Clinical Biomechanics 18 (2003) 685–693

www.elsevier.com/locate/clinbiomech
Review

Biomechanical and anatomical consequences of carpal tunnel release

Jeffrey J. Brooks, Jonathan R. Schiller, Scott D. Allen, Edward Akelman *

Department of Orthopaedics and Division of Engineering, Rhode Island Hospital/Brown University School of Medicine,
593 Eddy Street, Providence, RI 02905, USA
Received 26 February 2003; accepted 5 March 2003

Abstract
Carpal tunnel syndrome is an exceedingly common orthopaedic problem in the United States. When conservative management is
unsuccessful, most surgeons proceed to surgical treatment. Though the carpal tunnel release procedure is usually curative, many
patients experience postoperative complications, such as scar sensitivity, pillar pain, recurrent symptoms, and grip weakness, re-
gardless of whether the release was done through an open, mini-open, or endoscopic approach. The exact causes of these and other
complications of carpal tunnel release remain unclear. Release of the carpal tunnel has an effect on carpal anatomy and biome-
chanics, including an increase in carpal arch width, carpal tunnel volume, and changes in muscle and tendon mechanics. We set out
to review the morphological and biomechanical changes caused by carpal tunnel release with the goal of better understanding the
root causes of postoperative complications. This article first reviews normal carpal tunnel anatomy and anatomic variations, then
available surgical techniques for carpal tunnel release, and finally the literature on morphologic, physiologic and biomechanical
alterations in the wrist after carpal tunnel release.
Ó 2003 Elsevier Ltd. All rights reserved.

1. Introduction and clinical significance management of carpal tunnel syndrome and subsequent
treatment of poor postoperative outcomes.
Carpal tunnel syndrome is one of the most common
orthopedic conditions, with an estimated incidence of
nearly 1% annually in the United States (Einhorn and 2. Anatomy
Leddy, 1996), which translates into almost 2.8 million
new cases per year. If conservative therapy fails, surgical The carpal bones and intercarpal ligaments at its
release of the carpal tunnel is the preferred method of medial, lateral, and posterior borders form the carpal
treatment. Though the majority of patients have relief of tunnel. The anterior border is formed by the transverse
symptoms postoperatively, there are still a significant carpal ligament and flexor retinaculum (Hoppenfeld Sd,
number of patients who experience disabling postoper- 1984; Tanabe and Okutsu, 1997). The terms ‘‘flexor re-
ative symptoms. Disabling loss of grip strength and tinaculum’’ and ‘‘transverse carpal ligament’’ have been
pillar pain may be a result of anatomical or biome- considered synonyms, however, Cobb et al. demon-
chanical alterations caused by carpal tunnel release strated that they are distinct structures (Cobb et al.,
(Gartsman et al., 1986; Ludlow et al., 1997). 1993). The flexor retinaculum as a whole can be divided
This article will review normal carpal tunnel anatomy into three parts from proximal to distal. The antebra-
and anatomic variations. The available surgical tech- chial fascia forms the proximal part of the flexor reti-
niques for carpal tunnel release are reviewed. Finally, naculum. A superficial fascial layer is inseparable from
the literature on morphologic, physiologic, and biome- the thickened deep investing antebrachial fascia, which
chanical alterations in the wrist after carpal tunnel re- is anterior to the median nerve and continuous with the
lease will be summarized. An appreciation of these transverse carpal ligament distally. The two layers sep-
changes may improve the surgical and postoperative arate to enclose the flexor carpi radialis tendon radially
and the contents of GuyonÕs canal and the flexor carpi
ulnaris tendon ulnarly. Thus, the deep investing ante-
*
Corresponding author. brachial fascia at this level is volar to the contents of the
E-mail address: [email protected] (E. Akelman). carpal tunnel and dorsal to GuyonÕs canal (Fig. 1)
0268-0033/$ - see front matter Ó 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0268-0033(03)00052-4
686 J.J. Brooks et al. / Clinical Biomechanics 18 (2003) 685–693

Fig. 1. Schematic illustration of the three parts of the flexor retinac-

ulum (FR). D: The distal part is the aponeurosis from which the thenar
and hypothenar muscles take origin. M: The middle part, or the
transverse carpal ligament proper, also gives origin to the thenar and
hypothenar muslces. P: The proximal part of the flexor retinaculum,
courses deep to the flexor carpi ulnaris (FCU) and flexor carpi radialis
(FCR) tendons.

