High-resolution Ultrasonography of the Carpal Tunnel W. Buchberger, MD,• G. Schon, MD,• K. Strasser, MD,"' W. Jungwirth, Mot
Twenty-eight wrists of 25 patients with carpal tunnel syndrome (CTS) and 28 wrists of 14 normal control subjects were studied with high-frequency real-time ultrasonography. Three general findings could be ob· served in CTS, regardless of its cause: swelling of the median nerve at the entrance of the carpal tunnel; flattening of the median nerve in the distal carpal
C
arpal tunnel syndrome (CTS) is a relatively common disease characterized by burning pain, numbness, and paresthesia in the distribution of the median nerve. In advanced cases, thenar muscle weakness or atrophy is found. The disease occurs predominantly in patients in their fourth to sixth decades, and women are affected two to five times more often than men. Bilateral involve· ment is not uncommon. 1•2 The condition generally is attributed to compression of the median nerve in the carpal tunnel beneath the flexor retinaculum. However, the exact pathogenesis often remains unknown. 3 Operative treatment consists of decompression of the nerve by sectioning the transverse carpal ligament. 4 Conservative management includes splinting of the wrist, injection of steroids into the carpal tunnel, and administration of nonsteroidal anti-inflammatory drugs. 5 The diagnosis of CTS is mainly based on clinical findings, electromyography (EMG), and nerve conduction studies. These findings can be equivocal, and CTS can be confused with cervical root compression, thoReceived October 15, 1990, from the Departments of "Radiology and tp]astic and Reconstructive Surgery, University of Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria. Revised manuscript accepted for publication February 8, 1991. Address corre!lpondence and reprint requests to Dr. W. Buchberger: Department of Radiology, University of lnMbruck, Anichstrasse 35, A-6020 Innsbruck, Austria.
tunnel; and increased palmar flexion of the transverse carpal ligament. Quantitative analysis proved these findings to be significant. We conclude that high-resolution sonography is able to diagnose median nerve compression .in the carpal tunnel syndrome and to detect some of its potential causes.
racic outlet syndrome, and other peripheral nerve compression syndromes. Because there are many possible causes of median nerve compression, the etiologic basis of CTS cannot be evaluated preoperatively by means of clinical and EMG studies alone. Magnetic resonance imaging is the only imaging modality that has proved to be of diagnostic value in carpal tunnel syndrome. However, the method is expensive, relatively time consuming, and still not generally available. Early experiences with high-resolution real-time sonography in the imaging of the normal carpal tunnel were encouraging. The purpose of this study is to establish objective criteria for the ultrasound diagnosis of carpal tunnel syndrome.
PATIENTS AND METHODS Twenty-eight wrists of 25 nonselected patients with clinical symptoms of carpal tunnel syndrome were studied. All of the patients underwent EMG and measurement of nerve conduction delay. Seventeen patients were women and 8 were men; they ranged in age from 38 years up to 85 years (mean, 61 years). The right hand was affected in 16 cases and the left hand in 4 cases. Four patients had bilateral carpal tunnel syndrome; 4 had had operation on the contralateral wrist previously. Three patients with
© 1991 by the American Institute of Ultrasound in Medicine• J Ultrasound Med 10:531- 537, 1991•0278-4297/91/$3.50
532
J Ultrasound Med 10:531- 537, 1991
ULTRASONOGRAPHY OF THE CARPAL TUNNEL
unrelieved or recurrent CTS after surgical treatment were excluded from this study. For comparison with the normal population, 28 wrists of 14 asymptomatic volunteers (9 women and S men) were studied. All ultrasound examinations were performed with a real-time scanner (Picker LSC 9500) and a 7.5 MHz linear probe. A stand-off was used for better near-field resolution. Axial scans of the wrist at the level of the distal radius, pisiform, and hook of the hamate and sagittal sections through the carpal tunnel were recorded with a multiformat camera . At each level, the lateral and the anteroposterior diameters of the median nerve were measured, and the cross-sectional area was calculated, assuming an elliptic shape. To determine the bowing of the flexor retinaculum, a straight line was drawn between the attachments of the ligament to the tubercle of the trapezium and the hook of the hamate, and the distance from this line to the palmar apex of the flexor retinaculum was meas ~ ured (palmar displacement, PD) (Fig. 1).
A
ua
un
fr
mn
fer
fpl fds
p
fdp
Results s
Normal Ultrasound Anatomy The bony and ligamentous walls of the carpal tunnel and its contents are best studied on axial planes (Figs. 2 and 3). The dorsal border of the tunnel is formed by the scaphoid, triquetrum, lunate, and pisiform proximally and the trapezium, trapezoid, capitate, and hamate distally. The surfaces of the carpal bones appear as bright, hyperechoic lines with posterior shadow. The hyperechoic flexor retinaculum extends from the tubercle of the scaphoid to the pisiform and from the tubercle of the trapezium to the hook of the hamate. It is either straight or slightly bowed, and its thickest portion is seen at the level of the distal carpal tunnel. The superficial and the deep flexor tendons and the Figure 1 Schematic diagram showing the measuring points of palmar displacement (PD) of the flexor retinaculum. tm, Trapezium; td, trapezoid; c, capitate; h, hamate; ha, hook of the hamate; fr, flexor retinaculurn. fr
= ~ =---1 ,.;.;. .;.;. ,;.;. .;.;. ;.;.,.;.;. .;.;.,;.;. .; . .;.;. ;.~
-~~~~
PD ha
c
tr
B
Figure 2 Normal carpal tunnel. Axial ultrasound image (A) and schematic diagram (B) at the level of the pisifonn bone. S, Scaphoid; C, capitate; tr, triquetrum; p, pislform; fr, flexor retinaculum; fdp, flexor digitorum profundus tendons; fds, nexor digitorum superfidalis tendons; fpl, Aexor pollicis Ion· gus tendon; fer, flexor carpi radialis tendon; mn, median nerve; ua, ulnar artery; un, ulnar nerve. separate tendon of the flexor pollicis longus appear as round or ovoid shaped structures 4- 5 mm in diameter. Their echogenicity depends on the angle of incidence of the ultrasound beam. The tendons are surrounded by hypoechoic structures that are believed to represent all or part of the synovial sheath. The tendon of the flexor carpi radialis muscle lies in a separate fibro• osseous tunnel between a superficial and a deep layer of the transverse carpal ligament and within the vertical groove of the trapezium. The median nerve lies just beneath the flexor retlnaculum between the tendon of the flexor pollicis longus and the tendons of the flexor digitorum superficialis muscles. It is ovoid in cross section, with its long axis usually oriented parallel to the flexor retinaculum. The median nerve appears echogenic when the ultra sound beam is perpendicular to its surface (Fig. 2A). A slight obliquity of the beam results in a hypoecho· genicity of the nerve, which then seems to be surrounded by a narrow echogenic rim possibly rep ~ resenting the perineural fibrous tissue (Fig. 3A). Super-
J Ultrasound Med 10:531 - 537, 1991
BUCHBERGER ET AL
A ----ua ---un
---ha - - - fds fdp
Im
td
c
h
B
Figure 3 Normal carpal tunnel. Axial ultrasound image (A) and schematic diagram (B) at the level of the hook of the hamate. tm, trapezium; td, trapezoid; c, capitate bone; h, hamate; ha, hook of the hamate; fr, flexor retinaculum; ua, ulnar artery; un, ulnar nerve.
