- Email: [email protected]
Gliding resistance of the extensor pollicis brevis tendon and abductor pollicis longus tendon within the first dorsal compartment in fixed wrist positions
ELSEVIER
Journal of Ort hopaed ic Research
Journal of Orthopaedic Research 23 (2005) 243-248
www.elsevier.com/locate/orthres
Gliding resistance of the extensor pollicis brevis tendon and abductor pollicis longus tendon within the first dorsal compartment in fixed wrist positions Keiji Kutsumi, Peter C . Amadio
+,
Chunfeng Zhao, Mark E. Zobitz, Kai-Nan An
Orthopedic Biomechanics Luboratory, Mayo Clinic Rochester, 200 First Street S W. Rochester, MN 55905, USA
Received 16 lanuary 2004; accepted 18 June 2004
Abstract Purpose: While the etiology of de Quervain's disease is unknown, repetitive motion coupled with awkward wrist position and septation within the first dorsal compartment are considered causative factors. We hypothesize that these conditions might produce high gliding resistance, which could then induce micro-damage of the tendons and retinaculum. The purpose of this study was to measure the gliding resistance of the extensor pollicis brevis and abductor pollicis longus tendons within the first dorsal compartment in a human cadaver model. Methods: Fifteen human cadaver wrists, which included eight septation and seven non-septation wrists in the first dorsal compartment, were used. Gliding resistance of the extensor pollicis brevis and abductor pollicis longus tendons was measured in seven wrist positions: 60" extension, 30" extension, O", 30" flexion, 60" flexion in neutral deviation and 30" ulnar deviation, 15" radial deviation in neutral extensiodflexion. Results: The overall gliding resistance was not different between septation and non-septation wrists (0.21 versus 0.19 N for abductor pollicis longus and 0.21 versus 0.15 N for extensor pollicis brevis, respectively), but there was a significant effect on gliding resistance due to wrist position (p < 0.05) in both tendons. Interaction between wrist position and septation status was observed in the extensor pollicis brevis tendon (p < 0.05). With septation, the gliding resistance of the extensor pollicis brevis was significantly higher in 60" wrist flexion (0.51 N) compared to all other wrist positions tested (all less than 0.26 N) (p < 0.05). In the non-septation group, gliding resistance was significantly higher in 60" flexion (0.20 N) and 60" extension (0.22 N) compared to the other five wrist positions (all less than 0.15 N) (p < 0.05). Although no significant difference was observed, the extensor pollicis brevis tendon with septation tended to have higher gliding resistance than that without septation in wrist flexion. In 60" of wrist flexion the abductor pollicis longus tendon had significantly higher gliding resistance (0.33 N) than the other wrist positions (all less than 0.26 N) (p < 0.05). Conclusions: A combination of septation and wrist position significantly affected extensor pollicis brevis tendon gliding resistance in this cadaver model. These factors may contribute to the development of de Quervain's disease. 0 2004 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. Keywords: De Quervain's disease; Gliding resistance; Tendon: Extensor retinaculum; Biomechanics
Introduction
Corresponding author. Tel.: + I 507 538 1717; fax: + I 507 284 5392. E-mail address: [email protected](P.C. Amadio).
Tenosynovitis of the tendons in the first dorsal compartment of the wrist, the extensor pollicis brevis and abductor pollicis longus, was first mentioned in the thirteenth edition of Gray's Anatomy in 1893 as
0736-02666 - see front matter 0 2004 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. doi:10.10I6/J.orthres.2004.06.014
244
K. Kutsuini
el
al. I Journal of Orihopaedic Reseorcli 23 (2005) 243-248
washerwoman’s sprain [21]. In 1895, de Quervain published a report of five cases of chronic tenovaginitis in the first dorsal compartment [7]. After that report, many papers were published on the clinical characteristics, diagnosis and treatment of what quickly became known as de Quervain’s disease. The pathophysiology of de Quervain’s disease is characterized by non-inflammatory fibrosis of the tenosynovium. The etiology is presumed to be related to repetitive motion [9,14,16,18,20,25]. While the effect of wrist position on de Quervain’s disease is generally accepted, the specific position of risk is controversial. Some authors have stated that the repeated use of firm grasp, together with ulnar deviation of the wrist, could predispose to de Quervain’s disease [9,20,36]. Others have reported an association with radial deviation ~~251. Anatomic variations of the first compartment have been discussed frequently in relation to de Quervain’s disease. Variations in the number and insertion of the abductor pollicis longus tendon have been reported [10,15,18], but these have rarely been discussed in relation to the etiology. Septation between the extensor pollicis brevis and abductor pollicis longus tendons in the first dorsal compartment is common. The incidence of the septum in cadaver studies varies from 24% [l8] to 77.5% [23]. Most surgical series, however, report the presence of such a septum in every case and therefore some investigators have suggested that the septum might play an important role in the etiology of the condition [13,14,21,35]. Yet, despite these suggestions, to our knowledge there is no literature that explains how septation might predispose to de Quervain’s disease. We hypothesize that friction between the tendon and the extensor retinaculum in the first dorsal compartment might play a role in the development of de Quervain’s disease, and specifically that wrist positions and/or anatomic variations may result in higher friction levels within this compartment. Higher friction levels may, in turn, predispose the tenosynovium to mechanical injury. The purpose of this study was to investigate the gliding resistance of extensor pollicis brevis and abductor pollicis longus tendons within the first compartment, in a human cadaver model. Since the pathognomonic test for de Quervain’s disease, the Finkelstein test, holds the thumb trapeziometacarpal position fixed, the effect of this joint was not studied.
Methods We used a tendon frictional testing device modified from a previously described and validated gliding resistance testing machine [I ,32,40].The measurement system consists of one mechanical actuator with a linear potentiometer, two tensile load transducers, and a movable mechanical pulley to satisfy the need for three dimensional wrist
Fig. 1. The device for measuring gliding resistance. motion. With data collected from this device, the gliding resistance can be calculated, as the difference in force between the two transducers connected to the proximal and distal tendon ends. respectively. The specimen is mounted on the custom-made device by clamping the proximal end of the radius and ulna. The wrist position is maintained using a custom-built external fixator that has two rods connected by one lockable universal joint and two adjustable connectors. Each connector joins two threaded 3 mm rods that are driven into proximal part of the radius and the second metacarpal bone. The wrist angle can thus be changed to any selected angle (Fig. I). Fifteen fresh-frozen human cadaveric upper extremities, amputated 15 cm proximal to the wrist joint, were used for this study. The cadaver donors ranged in age from 56 to 97 (mean 76.4 years). Eight specimens with septation between the extensor pollicis brevis and abductor pollicis longus tendons and seven specimens without septation were used (presence of septation determined after testing was completed). Specimens were stored in a freezer at -20 “C, and thawed at room temperature immediately prior to testing. All specimens were X-rayed to exclude gross pathological evidence of injuries or major degenerative changes around the wrist. A longitudinal incision of the skin was made to expose the overall length of the extensor pollicis brevis and the abductor pollicis longus. Fascia over the extensor pollicis brevis and abductor pollicis longus muscles was removed. The extensor retinaculum was carefully preserved. The extensor pollicis brevis and abductor pollicis longus muscles were carefully separated. Excursion of the extensor pollicis brevis and abductor pollicis longus tendons was recorded in each fixed wrist position prior to the gliding resistance measurements. At first, the extensor pollicis brevis muscle belly was detached from its origin and connected to a 4.9 N weight through a Dacron cord to apply tension proximally on the muscle [32]. A mark was made on the tendon surface by suturing 6-0 monofilament nylon proximal enough to the retinaculum edge so as not to interfere with gliding. The distance between the mark and proximal retinaculum edge was measured in thumb interphalangeal, metacarpophalangeal and trapeziometacarpal joint extension, pulled by the weight attached to the muscle. Then, the distance was recorded again after passive thumb flexion to full interphalangeal, metacarpophalangeal and trapeziometacarpal joint flexion in each fixed wrist position. The range of excursion in each wrist position was identified using this method. The abductor pollicis longus tendon excursion was measured using the same method. Following excursion measurement, the extensor pollicis brevis tendon was cut at the level of the metacarpophalangeal joint. The distal tendon end and proximal muscle were connected to the distal ( F I ) and proximal (n) load transducers, respectively. The distal transducer (F1)was connected to a 4.9 N weight through a mechanical pulley and the position of the pulley was changed according to the wrist angle to represent the physiological line of action for the tendon. The 4.9 N weight was used to get a reliable excursion. as this weight has been shown to be the minimum load sufficient to remove slack from the muscle-tendon unit [I 21. The proximal transducer (n) was connected to the mechanical actuator that was placed to simulate the physiological muscle angle (Fig. I).