(Cobb et al., 1993). The transverse carpal ligament

proper represents the middle third of the flexor reti-
naculum and forms the palmar ‘‘roof’’ of the carpal
tunnel. It inserts into the scaphoid tuberosity and ridge
of the trapezium radially and the hamulus and pisiform Fig. 2. (a) Palmar and (b) axial views of the transverse carpal ligamne
ulnarly where it is narrowest between the hamate hook (TCL). (a) The three bands of the TCL are illustrated with their bony
and trapezial ridge (Fig. 2). The distal third is the apo- attachments. H ¼ hamate hook, P ¼ pisiform, T ¼ trapezium, S ¼
neurosis between the thenar and hypothenar muscles, scaphoid.
from which these muscles originate. Tanabe maintains
that the distal part of the flexor retinaculum described into four groups. Division of the motor branch can be
by Cobb is actually separated from the transverse carpal extraligamentous, subligamentous, transligamentous,
ligament by a layer of adipose tissue (Tanabe and supraligamentous, or can originate from the ulnar bor-
Okutsu, 1997). der of the median nerve. Furthermore, the motor branch
The proximal to distal extent of the transverse carpal may arise in the forearm, or may be separated by the
ligament proper runs from the level of the distal pisiform persistent median artery or an aberrant muscle only to
proximally (11 mm distal to the capitate–lunate joint) to join distal to the transverse carpal ligament (Lanz, 1977).
the meta-diaphyseal junction of the third metacarpal The transverse carpal ligament is that part of the
distally (10 mm distal to the carpometacarpal joint of flexor retinaculum in the mid-third that has bone-to-
the long finger). The thickness of the flexor retinaculum bone attachments. The middle third is the transverse
over the carpal tunnel is 10 times that of the antebra- carpal ligament proper. The distal part is the aponeu-
chial fascia (Cobb et al., 1993). rosis between the thenar and hypothenar muscles, from
Contained within the carpal tunnel are the median which these muscles originate.
nerve and the nine extrinsic flexor tendons of the thumb It has been hypothesized that median nerve com-
and fingers (flexor pollicis longus tendon, four flexor pression most likely occurs in wrist flexion at the prox-
digitorum superficialis tendons, and four flexor digito- imal edge of the transverse carpal ligament where it joins
rum profundus tendons). The median nerve normally the deep investing fascia of the forearm, the anatomic
divides into six branches at the distal end of the flexor explanation for PhalenÕs sign (1996). Alternately, the
retinaculum––the recurrent motor branch, three proper median nerve may be compressed where the carpal
digital nerves and two common digital nerves. The motor tunnel is narrowest at the level of the hook of the ha-
branch of the median nerve usually innervates the thenar mate by either synovial hypertrophy or a space-occu-
muscles and the index and middle finger lumbricals. pying lesion (Cobb et al., 1993). Additionally, imaging
Sensory branches usually innervate the thumb, index, studies of median nerve motion during wrist flexion
middle, and radial half of the ring fingers. Lanz classified have demonstrated that patients with carpal tunnel
variations of median nerve anatomy in the carpal tunnel syndrome are more likely than normal patients to have
 J.J. Brooks et al. / Clinical Biomechanics 18 (2003) 685–693 687

limited median nerve motion in the carpal canal. The failure of surgery to relieve symptoms of carpal tunnel
nerve in normal patients moved radially and posteriorly syndrome (Tanabe and Okutsu, 1997; Phalen, 1996;
to a position interposed between the flexor tendons Cobb and Cooney, 1994). Tanabe noted that release of
during wrist flexion. The nerve in patients with carpal only the transverse carpal ligament caused the edges of
tunnel syndrome was more likely to remain in position the cut ligament to spread apart by 1.3
0.2 mm but
at the flexor retinaculum. The limited motion of the that complete release of both the transverse carpal lig-
nerve in these cases may predispose the nerve to com- ament and the distal aponeurosis between the thenar
pression during wrist flexion leading to carpal tunnel and hypothenar muscles caused more than 6.6
0.2 mm
symptoms (Allmann et al., 1997). Another study on of separation of the transverse carpal ligament edges
patients with non-specific forearm pain associated with and a corresponding significant decrease in carpal tunnel
median nerve compression found that patients experi- pressure (Tanabe and Okutsu, 1997; Okutsu et al.,
encing pain were more likely to have decreased median 1996).
nerve motion during wrist flexion (Greening et al., Okutsu et al. did an elegant study in 1989 in which
1999). carpal tunnel pressures were measured in multiple posi-
tions through an endoscopically placed catheter inside
the carpal tunnel of patients under local anesthesia
3. Carpal tunnel anatomic variations (Okutsu et al., 1989). In patients without carpal tunnel
syndrome the normal carpal tunnel pressure at rest was
Phalen originally reported on his 17-year experience 14.3
10 mmHg. In patients with carpal tunnel syn-
in 1966, suggesting that thickening of the flexor syno- drome, pressure in the carpal tunnel averaged 43.0
17.2
vium was the most common direct cause of median mmHg preoperatively in the resting position, and
nerve compression and carpal tunnel syndrome in his 206.2
51.6 mmHg with active grip. With passive flexion
patients (Phalen, 1996). Tanzer, however, noted three of and extension the pressures were 191.9
64 and
21 patients with cystic masses in the carpal tunnel and 222.4
44.2 mmHg, respectively. This represents a five-
seven of these 21 hands had various congenital anom- fold increase in pressure during grip or extremes of mo-
alies (Tanzer, 1959). Singer retrospectively looked for tion in patients with carpal tunnel syndrome. In contrast,
anatomic variations in 147 hands undergoing carpal immediately postoperatively the pressures were 6.2
5.5
tunnel release via an extended exposure (Singer and mmHg in the resting position, and 88
65 mmHg with
Ashworth, 2001). The authors noted anatomic varia- active grip, pressures in other positions were similarly
tions in 60 hands (41% of hands), one-third of which decreased, documenting clearly the physiologic effect of
had anatomic variations intrinsic to the carpal tunnel. carpal tunnel release.
The intrinsic variations included anomalous lumbrical The pressure in GuyonÕs canal decreases after endo-
origins in nine hands (6%), which could crowd the car- scopic or open carpal tunnel release. In a study by Ab-
pal tunnel during grip and finger flexion, most com- love et al. (1996), the authors studied 20 patients with
monly the long finger lumbrical. Siegel found that documented carpal tunnel syndrome. GuyonÕs canal
proximal lumbrical origin and/or lumbrical hypertrophy pressures decreased from 12.7
7.1 to 4.2
4.6 mmHg
could cause median nerve compression by a space-oc- after carpal tunnel release alone. This is consistent with
cupying effect (Siegel et al., 1995). Cobb confirmed that a previous study noting that 90% of patients with both
lumbrical excision would reduce carpal tunnel pressure median and ulnar compressive neuropathies at the wrist
during simulated grip in cadavera (Cobb et al., 1995). have relief of the ulnar nerve symptoms after isolated
Singer noted a flexor digitorum superficialis muscle belly carpal tunnel release (Silver et al., 1985).
present in six of 147 hands (4%) of patients undergoing
carpal tunnel release (Singer and Ashworth, 2001).
These variants may not likely be appreciated during 5. Surgical techniques
limited-exposure carpal tunnel release.
To serve as background for understanding the bio-
mechanical and anatomic changes after different types of
4. Pressure changes and mechanism of relief after carpal carpal tunnel release, a brief review is presented here. A
tunnel release thorough review of the methods for carpal tunnel release
and their pros and cons is beyond the scope of this
When considering complete release of the carpal paper.
tunnel, one must consider the relevant anatomy (Tanabe
and Okutsu, 1997; Cobb et al., 1993); neglecting to re- 5.1. Open release
lease the distal part of the flexor retinaculum or the
proximal part described above may fail to completely Since the first carpal tunnel surgery in 1924 by Her-
relieve pressure on the median nerve or may result in bert Galloway, numerous advances have been made to
688 J.J. Brooks et al. / Clinical Biomechanics 18 (2003) 685–693

