be relieved and underlying causes should be addressed.
Fifty years since Phalen’s landmark article11 and more than 100 years since the initial description of carpal tunnel syndrome12, there is still no gold standard for its diagnosis. American Academy of Orthopaedic Surgeons (AAOS) Clinical Practice Guidelines give all Grade-C recommendations (poor-quality evidence), except for a single Grade-B recommendation (fair evidence), for carpal tunnel syndrome diagnostic methods13. The Grade-B recommendation was for obtaining electrodiagnostic studies if physical examination or provocative maneuvers were positive and surgical intervention was being considered. The diagnostic process remains a best-practice combination of history, physical examination, and electrodiagnostic testing.
Carpal tunnel syndrome symptoms arise from median nerve compression and the resultant dysfunction of its distal motor and sensory branches14. Symptoms that do not fit this anatomic distribution should raise suspicion for alternate abnormalities. Upton and McComas described the double-crush phenomenon in which one site of nerve compression makes the nerve more susceptible to a second site of compression because of inhibited axonal transport15. In carpal tunnel syndrome, this primarily refers to the possibility of simultaneous cervical radiculopathy and compression at the carpal canal. Cervical radiculopathy is more common in patients with carpal tunnel syndrome than in the general population16. Despite its popularity, this idea remains controversial because there is poor correlation between the electrically measured severity of carpal tunnel syndrome and the severity of cervical radiculopathy17. Kwon et al. suggested that physical examination should include the cervical spine and brachial plexus and that dedicated work-up of alternate sites of compression should be performed when examination and electrodiagnostic testing are not consistent with isolated carpal tunnel syndrome17.
Patients with primary carpal tunnel syndrome typically present with paresthesias in the thumb, the index and long fingers, and the radial side of the ring finger. Symptoms may be transient or constant and patients often describe awakening with symptoms. Patients often report that the entire hand is involved, even for isolated carpal tunnel syndrome. This is understandable because the median nerve innervates at least 80% of the hand’s palmar surface. Patients often describe having to “flick” the hand, known as the “flick sign,” or hang it down by their side or by the side of their bed at night to relieve symptoms. Thenar motor branch compression results in weakness and eventual atrophy of median-innervated muscles and may give symptoms of aching pain. In Phalen’s article, which included 439 patients with 654 affected hands, 79% (517 hands) had documented decreased sensation in some territory of the median nerve and 41% (270 hands) had thenar atrophy11. Night pain remains one of the most common findings and is up to 96% sensitive for a carpal tunnel syndrome diagnosis18.
Physical examination involves examination of the cervical spine including flexion, extension, lateral bending and palpation, and percussion of the ipsilateral brachial plexus, especially in patients with atypical presentation. Reproduction of symptoms with any of these maneuvers is consistent with primary or concomitant proximal disease.
Both hands are typically examined, even when reports are unilateral. Clinicians should inspect for atrophy, asymmetry, or traumatic sequelae. Thenar atrophy, a late finding of advanced disease, is best appreciated by comparing one hand with the other, often by looking in-plane with the palm. This allows the examiner to determine whether the amount of thenar bulk projecting palmarly is less than normal. Routine motor, sensory, and vascular examinations of both hands follow. Particular points to note are decreased palmar sensation in the thumb, the index and long fingers, and the radial side of the ring finger. In patients who report entire hand numbness, clinicians should compare the ring finger’s two sides or compare the small finger with the index finger to ensure no ulnar or more proximal nerve involvement. Semmes-Weinstein monofilament testing can be useful; an abnormal result (>2.83 monofilament) is reported to be 91% sensitive19. Two-point discrimination is less sensitive, at 33% when using >6 mm as a threshold. This indicates that if clinicians use static two-point discrimination as a sole diagnostic tool, they will miss many clinically important carpal tunnel syndrome cases. The key finding on motor examination is strength of the abductor pollicis brevis muscle. It is a purely median-innervated muscle, unless anastomotic variations exist with the ulnar nerve, and powers palmar abduction of the thumb. Placing the dorsum of the patient’s hand flat on the examination table and having the patient elevate the thumb away from the table tests the abductor pollicis brevis. Once the patient demonstrates strength against gravity, the examiner can test the patient’s abduction strength against resistance provided by the examiner.