245
K. Kutsumi et al. I Journul q/ Orthopedic Reseurch 23 (2005) 243-248
The tendon was pulled proximally by the actuator against the gliding resistance and the 4.9 N weight at a rate of 2.0 mm/s. This movement of the tendon toward the actuator was regarded as extension. Then, the actuator was reversed so that the weight pulled the tendon distally. This movement of the tendon toward the niechanical pulley was considered as flexion. F1, F2 and the corresponding excursion were recorded at a sampling rate of 10 Hz. After measuring the gliding resistance of the extensor pollicis brevis tendon, the superficial division of the abductor pollicis longus tendon 1331 was detached from the first metacarpal bone with a small bony fragment. Osteotomy of the bony fragment prevented damage and deformity to the abductor pollicis longus tendon and provided a structure for attachment to the load transducer. The deep division of the abductor pollicis longus tendon was cut at the insertion and connected to the superficial division using side-by-side suture. The bony fragment at the abductor pollicis longus insertion was Proximally, the abducconnected to the distal load transducer (A). tor pollicis longus muscle was connected to the proximal load transducer (F2). Then the gliding resistance of the abductor pollicis longus tendon was measured in the same manner. The trapeziometacarpal joint was fixed by a 1.5 nun Kirschner wire in a position where the posture of the joint would mimic activities such as tip pinch or grasp. Seven different wrist positions were tested 60" extension, 30" extension, O", 30" flexion, 60" flexion in neutral deviation and 30" ulnar deviation, 15O radial deviation in neutral extensionlflexion. The sequence of the wrist positrons was randomized. The angle made between the third metacarpal bone and long axis of radius on a lateral view was measured as the wrist angle using a goniometer. The specimen was kept moist throughout the testing procedure with saline. After completing all measurements, the extensor retinaculum was opened to determine the variations, if any, in the number of the tendons and the presence of septation in the first compartment. Data was collected as the tendon moved thorough its normal range of excursion. The tendons were in motion at the points determined to be the endpoints of excursion to eliminate any static friction effects. In a given direction of motion, the gliding resistance is the difference between the proximal and distal force sensors. The average gliding resistance of the results from extension to flexion of whole excursion was used in this study ((F2exlmsion - FI + should be the same as (FI flexion - F2aexion))/2.Because Fl F, flexion during the testing (4.9 N), the composite gliding resistance may be calculated as: Gliding Resistance = (F2rxlension - F!fiexion)/2. Bulk and tethering effects were eliminated with this calculation because direction of these effects are opposite between flexion and extension. This method was chosen in order to focus on surface friction. Two extension/flexion trials were performed for each condition. Only the data from the second trial was used for analysis to eliminate any positioning effects that may have been present during specimen set-up.
Results We tested eight septation and seven non-septation specimens. All of the septations were partial (3-1 5 mm), and located at the distal part of the first compartment. The number of tendon slips of the abductor pollicis longus were: two tendon slips ( n = 8 specimens), three tendon slips (n = 2 specimens), four tendon slips ( n = 4 specimens), five tendon slips (n = 1 specimen). The excursion from full extension to full flexion of the thumb interphalangeal, metacarpophalangeal, and trapeziometacarpal joint in each fixed wrist position is shown in Fig. 2. The average excursions of the extensor pollicis brevis and abductor pollicis longus tendons were 15.0 and 4.6 mm, respectively. From the main effects of the analysis of variance model there was no significant difference in extensor pollicis brevis gliding resistance between septation and nonseptation (p = 0.18) wrists, although there was a significant difference due to wrist position 0, < 0.05) and a significant interaction between wrist position and septation status (p < 0.05). Therefore, separate one-factor repeated measures analysis of variance models for the specimens with septation and without septation were run. With septation, the gliding resistance of the extensor pollicis brevis was significantly higher in 60" wrist flexion (0.51 N) compared to all other wrist positions tested (all less than 0.26 N) (p < 0.05) (Fig. 3). In the non-septation group, gliding resistance of the extensor pollicis brevis tendon was significantly higher in 60" flexi on (0.20 N) and 60" extension (0.22 N) compared to the other five wrist positions (all less than 0.15 N) (p < 0.05). Although not statistically significant, there was a diverging trend between septatiodnon-septation at 30" and 60" flexion, with septation having a higher gliding
Statistical methods The sample size requirements were determined by a power calculation using previous studies of tendon gliding resistance. A sample of 14 specimens (7 with septation and 7 without septation) will provide 80% power to detect a difference in gliding resistance between any two of the seven wrist positions equal to 0.12 N, and 80% power to detect a difference in mean gliding resistance between specimens with septation and those without septation equal to 0.24 N. Separate analyses were conducted for the extensor pollicis brevis tendon and the abductor pollicis longus tendon. For each tendon, two-factor analysis of variance with repeated measures on cine factor (wrist position) was used to compare the gliding resistance among the seven wrist positions and between specimens with sephtion and without septation. When significant interactions were identified, separate one-Factor repeated measures analysis of variance models were run for the specimens with septation and without septation. When wrist angle was found to be significant, further analysis was performed using the Ryan-EinotCrabriel-Welsch multiple comparisons procedure. All statistical tests were two-sided and p-values less than 0.05 were considered significant.
cxt.6W cxt.30'
neut.
fkx.30' 0 e x . W
u.dev.30'
&.
r.h.15'
Fig. 2. Mean values ? SD of the excursion of the extensor pollicis brevis and abductor pollicis longus tendons from full extension to full flexion of the thumb interphalangeal, metacarpophalangeal, and trapeziometacarpal joint in each wrist position. ext.: extension, flex.: flexion, neut.: neutral, u. dev.: ulnar deviation, r. dev.: radial deviation.
K. Kutsumi et al. I Journal of' Orthopaedic Research 23 (2005) 243-248
246
Discussion
eXt.60.
eXt.t.30.
ncllt.
flex.30. flCx.60.
U.deT.W
MU. r.dn. 15.