refine this very common procedure (Amadio, 1995). In weakness or atrophy, and pillar pain was found between
general, carpal tunnel release can be performed using the groups. Scar tenderness was similar for both pro-
open or endoscopic procedures. The classic open and cedures at three weeks postoperatively, but differed
limited-open carpal tunnel release techniques involve significantly at week 11 with 61% of the open patients
complete division of the transverse carpal ligament and complaining of mild to severe tenderness. The endo-
the deep fascia of the forearm under direct visualization scopic group returned to work around day 14, as op-
(Green et al., 1999). The lesser level of difficulty and posed to day 28 for the open group.
shortened operative time make the open release the Both open and endoscopic carpal tunnel release are
procedure of choice with numerous hand surgeons. In effective for relieving symptoms, however, endoscopic
the majority of patients, open release techniques lead to release of the transverse carpal ligament may shorten the
symptomatic relief with a low complication rate. How- recovery time, permitting an earlier return to work and
ever, scar tenderness and grip weakness may occur after resumption of activities of daily living.
open release. In order to minimize these postoperative
complications, a limited transverse incision (2.0 cm or
less) is made in the same location as the classic open 6. Potential biomechanical complications of carpal tunnel
release, but ends approximately 1 cm distal to the distal release
wrist crease. Carpal tunnel release through a limited
open incision was performed by Atik et al. (2001) Generally reported complications of carpal tunnel
without damage to any neurovascular structures, and release include incomplete release, neuropraxia or injury
when performed from distal to proximal allowed for to the median or ulnar nerve, and inadvertent entrance
precise identification of structures at risk of injury (the into GuyonÕs canal, injury to digital nerves, the ulnar
superficial palmar arch, third common digital nerve, and artery and the superficial palmar arch (Lee et al., 1992;
aberrant motor branches of the median nerve) and im- Seiler et al., 1992). However, there exist several biome-
proves the safety of carpal tunnel release. chanical changes after carpal tunnel release, which may
be considered ‘‘complications’’, not merely ‘‘expected’’
5.2. Endoscopic release postoperative, changes.
The phenomenon of pillar pain is a frequent com-
In an attempt to reduce postoperative morbidity, en- plication reported after both open and endoscopic sur-
doscopic techniques were developed and have been gical techniques with an incidence between 6% and 36%
evolving since the late 1980s. The Agee single portal and (Katz et al., 1995; Mirza et al., 1995). Division of the
Chow two portal carpal tunnel release utilize smaller in- transverse carpal ligament through open or endoscopic
cisions and require less dissection of the subcutaneous techniques may lead to an increase in carpal arch width
tissue and structures overlying the transverse carpal liga- and carpal tunnel volume (Fig. 3), both of which may be
ment. However, since these are endoscopic procedures, responsible for postoperative pillar pain, and may also
visibility may be limited and incomplete release of the cause changes in other intercarpal articulations (Fisk,
transverse carpal ligament, nerve injury (median, ulnar, 1984; Fuss and Wagner, 1996; Kiritsis and Kline, 1995;
digital, palmar cutaneous branch of the median nerve), Seradge and Seradge, 1989; Garcia-Elias et al., 1989a).
arterial injury (ulnar, superficial palmar arch), hematoma, ‘‘Pillar pain’’ can delay return to work and resumption
and flexor tendon injury have been reported (Lee et al., of activities of daily living, foster emotional distress,
1992; Van Heest et al., 1995; Rowland and Kleinert, 1994). increase cost to the health care system, and potentiate
With endoscopic release techniques, there may be less loss of grip strength (Seradge and Seradge, 1989). Pillar
scar tenderness, and proponents suggest an earlier re- pain has been characterized as pain or tenderness in the
turn to work and activities of daily living compared to thenar or hypothenar eminence, or radial and ulnar
the open procedure. Trumble published a prospective, tenderness, and has been reported to subside by the
randomized, multicenter trial comparing endoscopic third postoperative month (Ludlow et al., 1997). Pain
versus open release. The endoscopic group had a more most commonly originates in the piso-triquetral joint,
rapid improvement in strength and scar tenderness than possibly due to alteration of the forces over the joint
the open group in the first three months after surgery. and/or displacement of the pisiform after releasing the
However, at one year follow-up there was no difference transverse carpal ligament (Seradge and Seradge, 1989).
in patient satisfaction between the endoscopic and open The etiology of pillar pain remains uncertain. There is
release (Trumble et al., 2002). Brown et al. (1993) in a currently no evidence that carpal tunnel release causes
prospective, randomized trial of 145 patients found no changes in the piso-triquetral articulation, although
significant difference in patient satisfaction and symp- GuyonÕs canal does dramatically change in shape after
tom relief between the two portal endoscopic and the carpal tunnel release (Richman et al., 1989).
classic open release at 84 days postoperatively. Addi- Current work has suggested four potential etiologies
tionally, no significant difference in sensibility, thenar for pillar pain: muscular or ligamentous, neurogenic,
 J.J. Brooks et al. / Clinical Biomechanics 18 (2003) 685–693 689

