Christian L. Jantea (1), KaiNan An (2), Ronald L. Linscheid (2), P. Cooney III (2)
Priv.Doz. Dr. C. L. Jantea
Department of Orthopedic Surgery
University of Duesseldorf
Instability of the scaphoid is a common factor in instabilities
of the wrist. Most scientific interest has focused on the scapholunate
ligament with regard to the instability pattern of the scaphoid.
There are few clinical reports where a complete or partial disruption
of the scaphotrapezialtrapezoidal (STT) ligamentous complex
has occured. This may be the case in a subluxation"'5 or
a dislocation of the trapezium. The consequence of this condition
generally is the development of osteoarthritis of the STT joint
which may require operative treatment.
In an experimental descriptive paper on the STT ligaments, a gap between the trapezium and the distal pole of the scaphoid was observed in ulnar deviation when the STT ligaments were sectioned.' However a more precise biomechanical and kinematic. analysis of the STT ligament complex is desirable in order to understand the function of the STT ligaments.
The purpose of this study is to assess the role of the STT ligament complex in stabilization of the distal scaphoid. The central questions are:  do the STT ligaments contribute to the stability of the scaphoid during flexionextension motion (FEM) and radialulnar deviation (RUD) of the wrist and how is the motion pattem of the scaphoid altered when the STT ligaments are transsected ?  how does loading of the hand influence the kinematic behavior of the scaphoid with intact or transsected STT ligaments ?
A kinematic analysis was done in 10 fresh human cadaver wrists.
The premortem histories were studied to exclude diseases deletorius
to ligamentous integrity. Plain Xray and CT scans (General Electrics
model 9800) were obtained to exclude other bony and soft tissue
pathology. One specimen with a scapholunate dissociation had to
be excluded leaving 9 wrists of the 6 donors, 3 women and 3 men,
while 3 (2 females, 1 male) contributed both wrists. The age ranged
from 29 to 62 years with a mean of 42 years. These extensive preexperimental
studies were done to exclude carpal instabilities as the experiments
were done without opening the carpal joints.
The skin and subcutaneous fat was removed, fiberglass rods were inserted in the scaphoid, trapezium, 1st and 3rd metacarpals so as not to interfere with the tendon excursion during the kinematic experiments.
The radius was cemented in a custom device and rigidly fixed together with the source, which generated, the magnetic field to the same platform (Figure 3). The orientation of the sources coordinate system corresponds to the axis defined for the wrist motion (Figure 1).
Before implantation of the fiberglass rods a drill hole with a diameter of 2.2 mm was made in each bone without opening any of the carpal joints. Fiberglass rods do not interfere with the magnetic field as this would be the case with metal implants. Superglue (cyanoacrylate ester) was used to fix the rods in the bones to prevent axial rotation. Direct contact between the superglue and soft tissue was avoided in order not to alter the tissue properties. Small plastic platforms were fixed to the rods and the sensors of the Isotrack system were fixed to the latter by plastic screws (Figure 3).
The rod was inserted from the dorsum into the scaphoid between the tendons of the extensor pollicis longus and extensor carpi radialis muscles. The insertion area corresponds to the dorsal rim of the bone to which the dorsal transverse carpal ligament (ligament radiocarpeum arcuatum dorsum) running from the triquetrum over the midcarpal joint to the scaphoid is attached. This dorsal rim also is the insertion area of the radioscaphoid joint capsule. The dorsal STT joint capsule and ligaments insert on the distal part of this dorsal rim.
A second rod was inserted in the trapezium from a palmar direction through the origin of the tendon of the abductor pollicis longus muscle. This pin was drilled through both the trapezium and trapezoid in order to avoid motion between these bones. The trapezialtrapezoidal (TT) joint motion can be clinically neglected as there is only about 3 degrees of relative motion between these bones. A third rod was inserted into the first metacarpal adjacent to the carpometacarpal (CMC) joint of the thumb. The hand was moistened through the two hours average duration of the experiments with vaporized physiologic saline solution. The sensors in the first metacarpal and the trapezium were used to calculate the relative motion between them. This has to be zero by definition in order not to interfere with recording the scaphoid motion at the STT joint level. Kinematic data acquisition was obtained from the sensors while the wrist was moved by another rod inserted in the 3rd metacarpal. The latter glided in a special device along a frame which was aligned with the flexionextension motion (FEM) and radialuluar deviation (RUD) plane of the wrist (Figure 3). The kinematic data of the scaphoid in different experimental conditions is presented.