Fig. 3. Mean values k SD of the gliding resistance of the extensor pollicis brevis tendon. ext.: extension, flex.: flexion, neut.: neutral, u. dev.: ulnar deviation, r. dev.: radial deviation, A had significant difference with .. A had significant difference with 0.
resistance. At the other wrist positions, however, the gliding resistances were very similar. There was no significant difference in abductor pollicis longus tendon gliding resistance comparing the septation (0.21 N) and non-septation groups (0.19 N) (p = 0.67). Wrist position, however, had a significant effect (p < 0.05) on abductor pollicis longus tendon gliding resistance (Fig. 4). In 60" of wrist flexion the abductor pollicis longus tendon had significantly higher gliding resistance (0.33 N) than the other wrist positions (all less than 0.26 N) (p < 0.05). Additionally, in 60" wrist extension and 30" wrist flexion the abductor pollicis longus gliding resistance was significantly greater than neutral, 30" extension, 30" ulnar deviation, or 15" radial deviation. There was no significant interaction between septation status and wrist position in gliding resistance of the abductor pollicis longus tendon (p = 0.44).
O.7
0
r
ext.60. e x t W
ncllt.
flex.30. I k x . W
u.dev.30"
meut. r.dev. 15'
Fig. 4. Mean values 2 SD of the gliding resistance of the abductor pollicis longus tendon. ext.: extension, flex.: flexion, neut.: neutral, u. dev.: ulnar deviation, r. dev.: radial deviation, A had significant difference with 0 and 0. 0 had significant difference with 0.
In the first dorsal compartment of the wrist extensor retinaculum, the extensor pollicis brevis and abductor pollicis longus tendons pass under a fibrous sheath, which is fixed on either side of the tendons to the dorsal radius, thus forming a fibro-osseous tunnel. As the tendons are subjected to varying degrees of angulation with wrist motion, these tunnels function as a pulley, constraining the lateral motion of the tendons and preventing the tendons from bowstringing. But this function also acts to produce friction between the tendons and the retinaculum. In the first compartment, the extensor pollicis brevis tendon is located dorsally and the abductor pollicis longus volarly. In the presence of the septation, the extensor pollicis brevis tendon occupies a separate subcompartment. Leslie et al. [19] described that the septum began distally in the region of the radial styloid and extended proximally for a distance of 0.5-2 cm. The reported incidence of a septum in the first compartment in cadaver series varies from 24% [I81 to 77.5% [23]. In the extensor pollicis brevis tendon without septation, the volar surface contacts the abductor pollicis longus tendon. This interaction may have a different frictional coefficient than that with the surface of the septum. Moreover, as both tendons move synergistically in the physiological condition, relative displacement of the volar surface of the extensor pollicis brevis tendon without septation should be smaller than with septation. This may result in a less abrasive effect on the tendon without septation than with septation. We do not think that the increased friction would be caused by a narrowed canal, because the gliding resistance between septation and non-septation specimens was very similar at the other wrist positions. However, if swelling or edema occurred, the narrowed canal might further increase gliding resistance to a considerable extent. Variations in the anatomy of the first extensor compartment have been reported as being associated with the development of de Quervain's disease [9-1 lJ3, 21,35,37]. Our results seem to support these observations. Some extensor pollicis brevis tendons with septation had quite high gliding resistance in the wrist flexion position. This is the reason why the standard deviation of the extensor pollicis brevis tendon septation group was relatively high. While the values for gliding resistance in the neutral position, roughly 0.2 N, are similar to those seen in other normal intrasynovial tendons [24,39], the values for gliding resistance in the most adverse wrist positions approach those seen in healing, lacerated tendons [39]. We believe it is reasonable to suspect that tendons with such high gliding resistance, such as the extensor pollicis brevis tendon with septation, might be vulnerable to develop de Quervain's disease.
K. Kutsumi et ul. I Journal
of
Orthopuedic Researcli 23 (200.5) 243-248
Lubrication of the tendon is a critical factor in its gliding resistance. Higher friction, associated with movements in certain wrist positions, such as pinching in wrist flexion, especially in the presence of septation, may induce surface damage [40], which in turn may result in inflammation. Inflammatory enzymes may then destroy lubricating glycoproteins such as hyaluronan [31] and lubricin [27] on the surface of the tendon and extensor retinaculum, resulting in a vicious cycle of increasing friction, damage, and loss of lubricating glycoproteins. The histopathology of de Quervain’s disease has been variably described as showing inflammation [5,6,18,25], degenerative change [3], or some combination of the two [ 16,20,25,29,38], consistent with this hypothesis. The average excursion of the extensor pollicis brevis tendon in this study, which represents the excursion for thumb extension-flexion movement without wrist motion, was about three times greater than that of the abductor pollicis longus tendon. This observation, coupled with related reports by others, supports the hypothesis that the extensor pollicis brevis may be the tendon most affected by de Quervain’s disease. It is known that failure to open the septation can explain poor results of surgical treatment for de Quervain’s disease [22]. Selective steroid injection into the extensor pollicis brevis tenosynovium is reported as being effective therapy [28], and Yuasa and Kiyoshige [37] have reported good results by decompression of only the extensor pollicis brevis subcompartment. These studies suggest that the extensor pollicis brevis tendon may be more vulnerable to the disease than the abductor pollicis longus tendon. This is consistent with our observation, because the tendon with the longer excursion might easily be the one more susceptible to injury related to repetitive motion. One of the limitations of this study was that we used saline solution instead of synovial fluid for lubrication. This may affect the absolute gliding resistance of the tendon, although we believe that the effect is not so strong that it would change our results [31]. A second limitation is that the data was collected from one trial for each specimen (the second trial) at each wrist position. In preliminary testing we tested 3 times per position. The results were very stable so we tested each condition twice and analyzed the second run, considering the first trial to be a conditioning run. Other studies on gliding resistance after flexor tendon repair have also found that after the first cycle of motion the results are consistent over then next few trials. To minimize the risk of abrasion with repeated testing, we only made two trials at each position, we randomized the testing sequence among the specimens, and the tendon was kept moist between trials by spraying with saline. A third limitation was that we did not consider the effect of gender. De
241
Quervain’s disease is more common in women than in men, with reported ratios of women to men ranging from 4:l to 7:l [8,17,18,25,26,34,35]. Some investigators suggest that the wrists normally angulate further in women than in men and this is the reason why women are predisposed to the disease [4,20]. We did not test this hypothesis, as we only measured the gliding resistance at given wrist angles, not at the maximum angle. Even if there is a gender difference, however, we believe that the interaction of septation and wrist position that we observed is likely similar in both genders. A fourth limitation was that we did not analyze the number of abductor pollicis longus tendons and the size of the subcompartment. In our data, these factors did not seem to affect the friction of the tendons, although we did not have enough data for statistical analyses. The final limitation is that we only looked at the effect of wrist position and septation, and did not study the impact of forearm or thumb position. However, in a previous study performed in our laboratory (unpublished), there was little change in EPB or APL muscle-tendon excursion with forearm position. In the Finkelstein test, universally used to provoke the symptoms of de Quervain’s disease, the thumb is held still, while the wrist is moved. Based on these observations, we believe that the effect of forearm or thumb position on first compartment friction is likely to be small. However, we do plan to study this in the future. A strength of this study is that we were able to directly measure the gliding resistance in the first dorsal compartment of the wrist. To our knowledge, this is the first report to investigate the gliding resistance of the extensor tendons. Some authors suggest that the anatomical characteristic of the first compartment, in which the tendons form an acute angle, induces high friction and mechanical irritation [5,30]. Our validated gliding resistance testing machine also allowed us to measure the gliding resistance of the extensor tendons in various wrist positions, by adjusting the position of the mechanical pulley. In conclusion, we have directly measured the gliding resistance of the tendons of the human first dorsal wrist compartment in a cadaver model. We have shown that septation within the compartment and wrist position combine to affect friction on the extensor pollicis brevis. These observations support the hypothesis that a friction-induced tenosynovitis of the extensor pollicis brevis may be the primary source of pathology in de Quervain’s disease. Acknowledgement
This study was funded by grants from the NIH/ NIAMS (AR44391) and Mayo Clinic Rochester.