1989; Garcia-Elias et al., 1989b). Thus, division of this

ligament could alter the normal function of the wrist
and contribute to such postoperative complications as
pillar pain and grip weakness.

7.1. Carpal stiffness

Carpal stiffness is obtained by differentiating strain

energy with respect to displacement of centroids of ad-
jacent rigid bodies. By applying a dorsopalmar com-
pressive load, a calculated displacement force on
adjacent carpal bones can be described. Dorsopalmar
compression tests when the flexor retinaculum has been
sectioned demonstrate that carpal stiffness decreased by
an average of 7.8% (232–214 N/mm), indicating that the
flexor retinaculum plays a minor role in transverse sta-
bility of the carpus and should be regarded as a pulley
for the flexor tendons rather than an intercarpal stabi-
lizer (Garcia-Elias et al., 1989b). The transverse inter-
carpal ligaments, especially the capito-hamate ligament,
are essential in providing stability to the carpal tunnel,
since their division in the presence of an intact flexor
Fig. 3. (a) The carpal arch width depicted on an axial view of the carpal
retinaculum leads to dorsopalmar compressive instabil-
tunnel (H ¼ hamulus, P ¼ pisifrom, T ¼ trapezium, S ¼ scaphoid). ity of the carpus under applied stress.
(b) Carpal tunnel volume (dark area) and the contents of the carpal
tunnel––The tendons of the flexor digitorum profundus, flexor digi- 7.2. Muscular and tendon effects
torum superficialis, flexor pollicis longus, and the the median nerve.
The ulnar artery and nerve and the flexor carpi radialis tendon are
shown for orientation.
Fuss and Wagner dissected five fresh frozen forearms
and hands to demonstrate muscle attachments to the
flexor retinaculum prior to studying joint mechanics
edematous, and alteration of the carpal arch structure before and after carpal tunnel release. Bisection of the
(Ludlow et al., 1997). Muscular or ligamentous causes flexor retinaculum caused certain muscles to lose their
of pillar pain may be related to relaxation of the intrinsic anatomic attachments due to muscular pull shifted dis-
muscles of opposition and pinch with sectioning of the tally and radially or ulnarly (Fuss and Wagner, 1996).
transverse carpal ligament, and subsequent migration of Muscle shortening occurred in the superficial head of the
the transverse carpal ligament toward its osseous origins flexor pollicis brevis 25% (relative to rest length), the
(Hunter, 1991). Cutaneous nerve branches in the palmar ulnar part of the abductor pollicis brevis 20% (opposi-
and subcutaneous tissue may be injured when incising in tion and adduction), the opponens pollicis 20%, and the
the ‘‘critical pillar rectangle’’, which was defined by opponens digiti minimi 10%. This causes a loss of
Wilson as the wrist flexion crease proximally, 1 cm distal muscle length leading to a loss of muscle strength. Ad-
to the hook of hamate, ulnar border of the hamate ditionally, release of the transverse carpal ligament may
ulnarly, and the scaphoid tubercle radially (Wilson, disrupt the alignment and tracking of the piso-triquetral
1994). Postoperative edema superficial to the transverse joint causing pain (Seradge and Seradge, 1989). This
carpal ligament may also cause pain in the thenar or piso-triquetral tracking or alignment problem could be a
hypothenar areas. Resolution of this swelling is usually possible etiology of pillar pain.
coincident with relief of pillar pain (Green et al., 1999). Another change noted by many authors is that the
flexor tendons in the carpal tunnel displace palmarward
after carpal tunnel release (Amadio, 1995). The trans-
7. Biomechanical change after carpal tunnel release verse carpal ligament has been suggested as an impor-
tant pulley in the flexor tendon system that prevents
A number of studies have examined the effect of di- bowstringing. Division of the transverse carpal ligament
viding the transverse carpal ligament on the carpal arch. through an open, limited open, or endoscopic release,
The transverse carpal ligament has been shown to serve may significantly increase flexor tendon excursion during
three functions: anchor thenar and hypothenar muscu- wrist flexion. Increased excursion, or the distance the ten-
lature, provide transverse stability to the carpus, and act don travels relative to the jointÕs center of rotation, is
as a pulley for the flexor tendons (Seradge and Seradge, noted after 20–30° wrist flexion (Kiritsis and Kline, 1995).
690 J.J. Brooks et al. / Clinical Biomechanics 18 (2003) 685–693