ficial to the flexor retinaculum, the ulnar nerve and artery and the tendon of the palmaris longus muscle can be identified. On sagittal scans, the median nerve appears as a band-like structure just superficial to the Figure 4 Normal carpal tunnel, sagittal section. The median nerve (arrowfreads) lies superficial to the flexor tendons (FT)..
533
Figure 5 Carpal tunnel syndrome. Axial section at the level of the hook of the hamate showing increased flattening of the median nerve (arrows) owing to compression.
flexor tendons. Deep to the tendons, the contour of the radial epiphysis and the carpal bones is apparent (Fig. 4). General Fi11di11gs in CTS In all cases of carpal tunnel syndrome, a more or less marked flattening of the median nerve in the distal carpal tunnel owing to compression between the flexor tendons and the trans· verse carpal ligament was evident (Fig. 5). Sagittal scans usually showed either a diffuse thinning or a localized constriction of the nerve. ln 22 wrists, an obvious enlargement of the median nerve at the level of the radiocarpal joint and inside the proximal carpal tunnel in comparison to its size at the distal radius was evident (Fig. 6). In 12 cases the swelling was segmental, with a bulbous shape of the nerve suggesting a pseudoneuroma (Fig. 7). More dif• fuse enlargement of the nerve was observed in 10 cases. Figure 6 Carpal tunnel syndrome. Axial section through the proximal carpal tunnel. The median nerve (arrows) is markedly increased in size, and the transverse carpal ligament (arrowf1ead) $hows increased palmar bowing.
534
ULTRASONOGRAPHY OF TH E CARPAL TUNNEL
Figure 7 Carpal tunnel syndrome, sagittal section. Bulbous swelling of the median nerve (arrowf1cads) is seen at the proximal carpal tunn el and flattening of the nerve is apparent at the distal carpal tunnel.
Another relatively constant finding was an increased palmar bowing of the transverse carpal ligament, which is normally straight or only slightly bowed (Figs. l, 8, 9). Etiologic findings of median nerve compression were observed in 9 cases. In three wrists, the flexor tendons were surrounded by an unusually thick hypoechogenic halo, probably representing proliferated synovial tissue. The sono· graphic diagnosis of tenosynovitis was confirmed at operation. A ganglion cyst of the flexor tendons extend6ng into the proximal part of the carpal tunnel was demonstrable in two cases. In two patients, an inhomogeneous echogenic structure was evident at the bottom of the carpal tunnel, which was proved at operation to rep• resent an unusual amount of fatty tissue. Finally, in both wrists of one patient the muscle bellies of the flexor digitorum superficialis extended Figure 8 Carpal tunnel syndrome. Axial section at the level of the hook of the hamate showing increased palmar bowing of the transverse carpal ligament is seen. PD, palmar displacement.
J Ultrasound Med 10:531- 537, 1991
Figure 9 Normal carpal tunnel. Axial section at the level of hook of the hamate showing normal palmar bowing of the transverse carpal ligament. PD, palmar displacement.
deep into the carpal tunnel, which possibly contributed to his bUateral carpal tunnel syndrome. Quantitative Studies As described earlier, sonographic criteria of CTS are swelling of the median nerve at the level of the radiocarpal joint and inside the proximal part of the carpal tunnel, increased flattening of the nerve owing to compression within the distal part of the carpal tunnel, and increased palmar bowing of the flexor retinaculum. For better comparison of the find· ings in 28 patients with CTS and another 28 control subjects, quantitative analysis was performed. The mean cross-sectional area of the median nerve at the distal radius was 11.1 mml (range, 7.9 mm 2 15.0 mm 2; SD, 1.8 mm 2 ) for patients with CTS and 7.9 mm 2 (range, 5.4 mm 2- 7.7 mm 2; SD, 1.1 mm 2) in normal control subjects. The difference was not statistically significant when a Mann-Whitney U-test was applied. At the level of the pisiform, the mean cross-sectional area of the median nerve was 19 .4 mm 2 (range, 9 mm 237 .3 mm 2 ~ SD, 7.1 mm 2) in symptomatic patients compared to 8.1 mm 2 (range, 6.7 mm 2- 12.8 mm 2; SD, 1.3 mm 2) in asymptomatic volunteers. AU-test confirmed that the difference in size was significant (P < 0.001) (Fig. 10). When the mean cross-sectional area of the median nerve at the pisiform was compared to that at the distal radius, the increase in size was significant in patients with CTS (P < 0.001), whereas in normal persons the nerve showed no significant swelling. Even in the distal carpal tunnel, the cross-sectional area of the median nerve was significantly (P < 0.001) greater in patients with CTS (mean area, 13.6 mm 2; range, 7.3 mm 2- 22.9 mm 1; SD, 3.8 mm 2) than in the control group (mean area, 7.7 mm ~ ; range, 6,3 mm 21 l.6 mm 2; SD, 1.1 mm2 ) . Flattening of the median nerve was calculated as the ratio of its major to its minor axis rflattening ratio·). In normal subjects the mean ratio was 3.0 with a range
BUCHBERGER ET AL
JUltrasound Med 10:531 - 537, 1991
'E E
;
m patients wllh CTS
patients with CTS control subjects
l!I
40
•
Iii
a
Ill
Iii
~8
20
• i'ii ...• E
c
•
1
10
Iii
m
•• •
•
•~
1.