Provocative carpal tunnel syndrome tests are typically performed last. Phalen described the wrist flexion test in which the patient’s hands are allowed to fall passively into wrist flexion and are held for up to one minute. Reproduction of symptoms constitutes a positive finding. Phalen noted 74% (380 of 515 hands) had positive results with this maneuver11. Avoid elbow flexion when performing the Phalen test, as concomitant ulnar nerve paresthesias can occur in patients with cubital tunnel syndrome and may be misinterpreted as a positive Phalen test, especially if the examiner or patient is not specific about determining which digits are symptomatic.
In a carpal compression test, direct pressure is applied over the carpal canal, either with a sphygmomanometer or manually, for up to thirty seconds. Reproduction of symptoms is a positive test; Durkan noted 87% sensitivity after a mean time of sixteen seconds and a false-positive rate of 10%20. The Tinel sign involves percussion over the median nerve, with reproduction of paresthesias constituting a positive finding21. Patients with severe compression or constant symptoms may not have worsening of symptoms with provocative maneuvers and thus may have what could be considered false-negative provocative tests. Numerous studies have shown varying levels of sensitivity, specificity, and agreement between provocative tests. Szabo et al. found that an abnormal hand diagram and a positive Tinel sign were the most specific at >70%18. They found that the combination of an abnormal hand diagram, altered sensibility on Semmes-Weinstein testing, a positive compression test, and night pain had an 86% likelihood of correctly predicting carpal tunnel syndrome, defined by improvement after carpal tunnel release.
AAOS Clinical Guidelines provide Grade-C recommendations for obtaining electrodiagnostic tests to help to differentiate between diagnoses (3.1a) and in the presence of thenar atrophy and numbness (3.1b). There is a Grade-B recommendation to obtain electrodiagnostic tests if physical examination and provocative maneuvers are positive and surgical intervention is being considered (3.1c). This is the strongest recommendation in the guideline13; 90% of hand surgeons reported using electrodiagnostic testing22, although the utility of routine testing remains debatable23. Patients with less severe electrodiagnostic abnormalities note faster symptom improvement24, which can be useful in preoperative counseling. Subjective symptoms and electrodiagnostic studies typically correspond with severity in bilateral disease; discrepancies should raise concern for alternative diagnoses25.
Caregivers who address the pathophysiology of primary carpal tunnel syndrome use treatments including wrist-neutral braces, corticosteroid injections, or operative release of the transverse carpal ligament to decrease pressure within the carpal tunnel. Other treatments, including neurological medications, biofeedback, meditation, and other pain-control options, are meant to mitigate symptoms rather than to alleviate underlying pathophysiology. Debate continues about whether surgeons should routinely perform endoscopic or open carpal tunnel release for patients with idiopathic disease. Nearly all authors and surgeons agree that there are no significant or clinically important differences among various release techniques at six months and certainly by one year postoperatively. Surgeons disagree on whether there are important differences between techniques regarding recovery time, procedure value regarding the cost of the technique compared with the cost of time off work, and relative safety and recurrence rates. Published evidence has suggested that all techniques have similar efficacy and safety; therefore, surgeons should perform the technique with which they are most comfortable. This is consistent with the current AAOS Treatment Guidelines. Evidence suggests that endoscopic and limited open techniques have a slightly faster recovery time compared with traditional open release. The reason for this difference is unknown and may be physiologic or psychological, if a true difference exists. Surgeons also disagree as to whether this difference, if significant, is clinically relevant. Vasen et al. suggested that patients who have undergone endoscopic release would need to return to work activities at a mean time of three to four weeks earlier than patients who underwent open release to justify its use from a cost-benefit standpoint, not considering possible differences for Workers’ Compensation patients26. This timeframe is likely near that of the typical literature-reported difference in return to work between patients undergoing endoscopic release and those undergoing open release and is consistent with our clinical experience, although different investigators have reported larger and smaller differences in return-to-work time for various techniques. A meta-analysis of twenty-one randomized controlled trials revealed a mean earlier return-to-work difference of only nine days for patients who underwent endoscopic release compared with patients who underwent open release27. The type of work to which the patient is returning may be important as well.
