- Email: [email protected]
Changes in the carpal tunnel due to action of the flexor tendons: Visualization with magnetic resonance imaging
Changes in the Carpal Tunnel Due to Action of the Flexor Tendons: Visualization With Magnetic Resonance Imaging S. J. Ham, MD, Groningen, The Netherlands, W. F. A. Kolkman, MD, J. Heeres, MD, PhD, J. A. den Boer, P. A. M. Vierhout, MD, PhD, Enschede, The Netherlands Successive cross-sectional areas (CSA) of the carpal tunnel were measured with the fingers in both extension and full flexion in 12 healthy volunteers using magnetic resonance imaging. During flexion, lumbrical muscles could be observed to move into the carpal tunnel up to different levels in all volunteers. For each of the volunteers, the level of the hook of hamate was used as the reference level. The mean CSA measured at this level was considerably larger in flexion than in extension: 191 mm 2 (SD, + 26) and 169 mm 2 (SD, _+15), respectively (p = .004). In three volunteers, no difference in CSA between extension and flexion was measured at the hamate level, despite the presence of lumbrical muscles, whereas in these same volunteers at levels more distal, the CSA clearly increased during flexion. The mean CSA for extension and flexion distal and just proximal to the smallest level differed significantly, but the absence of expansion was noticed only at the smallest level. Other changes that were frequently observed during flexion were fat compression, flattening and displacement of the median nerve, and pressure on the superficial and deep flexor tendons. (J Hand Surg 1996;21A:997-1003.)
In the literature, much interest has been expressed in factors that could be of importance in producing or aggravating idiopathic carpal tunnel syndrome (CTS). I Hypertrophy of the lumbrical muscles is believed to be one of these factors, caused by occupation of an excessive amount of space in the carpal tunnel either in rest or during contraction of the deep
From the Department of Orthopedic Surgery, University Hospital, Groningen, the Netherlands, and the Departments of General Surgery and Radiology, Medical Spectrum Twente, Enschede, the Netherlands. Received for publication Nov. 9, 1994; accepted in revised form April 1, 1996. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Reprint requests: S. J. Ham, MD, Department of Orthopedic Surgery, University Hospital, RO. Box 30.001, 9700 RB Groningen, The Netherlands.
flexors of the fingers.1 For any further investigation of the etiologic or supporting role of the lumbrical muscles in the development and course of idiopathic CTS, knowledge about normal values and conditions in this anatomy are essential. Therefore, we investigated the position of the lumbrical muscles in healthy volunteers together with the changes occurring in the carpal tunnel both at extension and full flexion of the fingers by using magnetic resonance imaging (MRI). This method was used because it has proved to be a superior, noninvasive method of imaging and investigating the musculoskeletal system, including the carpal tunnel and its contents. 2-16
Materials and Methods In 12 healthy volunteers (6 men and 6 women) with a mean age of 30.5 years (range, 25-54 years) and with normal wrist and hand function, MRI studThe Journal of Hand Surgery
997
998
Ham et al. / Visualizing Carpal Tunnel Changes by MRI
ies of the carpal tunnel were made with the fingers both in extension and in full flexion. Full flexion (further referred to as flexion in this article) of the fingers was used to achieve maximal excursion of the profundus tendons. 17 The Gyroscan ACS whole-body system was used (Philips Inc., Eindhoven, The Netherlands). To exclude specific effects of dominance-related influences, the right, dominant hand was examined in six volunteers and the left, nondominant hand also in six volunteers, with equal distribution among the sexes. The volunteers were examined while they were in a prone position with their arm extended overhead. The palm of the hand was placed with the carpal tunnel located in the center of a 70-ram diameter flatsurface coil, with the distal part of the arm, with respect to the coil and the magnet, immobilized. Scans with 13 transverse slices of 4.0 mm and interslice gaps of 0.4 mm were made with the fingers both in extension and in flexion. These transverse slices were obtained at exactly fight angles to the carpal tunnel and forearm by using coronal and sagittal plane scans. Scans were done with the fingers in flexion and the thumb held in neutral because there is no lumbrical muscle attached to the flexor pollicis longus tendon, and to prevent the interference of this tendon with the shape and contents of the carpal tunnel. a5 Tl-weighted spin echo images (TR 519 ms, TE, 25 ms) were made with a field of view of 120 mm. Data were acquired using 410 x 512 matrices. Two excitations were used in all cases. A dynamic scan mode was used in order to minimize the time between scanning in extension and in flexion of the fingers. The total scan time was 11.26 minutes. Successive computer-enhanced cross-sectional areas (CSAs) of the carpal tunnel were measured in the images of each subject as described earlier by Richman et al. 9 The level in which the hook of hamate was visualized was used as level of reference (Fig. 1). Data acquired for the CSAs for both extension and flexion of the fingers were compared at this level and successive more proximal and distal levels using Wilcoxon's signed-rank test for statistical analysis. Reproducibility of the determination of the CSAs was validated by measuring one image five times. Volar bowing of the transverse ligament was measured using the volar bowing (VB) ratio as described by Mesgarzadeh et al. a0 The tendons used include the flexor digitorum superficialis (FDS), flexor digitorum profundus (FDP), and the flexor pollicis longus (FPL) in the index fin-
Figure 1. Transverse magnetic resonance imaging of the left carpal tunnel of one volunteer during extension of the fingers. The level of the hook of hamate (long, thin arrow) is visualized. The median nerve (short, thick arrow) is lying palmar to the flexor digitorum superficialis of the index and middle fingers. The flexor pollicis longus (long, thick arrow) is lying radial and slightly dorsal to the median nerve.
gers (II), middle fingers (III), ring fingers (IV), and little fingers (V).