248
K. Kutsumi et al. I Journal of Orthopaedic Research 23 (2005) 243-248
References [l] An KN, Berglund L, Uchiyama S, Coert JH. Measurement of friction between pulley and flexor tendon. Biomed Sci Instrum 1993;29:1-7. [2] Bums BH, Ellis VH. Stenosing tendovaginitis at the radial styloid process. Lancet 1936;1:717. [3] Clarke MT, Lyall HA, Grant JW, Matthewson MH. The histopathology of de Quervain’s disease. J Hand Surg [Br] 1998;23:7324. [4] Conklin JE, White WL. Stenosing tenosynovitis-and its possible relation to the carpal tunnel syndrome. Surg Clin North Am 196040:53140. [5] Cotton FJ, Morrison GM. Bradford CH. De Quervain’s diseaseradial styloid tendovaginitis. N Engl J Med 1938;219:120-3. [6] de Quervain F. Uber das Wesen und die Behandlung der stenosierenden Tendovaginitis am Processus styloideus radii. Miinchener Med Wochenschr 1912;59:5-6. [7l de Quervain F. Uber eine Form von chronischer Tendovaginitis. Corresp Blatt Schweizer Arzte 1895;25:389-94. [8] Faithfull DK, Lamb DW. De Quervain’s disease-a clinical review. Hand 1971;3:23-30. [9] Finkelstein H. Stenosing tenovaginitis at the radial styloid process. J Bone Joint Surg [Am] 1930;12:50940. [lo] Giles KW. Anatomical variations affecting the surgery of de Quervain’s disease. J Bone Joint Surg [Br] 1960;42:352-5. [Ill Gonzalez MH, Sohlberg R, Brown A, Weinzweig N. The first dorsal extensor compartment: an anatomic study. J Hand Surg [Am] 1995;20:657-9. [I21 Horii E, Lin GT, Cooney WP. Linscheid RL, An KN. Comparative flexor tendon excursion after passive mobilization: an in vitro study. J Hand Surg [Am] 1992;17:55946. [I31 Jackson WT, Viegas SF, Coon T. Anatomical variations in the first extensor compartment of the wrist. J Bone Joint Surg [Am] 1986;68:923-6. [I41 Keon-Cohen B. de Quervain’s disease. J Bone Joint Surg [Br] 1951;33:96-9. [15] Lacey T, Goldstein LA, Tobin CE. Anatomical and clinical study of the variations in the insertions of the abductor pollicis longus tendon, associated with stenosing tendovaginitis. J Bone Joint Surg [Am] 1951;33:347-50. [I61 Lamphier TA, Crooker C, Crooker JL. De Quervain’s disease. Ind Med Surg 1965;34:847-56. [I71 Lane LB, Boretz RS, Stuchin SA. Treatment of de Quervain’s disease: role of conservative management. J Hand Surg [Br] 2001 ;26:25860. [I81 LeHo L. de Quervain’s disease-a clinical and anatomical study. J Bone Joint Surg [Am] 1958;40:1063-70. [19] Leslie BM, Ericson Jr WB, Morehead JR. Incidence of a septum within the first dorsal compartment of the wrist. J Hand Surg [Am] 1990;15:88-91. [20] Lipscomb PR. Stenosing tenosynovitis at the radial styloid process (de Quervain’s disease). Ann Surg 1951;134:1l0-5. [21] Loomis LK. Variations of stenosing tenosynovitis at the radial styloid process. J Bone Joint Surg [Am] 1951;33:340-6.
[22] Louis DS. Incomplete release of the first dorsal compartment-a diagnostic test. J Hand Surg [Am] 1987;12:87-8. [23] Mahakkanukrauh P, Mahakkanukrauh C. Incidence of aseptum in the first dorsal compartment and its effects on therapy of de Quervain’s disease. Clin Anat 2000; 13:195-8. [24] Momose T, Amadio PC, Zhao C, Zobitz ME, Couvreur PJ, An KN. Suture techniques with high breaking strength and low gliding resistance: experiments in the dog flexor digitorum profundus tendon. Acta Orthop Scand 2001;72:635-41. [25] Muckart RD. Stenosing tendovaginitis of abductor pollicis longus and extensor pollicis brevis at the radial styloid (de Quervain’s disease). Clin Orthop 1964;33:201-8. [26] Otto N, Wehbe MA. Steroid injections for tenosynovitis in the hand. Orthop Rev 1986;15:290-3. [27] Rees SG, Davies JR, Tudor D, Flannery CR, Hughes CE, Dent CM, et al.. Immunolocalisation and expression of proteoglycan 4 (cartilage superficial zone proteoglycan) in tendon. Matrix Biol 2002;2 1593-602. [28] Sakai N. Selective corticosteroid injection into the extensor pollicis brevis tenosynovium for de Quervain’s disease. Orthopedics 2002;25:68-70. [29] Schned ES. De Quervain tenosynovitis in pregnant and postpartum women. Obstet Gynecol 1986;68:4114. [30] Stein AH, Ramsey RH, Key JA. Stenosing tendovaginitis at the radial styloid process (De Quervain’s disease). AMA Arch Surg 1951;63:21&28. [31] Uchiyama S, Amadio PC. Ishikawa J, An KN. Boundary lubrication between the tendon and the pulley in the finger. J Bone Joint Surg [Am] 1997;79:213-8. [32] Uchiyama S, Coert JH, Berglund L, Amadio PC. Method for the measurement of friction between tendon and pulley. J Orthop Res 1995;13:83-9. [33] van Oudenaarde E. Structure and function of the abductor pollicis longus muscle. J Anat 1991;174:221-7. [34] Weiss AP, Akelman E, Tabatabai M. Treatment of de Quervain’s disease. J Hand Surg [Am] 199419:595-8. [35] Witt J, Press G, Gelberman RH. Treatment of de Quervain tenosynovitis-a prospective study of the results of injection of steroids and immobilization in a splint. J Bone Joint Surg [Am] 1991;73:219-22. [36] Wood MB, Dobyns JH. Sports-related extraarticular wrist syndromes. Clin Orthop 1986;202:93-102. [37] Yuasa K, Kiyoshige Y. Limited surgical treatment of de Quervain’s disease: decompression of only the extensor pollicis brevis subcompartment. J Hand Surg [Am] 1998;23:840-3. [38] Yuhiko T, Hiroyuki H, Hajime I, Takashi 0. Clinicopathological analysis of de Quervain’s disease. Acta Medica Okayama 1994;48:7-15. [39] Zhao C, Amadio PC, Momose T. Zobitz ME, Couvreur P, An KN. Remodeling of the gliding surface after flexor tendon repair in a canine model in vivo. J Orthop Res 2002;20:85762. [40] Zhao C, Amadio PC, Zobitz ME, An KN. Gliding characteristics of tendon repair in canine flexor digitorum profundus tendons. J Orthop Res 2001;19:580-6.