0
30
20
10
0
conlrol subjects
ID
'E .!. 6
.!!!.
E c
•
8
m
30
ii-
535
El
Ill
4
a
El Ill
Ill
0 0
m
Ill
Ill
Iii
13 m
am
ma
a
El
•••
•
2
a
a
Ill
•• •• • Iii
••
10
• •
• 20
••• •• 30
Figure 10 Graph showing cross-sectional areas of the meT dian nerve at the level of the pisiform bone in patients with carpal tunnel syndrome and in normal control subjects.
Figure 12 Graph depicting measurements of palmar dis· placement of the transverse carpal ligament in patients with carpal tunnel syndrome and in normal control subjects.
of 1.6-3.9 (SD, 0.5) at the distal radius; 3.0 with a range of 1.7-4.0 (SD, 0.5) at the pisiform; and 3.2 with a range of 2.3-4.3 (SD, 0.5) at the hamate. The mean flattening ratio in patients with CTS was 3.1 (range, 2.5-3.8; SD, 0.4) at the disal radius; 3.2 (range, 2.36.2; SD, 0.8) at the pisiform; and 5.6 (range, 3.3-6.7; SD, 0.8) at the hamate level. A U-test confirmed a significantly increased flattening of the median nerve in the distal carpal tunnel (P < 0.001) (Fig. 11). Mean palmar displacement (PD) of the transverse carpal ligament was 4.8 mm with a range of 3.0 mm7.0 mm (SD, 1.0 mm) in patients with CTS and 2.1 mm with a range of 0 mm-3.1 mm (SD, 0.8 mm) in the control group. The difference was statistically significant (P < 0.001) (Fig. 12).
DISCUSSION
Figure 11 Graph depicting flattening of the median nerve at the level of the hamate in patients with carpal tunnel syndrome and in normal control subjects. m patients with CTS control subjects
•
7
mm
6
mm
a
Ill ID
5
m m m
ID
Iii
Iii
4
•
3
•• •• 0
•
••
•• • • •
• • ••• • • • 10
20
•• •• 30
Various attempts have been made to image the normal and pathologic anatomy of the carpal tunnel. Plain radiographs are largely noncontributory except for revealing alterations of the carpal bones, such as displaced fractures, malalignment, callus formation, or hypertrophic changes. The usefulness of CT in imaging the carpal tunnel is limited because of the small differences in attenuation values of its soft tissue contents. 6 •7 Measurement of the cross-sectional area of the carpal tunnel on CT scans has proved not to be appropriate for the diagnosis of carpal tunnel syndrome. 11 Recently, MR imaging has been shown to be able to depict clearly the normal anatomy of the carpal tunnel. Middleton et al 9 reported on preliminary observations in patients with carpal tunnel syndrome and described thinning of the median nerve under the flexor retinaculum and swelling of the nerve proximal to the carpal tunnel. Mesgarzadeh et alto performed quantitative studies in normal volunteers and patients with CTS. They found significantly greater swelling of the median nerve in the proximal carpal tunnel, flattening of the nerve in the distal carpal tunnel, and palmar bowing of the flexor retinaculum in patients with CTS. Real ~ time high-frequency ultrasonography has been used successfully in the diagnosis of soft tissue lesions of the extremities. Fomage et al 11 • 12 described the normal and pathologic ultrasound anatomy of the hand. Early experiences with the sonographic examination of peripheral nerves have been published by Solbiati et a1. t3 Fornaget 4 examined peripheral nerves with ultrasound and described the sonographic aspect of the median nerve as echogenic with a fibrillary structure . Calleja Cancho et alts described in detail the normal ultrasound anatomy of the carpal tunnel and its con·
536
ULTRASONOGRAPHY OF THE CARPAL TUNNEL
tents. They discussed the sonographic characteristics of the median nerve and demonstrated the possibilities of its identification on axial and sagittal sections. In concordance with that study, we found that highfrequency ultrasonography allowed differentiation of the median nerve and the flexor tendons in all cases. The accuracy of sonographic measurements is limited by the axial and lateral resolution of the transducer. In addition, the nerve is sometimes not easy to distinguish from its surroundings, especially in the distal carpal tunnel. Despite these restrictions, our results correspond to a great extent with those of anatomic examinations and previous ultrasound and MR studies. 16 Consistent with the study of Mesgarzadeh et al, who compared MRI findings in the wrists of patients with CTS and normal control subjects, our results indicate that there are characteristic findings in carpal tunnel syndrome regardless of its cause. Comparisons of the mean flattening ratios of the median nerve in symptomatic and asymptomatic subjects showed that increased flattening of the nerve in the distal carpal tunnel is a significant sign of CTS. Quantitative analysis of the cross-sectional area of the median nerve in patients with CTS proved that the nerve showed a significant increase in size at the level of the proximal carpal tunnel and to a lesser extent also in the distal carpal tunnel. These findings are supported by the observations of Tanzer" ' and Rietz and Onne, 111 who found at operation a marked swelling of the median nerve in 43% and 66% of cases, respectively. Another constant finding in carpal tunnel syndrome was an increased palmar bowing of the flexor retinaculum. Mean palmar displacement of the ligament was proved to be significantly greater in patients with CTS than in the normal population. This is probably attributable to an increased volume of contents of the carpal tunnel in CTS. Many potential causes of carpal tunnel syndrome have been described; including malalignment of the carpal bones, fractures and other injuries, tendon sheath enlargement caused by acute tenovaginitis or chronic fibrosis; space-occupying lesions such as ganglion cysts, hemangioma, or lipoma; deposits of calcium, uric add crystals, or amyloid; endocrinopathies such as myxedema or acromegaly; and finally venostasis and vasodilatation. 18" 2° Congenital abnormalities, such as aberrant muscles 21 or a persistent median artery, 22 are thought to be contributory to the pathogenesis of CTS. As our preliminary results indicate, sonography is able to detect preoperatively some of the etioJogic factors of CTS. Ganglion cysts, excessive fat at the bottom of the carpal tunnel, distal prolongation of the belly of the flexor digitorum superfkialis muscle, and tenovaginitis have been diagnosed correctly. As