Results Changes in Cross-sectional Area Figure 2 represents the mean CSA of the carpal tunnel measured during extension and flexion of the fingers. During extension of the fingers, one can clearly see the decrease in CSA from the proximal entrance of the carpal tunnel to the level where the hook of hamate was always visualized, lying about 2.0 cm distal to this entrance. Further distally, the mean CSA of the carpal tunnel increases. The effect is a canal with a slightly narrowed waist. During flexion of the fingers, there is an initial decrease in CSA followed by an increase up to the level where the hook of hamate was visualized, which was used as reference level, as described earlier. At this level, the mean CSA is considerably smaller than that of the level proximal to it. Again, rapid increase in CSA takes place distal to the reference level. The bony landmarks of each transverse level of the carpal tunnel that were observed in the volunteers are shown in Table 1. During extension of the fingers, proximal portions of lumbrical muscles were observed in the carpal tunnel in two volunteers, starting at the reference level in one, and at two levels distal to the reference
The Journal of Hand Surgery/Vol. 21A No. 6 November 1996 999 ram2 220
200
18o
16o 0
~ 1
~ 2
_1 3
~ 4
,
I
,
5
6
7
8
slices 9 Series 1 4- Series 2 Figure
2. Mean cross-sectional areas of the carpal tunnel during extension (series 1) and flexion (series 2).
level in the other volunteer. During flexion of the fingers, however, lumbrical muscles were observed at the reference level and at levels distal to it in all volunteers and were observed to extend proximal to the reference level in 11 of the 12 volunteers. Like Middleton et al., we found that the lumbrical muscles, although attached to the deep flexor tendons of the fingers, often appeared contiguous to both the deep and superficial flexor tendons on MRI. 8 The mean CSA measured at the reference level was considerably larger in flexion of the fingers than in extension: 191 mm 2 (SD, _+26) and 169 mm 2 (SD, +_ 15), respectively (p = .004). The mean CSAs for
Table 1. Relation Between Different Levels and Bony Landmarks of the Carpal Tunnel as Visualized With Magnetic Resonance Imaging Level
4 proximal 3 proximal 2 proximal 1 proximal Reference level 1 distal 2 distal
Bony Landmarks
Scaphoid; transition between lunate and capitate; triquetrum Scaphoid; capitate; proximal hamate; triquetrum; pisiform Transition between scaphoid and trapezium; capitate; transition between hamate and triquetrum; pisiform Trapezium; trapezoideum; capitate; hamate Hook of hamate Transition between distal carpal row and proximal part of metacarpals Proximal part of metacarpals
extension and flexion distal and just proximal to the reference level were also significantly different (Table 2). In three volunteers, however, no difference in CSA between extension and flexion was measured at the reference level, despite the presence of lumbrical muscles, whereas in these same volunteers, the CSA clearly increased during flexion at levels distal to it. During flexion, portions of lumbrical muscles were observed in a number of volunteers in the three levels proximal to the reference level. At the second and third level proximal to the reference level, the difference in mean CSA measured between extension and flexion was significantly larger when portions of lumbrical muscles were present (Table 3). Topographic C h a n g e s in the Carpal Tunnel During extension of the fingers, the median nerve was found to be palmar to FDS II and palmar and slightly radial to FDS III in 9 volunteers. The FPL was lying radial and dorsal to the median nerve before shifting to a more palmar position on distal sections. During flexion of the fingers, these topographic relations were unchanged in 4 volunteers. In 4 other volunteers, there was an ulnar displacement of the median nerve, especially on more distal sections (Fig. 3). In 1 volunteer, the FDS II lay immediately ulnar to the median nerve in the presence of lumbrical muscles (Fig. 4). In another volunteer,
1000
Ham et al./Visualizing Carpal Tunnel Changes by MRI Table 2. Mean Cross-sectional Areas Measured at Different Levels in the Carpal Tunnel in 12 Volunteers Level
Extension (mm2) (SD)
4 p. 3 p.* 2 p.* 1 p. Reference level 1 d. 2 d.
197 (_+ 191 (_+ 188 (_+ 179 (_+ 169 (_+ 178 (+ 191 (_+
20) 21) 19) 16) 15) 17) 20)
Flexion (mm2) (SD)
200 (+_ 195 (_+ 197 (+ 200 (+_ 191 (+_ 208 (+ 218 (_+
20) 21) 19) 32) 26) 28) 24)
Difference (mm2)
Significance (p < .05)
3 4 9 21 22 30 27
.220 .656 .083 .016 .004 .000 .001
p., proximal to the reference level; d., distal to the reference level. *See also Table 3.
the median nerve was interposed posterolaterally between the FDS II and the FPL as described by other authors.S, la During flexion, the same position was retained. In 2 volunteers, the median nerve b e c a m e interposed between the FDS III and FDS IV, with the FDP III lying dorsal to it, during extension. On the most distal sections, the nerve was located ulnar and palmar to the FDS III. During flexion,
there was radial displacement of the median nerve on distal sections in 1 of these 2 volunteers. In the other volunteer, the median nerve never became interposed between the FDS III and FDS IV during flexion, because of palmar displacement of the FDS III in the presence of lumbrical muscles lying dorsal to it in the carpal tunnel. The median nerve remained locked between the FPL lying radial to it, the FDS II i m m e diately dorsal, and the FDS III lying immediately ulnar to it (Fig. 5). In all 12 volunteers, increase in nerve flattening was observed in the presence of lumbrical muscles. The F D P II was always located dorsal to the FDS II and FDS III during extension, before shifting radially. During flexion, the F D P II was pressed in the radiodorsal c o m e r of the carpal tunnel (in 6 volunteers) or against the dorsal wall of the tunnel (in 4). Two times, there was no difference between extension and flexion. During extension of the fingers, the FDP III was always dorsal and more or less ulnar to
A
B
Table 3. Difference in Cross-sectional Areas Between Extension and Full Flexion Mean Difference (mm2) Level
3 p. 2 p.