Thus, as the distance increases between the tendon and levels of the carpal tunnel significant error could be in-
the center of rotation in the wrist joint due to bow- troduced into the measurements of carpal arch width.
stringing, the distance the tendon travels with chang- The transverse carpal ligament normally attaches on
ing wrist position increases. The palmar displacement of the pisiform, hook of the hamate, scaphoid and trapez-
the flexor tendons with wrist flexion could be another ium, maintaining a concavity to the carpal arch and
potential cause of postoperative wrist flexion weakness, carpal tunnel. When this ligament is divided, the concave
and potentially reinforces the need for a short period of arch flattens and the distance between the carpal inser-
immobilization to allow sufficient time for intrinsic re- tions increases. Fisk demonstrated an average increase of
constitution of the transverse carpal ligament, although 3 mm between the scaphoid and pisiform after carpal
splinting after carpal tunnel release has not been proven tunnel release (Fisk, 1984). Garcia-Elias measured the
beneficial (Bury et al., 1995). Conversely, functional width of the carpal arch using the trapezio-hook of ha-
gripping activities and maximum grip strength are ac- mate distance before and after transverse carpal ligament
complished with the wrist in an extended position, thus release. When the ligament was intact, the distance de-
palmar displacement of the flexor tendons during wrist creased between these two bones in both wrist flexion
flexion is unlikely clinically significant. and extension. However, when the transverse carpal
ligament was released, the distance significantly in-
creased an average of 11% (Garcia-Elias et al., 1992).
8. Morphologic changes following carpal tunnel release Gartsman et al. (1986) did a retrospective review of
patients who had undergone open carpal tunnel release
8.1. Carpal arch and carpal tunnel volume and either had standardized carpal tunnel view radio-
graphs taken before and after the procedure was done,
Cobb et al. (1993) determined normal carpal tunnel or a radiograph of the unoperated side was used as a
dimensions by injecting contrast material into the carpal control. In this study, the lack of preoperative assess-
tunnel and examining antero-posterior (AP) radio- ment of the operative wrist is the major criticism. The
graphs. On the AP view, the carpal tunnel is shaped like carpal arch width was measured between the palmar tips
an hourglass (Fig. 4), with the narrowest part at the level of the trapezial ridge and hook of the hamate with the
of the hook of the hamate. The mean width of the carpal wrist extended 50°.
tunnel was 25
1.2 mm proximally, 20
1.2 mm at the Range of motion did not change significantly after
hook of the hamate, and 25
1.5 mm at its distal extent, carpal tunnel release using the opposite side as a control.
a significant difference ðP < 0:0001Þ. This is an impor- Of 50 operated wrists, 47 had widening of the carpal
tant feature to note, as it is of utmost importance when arch width and three had no change, ranging from 0% to
measuring carpal arch width pre and postcarpal tunnel 52% widening with an average widening of 13.6%, or 2.9
release––if axial slices are at different proximal–distal mm. The range of increase in absolute widening of the
carpal arch was 0–8.5 mm. However, the authors did not
examine variability or repeatability of measurements.
This study differs from others in that it found a greater
magnitude of arch widening after carpal tunnel release,
perhaps because of the wrist position (50° extension)
being different from wrist position in other studies using
magnetic resonance imaging (MRI) or computed to-
mography (CT) in neutral or only slight extension po-
sitions.
Gartsman found that an increase in carpal arch width
greater than 20% over control correlated with a statis-
tically significant decrease in grip strength of 25.85%.
Patients with greater widening did not have an increased
incidence of postoperative wrist pain. ‘‘Columnar pain’’
did not correlate with arch widening either.
In 1992, Viegas et al. (1992) examined carpal tunnel
radiographs before and after endoscopic carpal tunnel
release using the same technique for measurement and
radiographic documentation as Gartsman et al. The
authors noted an increase of 1.7 mm (7% increase) in
carpal arch width 10 days after surgery. In contrast to
Fig. 4. The carpal tunnel is hourglass-shaped, with the narrowest GartsmanÕs findings, most patients had an increase in
portion located between the hook of the hamate and trapezial ridge. arch width of 0–10%, with very few wrists increasing
 J.J. Brooks et al. / Clinical Biomechanics 18 (2003) 685–693 691