JUltrasound Med 10:531 - 537, 1991
Pahlen:z3 pointed out, the most common finding at surgery is thickening or fibrosis of the flexor tendon sheaths. Sonographically, tendon sheath thickening is not easy to evaluate. Mild or moderate degrees of thickening may not be demonstrable by sonography. Comparison with the contralateral wrist is not reliable, because bilateral involvement is common. In summary, high-frequency sonography seems to be an adequate method for displaying the normal and pathologic anatomy of the carpal tunnel. Further clinical experience is needed to evaluate the exact role of sonography in the diagnosis of carpal tunnel syndrome as well as in postoperative follow -up examinations.
REFERENCES 1. Pahlen G, Gardner WJ, Lalonde AA: Neuropathy of the median nerve due to compression of the median nerve beneath the transverse carpal ligament. J Bone Joint Surg 32A:109, 1950 2. Hurst LC, Weissberg D, Carroll RE: The relationship of the double crush to carpal tunnel syndrome (analysis of 1,000 cases of carpal tunnel syndrome). J Hand Surg 10B:202, 1985 3. Robbins H: Anatomical study of the median nerve in the carpal tunnel and etiologies o.f the carpal tunnel syndrome. J Bone Joint Surg 45A:953, 1963 4. Tanzer BC: The carpal tunnel syndrome: A clinical and anatomical study. J Bone Joint Surg 4l A:626, 1959 5. r ahlen GS: The CilrpaH unne] syndrome: Seventeen years' experience in diagnosis and treatment of six hundred fifty-four hands. J Bone Joint Surg 48A:211, 1966 6. Zucker-Pinchoff B, Hermann G, Srinivasan R: Computed tomography of the carpal tunnel: A radioanatomical study. J Comput Assist Tomogr 5:525, 1982 7. John V, Nau HE, Nahsen HC, et al: CT of carpal tunnel syndrome. AJNR 4:770, 1983 8. Marhar GL, Clark RA, Schneider HJ: High resolution CT scans of the wrist in patients with carpal tunnel syndrome (abstr). Radiology 157:30, 1985 9. Middleton WO, Kneeland JB, Kellman GM, et al: MR imaging of the carpal tunnel: Normal anatomy and pre& liminary findings in the carpal tunnel syndrome. AJR 148;307, 1987 10. Mesgarzadeh M, Schneck CD, Bonakdarpour A, et al: Carpal tunnel: MR imaging. Part II. Carpal tunnel syndrome. Radiology 171:749, 1989 11. Fomage BO, Schemberg FL, Rifkin MD: Ultrasound ex• amination of the hand. Radiology 160:853, 1986 12. Fomage BO Rifkin MD: Ultrasound examination of the hand and foot, Radiol Clin North Am 26;109, 1988 13. Solbiati L, De Pra L, Ierace T, et al: High-resolution sonography of the recurrent laryngeal nerve: Anat~mic and pathologic considerations. AJR 145;989, 1985
J Ultrasound Med 10:531 - 537, 1991
14. Fomage BO: Peripheral nerves of the extremities: Imaging with US. Radiology 167:179, 1988 15. Calleja Cancho E, Schawe-Calleja Cancho M, Milbradt H, et al: Ultrasonic anatomy and study technique of the normal carpal tunnel and the distal median nerve. (Engl Abstr). ROFO 151:414, 1989 16. Mesgarzadeh M, Schneck CD, Bonakdarpour A: Carpal tunnel: MR imaging. Part I. Normal anatomy. Radiology 171:743, 1989 17. Tanzer RC: The carpal tunnel syndrome. A clinical and anatomical study. J Bone Joint Surg 41A:626, 1959 18. Rietz KA, Onne L. Analysis of sixty-five operated cases of carpal tunnel syndrome. Acta Chir Scand 133:443, 1967
BUCHBERGER ET AL
537
19. Fenves AZ, Emmet M, White MG, et al: Carpal tunnel ~ syndrome with cystic bone lesions secondary to amyloi dosis in chronic hemodialysis patients. Am J Kidney Dis 7:130, 1986 20. Schiller F, Kolb FO: Carpal tunnel syndrome in acromegaly. Neurology 4:271, 1954 21. Asai M, Wong ACW, Matsunaga T: Carpal tunnel syndrome caused by aberrant lumbrical muscles associated with cystic degeneration of tenosynovium. J Hand Surg 11A:218, 1986 22. Luyendijik W: The carpal tunnel syndrome: The role of a persistent median artery. Acta Neurol 79:52, 1986 23. Pahlen GS: The carpal-tunnel syndrome: Clinical evaluation of 598 hands. Clin Orthop 83:29, 1972
Twenty-eight wrists of 25 patients with carpal tunnel syndrome (CTS) and 28 wrists of 14 normal control subjects were studied with high-frequency real-time ultrasonography. Three general findings could be ob· served in CTS, regardless of its cause: swelling of the median nerve at the entrance of the carpal tunnel; flattening of the median nerve in the distal carpal
C
arpal tunnel syndrome (CTS) is a relatively common disease characterized by burning pain, numbness, and paresthesia in the distribution of the median nerve. In advanced cases, thenar muscle weakness or atrophy is found. The disease occurs predominantly in patients in their fourth to sixth decades, and women are affected two to five times more often than men. Bilateral involve· ment is not uncommon. 1•2 The condition generally is attributed to compression of the median nerve in the carpal tunnel beneath the flexor retinaculum. However, the exact pathogenesis often remains unknown. 3 Operative treatment consists of decompression of the nerve by sectioning the transverse carpal ligament. 4 Conservative management includes splinting of the wrist, injection of steroids into the carpal tunnel, and administration of nonsteroidal anti-inflammatory drugs. 5 The diagnosis of CTS is mainly based on clinical findings, electromyography (EMG), and nerve conduction studies. These findings can be equivocal, and CTS can be confused with cervical root compression, thoReceived October 15, 1990, from the Departments of "Radiology and tp]astic and Reconstructive Surgery, University of Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria. Revised manuscript accepted for publication February 8, 1991. Address corre!lpondence and reprint requests to Dr. W. Buchberger: Department of Radiology, University of lnMbruck, Anichstrasse 35, A-6020 Innsbruck, Austria.