Lumbrical Muscles Present
Lumbrical Muscles Absent
Significance (p <. 05)*
22 (n = 4) 21 (n = 6)
-6 (n = 8) -3 (n = 6)
.006 .004
p., proximal to the reference level. *Mann-Whitney test for statistical analysis.
Figure 3. Transverse magnetic resonance images of the right carpal tunnel of one volunteer visualized at the level of the hook of hamate during (A) extension and (B) flexion of the fingers. During flexion of the fingers, ulnar displacement of the median nerve (short, thick arrow) and fat compression (long, thin arrow) dorsal in the carpal tunnel in the presence of lumbrical muscles (short, thin arrows) can be observed, with a lumbrical muscle dividing the median nerve from the flexor pollicis longus (long, thick arrow).
The Journal of Hand Surgery / Vol. 21A No. 6 November 1996
1001
A B Figure 4. Transverse magnetic resonance images of the left carpal tunnel of one volunteer visualized at the level of the hook of hamate during (A) extension and (B) flexion of the fingers. During flexion, there is ulnar displacement of the median nerve (short, thick arrow) in the presence of lumbrical muscles (short, thin arrows), with the flexor digitorum superficialis of the index finger (long, thin arrow) being interposed between the flexor pollicis longus (long, thick arrow) tendon radial to it and the median nerve ulnar to it.
the FDS IV, before shifting radially, but always lay ulnar and dorsal to the FDS III. During flexion, the same topographic relation of these tendons was maintained, with the FDP III being pressed against the dorsal roof of the carpal tunnel. The FDS IV and V always lay in the palmar and ulnar comer of the carpal tunnel and were pressed in this comer during flexion. Discrimination of the FDP IV and V lying in the dorsal and ulnar corner of the carpal tunnel was more pronounced during flexion, owing to the presence of a lumbrical muscle, which divided them on distal sections (Fig. 4). Fat compression was seen in all volunteers in the presence of lumbrical muscles (Fig. 3). Volar bowing
at the reference level was observed in all volunteers during extension, with a mean VB ratio of 9% (range, 6%-13%; SD, + 2.0). During flexion, this ratio clearly increased in only five volunteers at this level, with a mean VB ratio of 11% (range, 7%-17%; SD, + 2.4).
Discussion The role of CSAs and volume of the carpal tunnel in the development of CTS remains unclear. Both Dekel et al. and Bleecker used computed tomography to show a decrease in the CSAs of CTS patients, although later, Merhar et al. found no significant dif-
~N
A
B
Figure 5. Transverse magnetic resonance images of the fight carpal tunnel of one volunteer visualized at the level of the hook of hamate during (A) extension and (B) flexion of the fingers. During extension, the median nerve (short, thick arrow) is interposed between the flexor digitorum superficialis of the middle finger (long, thick arrow) and the FDS of the ring finger (long, thin arrow). During flexion, the ulnar displacement of the median nerve does not occur. The FDS of the middle finger has been displaced anteriorly in the presence of lumbrical muscles (short, thin arrows) lying dorsal to it.
1002
Ham et al./Visualizing Carpal Tunnel Changes by MRI
ference in CSA measurements between wrists with and without symptoms. 18-20 However, CSA measurements of the carpal tunnel vary, depending on the position of the wrist and of the fingers, a2 During flexion of the fingers, we observed lumbrical muscles to move into the carpal tunnel in all our volunteers, resulting in considerable increase in CSA in the presence of these muscles. At the level of the hook of hamate, where the tunnel is smallest, lumbrical muscles were observed in all our volunteers during flexion of the fingers, with a considerable increase in C S A in 9. In 3 volunteers, however, no increase in CSA was observed at this level during flexion. According to Cobb et al., the level of the hook of hamate is anatomically a potential site for median nerve compression. 1~ At this level, the tunnel is rigidly bounded on three sides by bony structures and roofed by a thickened transverse carpal ligament, and apparently, there is not always the possibility to respond to the increase in volume of structures by expansion of the CSA. Furthermore, functionally, this site may rely on any space-occupying condition to exceed the critical volume of the tunnel at this level, leading to median nerve compression. 24 Abnormalities of the lumbrical muscles, therefore, may play a role in the etiology of the CTS in some patients. However, further work will be needed to establish this relationship. In our study, the alignment and shape of the median nerve in the carpal tunnel and its relationship to the flexor tendons were variable and depended on the presence, amount, shape, and size of the lumbrical muscles. Transverse sliding of the median nerve was clearly seen in 7 volunteers, with ulnar displacement in 5 and radial displacement in 2. Nakamichi and Tachibana demonstrated ulnar sliding of the median nerve beneath the flexor retinaculum during active-resistant flexion of the fingers in healthy volunteers by using sonography.25 In their opinion, this sliding was caused by tensed overlying flexor tendons and could be a reason for mechanical nerve deformation. In our study, the ulnar and radial displacement of the median nerve was caused by the rearrangement of structures in the carpal canal due to the presence of lumbrical muscles. This presence of lumbrical muscles during flexion of the fingers also seemed to be responsible for the displacement of the flexor tendons occasionally observed in the carpal tunnel and for the alterations in configuration of the flexor tendons, which became more rounded during flexion of the fingers.