Magnetic field technology provides a unique method to register the data required for analysis of the position and motion of a coordinate system attached to a moveable sensor with a fixed spherical coordinate system. The relative position and motion of the sensor to a source has to be described in terms of translation and rotation of the coordinate system located in the sensor. These variables (translation and rotation matrix) are necessary to describe the 6 degrees of freedom (DOF). These 6 variables were registered during the guided motion of the wrist in the different experimental conditions. The analytical method for the description of 6 DOF spatial motion used in this study is the screw displacement axis (SDA).'2529 The relative displacement of a moving segment from one position to another can be described in terms of a rotation around and a translation along a unique axis called the SDA which is fixed in the segment (Figure 2). The advantage of using a screw axis is that the orientation of the SDA remains invariant, regardless of the reference coordinate axes used. The SDA is a true vector quantity; its magnitude can be decomposed along any coordinate axes used for analysis. So the description of the motion pattern of the scaphoid around the 3 defined X, Y and Zaxes of the cartesian coordinate system is possible (Figure 1).
Magnetic field technology allows a continuous data acquisition from the sensors used through the whole range of motion. It has proved its accuracy (0.1 degree at 60 measurements per second) in other experimental kinesiologic studies. 2 The main advantage of this technology is that the performed motion does not require interruption for data acquisition. In this study a frequency of 15 Hz for data acquisition was chosen during which FEM and RUD of the wrist were performed over 10 seconds interval for one motion cycle.
The following experiments were performed in order to assess the role of the STT ligaments in scaphoid motion.
Experiment "NLL": normal/intact STT ligaments, hand low loaded. Normal muscle tension was simulated by loading all tendons of the hand in a low loading condition ("LL") with intact STT ligaments ("NLL"). Force distribution to each tendon was 100g, resulting in a total of 1.4 kg for all tendons of the hand (EPL 100g, EPB 100g, APB 100g, FPL 100g, EXT DIG COM 200g, FLEX DIG SUP 100g, FLEX DIG PROF 100g, ECRB 100g, ECRL 100g, ECU 200g, FCR 100g, FCU 100g). The kinematic data obtained from this experiment is the reference for the normal scaphoid motion, which may simulate the physiologic situation with intact STT ligaments.
Experiment "NHL": normal/intact STT ligaments, hand high loaded. This experiment was to determinate how a high loading condition changes the normal motion pattern of the scaphoid with the STT ligaments intact. A total weight of 23.5 kg was distributed to the tendons of the hand (EPL 1.0 kg, EPB 0.5 kg, APB 0.5 kg, FPL 1.5 kg, EXT DIG COM 4.0 kg, FLEX DIG SUP 2.0 kg, FLEX DIG PROF 2.0 kg, ECRB 0.75 kg, ECRL 0.75 kg, ECU 3.5 kg, FCR 3.5 kg, FCU 3.5 kg).
Experiment "2LL": STT ligaments dissected, hand low loaded. In this experiment an isolated STT dissociation was simulated by circumferentially cutting of the capsule and ligaments of the STT joint. The specimen was left attached to the fixation device. A needle was passed through the STT joint from dorsal to palmar in order to identify the STT ligaments accurately. The floor of the flexor carpi radialis (FCR) tendon sheath covers the palmar aspect of the STT ligaments and it is not possible to separate the two structures from each another. The flexor retinaculum covering the FCR sheath was left intact while sectioning the underlying STT ligaments. The tensile load on the tendons of the hand crossing the wrist was the same as in the experiment "NLL", so a direct comparison between both experiments is possible.
Experiment "2HL": STT ligaments dissected, hand high loaded. In this experiment the load was increased to 23.5 kg while the STT ligaments were sectioned to simulate an instability pattern of the scaphoid under a high axial load of the hand.
Analysis of the results. Differences in kinematics between the four experimental conditions ("NLL", "NHL", "2LL", and "2HL") were registered and statistical analysis of variance (ANOVA) using commercially available software was performed.