Journal of Ort hopaed ic Research
Journal of Orthopaedic Research 23 (2005) 243-248
www.elsevier.com/locate/orthres
Gliding resistance of the extensor pollicis brevis tendon and abductor pollicis longus tendon within the first dorsal compartment in fixed wrist positions Keiji Kutsumi, Peter C . Amadio
+,
Chunfeng Zhao, Mark E. Zobitz, Kai-Nan An
Orthopedic Biomechanics Luboratory, Mayo Clinic Rochester, 200 First Street S W. Rochester, MN 55905, USA
Received 16 lanuary 2004; accepted 18 June 2004
Abstract Purpose: While the etiology of de Quervain's disease is unknown, repetitive motion coupled with awkward wrist position and septation within the first dorsal compartment are considered causative factors. We hypothesize that these conditions might produce high gliding resistance, which could then induce micro-damage of the tendons and retinaculum. The purpose of this study was to measure the gliding resistance of the extensor pollicis brevis and abductor pollicis longus tendons within the first dorsal compartment in a human cadaver model. Methods: Fifteen human cadaver wrists, which included eight septation and seven non-septation wrists in the first dorsal compartment, were used. Gliding resistance of the extensor pollicis brevis and abductor pollicis longus tendons was measured in seven wrist positions: 60" extension, 30" extension, O", 30" flexion, 60" flexion in neutral deviation and 30" ulnar deviation, 15" radial deviation in neutral extensiodflexion. Results: The overall gliding resistance was not different between septation and non-septation wrists (0.21 versus 0.19 N for abductor pollicis longus and 0.21 versus 0.15 N for extensor pollicis brevis, respectively), but there was a significant effect on gliding resistance due to wrist position (p < 0.05) in both tendons. Interaction between wrist position and septation status was observed in the extensor pollicis brevis tendon (p < 0.05). With septation, the gliding resistance of the extensor pollicis brevis was significantly higher in 60" wrist flexion (0.51 N) compared to all other wrist positions tested (all less than 0.26 N) (p < 0.05). In the non-septation group, gliding resistance was significantly higher in 60" flexion (0.20 N) and 60" extension (0.22 N) compared to the other five wrist positions (all less than 0.15 N) (p < 0.05). Although no significant difference was observed, the extensor pollicis brevis tendon with septation tended to have higher gliding resistance than that without septation in wrist flexion. In 60" of wrist flexion the abductor pollicis longus tendon had significantly higher gliding resistance (0.33 N) than the other wrist positions (all less than 0.26 N) (p < 0.05). Conclusions: A combination of septation and wrist position significantly affected extensor pollicis brevis tendon gliding resistance in this cadaver model. These factors may contribute to the development of de Quervain's disease. 0 2004 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. Keywords: De Quervain's disease; Gliding resistance; Tendon: Extensor retinaculum; Biomechanics
Introduction
Corresponding author. Tel.: + I 507 538 1717; fax: + I 507 284 5392. E-mail address: [email protected](P.C. Amadio).
Tenosynovitis of the tendons in the first dorsal compartment of the wrist, the extensor pollicis brevis and abductor pollicis longus, was first mentioned in the thirteenth edition of Gray's Anatomy in 1893 as
0736-02666 - see front matter 0 2004 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. doi:10.10I6/J.orthres.2004.06.014
244
K. Kutsuini
el
al. I Journal of Orihopaedic Reseorcli 23 (2005) 243-248
washerwoman’s sprain [21]. In 1895, de Quervain published a report of five cases of chronic tenovaginitis in the first dorsal compartment [7]. After that report, many papers were published on the clinical characteristics, diagnosis and treatment of what quickly became known as de Quervain’s disease. The pathophysiology of de Quervain’s disease is characterized by non-inflammatory fibrosis of the tenosynovium. The etiology is presumed to be related to repetitive motion [9,14,16,18,20,25]. While the effect of wrist position on de Quervain’s disease is generally accepted, the specific position of risk is controversial. Some authors have stated that the repeated use of firm grasp, together with ulnar deviation of the wrist, could predispose to de Quervain’s disease [9,20,36]. Others have reported an association with radial deviation ~~251. Anatomic variations of the first compartment have been discussed frequently in relation to de Quervain’s disease. Variations in the number and insertion of the abductor pollicis longus tendon have been reported [10,15,18], but these have rarely been discussed in relation to the etiology. Septation between the extensor pollicis brevis and abductor pollicis longus tendons in the first dorsal compartment is common. The incidence of the septum in cadaver studies varies from 24% [l8] to 77.5% [23]. Most surgical series, however, report the presence of such a septum in every case and therefore some investigators have suggested that the septum might play an important role in the etiology of the condition [13,14,21,35]. Yet, despite these suggestions, to our knowledge there is no literature that explains how septation might predispose to de Quervain’s disease. We hypothesize that friction between the tendon and the extensor retinaculum in the first dorsal compartment might play a role in the development of de Quervain’s disease, and specifically that wrist positions and/or anatomic variations may result in higher friction levels within this compartment. Higher friction levels may, in turn, predispose the tenosynovium to mechanical injury. The purpose of this study was to investigate the gliding resistance of extensor pollicis brevis and abductor pollicis longus tendons within the first compartment, in a human cadaver model. Since the pathognomonic test for de Quervain’s disease, the Finkelstein test, holds the thumb trapeziometacarpal position fixed, the effect of this joint was not studied.
Methods We used a tendon frictional testing device modified from a previously described and validated gliding resistance testing machine [I ,32,40].The measurement system consists of one mechanical actuator with a linear potentiometer, two tensile load transducers, and a movable mechanical pulley to satisfy the need for three dimensional wrist
Fig. 1. The device for measuring gliding resistance. motion. With data collected from this device, the gliding resistance can be calculated, as the difference in force between the two transducers connected to the proximal and distal tendon ends. respectively. The specimen is mounted on the custom-made device by clamping the proximal end of the radius and ulna. The wrist position is maintained using a custom-built external fixator that has two rods connected by one lockable universal joint and two adjustable connectors. Each connector joins two threaded 3 mm rods that are driven into proximal part of the radius and the second metacarpal bone. The wrist angle can thus be changed to any selected angle (Fig. I). Fifteen fresh-frozen human cadaveric upper extremities, amputated 15 cm proximal to the wrist joint, were used for this study. The cadaver donors ranged in age from 56 to 97 (mean 76.4 years). Eight specimens with septation between the extensor pollicis brevis and abductor pollicis longus tendons and seven specimens without septation were used (presence of septation determined after testing was completed). Specimens were stored in a freezer at -20 “C, and thawed at room temperature immediately prior to testing. All specimens were X-rayed to exclude gross pathological evidence of injuries or major degenerative changes around the wrist. A longitudinal incision of the skin was made to expose the overall length of the extensor pollicis brevis and the abductor pollicis longus. Fascia over the extensor pollicis brevis and abductor pollicis longus muscles was removed. The extensor retinaculum was carefully preserved. The extensor pollicis brevis and abductor pollicis longus muscles were carefully separated. Excursion of the extensor pollicis brevis and abductor pollicis longus tendons was recorded in each fixed wrist position prior to the gliding resistance measurements. At first, the extensor pollicis brevis muscle belly was detached from its origin and connected to a 4.9 N weight through a Dacron cord to apply tension proximally on the muscle [32]. A mark was made on the tendon surface by suturing 6-0 monofilament nylon proximal enough to the retinaculum edge so as not to interfere with gliding. The distance between the mark and proximal retinaculum edge was measured in thumb interphalangeal, metacarpophalangeal and trapeziometacarpal joint extension, pulled by the weight attached to the muscle. Then, the distance was recorded again after passive thumb flexion to full interphalangeal, metacarpophalangeal and trapeziometacarpal joint flexion in each fixed wrist position. The range of excursion in each wrist position was identified using this method. The abductor pollicis longus tendon excursion was measured using the same method. Following excursion measurement, the extensor pollicis brevis tendon was cut at the level of the metacarpophalangeal joint. The distal tendon end and proximal muscle were connected to the distal ( F I ) and proximal (n) load transducers, respectively. The distal transducer (F1)was connected to a 4.9 N weight through a mechanical pulley and the position of the pulley was changed according to the wrist angle to represent the physiological line of action for the tendon. The 4.9 N weight was used to get a reliable excursion. as this weight has been shown to be the minimum load sufficient to remove slack from the muscle-tendon unit [I 21. The proximal transducer (n) was connected to the mechanical actuator that was placed to simulate the physiological muscle angle (Fig. I).