Within the group of surgeons who routinely perform endoscopic release are those who prefer the single-portal technique and those who prefer the double-portal technique. Some surgeons who favor open release perform what could be considered full-open skin incisions, which span at least the transverse carpal ligament’s entire length; others perform some version of limited-incision open release, sometimes with specialized equipment. For some of these limited-incision techniques, the total skin incision length compared with that for endoscopic release is probably not different. Some limited-incision open techniques involve blind or semi-blind transverse carpal ligament release, whereas endoscopic techniques permit visualization of the ligament’s undersurface but not the palmar surface. The goal for all procedures is complete transverse carpal ligament release.
The importance of learning and confirming mastery of hand anatomy and surgical technique cannot be overstated in the context of median nerve decompression at the carpal tunnel. One surgeon reported an enlightening experience with surgical residents who were told to prepare for observing and performing open carpal tunnel release28. At the operation, the attending surgeon asked the residents to draw their planned incision. There was a wide and dangerous variation in planned incisions, including some that were marked dorsally. This report made the point that new surgeons must learn in the operating room and/or anatomy laboratory with teaching from experienced surgeons for even what would be considered the most basic aspects of this procedure.
For routine open-release skin incisions, the surgeon’s main goals are to avoid injuring the median nerve’s palmar cutaneous branch and to gain adequate access to completely release the transverse carpal ligament. Surgeons can avoid the palmar cutaneous nerve branch by staying ulnar to the palmaris longus tendon during superficial dissection because this nerve most commonly courses between the palmaris longus and flexor carpi radialis. The transverse carpal ligament’s proximal edge is at the level of the distal volar wrist crease. Therefore, the skin incision can start at the intersection of the palmaris longus’ ulnar edge and the wrist crease. If there is no palmaris longus, surgeons may mark the intersection between the wrist crease and a longitudinal line that runs just radial to the hamate hook. Many surgeons instruct that the ring finger in flexion may be used as a landmark to guide the incision location, but this standard is more variable to use than the hamate hook. From this proximal starting point, the surgeon incises along the ulnarmost longitudinal palmar crease, which helps with incision or scar concealment. Some surgeons recommend keeping the incision even ulnar to this point to avoid potential terminal branches of the palmar cutaneous nerve in the skin. Once the skin is incised, the palmar fascia is released longitudinally, again along a line that courses just ulnar to the palmaris longus or just radial to the hamate hook. Some surgeons release the forearm fascia proximal to the level of the distal volar wrist crease, either with an extension of the skin incision proximally or with subcutaneous release29. If the surgeon extends the skin incision proximally, he or she should make a z-type extension or at least extend at an oblique angle at the level of the distal wrist crease to avoid a longitudinal scar, which could lead to a flexion contracture. If the surgeon chooses to release the forearm fascia proximal to the wrist crease, it is done so longitudinally just ulnar and deep to the palmaris longus. Finally, the transverse carpal ligament is completely released along the same line on which the palmar fascia was released. As the surgeon proceeds distally, he or she will typically encounter fatty tissue just before the transverse carpal ligament’s distal edge. This sentinel fat indicates that the surgeon will soon reach the ligament’s distal end. This fat pad also warns that the ulnar nerve and superficial arterial arch will soon be encountered if dissection continues distally beyond the ligament’s edge.