Finally, VB of the flexor retinaculum at the level of the hook of hamate was seen in all volunteers during extension of the fingers. The mean VB ratio at this level was slightly larger than that found by Mesgarzadeh et al. (9% versus 5.8%, respectively). 1~ During flexion of the fingers, no difference in this ratio was seen in 7 volunteers, including the 3 volunteers in whom the CSA was the same during extension and flexion of the fingers. The mean VB ratio was only 2% larger at this level compared to extension, despite the presence of lumbrical muscles in all volunteers. This could represent the thickness and stiffness of the flexor retinaculum at this level. However, in this ratio, only the most anterior point of the flexor retinaculum is used, and a more general bowing, which was often observed, is not measured. We observed anatomic changes in the carpal tunnel in 12 volunteers, with alteration of anatomy between finger flexion and extension. During flexion of the fingers, there is excursion of the lumbrical muscles into the carpal tunnel, together with an increase in CSA of the carpal tunnel. The exception to this is at the level of the hook of hamate, where we observed no increase in CSA during flexion in three volunteers, despite the presence of lumbrical muscles at this level. Further research will need to be done to establish the dynamic interaction between median nerve, flexor tendons, and lumbrical muscles in the carpal tunnel and its effect on CTS. We thank J. Klaase, MD, PhD, for statistical analysis.
References 1. Jessurun W. Revision of the carpal tunnel syndrome (a computed tomographic study). Thesis, State University of Groningen, the Netherlands, 1983. 2. Bassett LW, Gold RH. Magnetic resonance imaging of the musculoskeletal system: an overview. Clin Orthop 1989; 244:17-28. 3. Seeger LL. Physical principles of magnetic resonance imaging. Clin Orthop 1989;244:7-16. 4. Koenig H, Lucas D, Meissner R. The wrist: a preliminary report on high-resolution MR imaging. Radiology 1986; 160:463-467. 5. Baker LL, Hajek PC, Bjorkengren A et al. High-resolution magnetic resonance imaging of the wrist: normal anatomy. Skeletal Radiol 1987;16:128-132. 6. Weiss KL, Beltran J, Shamam OM, Stilla RF, Levey M. High-field MR surface-coil imaging of the hand and wrist: part 1. Normal anatomy. Radiology 1986;160:143-146. 7. Weiss KL, Beltran J, Lubbers LM. High-field MR surfacecoil imaging of the hand and wrist: part 2. Pathological
The Journal of Hand Surgery / Vol. 21A No. 6 November 1996
8.
9.
10.
11.
12.
13.
14.
15.
correlations and clinical relevance. Radiology 1986;160: 147-152. Middleton WD, Kneeland JB, Kellmann GM et al. MR imaging of the carpal tunnel: normal anatomy and preliminary findings in the carpal tunnel syndrome. Am J Roentgenol 1987;148:307-316. Richman JA, Gelberman RH, Rydevik BL et al. Carpal tunnel volume determination by magnetic resonance imaging and three-dimensional reconstruction. J Hand Surg 1987;12A:712-717. Mesgarzadeh M, Schneck CD, Bonakdarpour A. Carpal tunnel: MR imaging. Part I. Normal anatomy. Radiology 1989; 171:743-748. Mesgarzadeh M, Schneck CD, Bonakdarpour A, Mitra A, Conaway D. Carpal tunnel: MR imaging. Part II. Carpal tunnel syndrome. Radiology 1989; 171:749-754. Skie M, Zeiss J, Ebraheim NA, Jackson WTH. Carpal tunnel changes and median nerve compression during wrist flexion and extension seen by magnetic resonance imaging. J Hand Surg 1990;15A:934-939. Healy C, Watson JD, LongstaffA, Campbell MJ. Magnetic resonance imaging of the carpal tunnel. J Hand Surg 1990; 15B :243-248. Reicher MA, Kellerhouse LE. Carpal tunnel disease, flexor and extensor tendon disorders. In: Reicher MA, Kellerhouse LE, eds. MRI of the wrist and hand. New York: Raven Press, 1990: 49-68. Ham SJ, Konings JG, Wolf RFE, Mooyaart EL. Functional anatomy of the soft tissues of the hand and the wrist: in vivo excursion measurement of the flexor pollicis longus
16.
17.
18.
19.
20. 21.
22.
23.
24. 25.
1003
tendon using MRI. Magn Reson Imaging 1993;11: 163-167. Murphy RX, Chernofsky MA, Osborne MA, Wolson AH. Magnetic resonance imaging in the evaluation of persistent carpal tunnel syndrome. J Hand Surg 1993;18A: 113-120. Wehb6 MA, Hunter JM. Flexor tendon gliding in the hand: part II. Differential gliding. J Hand Surg 1985;10A: 575-579. Dekel S, Papaioannou T, Rushworth G et al. Idiopathic carpal tunnel syndrome caused by carpal stenosis. Br Med J 1980;1297-1299. Merhar GL, Clark RA, Schneider HJ. High-resolution CT scans of the wrist in patients with carpal tunnel syndrome. Radiology 1985;157:30. Bleecker ML. Medical surveillance for carpal tunnel syndrome in workers. J Hand Surg 1987;12A:845-848. Robbins H. Anatomic study of the median nerve in the carpal tunnel and etiologies of the carpal-tunnel syndrome. J Bone Joint Surg 1963;45A:953-966. Schmidt HM, Moser TH, Lucas D. Klinisch-anatomische Untersuchungen des Karpaltunnels der menschlichen Hand. Handchirurgie 1987;19:145-152. Bleecker ML, Bohlman M, Moreland R, Tipton A. Carpal tunnel syndrome: role of carpal canal size. Neurology 1985;35:1599-1604. Cobb TK, Dalley BK, Posteraro RH, Lewis RC. Anatomy of the flexor retinaculum. J Hand Surg 1993; 18-A:91-99. Nakamichi K, Tachibana S. Transverse sliding of the median nerve beneath the flexor retinaculum. J Hand Surg 1992;17B:213-216.