245
K. Kutsumi et al. I Journul q/ Orthopedic Reseurch 23 (2005) 243-248
The tendon was pulled proximally by the actuator against the gliding resistance and the 4.9 N weight at a rate of 2.0 mm/s. This movement of the tendon toward the actuator was regarded as extension. Then, the actuator was reversed so that the weight pulled the tendon distally. This movement of the tendon toward the niechanical pulley was considered as flexion. F1, F2 and the corresponding excursion were recorded at a sampling rate of 10 Hz. After measuring the gliding resistance of the extensor pollicis brevis tendon, the superficial division of the abductor pollicis longus tendon 1331 was detached from the first metacarpal bone with a small bony fragment. Osteotomy of the bony fragment prevented damage and deformity to the abductor pollicis longus tendon and provided a structure for attachment to the load transducer. The deep division of the abductor pollicis longus tendon was cut at the insertion and connected to the superficial division using side-by-side suture. The bony fragment at the abductor pollicis longus insertion was Proximally, the abducconnected to the distal load transducer (A). tor pollicis longus muscle was connected to the proximal load transducer (F2). Then the gliding resistance of the abductor pollicis longus tendon was measured in the same manner. The trapeziometacarpal joint was fixed by a 1.5 nun Kirschner wire in a position where the posture of the joint would mimic activities such as tip pinch or grasp. Seven different wrist positions were tested 60" extension, 30" extension, O", 30" flexion, 60" flexion in neutral deviation and 30" ulnar deviation, 15O radial deviation in neutral extensionlflexion. The sequence of the wrist positrons was randomized. The angle made between the third metacarpal bone and long axis of radius on a lateral view was measured as the wrist angle using a goniometer. The specimen was kept moist throughout the testing procedure with saline. After completing all measurements, the extensor retinaculum was opened to determine the variations, if any, in the number of the tendons and the presence of septation in the first compartment. Data was collected as the tendon moved thorough its normal range of excursion. The tendons were in motion at the points determined to be the endpoints of excursion to eliminate any static friction effects. In a given direction of motion, the gliding resistance is the difference between the proximal and distal force sensors. The average gliding resistance of the results from extension to flexion of whole excursion was used in this study ((F2exlmsion - FI + should be the same as (FI flexion - F2aexion))/2.Because Fl F, flexion during the testing (4.9 N), the composite gliding resistance may be calculated as: Gliding Resistance = (F2rxlension - F!fiexion)/2. Bulk and tethering effects were eliminated with this calculation because direction of these effects are opposite between flexion and extension. This method was chosen in order to focus on surface friction. Two extension/flexion trials were performed for each condition. Only the data from the second trial was used for analysis to eliminate any positioning effects that may have been present during specimen set-up.
Results We tested eight septation and seven non-septation specimens. All of the septations were partial (3-1 5 mm), and located at the distal part of the first compartment. The number of tendon slips of the abductor pollicis longus were: two tendon slips ( n = 8 specimens), three tendon slips (n = 2 specimens), four tendon slips ( n = 4 specimens), five tendon slips (n = 1 specimen). The excursion from full extension to full flexion of the thumb interphalangeal, metacarpophalangeal, and trapeziometacarpal joint in each fixed wrist position is shown in Fig. 2. The average excursions of the extensor pollicis brevis and abductor pollicis longus tendons were 15.0 and 4.6 mm, respectively. From the main effects of the analysis of variance model there was no significant difference in extensor pollicis brevis gliding resistance between septation and nonseptation (p = 0.18) wrists, although there was a significant difference due to wrist position 0, < 0.05) and a significant interaction between wrist position and septation status (p < 0.05). Therefore, separate one-factor repeated measures analysis of variance models for the specimens with septation and without septation were run. With septation, the gliding resistance of the extensor pollicis brevis was significantly higher in 60" wrist flexion (0.51 N) compared to all other wrist positions tested (all less than 0.26 N) (p < 0.05) (Fig. 3). In the non-septation group, gliding resistance of the extensor pollicis brevis tendon was significantly higher in 60" flexi on (0.20 N) and 60" extension (0.22 N) compared to the other five wrist positions (all less than 0.15 N) (p < 0.05). Although not statistically significant, there was a diverging trend between septatiodnon-septation at 30" and 60" flexion, with septation having a higher gliding
Statistical methods The sample size requirements were determined by a power calculation using previous studies of tendon gliding resistance. A sample of 14 specimens (7 with septation and 7 without septation) will provide 80% power to detect a difference in gliding resistance between any two of the seven wrist positions equal to 0.12 N, and 80% power to detect a difference in mean gliding resistance between specimens with septation and those without septation equal to 0.24 N. Separate analyses were conducted for the extensor pollicis brevis tendon and the abductor pollicis longus tendon. For each tendon, two-factor analysis of variance with repeated measures on cine factor (wrist position) was used to compare the gliding resistance among the seven wrist positions and between specimens with sephtion and without septation. When significant interactions were identified, separate one-Factor repeated measures analysis of variance models were run for the specimens with septation and without septation. When wrist angle was found to be significant, further analysis was performed using the Ryan-EinotCrabriel-Welsch multiple comparisons procedure. All statistical tests were two-sided and p-values less than 0.05 were considered significant.
cxt.6W cxt.30'
neut.
fkx.30' 0 e x . W
u.dev.30'
&.
r.h.15'
Fig. 2. Mean values ? SD of the excursion of the extensor pollicis brevis and abductor pollicis longus tendons from full extension to full flexion of the thumb interphalangeal, metacarpophalangeal, and trapeziometacarpal joint in each wrist position. ext.: extension, flex.: flexion, neut.: neutral, u. dev.: ulnar deviation, r. dev.: radial deviation.
K. Kutsumi et al. I Journal of' Orthopaedic Research 23 (2005) 243-248
246
Discussion
eXt.60.
eXt.t.30.
ncllt.
flex.30. flCx.60.
U.deT.W
MU. r.dn. 15.