For endoscopic release, patients should be counseled that conversion to open release might be necessary if the procedure cannot be performed safely or effectively. Once the portals are established, the principles for transverse carpal ligament release are similar for single or double-portal release. This review focuses on the single-portal technique. A transverse incision is made just proximal to the distal wrist crease and just ulnar to the palmaris longus in one of the transverse skin creases. If the incision is too distal at the wrist crease, adipose tissue can make visualization difficult. If the incision is too proximal, tools can pierce the thinner fascia of the distal aspect of the forearm during instrumentation and result in the endoscopic tools being passed superficial rather than deep to the transverse carpal ligament distally. After making the skin incision, one should be sure to obtain a clear view of the forearm fascia. The area to gain access through the fascia is just deep and ulnar to the palmaris longus, if present. Once the fascia is incised, the surgeon can release as much proximal forearm fascia as he or she wants and as can also be safely released. A hook is placed under the fascia’s distal edge. As long as the surgeon stays deep to the fascia, he or she will also be deep to the transverse carpal ligament. The ligament’s undersurface is cleared with one of the endoscopic tools, a series of dilators are inserted, and the endoscope is inserted. A clear view of the ligament’s undersurface should be obtained before proceeding. The median nerve can be visualized by rolling the endoscope radially and can then be excluded for safety during ligament release. It is important to release the ligament with the endoscope just radial to the hamate hook and in a longitudinal orientation. Straying more than 15° from longitudinal may increase the chance for injury to other structures30. Many surgeons recommend releasing the ligament’s distal half first and then reinserting the endoscope with the cutting-blade retracted to ensure complete distal release. If the ligament is released in one complete pass, it is more difficult to assess the distal release afterward due to substantial fatty tissue falling into the field of view. Proximal ligament release is then completed. At the proximal end of the release, as tension is lost on the ligament, the cutting-blade may push the tissue palmarly rather than incise it, so surgeons should ensure that the ligament’s proximal end and forearm fascia near the skin incision are completely released. If any of these steps cannot be performed safely and definitively, the surgeon and patient should be prepared to convert to open release.
AAOS Treatment Guidelines
The AAOS conducted an independent carpal tunnel syndrome literature review in 2008 and provided a reissue statement in 201131,32. The committee concluded31 that there was strong evidence for “recommending surgical treatment of carpal tunnel syndrome by complete division of the flexor retinaculum regardless of the specific surgical technique.” This recommendation was based on Grade-A, Level-I and II evidence from two systematic reviews and six randomized controlled studies. The 2011 reissue statement addressed open release compared with endoscopic release and concluded that of five outcomes, four favored endoscopic release in the short term32. However, the AAOS concluded that the new evidence did not change the recommendation that any technique may be used for surgical release.
Caregivers debate the relative risks, costs, and benefits of endoscopic release compared with open carpal tunnel release. Endoscopic release may result in less pain and in faster healing and earlier return to work and other activities. However, many surgeons have concerns regarding the endoscopic method’s safety and efficacy. Detractors have also cited endoscopic release as having increased cost of equipment and greater required surgical training compared with open release33.
Relative outcomes of endoscopic release compared with open release have been investigated in several randomized controlled trials and meta-analyses. Sayegh and Strauch compared symptom relief, validated outcome measures, sensibility, complications, and operative time from twenty-one studies and found that operative time is a mean of five minutes less for endoscopic compared with open and results in a faster return to work by a mean time of nine days27. However, symptom relief was equally likely after either procedure. No difference was found in the likelihood of persistent pain, numbness, paresthesias, subjective weakness, or nocturnal symptoms. Endoscopic and open release improved pain scores and digit sensibility to a similar extent. Boston Carpal Tunnel Questionnaire scores showed a similar improvement in symptom severity and functional status between the procedures34.
The most recent Cochrane Review on this topic also concluded that there was no difference on symptom or functional severity scales in the short term, defined as follow-up of three months or less35. There was also no difference in the rate of residual numbness between the procedures. The authors found a small but clinically unimportant increase in grip strength among patients who underwent endoscopic release. When including only studies with a low risk of attrition bias, this difference in short-term grip strength was not present. Comparing the procedures at more than three months did not reveal any difference in functional or symptomatic outcome except for an increase in grip strength among patients undergoing endoscopic release, with a mean difference of 11 kg35. However, Sayegh and Strauch found no difference in grip or pinch strength when including only follow-up data at six months or more27. It remains unclear whether there is a clinically important difference between these procedures for patients’ strength at different time points in the postoperative course.