In the literature, much interest has been expressed in factors that could be of importance in producing or aggravating idiopathic carpal tunnel syndrome (CTS). I Hypertrophy of the lumbrical muscles is believed to be one of these factors, caused by occupation of an excessive amount of space in the carpal tunnel either in rest or during contraction of the deep
From the Department of Orthopedic Surgery, University Hospital, Groningen, the Netherlands, and the Departments of General Surgery and Radiology, Medical Spectrum Twente, Enschede, the Netherlands. Received for publication Nov. 9, 1994; accepted in revised form April 1, 1996. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Reprint requests: S. J. Ham, MD, Department of Orthopedic Surgery, University Hospital, RO. Box 30.001, 9700 RB Groningen, The Netherlands.
flexors of the fingers.1 For any further investigation of the etiologic or supporting role of the lumbrical muscles in the development and course of idiopathic CTS, knowledge about normal values and conditions in this anatomy are essential. Therefore, we investigated the position of the lumbrical muscles in healthy volunteers together with the changes occurring in the carpal tunnel both at extension and full flexion of the fingers by using magnetic resonance imaging (MRI). This method was used because it has proved to be a superior, noninvasive method of imaging and investigating the musculoskeletal system, including the carpal tunnel and its contents. 2-16
Materials and Methods In 12 healthy volunteers (6 men and 6 women) with a mean age of 30.5 years (range, 25-54 years) and with normal wrist and hand function, MRI studThe Journal of Hand Surgery
997
998
Ham et al. / Visualizing Carpal Tunnel Changes by MRI
ies of the carpal tunnel were made with the fingers both in extension and in full flexion. Full flexion (further referred to as flexion in this article) of the fingers was used to achieve maximal excursion of the profundus tendons. 17 The Gyroscan ACS whole-body system was used (Philips Inc., Eindhoven, The Netherlands). To exclude specific effects of dominance-related influences, the right, dominant hand was examined in six volunteers and the left, nondominant hand also in six volunteers, with equal distribution among the sexes. The volunteers were examined while they were in a prone position with their arm extended overhead. The palm of the hand was placed with the carpal tunnel located in the center of a 70-ram diameter flatsurface coil, with the distal part of the arm, with respect to the coil and the magnet, immobilized. Scans with 13 transverse slices of 4.0 mm and interslice gaps of 0.4 mm were made with the fingers both in extension and in flexion. These transverse slices were obtained at exactly fight angles to the carpal tunnel and forearm by using coronal and sagittal plane scans. Scans were done with the fingers in flexion and the thumb held in neutral because there is no lumbrical muscle attached to the flexor pollicis longus tendon, and to prevent the interference of this tendon with the shape and contents of the carpal tunnel. a5 Tl-weighted spin echo images (TR 519 ms, TE, 25 ms) were made with a field of view of 120 mm. Data were acquired using 410 x 512 matrices. Two excitations were used in all cases. A dynamic scan mode was used in order to minimize the time between scanning in extension and in flexion of the fingers. The total scan time was 11.26 minutes. Successive computer-enhanced cross-sectional areas (CSAs) of the carpal tunnel were measured in the images of each subject as described earlier by Richman et al. 9 The level in which the hook of hamate was visualized was used as level of reference (Fig. 1). Data acquired for the CSAs for both extension and flexion of the fingers were compared at this level and successive more proximal and distal levels using Wilcoxon's signed-rank test for statistical analysis. Reproducibility of the determination of the CSAs was validated by measuring one image five times. Volar bowing of the transverse ligament was measured using the volar bowing (VB) ratio as described by Mesgarzadeh et al. a0 The tendons used include the flexor digitorum superficialis (FDS), flexor digitorum profundus (FDP), and the flexor pollicis longus (FPL) in the index fin-
Figure 1. Transverse magnetic resonance imaging of the left carpal tunnel of one volunteer during extension of the fingers. The level of the hook of hamate (long, thin arrow) is visualized. The median nerve (short, thick arrow) is lying palmar to the flexor digitorum superficialis of the index and middle fingers. The flexor pollicis longus (long, thick arrow) is lying radial and slightly dorsal to the median nerve.
gers (II), middle fingers (III), ring fingers (IV), and little fingers (V).