Fig. 3. Mean values k SD of the gliding resistance of the extensor pollicis brevis tendon. ext.: extension, flex.: flexion, neut.: neutral, u. dev.: ulnar deviation, r. dev.: radial deviation, A had significant difference with .. A had significant difference with 0.
resistance. At the other wrist positions, however, the gliding resistances were very similar. There was no significant difference in abductor pollicis longus tendon gliding resistance comparing the septation (0.21 N) and non-septation groups (0.19 N) (p = 0.67). Wrist position, however, had a significant effect (p < 0.05) on abductor pollicis longus tendon gliding resistance (Fig. 4). In 60" of wrist flexion the abductor pollicis longus tendon had significantly higher gliding resistance (0.33 N) than the other wrist positions (all less than 0.26 N) (p < 0.05). Additionally, in 60" wrist extension and 30" wrist flexion the abductor pollicis longus gliding resistance was significantly greater than neutral, 30" extension, 30" ulnar deviation, or 15" radial deviation. There was no significant interaction between septation status and wrist position in gliding resistance of the abductor pollicis longus tendon (p = 0.44).
O.7
0
r
ext.60. e x t W
ncllt.
flex.30. I k x . W
u.dev.30"
meut. r.dev. 15'
Fig. 4. Mean values 2 SD of the gliding resistance of the abductor pollicis longus tendon. ext.: extension, flex.: flexion, neut.: neutral, u. dev.: ulnar deviation, r. dev.: radial deviation, A had significant difference with 0 and 0. 0 had significant difference with 0.
In the first dorsal compartment of the wrist extensor retinaculum, the extensor pollicis brevis and abductor pollicis longus tendons pass under a fibrous sheath, which is fixed on either side of the tendons to the dorsal radius, thus forming a fibro-osseous tunnel. As the tendons are subjected to varying degrees of angulation with wrist motion, these tunnels function as a pulley, constraining the lateral motion of the tendons and preventing the tendons from bowstringing. But this function also acts to produce friction between the tendons and the retinaculum. In the first compartment, the extensor pollicis brevis tendon is located dorsally and the abductor pollicis longus volarly. In the presence of the septation, the extensor pollicis brevis tendon occupies a separate subcompartment. Leslie et al. [19] described that the septum began distally in the region of the radial styloid and extended proximally for a distance of 0.5-2 cm. The reported incidence of a septum in the first compartment in cadaver series varies from 24% [I81 to 77.5% [23]. In the extensor pollicis brevis tendon without septation, the volar surface contacts the abductor pollicis longus tendon. This interaction may have a different frictional coefficient than that with the surface of the septum. Moreover, as both tendons move synergistically in the physiological condition, relative displacement of the volar surface of the extensor pollicis brevis tendon without septation should be smaller than with septation. This may result in a less abrasive effect on the tendon without septation than with septation. We do not think that the increased friction would be caused by a narrowed canal, because the gliding resistance between septation and non-septation specimens was very similar at the other wrist positions. However, if swelling or edema occurred, the narrowed canal might further increase gliding resistance to a considerable extent. Variations in the anatomy of the first extensor compartment have been reported as being associated with the development of de Quervain's disease [9-1 lJ3, 21,35,37]. Our results seem to support these observations. Some extensor pollicis brevis tendons with septation had quite high gliding resistance in the wrist flexion position. This is the reason why the standard deviation of the extensor pollicis brevis tendon septation group was relatively high. While the values for gliding resistance in the neutral position, roughly 0.2 N, are similar to those seen in other normal intrasynovial tendons [24,39], the values for gliding resistance in the most adverse wrist positions approach those seen in healing, lacerated tendons [39]. We believe it is reasonable to suspect that tendons with such high gliding resistance, such as the extensor pollicis brevis tendon with septation, might be vulnerable to develop de Quervain's disease.
K. Kutsumi et ul. I Journal
of
Orthopuedic Researcli 23 (200.5) 243-248
Lubrication of the tendon is a critical factor in its gliding resistance. Higher friction, associated with movements in certain wrist positions, such as pinching in wrist flexion, especially in the presence of septation, may induce surface damage [40], which in turn may result in inflammation. Inflammatory enzymes may then destroy lubricating glycoproteins such as hyaluronan [31] and lubricin [27] on the surface of the tendon and extensor retinaculum, resulting in a vicious cycle of increasing friction, damage, and loss of lubricating glycoproteins. The histopathology of de Quervain’s disease has been variably described as showing inflammation [5,6,18,25], degenerative change [3], or some combination of the two [ 16,20,25,29,38], consistent with this hypothesis. The average excursion of the extensor pollicis brevis tendon in this study, which represents the excursion for thumb extension-flexion movement without wrist motion, was about three times greater than that of the abductor pollicis longus tendon. This observation, coupled with related reports by others, supports the hypothesis that the extensor pollicis brevis may be the tendon most affected by de Quervain’s disease. It is known that failure to open the septation can explain poor results of surgical treatment for de Quervain’s disease [22]. Selective steroid injection into the extensor pollicis brevis tenosynovium is reported as being effective therapy [28], and Yuasa and Kiyoshige [37] have reported good results by decompression of only the extensor pollicis brevis subcompartment. These studies suggest that the extensor pollicis brevis tendon may be more vulnerable to the disease than the abductor pollicis longus tendon. This is consistent with our observation, because the tendon with the longer excursion might easily be the one more susceptible to injury related to repetitive motion. One of the limitations of this study was that we used saline solution instead of synovial fluid for lubrication. This may affect the absolute gliding resistance of the tendon, although we believe that the effect is not so strong that it would change our results [31]. A second limitation is that the data was collected from one trial for each specimen (the second trial) at each wrist position. In preliminary testing we tested 3 times per position. The results were very stable so we tested each condition twice and analyzed the second run, considering the first trial to be a conditioning run. Other studies on gliding resistance after flexor tendon repair have also found that after the first cycle of motion the results are consistent over then next few trials. To minimize the risk of abrasion with repeated testing, we only made two trials at each position, we randomized the testing sequence among the specimens, and the tendon was kept moist between trials by spraying with saline. A third limitation was that we did not consider the effect of gender. De
241
Quervain’s disease is more common in women than in men, with reported ratios of women to men ranging from 4:l to 7:l [8,17,18,25,26,34,35]. Some investigators suggest that the wrists normally angulate further in women than in men and this is the reason why women are predisposed to the disease [4,20]. We did not test this hypothesis, as we only measured the gliding resistance at given wrist angles, not at the maximum angle. Even if there is a gender difference, however, we believe that the interaction of septation and wrist position that we observed is likely similar in both genders. A fourth limitation was that we did not analyze the number of abductor pollicis longus tendons and the size of the subcompartment. In our data, these factors did not seem to affect the friction of the tendons, although we did not have enough data for statistical analyses. The final limitation is that we only looked at the effect of wrist position and septation, and did not study the impact of forearm or thumb position. However, in a previous study performed in our laboratory (unpublished), there was little change in EPB or APL muscle-tendon excursion with forearm position. In the Finkelstein test, universally used to provoke the symptoms of de Quervain’s disease, the thumb is held still, while the wrist is moved. Based on these observations, we believe that the effect of forearm or thumb position on first compartment friction is likely to be small. However, we do plan to study this in the future. A strength of this study is that we were able to directly measure the gliding resistance in the first dorsal compartment of the wrist. To our knowledge, this is the first report to investigate the gliding resistance of the extensor tendons. Some authors suggest that the anatomical characteristic of the first compartment, in which the tendons form an acute angle, induces high friction and mechanical irritation [5,30]. Our validated gliding resistance testing machine also allowed us to measure the gliding resistance of the extensor tendons in various wrist positions, by adjusting the position of the mechanical pulley. In conclusion, we have directly measured the gliding resistance of the tendons of the human first dorsal wrist compartment in a cadaver model. We have shown that septation within the compartment and wrist position combine to affect friction on the extensor pollicis brevis. These observations support the hypothesis that a friction-induced tenosynovitis of the extensor pollicis brevis may be the primary source of pathology in de Quervain’s disease. Acknowledgement
This study was funded by grants from the NIH/ NIAMS (AR44391) and Mayo Clinic Rochester.