tunnel; and increased palmar flexion of the transverse carpal ligament. Quantitative analysis proved these findings to be significant. We conclude that high-resolution sonography is able to diagnose median nerve compression .in the carpal tunnel syndrome and to detect some of its potential causes.
racic outlet syndrome, and other peripheral nerve compression syndromes. Because there are many possible causes of median nerve compression, the etiologic basis of CTS cannot be evaluated preoperatively by means of clinical and EMG studies alone. Magnetic resonance imaging is the only imaging modality that has proved to be of diagnostic value in carpal tunnel syndrome. However, the method is expensive, relatively time consuming, and still not generally available. Early experiences with high-resolution real-time sonography in the imaging of the normal carpal tunnel were encouraging. The purpose of this study is to establish objective criteria for the ultrasound diagnosis of carpal tunnel syndrome.
PATIENTS AND METHODS Twenty-eight wrists of 25 nonselected patients with clinical symptoms of carpal tunnel syndrome were studied. All of the patients underwent EMG and measurement of nerve conduction delay. Seventeen patients were women and 8 were men; they ranged in age from 38 years up to 85 years (mean, 61 years). The right hand was affected in 16 cases and the left hand in 4 cases. Four patients had bilateral carpal tunnel syndrome; 4 had had operation on the contralateral wrist previously. Three patients with
© 1991 by the American Institute of Ultrasound in Medicine• J Ultrasound Med 10:531- 537, 1991•0278-4297/91/$3.50
532
J Ultrasound Med 10:531- 537, 1991
ULTRASONOGRAPHY OF THE CARPAL TUNNEL
unrelieved or recurrent CTS after surgical treatment were excluded from this study. For comparison with the normal population, 28 wrists of 14 asymptomatic volunteers (9 women and S men) were studied. All ultrasound examinations were performed with a real-time scanner (Picker LSC 9500) and a 7.5 MHz linear probe. A stand-off was used for better near-field resolution. Axial scans of the wrist at the level of the distal radius, pisiform, and hook of the hamate and sagittal sections through the carpal tunnel were recorded with a multiformat camera . At each level, the lateral and the anteroposterior diameters of the median nerve were measured, and the cross-sectional area was calculated, assuming an elliptic shape. To determine the bowing of the flexor retinaculum, a straight line was drawn between the attachments of the ligament to the tubercle of the trapezium and the hook of the hamate, and the distance from this line to the palmar apex of the flexor retinaculum was meas ~ ured (palmar displacement, PD) (Fig. 1).
A
ua
un
fr
mn
fer
fpl fds
p
fdp
Results s
Normal Ultrasound Anatomy The bony and ligamentous walls of the carpal tunnel and its contents are best studied on axial planes (Figs. 2 and 3). The dorsal border of the tunnel is formed by the scaphoid, triquetrum, lunate, and pisiform proximally and the trapezium, trapezoid, capitate, and hamate distally. The surfaces of the carpal bones appear as bright, hyperechoic lines with posterior shadow. The hyperechoic flexor retinaculum extends from the tubercle of the scaphoid to the pisiform and from the tubercle of the trapezium to the hook of the hamate. It is either straight or slightly bowed, and its thickest portion is seen at the level of the distal carpal tunnel. The superficial and the deep flexor tendons and the Figure 1 Schematic diagram showing the measuring points of palmar displacement (PD) of the flexor retinaculum. tm, Trapezium; td, trapezoid; c, capitate; h, hamate; ha, hook of the hamate; fr, flexor retinaculurn. fr
= ~ =---1 ,.;.;. .;.;. ,;.;. .;.;. ;.;.,.;.;. .;.;.,;.;. .; . .;.;. ;.~
-~~~~
PD ha
c
tr
B
Figure 2 Normal carpal tunnel. Axial ultrasound image (A) and schematic diagram (B) at the level of the pisifonn bone. S, Scaphoid; C, capitate; tr, triquetrum; p, pislform; fr, flexor retinaculum; fdp, flexor digitorum profundus tendons; fds, nexor digitorum superfidalis tendons; fpl, Aexor pollicis Ion· gus tendon; fer, flexor carpi radialis tendon; mn, median nerve; ua, ulnar artery; un, ulnar nerve. separate tendon of the flexor pollicis longus appear as round or ovoid shaped structures 4- 5 mm in diameter. Their echogenicity depends on the angle of incidence of the ultrasound beam. The tendons are surrounded by hypoechoic structures that are believed to represent all or part of the synovial sheath. The tendon of the flexor carpi radialis muscle lies in a separate fibro• osseous tunnel between a superficial and a deep layer of the transverse carpal ligament and within the vertical groove of the trapezium. The median nerve lies just beneath the flexor retlnaculum between the tendon of the flexor pollicis longus and the tendons of the flexor digitorum superficialis muscles. It is ovoid in cross section, with its long axis usually oriented parallel to the flexor retinaculum. The median nerve appears echogenic when the ultra sound beam is perpendicular to its surface (Fig. 2A). A slight obliquity of the beam results in a hypoecho· genicity of the nerve, which then seems to be surrounded by a narrow echogenic rim possibly rep ~ resenting the perineural fibrous tissue (Fig. 3A). Super-
J Ultrasound Med 10:531 - 537, 1991
BUCHBERGER ET AL
A ----ua ---un
---ha - - - fds fdp
Im
td
c
h
B
Figure 3 Normal carpal tunnel. Axial ultrasound image (A) and schematic diagram (B) at the level of the hook of the hamate. tm, trapezium; td, trapezoid; c, capitate bone; h, hamate; ha, hook of the hamate; fr, flexor retinaculum; ua, ulnar artery; un, ulnar nerve.