Results Changes in Cross-sectional Area Figure 2 represents the mean CSA of the carpal tunnel measured during extension and flexion of the fingers. During extension of the fingers, one can clearly see the decrease in CSA from the proximal entrance of the carpal tunnel to the level where the hook of hamate was always visualized, lying about 2.0 cm distal to this entrance. Further distally, the mean CSA of the carpal tunnel increases. The effect is a canal with a slightly narrowed waist. During flexion of the fingers, there is an initial decrease in CSA followed by an increase up to the level where the hook of hamate was visualized, which was used as reference level, as described earlier. At this level, the mean CSA is considerably smaller than that of the level proximal to it. Again, rapid increase in CSA takes place distal to the reference level. The bony landmarks of each transverse level of the carpal tunnel that were observed in the volunteers are shown in Table 1. During extension of the fingers, proximal portions of lumbrical muscles were observed in the carpal tunnel in two volunteers, starting at the reference level in one, and at two levels distal to the reference
The Journal of Hand Surgery/Vol. 21A No. 6 November 1996 999 ram2 220
200
18o
16o 0
~ 1
~ 2
_1 3
~ 4
,
I
,
5
6
7
8
slices 9 Series 1 4- Series 2 Figure
2. Mean cross-sectional areas of the carpal tunnel during extension (series 1) and flexion (series 2).
level in the other volunteer. During flexion of the fingers, however, lumbrical muscles were observed at the reference level and at levels distal to it in all volunteers and were observed to extend proximal to the reference level in 11 of the 12 volunteers. Like Middleton et al., we found that the lumbrical muscles, although attached to the deep flexor tendons of the fingers, often appeared contiguous to both the deep and superficial flexor tendons on MRI. 8 The mean CSA measured at the reference level was considerably larger in flexion of the fingers than in extension: 191 mm 2 (SD, _+26) and 169 mm 2 (SD, +_ 15), respectively (p = .004). The mean CSAs for
Table 1. Relation Between Different Levels and Bony Landmarks of the Carpal Tunnel as Visualized With Magnetic Resonance Imaging Level
4 proximal 3 proximal 2 proximal 1 proximal Reference level 1 distal 2 distal
Bony Landmarks
Scaphoid; transition between lunate and capitate; triquetrum Scaphoid; capitate; proximal hamate; triquetrum; pisiform Transition between scaphoid and trapezium; capitate; transition between hamate and triquetrum; pisiform Trapezium; trapezoideum; capitate; hamate Hook of hamate Transition between distal carpal row and proximal part of metacarpals Proximal part of metacarpals
extension and flexion distal and just proximal to the reference level were also significantly different (Table 2). In three volunteers, however, no difference in CSA between extension and flexion was measured at the reference level, despite the presence of lumbrical muscles, whereas in these same volunteers, the CSA clearly increased during flexion at levels distal to it. During flexion, portions of lumbrical muscles were observed in a number of volunteers in the three levels proximal to the reference level. At the second and third level proximal to the reference level, the difference in mean CSA measured between extension and flexion was significantly larger when portions of lumbrical muscles were present (Table 3). Topographic C h a n g e s in the Carpal Tunnel During extension of the fingers, the median nerve was found to be palmar to FDS II and palmar and slightly radial to FDS III in 9 volunteers. The FPL was lying radial and dorsal to the median nerve before shifting to a more palmar position on distal sections. During flexion of the fingers, these topographic relations were unchanged in 4 volunteers. In 4 other volunteers, there was an ulnar displacement of the median nerve, especially on more distal sections (Fig. 3). In 1 volunteer, the FDS II lay immediately ulnar to the median nerve in the presence of lumbrical muscles (Fig. 4). In another volunteer,
1000
Ham et al./Visualizing Carpal Tunnel Changes by MRI Table 2. Mean Cross-sectional Areas Measured at Different Levels in the Carpal Tunnel in 12 Volunteers Level
Extension (mm2) (SD)
4 p. 3 p.* 2 p.* 1 p. Reference level 1 d. 2 d.
197 (_+ 191 (_+ 188 (_+ 179 (_+ 169 (_+ 178 (+ 191 (_+
20) 21) 19) 16) 15) 17) 20)
Flexion (mm2) (SD)
200 (+_ 195 (_+ 197 (+ 200 (+_ 191 (+_ 208 (+ 218 (_+
20) 21) 19) 32) 26) 28) 24)
Difference (mm2)
Significance (p < .05)
3 4 9 21 22 30 27
.220 .656 .083 .016 .004 .000 .001
p., proximal to the reference level; d., distal to the reference level. *See also Table 3.
the median nerve was interposed posterolaterally between the FDS II and the FPL as described by other authors.S, la During flexion, the same position was retained. In 2 volunteers, the median nerve b e c a m e interposed between the FDS III and FDS IV, with the FDP III lying dorsal to it, during extension. On the most distal sections, the nerve was located ulnar and palmar to the FDS III. During flexion,
there was radial displacement of the median nerve on distal sections in 1 of these 2 volunteers. In the other volunteer, the median nerve never became interposed between the FDS III and FDS IV during flexion, because of palmar displacement of the FDS III in the presence of lumbrical muscles lying dorsal to it in the carpal tunnel. The median nerve remained locked between the FPL lying radial to it, the FDS II i m m e diately dorsal, and the FDS III lying immediately ulnar to it (Fig. 5). In all 12 volunteers, increase in nerve flattening was observed in the presence of lumbrical muscles. The F D P II was always located dorsal to the FDS II and FDS III during extension, before shifting radially. During flexion, the F D P II was pressed in the radiodorsal c o m e r of the carpal tunnel (in 6 volunteers) or against the dorsal wall of the tunnel (in 4). Two times, there was no difference between extension and flexion. During extension of the fingers, the FDP III was always dorsal and more or less ulnar to
A
B
Table 3. Difference in Cross-sectional Areas Between Extension and Full Flexion Mean Difference (mm2) Level
3 p. 2 p.
Lumbrical Muscles Present
Lumbrical Muscles Absent
Significance (p <. 05)*
22 (n = 4) 21 (n = 6)
-6 (n = 8) -3 (n = 6)
.006 .004
p., proximal to the reference level. *Mann-Whitney test for statistical analysis.