248
K. Kutsumi et al. I Journal of Orthopaedic Research 23 (2005) 243-248
References [l] An KN, Berglund L, Uchiyama S, Coert JH. Measurement of friction between pulley and flexor tendon. Biomed Sci Instrum 1993;29:1-7. [2] Bums BH, Ellis VH. Stenosing tendovaginitis at the radial styloid process. Lancet 1936;1:717. [3] Clarke MT, Lyall HA, Grant JW, Matthewson MH. The histopathology of de Quervain’s disease. J Hand Surg [Br] 1998;23:7324. [4] Conklin JE, White WL. Stenosing tenosynovitis-and its possible relation to the carpal tunnel syndrome. Surg Clin North Am 196040:53140. [5] Cotton FJ, Morrison GM. Bradford CH. De Quervain’s diseaseradial styloid tendovaginitis. N Engl J Med 1938;219:120-3. [6] de Quervain F. Uber das Wesen und die Behandlung der stenosierenden Tendovaginitis am Processus styloideus radii. Miinchener Med Wochenschr 1912;59:5-6. [7l de Quervain F. Uber eine Form von chronischer Tendovaginitis. Corresp Blatt Schweizer Arzte 1895;25:389-94. [8] Faithfull DK, Lamb DW. De Quervain’s disease-a clinical review. Hand 1971;3:23-30. [9] Finkelstein H. Stenosing tenovaginitis at the radial styloid process. J Bone Joint Surg [Am] 1930;12:50940. [lo] Giles KW. Anatomical variations affecting the surgery of de Quervain’s disease. J Bone Joint Surg [Br] 1960;42:352-5. [Ill Gonzalez MH, Sohlberg R, Brown A, Weinzweig N. The first dorsal extensor compartment: an anatomic study. J Hand Surg [Am] 1995;20:657-9. [I21 Horii E, Lin GT, Cooney WP. Linscheid RL, An KN. Comparative flexor tendon excursion after passive mobilization: an in vitro study. J Hand Surg [Am] 1992;17:55946. [I31 Jackson WT, Viegas SF, Coon T. Anatomical variations in the first extensor compartment of the wrist. J Bone Joint Surg [Am] 1986;68:923-6. [I41 Keon-Cohen B. de Quervain’s disease. J Bone Joint Surg [Br] 1951;33:96-9. [15] Lacey T, Goldstein LA, Tobin CE. Anatomical and clinical study of the variations in the insertions of the abductor pollicis longus tendon, associated with stenosing tendovaginitis. J Bone Joint Surg [Am] 1951;33:347-50. [I61 Lamphier TA, Crooker C, Crooker JL. De Quervain’s disease. Ind Med Surg 1965;34:847-56. [I71 Lane LB, Boretz RS, Stuchin SA. Treatment of de Quervain’s disease: role of conservative management. J Hand Surg [Br] 2001 ;26:25860. [I81 LeHo L. de Quervain’s disease-a clinical and anatomical study. J Bone Joint Surg [Am] 1958;40:1063-70. [19] Leslie BM, Ericson Jr WB, Morehead JR. Incidence of a septum within the first dorsal compartment of the wrist. J Hand Surg [Am] 1990;15:88-91. [20] Lipscomb PR. Stenosing tenosynovitis at the radial styloid process (de Quervain’s disease). Ann Surg 1951;134:1l0-5. [21] Loomis LK. Variations of stenosing tenosynovitis at the radial styloid process. J Bone Joint Surg [Am] 1951;33:340-6.
[22] Louis DS. Incomplete release of the first dorsal compartment-a diagnostic test. J Hand Surg [Am] 1987;12:87-8. [23] Mahakkanukrauh P, Mahakkanukrauh C. Incidence of aseptum in the first dorsal compartment and its effects on therapy of de Quervain’s disease. Clin Anat 2000; 13:195-8. [24] Momose T, Amadio PC, Zhao C, Zobitz ME, Couvreur PJ, An KN. Suture techniques with high breaking strength and low gliding resistance: experiments in the dog flexor digitorum profundus tendon. Acta Orthop Scand 2001;72:635-41. [25] Muckart RD. Stenosing tendovaginitis of abductor pollicis longus and extensor pollicis brevis at the radial styloid (de Quervain’s disease). Clin Orthop 1964;33:201-8. [26] Otto N, Wehbe MA. Steroid injections for tenosynovitis in the hand. Orthop Rev 1986;15:290-3. [27] Rees SG, Davies JR, Tudor D, Flannery CR, Hughes CE, Dent CM, et al.. Immunolocalisation and expression of proteoglycan 4 (cartilage superficial zone proteoglycan) in tendon. Matrix Biol 2002;2 1593-602. [28] Sakai N. Selective corticosteroid injection into the extensor pollicis brevis tenosynovium for de Quervain’s disease. Orthopedics 2002;25:68-70. [29] Schned ES. De Quervain tenosynovitis in pregnant and postpartum women. Obstet Gynecol 1986;68:4114. [30] Stein AH, Ramsey RH, Key JA. Stenosing tendovaginitis at the radial styloid process (De Quervain’s disease). AMA Arch Surg 1951;63:21&28. [31] Uchiyama S, Amadio PC. Ishikawa J, An KN. Boundary lubrication between the tendon and the pulley in the finger. J Bone Joint Surg [Am] 1997;79:213-8. [32] Uchiyama S, Coert JH, Berglund L, Amadio PC. Method for the measurement of friction between tendon and pulley. J Orthop Res 1995;13:83-9. [33] van Oudenaarde E. Structure and function of the abductor pollicis longus muscle. J Anat 1991;174:221-7. [34] Weiss AP, Akelman E, Tabatabai M. Treatment of de Quervain’s disease. J Hand Surg [Am] 199419:595-8. [35] Witt J, Press G, Gelberman RH. Treatment of de Quervain tenosynovitis-a prospective study of the results of injection of steroids and immobilization in a splint. J Bone Joint Surg [Am] 1991;73:219-22. [36] Wood MB, Dobyns JH. Sports-related extraarticular wrist syndromes. Clin Orthop 1986;202:93-102. [37] Yuasa K, Kiyoshige Y. Limited surgical treatment of de Quervain’s disease: decompression of only the extensor pollicis brevis subcompartment. J Hand Surg [Am] 1998;23:840-3. [38] Yuhiko T, Hiroyuki H, Hajime I, Takashi 0. Clinicopathological analysis of de Quervain’s disease. Acta Medica Okayama 1994;48:7-15. [39] Zhao C, Amadio PC, Momose T. Zobitz ME, Couvreur P, An KN. Remodeling of the gliding surface after flexor tendon repair in a canine model in vivo. J Orthop Res 2002;20:85762. [40] Zhao C, Amadio PC, Zobitz ME, An KN. Gliding characteristics of tendon repair in canine flexor digitorum profundus tendons. J Orthop Res 2001;19:580-6.