ficial to the flexor retinaculum, the ulnar nerve and artery and the tendon of the palmaris longus muscle can be identified. On sagittal scans, the median nerve appears as a band-like structure just superficial to the Figure 4 Normal carpal tunnel, sagittal section. The median nerve (arrowfreads) lies superficial to the flexor tendons (FT)..
533
Figure 5 Carpal tunnel syndrome. Axial section at the level of the hook of the hamate showing increased flattening of the median nerve (arrows) owing to compression.
flexor tendons. Deep to the tendons, the contour of the radial epiphysis and the carpal bones is apparent (Fig. 4). General Fi11di11gs in CTS In all cases of carpal tunnel syndrome, a more or less marked flattening of the median nerve in the distal carpal tunnel owing to compression between the flexor tendons and the trans· verse carpal ligament was evident (Fig. 5). Sagittal scans usually showed either a diffuse thinning or a localized constriction of the nerve. ln 22 wrists, an obvious enlargement of the median nerve at the level of the radiocarpal joint and inside the proximal carpal tunnel in comparison to its size at the distal radius was evident (Fig. 6). In 12 cases the swelling was segmental, with a bulbous shape of the nerve suggesting a pseudoneuroma (Fig. 7). More dif• fuse enlargement of the nerve was observed in 10 cases. Figure 6 Carpal tunnel syndrome. Axial section through the proximal carpal tunnel. The median nerve (arrows) is markedly increased in size, and the transverse carpal ligament (arrowf1ead) $hows increased palmar bowing.
534
ULTRASONOGRAPHY OF TH E CARPAL TUNNEL
Figure 7 Carpal tunnel syndrome, sagittal section. Bulbous swelling of the median nerve (arrowf1cads) is seen at the proximal carpal tunn el and flattening of the nerve is apparent at the distal carpal tunnel.
Another relatively constant finding was an increased palmar bowing of the transverse carpal ligament, which is normally straight or only slightly bowed (Figs. l, 8, 9). Etiologic findings of median nerve compression were observed in 9 cases. In three wrists, the flexor tendons were surrounded by an unusually thick hypoechogenic halo, probably representing proliferated synovial tissue. The sono· graphic diagnosis of tenosynovitis was confirmed at operation. A ganglion cyst of the flexor tendons extend6ng into the proximal part of the carpal tunnel was demonstrable in two cases. In two patients, an inhomogeneous echogenic structure was evident at the bottom of the carpal tunnel, which was proved at operation to rep• resent an unusual amount of fatty tissue. Finally, in both wrists of one patient the muscle bellies of the flexor digitorum superficialis extended Figure 8 Carpal tunnel syndrome. Axial section at the level of the hook of the hamate showing increased palmar bowing of the transverse carpal ligament is seen. PD, palmar displacement.
J Ultrasound Med 10:531- 537, 1991
Figure 9 Normal carpal tunnel. Axial section at the level of hook of the hamate showing normal palmar bowing of the transverse carpal ligament. PD, palmar displacement.
deep into the carpal tunnel, which possibly contributed to his bUateral carpal tunnel syndrome. Quantitative Studies As described earlier, sonographic criteria of CTS are swelling of the median nerve at the level of the radiocarpal joint and inside the proximal part of the carpal tunnel, increased flattening of the nerve owing to compression within the distal part of the carpal tunnel, and increased palmar bowing of the flexor retinaculum. For better comparison of the find· ings in 28 patients with CTS and another 28 control subjects, quantitative analysis was performed. The mean cross-sectional area of the median nerve at the distal radius was 11.1 mml (range, 7.9 mm 2 15.0 mm 2; SD, 1.8 mm 2 ) for patients with CTS and 7.9 mm 2 (range, 5.4 mm 2- 7.7 mm 2; SD, 1.1 mm 2) in normal control subjects. The difference was not statistically significant when a Mann-Whitney U-test was applied. At the level of the pisiform, the mean cross-sectional area of the median nerve was 19 .4 mm 2 (range, 9 mm 237 .3 mm 2 ~ SD, 7.1 mm 2) in symptomatic patients compared to 8.1 mm 2 (range, 6.7 mm 2- 12.8 mm 2; SD, 1.3 mm 2) in asymptomatic volunteers. AU-test confirmed that the difference in size was significant (P < 0.001) (Fig. 10). When the mean cross-sectional area of the median nerve at the pisiform was compared to that at the distal radius, the increase in size was significant in patients with CTS (P < 0.001), whereas in normal persons the nerve showed no significant swelling. Even in the distal carpal tunnel, the cross-sectional area of the median nerve was significantly (P < 0.001) greater in patients with CTS (mean area, 13.6 mm 2; range, 7.3 mm 2- 22.9 mm 1; SD, 3.8 mm 2) than in the control group (mean area, 7.7 mm ~ ; range, 6,3 mm 21 l.6 mm 2; SD, 1.1 mm2 ) . Flattening of the median nerve was calculated as the ratio of its major to its minor axis rflattening ratio·). In normal subjects the mean ratio was 3.0 with a range
BUCHBERGER ET AL
JUltrasound Med 10:531 - 537, 1991
'E E
;
m patients wllh CTS
patients with CTS control subjects
l!I
40
•
Iii
a
Ill
Iii
~8
20
• i'ii ...• E
c
•
1
10
Iii
m
•• •
•
•~
1.