Figure 3. Transverse magnetic resonance images of the right carpal tunnel of one volunteer visualized at the level of the hook of hamate during (A) extension and (B) flexion of the fingers. During flexion of the fingers, ulnar displacement of the median nerve (short, thick arrow) and fat compression (long, thin arrow) dorsal in the carpal tunnel in the presence of lumbrical muscles (short, thin arrows) can be observed, with a lumbrical muscle dividing the median nerve from the flexor pollicis longus (long, thick arrow).
The Journal of Hand Surgery / Vol. 21A No. 6 November 1996
1001
A B Figure 4. Transverse magnetic resonance images of the left carpal tunnel of one volunteer visualized at the level of the hook of hamate during (A) extension and (B) flexion of the fingers. During flexion, there is ulnar displacement of the median nerve (short, thick arrow) in the presence of lumbrical muscles (short, thin arrows), with the flexor digitorum superficialis of the index finger (long, thin arrow) being interposed between the flexor pollicis longus (long, thick arrow) tendon radial to it and the median nerve ulnar to it.
the FDS IV, before shifting radially, but always lay ulnar and dorsal to the FDS III. During flexion, the same topographic relation of these tendons was maintained, with the FDP III being pressed against the dorsal roof of the carpal tunnel. The FDS IV and V always lay in the palmar and ulnar comer of the carpal tunnel and were pressed in this comer during flexion. Discrimination of the FDP IV and V lying in the dorsal and ulnar corner of the carpal tunnel was more pronounced during flexion, owing to the presence of a lumbrical muscle, which divided them on distal sections (Fig. 4). Fat compression was seen in all volunteers in the presence of lumbrical muscles (Fig. 3). Volar bowing
at the reference level was observed in all volunteers during extension, with a mean VB ratio of 9% (range, 6%-13%; SD, + 2.0). During flexion, this ratio clearly increased in only five volunteers at this level, with a mean VB ratio of 11% (range, 7%-17%; SD, + 2.4).
Discussion The role of CSAs and volume of the carpal tunnel in the development of CTS remains unclear. Both Dekel et al. and Bleecker used computed tomography to show a decrease in the CSAs of CTS patients, although later, Merhar et al. found no significant dif-
~N
A
B
Figure 5. Transverse magnetic resonance images of the fight carpal tunnel of one volunteer visualized at the level of the hook of hamate during (A) extension and (B) flexion of the fingers. During extension, the median nerve (short, thick arrow) is interposed between the flexor digitorum superficialis of the middle finger (long, thick arrow) and the FDS of the ring finger (long, thin arrow). During flexion, the ulnar displacement of the median nerve does not occur. The FDS of the middle finger has been displaced anteriorly in the presence of lumbrical muscles (short, thin arrows) lying dorsal to it.
1002
Ham et al./Visualizing Carpal Tunnel Changes by MRI
ference in CSA measurements between wrists with and without symptoms. 18-20 However, CSA measurements of the carpal tunnel vary, depending on the position of the wrist and of the fingers, a2 During flexion of the fingers, we observed lumbrical muscles to move into the carpal tunnel in all our volunteers, resulting in considerable increase in CSA in the presence of these muscles. At the level of the hook of hamate, where the tunnel is smallest, lumbrical muscles were observed in all our volunteers during flexion of the fingers, with a considerable increase in C S A in 9. In 3 volunteers, however, no increase in CSA was observed at this level during flexion. According to Cobb et al., the level of the hook of hamate is anatomically a potential site for median nerve compression. 1~ At this level, the tunnel is rigidly bounded on three sides by bony structures and roofed by a thickened transverse carpal ligament, and apparently, there is not always the possibility to respond to the increase in volume of structures by expansion of the CSA. Furthermore, functionally, this site may rely on any space-occupying condition to exceed the critical volume of the tunnel at this level, leading to median nerve compression. 24 Abnormalities of the lumbrical muscles, therefore, may play a role in the etiology of the CTS in some patients. However, further work will be needed to establish this relationship. In our study, the alignment and shape of the median nerve in the carpal tunnel and its relationship to the flexor tendons were variable and depended on the presence, amount, shape, and size of the lumbrical muscles. Transverse sliding of the median nerve was clearly seen in 7 volunteers, with ulnar displacement in 5 and radial displacement in 2. Nakamichi and Tachibana demonstrated ulnar sliding of the median nerve beneath the flexor retinaculum during active-resistant flexion of the fingers in healthy volunteers by using sonography.25 In their opinion, this sliding was caused by tensed overlying flexor tendons and could be a reason for mechanical nerve deformation. In our study, the ulnar and radial displacement of the median nerve was caused by the rearrangement of structures in the carpal canal due to the presence of lumbrical muscles. This presence of lumbrical muscles during flexion of the fingers also seemed to be responsible for the displacement of the flexor tendons occasionally observed in the carpal tunnel and for the alterations in configuration of the flexor tendons, which became more rounded during flexion of the fingers.