more than 20% in arch width postoperatively. One yonÕs canal was also noted by Ablove et al. (1994). Silver
might conclude from comparing ViegasÕ and GartsmanÕs noted resolution of ulnar compressive neuropathy at the
studies that fewer patients would be expected to have wrist after isolated carpal tunnel decompression (Silver
impaired grip after endoscopic carpal tunnel release et al., 1985).
compared with open carpal tunnel release, since fewer of Kato and colleagues used MRI before and after en-
the wrists in the former group had increases in arch doscopic carpal tunnel release in 10 hands to determine
width of more than 10%. morphologic changes in the carpal tunnel after endo-
In 1987 Richman et al. (1987) introduced MRI as a scopic release (Kato et al., 1994). The authors divided
means of evaluating carpal tunnel volume. The study the carpal tunnel into a palmar part and a dorsal part
determined that there was no significant difference be- using a line from the beak of the trapezium to the ha-
tween MRI and direct cadaveric measurements of carpal mulus, the ‘‘H–T line’’. Of note, the majority of the in-
tunnel volume (with a silicone injection technique) and crease in volume postoperatively was found to come
carpal arch width (measured using calipers accurate to from an increase in the palmar area, indicating that
0.5 mm), using a standard protocol for direct measure- widening the carpal arch is not a major mechanism for
ments on cadavera and MRI data from cadaver wrists. increase in carpal tunnel volume after endoscopic carpal
This indicated excellent intra and inter-observer reli- tunnel release.
ability of this MRI measurement technique, and excel- Specifically, the dorsal cross sectional area increased
lent reproduction by MRI of true direct measurements by about 3% (from 200
20 to 206
30 mm2 ). The
of volume and width of the carpal tunnel. palmar cross sectional area accounted for most of the
Two years later, Richman examined the effect of open volume increase: it was 32
8 mm2 preoperatively, and
carpal tunnel release on carpal canal volume, carpal 114
13 mm2 postoperatively, and increase of 360%.
arch width, and position of the canal contents in 15 This data implies that the volume increase after endo-
wrists assessed by MRI (Richman et al., 1989). Carpal scopic carpal tunnel release comes from a palmar
canal volume in treated hands increased from 6.3
1.0 opening of the edges of the transverse carpal ligament,
to 7.8
1.5 ml after surgery, while untreated hands had increasing the palmar area of the carpal tunnel, without
no change in volume of the carpal tunnel. These mea- much widening of the carpal arch.
surements were significant ðP < 0:001Þ. The findings of Kato et al. are different from previous
These volume differences persisted at an eight-month authors (Gartsman et al., 1986) who did note an increase
examination in eight hands as well. The volume increase in carpal arch width (increased distance between the
seemed to come from a more circular shape of the carpal hamulus and trapezial ridge). Perhaps the carpal arch
tunnel on axial images, with the transverse carpal liga- does widen a small amount, as other studies suggest, but
ment becoming more convex, while the distance between KatoÕs sample size (10 hands) was too small to detect a
the trapezial ridge and the hook of the hamate increased significant difference. The authors did note and increase
to a lesser degree (more increase in the antero-posterior in arch width (H–T line) from 22.1 mm preoperatively to
than in the medial–lateral dimension of the carpal tun- 23.8 mm postoperatively without statistical significance.
nel). The median nerve displaced an average of 3.5
1.9 Ablove et al. (1994) reported the MRI-documented
mm anteriorly postoperatively when compared with morphologic changes of 18 wrists in 17 patients after
control wrists, which was significant. The carpal arch Agee endoscopic (Agee et al., 1992) or subcutaneous,
width increased a small amount after carpal tunnel re- two-incision (Chow, 1989) (non-endoscopic) carpal
lease, by 6.3
4.6% which was statistically significant. tunnel release. The authors measured carpal arch width
However, in the eight hands studied eight months at the same level as other studies, at the level of the
postoperatively, there was no difference in carpal arch trapezial ridge and hamate hook. Carpal canal volume
width from preoperative values, implying that the carpal was nearly identical pre and postoperatively when en-
arch recoils somewhat by eight months postoperatively. doscopically released wrists are compared with the re-
However, the authors did not mention if the identical sults of Richman et al. (6.1
2.2 preoperatively and
proximal–distal level was used for measurement of the 7.5
2.5 ml postoperatively). Carpal arch width in-
pre and postoperative carpal arch width, introducing the creased from 22.4
2.5 to 23.3
3.3 mm in the endo-
possibility of error from comparing axial MRI cuts from scopic group and from 23.5
2.0 to 24.1
2.1 mm in the
different levels in the carpal tunnel which could hide or subcutaneous group at final follow-up examination two
exaggerate differences in this parameter. The images in years postoperatively, but the differences were not sta-
RichmanÕs paper clearly demonstrated a difference in tistically significant. Of note, this study also noted pal-
shape of GuyonÕs canal from flat to oval after release of mar displacement of the median nerve measuring 0.6–
the transverse carpal ligament, a finding which is sup- 1.0 mm after endoscopic or subcutaneous carpal tunnel
ported by the measurement of decreased pressure in release, confirming the findings of Richman et al.
GuyonÕs canal after carpal tunnel release (Ablove et al., However, the palmar displacement of the median nerve
1996; Silver et al., 1985). The change in shape of Gu- is less than that noted after open carpal tunnel release.
692 J.J. Brooks et al. / Clinical Biomechanics 18 (2003) 685–693