0
30
20
10
0
conlrol subjects
ID
'E .!. 6
.!!!.
E c
•
8
m
30
ii-
535
El
Ill
4
a
El Ill
Ill
0 0
m
Ill
Ill
Iii
13 m
am
ma
a
El
•••
•
2
a
a
Ill
•• •• • Iii
••
10
• •
• 20
••• •• 30
Figure 10 Graph showing cross-sectional areas of the meT dian nerve at the level of the pisiform bone in patients with carpal tunnel syndrome and in normal control subjects.
Figure 12 Graph depicting measurements of palmar dis· placement of the transverse carpal ligament in patients with carpal tunnel syndrome and in normal control subjects.
of 1.6-3.9 (SD, 0.5) at the distal radius; 3.0 with a range of 1.7-4.0 (SD, 0.5) at the pisiform; and 3.2 with a range of 2.3-4.3 (SD, 0.5) at the hamate. The mean flattening ratio in patients with CTS was 3.1 (range, 2.5-3.8; SD, 0.4) at the disal radius; 3.2 (range, 2.36.2; SD, 0.8) at the pisiform; and 5.6 (range, 3.3-6.7; SD, 0.8) at the hamate level. A U-test confirmed a significantly increased flattening of the median nerve in the distal carpal tunnel (P < 0.001) (Fig. 11). Mean palmar displacement (PD) of the transverse carpal ligament was 4.8 mm with a range of 3.0 mm7.0 mm (SD, 1.0 mm) in patients with CTS and 2.1 mm with a range of 0 mm-3.1 mm (SD, 0.8 mm) in the control group. The difference was statistically significant (P < 0.001) (Fig. 12).
DISCUSSION
Figure 11 Graph depicting flattening of the median nerve at the level of the hamate in patients with carpal tunnel syndrome and in normal control subjects. m patients with CTS control subjects
•
7
mm
6
mm
a
Ill ID
5
m m m
ID
Iii
Iii
4
•
3
•• •• 0
•
••
•• • • •
• • ••• • • • 10
20
•• •• 30
Various attempts have been made to image the normal and pathologic anatomy of the carpal tunnel. Plain radiographs are largely noncontributory except for revealing alterations of the carpal bones, such as displaced fractures, malalignment, callus formation, or hypertrophic changes. The usefulness of CT in imaging the carpal tunnel is limited because of the small differences in attenuation values of its soft tissue contents. 6 •7 Measurement of the cross-sectional area of the carpal tunnel on CT scans has proved not to be appropriate for the diagnosis of carpal tunnel syndrome. 11 Recently, MR imaging has been shown to be able to depict clearly the normal anatomy of the carpal tunnel. Middleton et al 9 reported on preliminary observations in patients with carpal tunnel syndrome and described thinning of the median nerve under the flexor retinaculum and swelling of the nerve proximal to the carpal tunnel. Mesgarzadeh et alto performed quantitative studies in normal volunteers and patients with CTS. They found significantly greater swelling of the median nerve in the proximal carpal tunnel, flattening of the nerve in the distal carpal tunnel, and palmar bowing of the flexor retinaculum in patients with CTS. Real ~ time high-frequency ultrasonography has been used successfully in the diagnosis of soft tissue lesions of the extremities. Fomage et al 11 • 12 described the normal and pathologic ultrasound anatomy of the hand. Early experiences with the sonographic examination of peripheral nerves have been published by Solbiati et a1. t3 Fornaget 4 examined peripheral nerves with ultrasound and described the sonographic aspect of the median nerve as echogenic with a fibrillary structure . Calleja Cancho et alts described in detail the normal ultrasound anatomy of the carpal tunnel and its con·
536
ULTRASONOGRAPHY OF THE CARPAL TUNNEL
tents. They discussed the sonographic characteristics of the median nerve and demonstrated the possibilities of its identification on axial and sagittal sections. In concordance with that study, we found that highfrequency ultrasonography allowed differentiation of the median nerve and the flexor tendons in all cases. The accuracy of sonographic measurements is limited by the axial and lateral resolution of the transducer. In addition, the nerve is sometimes not easy to distinguish from its surroundings, especially in the distal carpal tunnel. Despite these restrictions, our results correspond to a great extent with those of anatomic examinations and previous ultrasound and MR studies. 16 Consistent with the study of Mesgarzadeh et al, who compared MRI findings in the wrists of patients with CTS and normal control subjects, our results indicate that there are characteristic findings in carpal tunnel syndrome regardless of its cause. Comparisons of the mean flattening ratios of the median nerve in symptomatic and asymptomatic subjects showed that increased flattening of the nerve in the distal carpal tunnel is a significant sign of CTS. Quantitative analysis of the cross-sectional area of the median nerve in patients with CTS proved that the nerve showed a significant increase in size at the level of the proximal carpal tunnel and to a lesser extent also in the distal carpal tunnel. These findings are supported by the observations of Tanzer" ' and Rietz and Onne, 111 who found at operation a marked swelling of the median nerve in 43% and 66% of cases, respectively. Another constant finding in carpal tunnel syndrome was an increased palmar bowing of the flexor retinaculum. Mean palmar displacement of the ligament was proved to be significantly greater in patients with CTS than in the normal population. This is probably attributable to an increased volume of contents of the carpal tunnel in CTS. Many potential causes of carpal tunnel syndrome have been described; including malalignment of the carpal bones, fractures and other injuries, tendon sheath enlargement caused by acute tenovaginitis or chronic fibrosis; space-occupying lesions such as ganglion cysts, hemangioma, or lipoma; deposits of calcium, uric add crystals, or amyloid; endocrinopathies such as myxedema or acromegaly; and finally venostasis and vasodilatation. 