Vasiliadis et al. evaluated the safety of endoscopic release compared with open release in twenty-three studies35. The rate of major complications was the same for both procedures (0.9%) among 1508 participants across fifteen studies. Sayegh and Strauch found a higher number of neurapraxias and minor complications among patients who underwent endoscopic release27; however, Vasiliadis et al. reported a lower rate of minor complications in the endoscopic release group when considering wound problems as well35. Their overall reported minor complication rate was 5.5% for the endoscopic group and 8.7% for the open release group35. Recurrence and reoperation rates were equal between the two procedures, and there was no difference in pillar pain27,35.
Major nerve, tendon, and vessel injuries have been reported for open and endoscopic release. Open release requires detailed knowledge of and experience with the palmar anatomy of the hand and wrist. If incision and dissection are too ulnar, the ulnar neurovascular structures are at risk; if incision and dissection are too distal, the superficial arterial arch is at risk; and if incision and dissection are too radial, the palmar cutaneous branch of the median nerve and thenar motor branch are at risk. These structures are also at risk with endoscopic techniques with careless instrumentation use and/or blade deployment. Endoscopic technique detractors point to potential injuries due to lack of visualization. There is the possibility of injuring an aberrant thenar motor branch of the median nerve that could be traversing the transverse carpal ligament, that could be beneath the ligament, or that could be more ulnar than usual and could not be visualized during endoscopic release. Open release may similarly suffer from lack of visualization, as only endoscopic release provides the opportunity to visualize the median nerve and any aberrant branches before dividing the transverse carpal ligament. However, most surgeons who perform endoscopic or open release simply keep their transverse carpal ligament incision longitudinal and just radial to the hamate hook, thus staying away from any thenar motor branch variations, rather than routinely identifying the motor branch before ligament release. If, during open release, the thenar muscles are seen originating on top of the ligament more ulnarly than usual, this can be an indicator of an aberrant thenar motor branch course and, in these cases, the ligament is released ulnar to the muscle origins. This variation in technique to potentially decrease the risk to an aberrant thenar motor branch is not possible with endoscopic procedures. However, we are not aware of any studies specifically showing relative risks to aberrant thenar motor branches with one technique compared with another technique.
Another known potential complication of carpal tunnel release, regardless of technique, is irritation of or injury to the palmar cutaneous branch of the median nerve or its terminal branches. Staying ulnar to the palmaris longus for the skin incision during either endoscopic or open release should avoid direct injury to these nerve branches. If injured, a painful neuroma may form, which can be difficult to treat2,4,8. Nonoperative measures are typically tried first, including therapy desensitization and other coping strategies. Nerve medications such as gabapentin or pregabalin can also improve symptoms. Surgical options include repair or resection and burial. However, as with all neuroma surgical procedures, there is a possibility of entering into an endless cycle of surgical procedure, short-term relief, and symptom recurrence.
The most common complication following carpal tunnel release is continued median nerve symptoms. It is important to try to determine if (1) symptoms were improved or were resolved and have recurred, or (2) symptoms never improved or even worsened following release. If symptoms never improved or even worsened following release, then it is typically thought that an incomplete release or direct injury to the median nerve at the time of the initial surgical procedure is more likely. One should also consider the possibility that other disease more proximal to the wrist may be responsible for patients’ symptoms. Finally, some patients, despite having a correct diagnosis of carpal tunnel syndrome, undergoing an appropriate release procedure, and having no other identifiable confounding issues, will continue to have median nerve symptoms after surgical release. For these patients, most surgeons typically recommend waiting four months after the surgical procedure before obtaining repeat electrodiagnostics unless there are progressive symptom or examination changes or there is concern that an intraoperative nerve injury occurred. If no identifiable and correctable issues are discovered, therapy and medications may help with residual symptoms.
Investigation performed at the Curtis National Hand Center, MedStar Union Memorial Hospital, Baltimore, Maryland
Disclosure: There was no external funding source for this study. The Disclosure of Potential Conflicts of Interest forms are provided with the online version of the article.
- Copyright © 2016 by The Journal of Bone and Joint Surgery, Incorporated