Roger et al. (1985) studied axial CT scans at three Acknowledgement

levels and in three wrist positions in 34 wrists immedi-
ately before and two to three months after carpal tunnel We would like to thank Ted Trafton for his tireless
release. They concluded that carpal tunnel release in- illustrative effort.
creased the carpal canal volume at all positions of flex-
ion and extension, while the bony anatomy remained
unchanged. Thus, the mechanism of increased volume of
the carpal tunnel was by an increase in convexity of the References
transverse carpal ligament and not by an increase in
carpal arch width. Similar CT scan results were noted by Ablove, R.H., Moy, O.J., Peimer, C.A., Wheeler, D.R., Diao, E., 1996.
Schmitt et al. (1988), who noted an anterior ‘‘prolapsed’’ Pressure changes in GuyonÕs canal after carpal tunnel release. J.
of the volar boundary of the carpal tunnel in most pa- Hand Surg. [Br.] 21, 664–665.
tients to account for the increase in carpal tunnel vol- Ablove, R.H., Peimer, C.A., Diao, E., Oliverio, R., Kuhn, J.P., 1994.
Morphologic changes following endoscopic and two-portal subcu-
ume after carpal tunnel release in 90% of patients in taneous carpal tunnel release. J. Hand Surg. [Am.] 19, 821–826.
their study. However, 10% of patients also had an in- Agee, J.M., McCarroll Jr., H.R., Tortosa, R.D., Berry, D.A., Szabo,
crease in carpal arch area (increased distance of a R.M., Peimer, C.A., 1992. Endoscopic release of the carpal tunnel:
line from the hamate hook to the trapezial ridge on axial a randomized prospective multicenter study. J. Hand Surg. [Am.]
CT scan cuts), suggesting that there is a small minority 17, 987–995.
Allmann, K.H., Horch, R., Uhl, M., et al., 1997. MR imaging of the
of patients who experience increase in carpal tunnel carpal tunnel. Eur. J. Radiol. 25, 141–145.
volume by this additional mechanism. As Gartsman Amadio, P.C., 1995. The first carpal tunnel release? J. Hand Surg. [Br.]
noted a significant decrease in grip strength in patients 20, 40–41.
whose carpal arch width increases more than 20%, Atik, T.L., Smith, B., Baratz, M.E., 2001. Risk of neurovascular injury
perhaps there is a subset of patients at risk for this with limited-open carpal tunnel release: defining the ‘‘safe-zone’’. J.
Hand Surg. [Br.] 26, 484–487.
‘‘pathologic’’ widening of the arch. Brown, R.A., Gelberman, R.H., Seiler III, J.G., et al., 1993. Carpal
tunnel release A prospective, randomized assessment of open and
endoscopic methods. J. Bone Joint Surg. [Am.] 75, 1265–1275.
9. Conclusion Bury, T.F., Akelman, E., Weiss, A.P., 1995. Prospective, randomized
trial of splinting after carpal tunnel release. Ann. Plast. Surgery 35,
19–22.
Carpal tunnel surgery has been performed for nearly Chow, J.C., 1989. Endoscopic release of the carpal ligament: a new
80 years, yet despite the majority of patients who re- technique for carpal tunnel syndrome. Arthroscopy 5, 19–24.
cover without complication, there are still a significant Cobb, T.K., Cooney, W.P., 1994. Significance of incomplete release of
number that suffer postoperative pain and weakness. A the distal portion of the flexor retinaculum. Implications for
successful decompression of the median nerve provides endoscopic carpal tunnel surgery. J. Hand Surg. [Br.] 19, 283–285.
Cobb, T.K., An, K.N., Cooney, W.P., 1995. Effect of lumbrical muscle
symptomatic relief of carpal tunnel symptoms, and incursion within the carpal tunnel on carpal tunnel pressure: a
generally allows patients to return to activities of daily cadaveric study. J. Hand Surg. [Am.] 20, 186–192.
living and employment. However, complete release of Cobb, T.K., Dalley, B.K., Posteraro, R.H., Lewis, R.C., 1993.
the carpal tunnel probably has a small but significant Anatomy of the flexor retinaculum. J. Hand Surg. [Am.] 18, 91–99.
effect on carpal biomechanics and morphology. The Einhorn, N., Leddy, J.P., 1996. Pitfalls of endoscopic carpal tunnel
release. Orthop. Clin. North Am. 27, 373–380.
small but appreciable increase in carpal arch width, as Fisk, G.R., 1984. The influence of the transverse carpal ligament
well as increased tendon excursion may have conse- (flexor retinaculum) on carpal stability. Ann. Chir. Main. 3, 297–
quences on carpal kinematics. Postoperative complica- 299.
tions of pillar pain and grip weakness may be Fuss, F.K., Wagner, T.F., 1996. Biomechanical alterations in the
attributed to a change in the carpal articulation or carpal arch and hand muscles after carpal tunnel release: a further
approach toward understanding the function of the flexor retinac-
changes in the flexor tendon pulley system. Do the ulum and the cause of postoperative grip weakness. Clin. Anat. 9,
changes described warrant repair of the transverse car- 100–108.
pal ligament or should a limited release of the carpal Garcia-Elias, M., An, K.N., Cooney III, W.P., Linscheid, R.L., Chao,
tunnel be performed to prevent troublesome postoper- E.Y., 1989a. Stability of the transverse carpal arch: an experimental
ative complications? It is generally accepted that trans- study. J. Hand Surg. [Am.] 14, 277–282.
Garcia-Elias, M., An, K.N., Cooney, W.P., Linscheid, R.L., Chao,
verse carpal ligament lengthening, rather than E.Y., 1989b. Transverse stability of the carpus. An analytical study.
simple division, is unnecessary to stabilize the carpal J. Orthop. Res. 7, 738–743.
arch, however, it may be reasonable to prevent flexor Garcia-Elias, M., Sanchez-Freijo, J.M., Salo, J.M., Lluch, A.L., 1992.
tendon bowstringing. Perhaps excessive carpal mor- Dynamic changes of the transverse carpal arch during flexion-
phologic change is preventable by ligament-preserving extension of the wrist: effects of sectioning the transverse carpal
ligament. J. Hand Surg. [Am.] 17, 1017–1019.
procedures? A study of carpal kinematics after car- Gartsman, G.M., Kovach, J.C., Crouch, C.C., Noble, P.C., Bennett,
pal tunnel release is warranted to answer these ques- J.B., 1986. Carpal arch alteration after carpal tunnel release. J.
tions. Hand Surg. [Am.] 11, 372–374.
 J.J. Brooks et al. / Clinical Biomechanics 18 (2003) 685–693 693