18" 2° Congenital abnormalities, such as aberrant muscles 21 or a persistent median artery, 22 are thought to be contributory to the pathogenesis of CTS. As our preliminary results indicate, sonography is able to detect preoperatively some of the etioJogic factors of CTS. Ganglion cysts, excessive fat at the bottom of the carpal tunnel, distal prolongation of the belly of the flexor digitorum superfkialis muscle, and tenovaginitis have been diagnosed correctly. As
JUltrasound Med 10:531 - 537, 1991
Pahlen:z3 pointed out, the most common finding at surgery is thickening or fibrosis of the flexor tendon sheaths. Sonographically, tendon sheath thickening is not easy to evaluate. Mild or moderate degrees of thickening may not be demonstrable by sonography. Comparison with the contralateral wrist is not reliable, because bilateral involvement is common. In summary, high-frequency sonography seems to be an adequate method for displaying the normal and pathologic anatomy of the carpal tunnel. Further clinical experience is needed to evaluate the exact role of sonography in the diagnosis of carpal tunnel syndrome as well as in postoperative follow -up examinations.
REFERENCES 1. Pahlen G, Gardner WJ, Lalonde AA: Neuropathy of the median nerve due to compression of the median nerve beneath the transverse carpal ligament. J Bone Joint Surg 32A:109, 1950 2. Hurst LC, Weissberg D, Carroll RE: The relationship of the double crush to carpal tunnel syndrome (analysis of 1,000 cases of carpal tunnel syndrome). J Hand Surg 10B:202, 1985 3. Robbins H: Anatomical study of the median nerve in the carpal tunnel and etiologies o.f the carpal tunnel syndrome. J Bone Joint Surg 45A:953, 1963 4. Tanzer BC: The carpal tunnel syndrome: A clinical and anatomical study. J Bone Joint Surg 4l A:626, 1959 5. r ahlen GS: The CilrpaH unne] syndrome: Seventeen years' experience in diagnosis and treatment of six hundred fifty-four hands. J Bone Joint Surg 48A:211, 1966 6. Zucker-Pinchoff B, Hermann G, Srinivasan R: Computed tomography of the carpal tunnel: A radioanatomical study. J Comput Assist Tomogr 5:525, 1982 7. John V, Nau HE, Nahsen HC, et al: CT of carpal tunnel syndrome. AJNR 4:770, 1983 8. Marhar GL, Clark RA, Schneider HJ: High resolution CT scans of the wrist in patients with carpal tunnel syndrome (abstr). Radiology 157:30, 1985 9. Middleton WO, Kneeland JB, Kellman GM, et al: MR imaging of the carpal tunnel: Normal anatomy and pre& liminary findings in the carpal tunnel syndrome. AJR 148;307, 1987 10. Mesgarzadeh M, Schneck CD, Bonakdarpour A, et al: Carpal tunnel: MR imaging. Part II. Carpal tunnel syndrome. Radiology 171:749, 1989 11. Fomage BO, Schemberg FL, Rifkin MD: Ultrasound ex• amination of the hand. Radiology 160:853, 1986 12. Fomage BO Rifkin MD: Ultrasound examination of the hand and foot, Radiol Clin North Am 26;109, 1988 13. Solbiati L, De Pra L, Ierace T, et al: High-resolution sonography of the recurrent laryngeal nerve: Anat~mic and pathologic considerations. AJR 145;989, 1985
J Ultrasound Med 10:531 - 537, 1991
14. Fomage BO: Peripheral nerves of the extremities: Imaging with US. Radiology 167:179, 1988 15. Calleja Cancho E, Schawe-Calleja Cancho M, Milbradt H, et al: Ultrasonic anatomy and study technique of the normal carpal tunnel and the distal median nerve. (Engl Abstr). ROFO 151:414, 1989 16. Mesgarzadeh M, Schneck CD, Bonakdarpour A: Carpal tunnel: MR imaging. Part I. Normal anatomy. Radiology 171:743, 1989 17. Tanzer RC: The carpal tunnel syndrome. A clinical and anatomical study. J Bone Joint Surg 41A:626, 1959 18. Rietz KA, Onne L. Analysis of sixty-five operated cases of carpal tunnel syndrome. Acta Chir Scand 133:443, 1967
BUCHBERGER ET AL
537
19. Fenves AZ, Emmet M, White MG, et al: Carpal tunnel ~ syndrome with cystic bone lesions secondary to amyloi dosis in chronic hemodialysis patients. Am J Kidney Dis 7:130, 1986 20. Schiller F, Kolb FO: Carpal tunnel syndrome in acromegaly. Neurology 4:271, 1954 21. Asai M, Wong ACW, Matsunaga T: Carpal tunnel syndrome caused by aberrant lumbrical muscles associated with cystic degeneration of tenosynovium. J Hand Surg 11A:218, 1986 22. Luyendijik W: The carpal tunnel syndrome: The role of a persistent median artery. Acta Neurol 79:52, 1986 23. Pahlen GS: The carpal-tunnel syndrome: Clinical evaluation of 598 hands. Clin Orthop 83:29, 1972