Finally, VB of the flexor retinaculum at the level of the hook of hamate was seen in all volunteers during extension of the fingers. The mean VB ratio at this level was slightly larger than that found by Mesgarzadeh et al. (9% versus 5.8%, respectively). 1~ During flexion of the fingers, no difference in this ratio was seen in 7 volunteers, including the 3 volunteers in whom the CSA was the same during extension and flexion of the fingers. The mean VB ratio was only 2% larger at this level compared to extension, despite the presence of lumbrical muscles in all volunteers. This could represent the thickness and stiffness of the flexor retinaculum at this level. However, in this ratio, only the most anterior point of the flexor retinaculum is used, and a more general bowing, which was often observed, is not measured. We observed anatomic changes in the carpal tunnel in 12 volunteers, with alteration of anatomy between finger flexion and extension. During flexion of the fingers, there is excursion of the lumbrical muscles into the carpal tunnel, together with an increase in CSA of the carpal tunnel. The exception to this is at the level of the hook of hamate, where we observed no increase in CSA during flexion in three volunteers, despite the presence of lumbrical muscles at this level. Further research will need to be done to establish the dynamic interaction between median nerve, flexor tendons, and lumbrical muscles in the carpal tunnel and its effect on CTS. We thank J. Klaase, MD, PhD, for statistical analysis.
References 1. Jessurun W. Revision of the carpal tunnel syndrome (a computed tomographic study). Thesis, State University of Groningen, the Netherlands, 1983. 2. Bassett LW, Gold RH. Magnetic resonance imaging of the musculoskeletal system: an overview. Clin Orthop 1989; 244:17-28. 3. Seeger LL. Physical principles of magnetic resonance imaging. Clin Orthop 1989;244:7-16. 4. Koenig H, Lucas D, Meissner R. The wrist: a preliminary report on high-resolution MR imaging. Radiology 1986; 160:463-467. 5. Baker LL, Hajek PC, Bjorkengren A et al. High-resolution magnetic resonance imaging of the wrist: normal anatomy. Skeletal Radiol 1987;16:128-132. 6. Weiss KL, Beltran J, Shamam OM, Stilla RF, Levey M. High-field MR surface-coil imaging of the hand and wrist: part 1. Normal anatomy. Radiology 1986;160:143-146. 7. Weiss KL, Beltran J, Lubbers LM. High-field MR surfacecoil imaging of the hand and wrist: part 2. Pathological
The Journal of Hand Surgery / Vol. 21A No. 6 November 1996
8.
9.
10.
11.
12.
13.
14.
15.
correlations and clinical relevance. Radiology 1986;160: 147-152. Middleton WD, Kneeland JB, Kellmann GM et al. MR imaging of the carpal tunnel: normal anatomy and preliminary findings in the carpal tunnel syndrome. Am J Roentgenol 1987;148:307-316. Richman JA, Gelberman RH, Rydevik BL et al. Carpal tunnel volume determination by magnetic resonance imaging and three-dimensional reconstruction. J Hand Surg 1987;12A:712-717. Mesgarzadeh M, Schneck CD, Bonakdarpour A. Carpal tunnel: MR imaging. Part I. Normal anatomy. Radiology 1989; 171:743-748. Mesgarzadeh M, Schneck CD, Bonakdarpour A, Mitra A, Conaway D. Carpal tunnel: MR imaging. Part II. Carpal tunnel syndrome. Radiology 1989; 171:749-754. Skie M, Zeiss J, Ebraheim NA, Jackson WTH. Carpal tunnel changes and median nerve compression during wrist flexion and extension seen by magnetic resonance imaging. J Hand Surg 1990;15A:934-939. Healy C, Watson JD, LongstaffA, Campbell MJ. Magnetic resonance imaging of the carpal tunnel. J Hand Surg 1990; 15B :243-248. Reicher MA, Kellerhouse LE. Carpal tunnel disease, flexor and extensor tendon disorders. In: Reicher MA, Kellerhouse LE, eds. MRI of the wrist and hand. New York: Raven Press, 1990: 49-68. Ham SJ, Konings JG, Wolf RFE, Mooyaart EL. Functional anatomy of the soft tissues of the hand and the wrist: in vivo excursion measurement of the flexor pollicis longus
16.
17.
18.
19.
20. 21.
22.
23.
24. 25.
1003
tendon using MRI. Magn Reson Imaging 1993;11: 163-167. Murphy RX, Chernofsky MA, Osborne MA, Wolson AH. Magnetic resonance imaging in the evaluation of persistent carpal tunnel syndrome. J Hand Surg 1993;18A: 113-120. Wehb6 MA, Hunter JM. Flexor tendon gliding in the hand: part II. Differential gliding. J Hand Surg 1985;10A: 575-579. Dekel S, Papaioannou T, Rushworth G et al. Idiopathic carpal tunnel syndrome caused by carpal stenosis. Br Med J 1980;1297-1299. Merhar GL, Clark RA, Schneider HJ. High-resolution CT scans of the wrist in patients with carpal tunnel syndrome. Radiology 1985;157:30. Bleecker ML. Medical surveillance for carpal tunnel syndrome in workers. J Hand Surg 1987;12A:845-848. Robbins H. Anatomic study of the median nerve in the carpal tunnel and etiologies of the carpal-tunnel syndrome. J Bone Joint Surg 1963;45A:953-966. Schmidt HM, Moser TH, Lucas D. Klinisch-anatomische Untersuchungen des Karpaltunnels der menschlichen Hand. Handchirurgie 1987;19:145-152. Bleecker ML, Bohlman M, Moreland R, Tipton A. Carpal tunnel syndrome: role of carpal canal size. Neurology 1985;35:1599-1604. Cobb TK, Dalley BK, Posteraro RH, Lewis RC. Anatomy of the flexor retinaculum. J Hand Surg 1993; 18-A:91-99. Nakamichi K, Tachibana S. Transverse sliding of the median nerve beneath the flexor retinaculum. J Hand Surg 1992;17B:213-216.