Green, D, Hotchkiss, R.N., Pederson, W.C., 1999. In: GreenÕs nation by magnetic resonance imaging three-dimensional recon-
Operative Hand Surgery, vol. 2. Churchill Livingstone, Philadel- struction. J. Hand Surg. [Am.] 12, 712–717.
phia, pp. 1406–1417. Roger, B., Chaise, F., Laval-Jeantet, M., 1985. X-ray computed
Greening, J., Smart, S., Leary, R., Hall-Craggs, M., OÕHiggins, P., tomographic analysis of structural changes of the carpal bones,
Lynn, B., 1999. Reduced movement of median nerve in carpal after section of the anterior annular ligament. J. Radiol. 66, 693–
tunnel during wrist flexion in patients with non-specific arm pain. 697.
Lancet 354, 217–218. Rowland, E.B., Kleinert, J.M., 1994. Endoscopic carpal-tunnel release
Hoppenfeld Sd, P., 1984. Surgical Exposures in Orthopaedics, The in cadavera. An investigation of the results of twelve surgeons with
Anatomic Approach. JB Lippincott, Philadelphia, pp. 162–165. this training model. J. Bone Joint Surg. [Am.] 76, 266–268.
Hunter, J.M., 1991. Recurrent carpal tunnel syndrome, epineural Schmitt, R., Lucas, D., Buhmann, S., Lanz, U., Schindler, G., 1988.
fibrous fixation, and traction neuropathy. Hand Clin. 7, 491–504. Morphometry using computerized tomography of the wrist in
Kato, T., Kuroshima, N., Okutsu, I., Ninomiya, S., 1994. Effects of idiopathic carpal tunnel syndrome. Comparison of pre- and
endoscopic release of the transverse carpal ligament on carpal canal postoperative findings. Handchir. Mikrochir. Plast. Chir. 20, 41–
volume. J. Hand Surg. [Am.] 19, 416–419. 46.
Katz, J.N., Fossel, K.K., Simmons, B.P., Swartz, R.A., Fossel, A.H., Seiler III, J.G., Barnes, K., Gelberman, R.H., Chalidapong, P., 1992.
Koris, M.J., 1995. Symptoms, functional status, and neuromuscu- Endoscopic carpal tunnel release: an anatomic study of the two-
lar impairment following carpal tunnel release. J. Hand Surg. [Am.] incision method in human cadavers. J. Hand Surg. [Am.] 17, 996–
20, 549–555. 1002.
Kiritsis, P.G., Kline, S.C., 1995. Biomechanical changes after carpal Seradge, H., Seradge, E., 1989. Piso-triquetral pain syndrome after
tunnel release: a cadaveric model for comparing open, endoscopic, carpal tunnel release. J. Hand Surg. [Am.] 14, 858–862.
and step-cut lengthening techniques. J. Hand Surg. [Am.] 20, 173– Siegel, D.B., Kuzma, G., Eakins, D., 1995. Anatomic investigation of
180. the role of the lumbrical muscles in carpal tunnel syndrome. J.
Lanz, U., 1977. Anatomical variations of the median nerve in the Hand Surg. [Am.] 20, 860–863.
carpal tunnel. J. Hand Surg. [Am.] 2, 44–53. Silver, M.A., Gelberman, R.H., Gellman, H., Rhoades, C.E., 1985.
Lee, D.H., Masear, V.R., Meyer, R.D., Stevens, D.M., Colgin, S., Carpal tunnel syndrome: associated abnormalities in ulnar nerve
1992. Endoscopic carpal tunnel release: a cadaveric study. J. Hand function and the effect of carpal tunnel release on these abnormal-
Surg. [Am.] 17, 1003–1008. ities. J. Hand Surg. [Am.] 10, 710–713.
Ludlow, K.S., Merla, J.L., Cox, J.A., Hurst, L.N., 1997. Pillar pain as Singer, G., Ashworth, C.R., 2001. Anatomic variations and carpal
a postoperative complication of carpal tunnel release: a review of tunnel syndrome: 10-year clinical experience. Clin. Orthop., 330–
the literature. J. Hand Ther. 10, 277–282. 340.
Mirza, M.A., King Jr., E.T., Tanveer, S., 1995. Palmar uniportal Tanabe, T., Okutsu, I., 1997. An anatomical study of the palmar
extrabursal endoscopic carpal tunnel release. Arthroscopy 11, 82– ligamentous structures of the carpal canal. J. Hand Surg. [Br.] 22,
90. 754–757.
Okutsu, I., Hamanaka, I., Tanabe, T., Takatori, Y., Ninomiya, S., Tanzer, R., 1959. The carpal-tunnel syndrome: A clinical and
1996. Complete endoscopic carpal tunnel release in long-term anatomical study. J. Bone Joint Surg. [Am.] 41A, 626–634.
haemodialysis patients. J. Hand Surg. [Br.] 21, 668–671. Trumble, T.E., Diao, E., Abrams, R.A., Gilbert-Anderson, M.M.,
Okutsu, I., Ninomiya, S., Hamanaka, I., Kuroshima, N., Inanami, H., 2002. Single-portal endoscopic carpal tunnel release compared with
1989. Measurement of pressure in the carpal canal before and after open release: a prospective, randomized trial. J. Bone Joint Surg.
endoscopic management of carpal tunnel syndrome. J. Bone Joint [Am.] 84-A, 1107–1115.
Surg. [Am.] 71, 679–683. Van Heest, A., Waters, P., Simmons, B., Schwartz, J.T., 1995. A
Phalen, G.S., 1996. The carpal-tunnel syndrome. Seventeen yearsÕ cadaveric study of the single-portal endoscopic carpal tunnel
experience in diagnosis and treatment of six hundred fifty-four release. J. Hand Surg. [Am.] 20, 363–366.
hands. J. Bone Joint Surg. [Am.] 48, 211–228. Viegas, S.F., Pollard, A., Kaminksi, K., 1992. Carpal arch alteration
Richman, J.A., Gelberman, R.H., Rydevik, B.L., et al., 1989. Carpal and related clinical status after endoscopic carpal tunnel release. J.
tunnel syndrome: morphologic changes after release of the trans- Hand Surg. [Am.] 17, 1012–1016.
verse carpal ligament. J. Hand Surg. [Am.] 14, 852–857. Wilson, K.M., 1994. Double incision open technique for carpal tunnel
Richman, J.A., Gelberman, R.H., Rydevik, B.L., Gylys-Morin, V.M., release: an alternative to endoscopic release. J. Hand Surg. [Am.]
Hajek, P.C., Sartoris, D.J., 1987. Carpal tunnel volume determi- 19, 907–912.