The 3 motioncomponents of the scaphoid are illustrated as rotation around the 3 axes of the cartesian coordinate system (Figure 1). The curves for all specimens were interpolated to the same increment (of 1 degree). For FEM of the wrist an interval from 20 degrees of extension to +40 degrees of flexion was chosen. For RUD the interval ranged from 20 degrees for ulnar deviation to +15 degrees for radial deviation. The abscisse of each graph represents the motion of the wrist while the ordinate represents the scaphoid motion in respect to each axis of the cartesian coordinate system. Figure 4 represents the average motion of the scaphoid for all specimens when FEM of the wrist was performed:  when the wrist is moved from extension to flexion the main motion component of the scaphoid is extensionflexion; in 20 degrees of wrist extension the scaphoid is extended in 22 degrees; at the neutral position of the wrist (point 0/0) the scaphoid is in a slightly extended position of 6 degrees; with the wrist flexed in 40 degrees the scaphoid flexes to 28 degrees (SD= 4.2 degrees);  the second motion component of the scaphoid during the FEM of the wrist is pronation and supination (rotation around the Xaxis); the position of the scaphoid is 6 degrees of supination with the wrist extended, 2.3 degrees of supination in the neutral position of the wrist and 2 degrees of pronation, at 40 degrees of wrist flexion (SD= 2.2 degrees);  rotation around~the Zaxis which describes the RUD of the scaphoid is less during FEM of the wrist. In extension the scaphoid is in 2 degrees of ulnar deviation and rotates to the neutral position of O degree of rotation (SD= 2.3 degrees) during flexion of the wrist.
Figure 5 describes the motion pattem of the scaphoid around the 3 defined axes (Figure 1) during RUD of the wrist:  with the wrist in ulnar deviation of 20 degrees, the scaphoid is extended to 14 degrees; when the wrist is radially deviated to 15 degrees, the scaphoid is flexed to nearly 7 degrees (SD= 3.8 degrees);  the scaphoid rotates around the Xaxis from 8 degrees of supination to 6 degrees of pronation (SD= 2.1 degrees) during RUD of the wrist;  the smallest amount of rotation of the scaphoid occurs around the Zaxis during RUD of the wrist; during an ulnar to radial deviation of the wrist, the scaphoid is ulnar deviated to 1.2 degrees and rotates to a radially deviated position of 0.8 degree.
The effects of sectioning of the STT ligaments and loading are presented separately for all 3 motion components of the scaphoid (Figure 6, Figure 7, Figure 8). The FEM motion component of the scaphoid around the Yaxis when FEM of the wrist is performed shows no statistically significant difference (p= 0.098) between experiments. All curves show similar slope and configuration; however, loading sligthly decreases the FEM component of the scaphoid (Figure 6). Figure 7 illustrates the RUD scaphoid rotation around the Zaxis during FEM of the wrist. Sectioning of the STT ligaments (experiment "2LL") decreased that range of motion. In a high loading condition (experiment "2HL") the motion pattern is similar to the low loading condition (experiment "2LL"), but displaced upwards, showing that the scaphoid remains ulnardeviated during the FEM motion of the wrist. In summary, loading and sectioning of the STT ligaments significantly decrease the range for RUD of the scaphoid during FEM of the wrist. Figure 8 illustrates the changes in the pronation supination motion component of the scaphoid during FEM of the wrist. In the normal situation (experiment "NLL") and after sectioning of the STT ligaments in a low loading condition (experiment "2LL") the curves have the same configuration and slope. However in the latter the amount of rotation of the scaphoid is considerably reduced. When the wrist is loaded (experiment "NHL" and "2HL") there is basically no pronation and supination of the scaphoid: both curves parallel the abscisses. In summary, when a qualitative comparison of the 3 motion components for the scaphoid is done during FEM of the wrist (Figure 6, Figure 7, Figure 8). the pronation supination motion component of the scaphoid is influenced most byloading the wrist or sectioning the STT ligaments.
Figure 5 illustrates the 3 motion components of the scaphoid during RUD of the wrist with the STT ligaments intact and the muscles of the hand loaded with a total weight of 1.4 kg (experiment "NLL"). The main motion component of the scaphoid is the FEM, followed by the pronation and supination during RUD of the wrist. This motion pattern is also observed in the other experimental conditions (figure 9, figure 10, figure 11). Figure 9 demonstrates that sectioning the ligaments (experiment "2LL") has the most dramatic effect on the FEM component of scaphoid motion, as the motion arc is reduced to 70 % (11.5 degrees) compared to the normal (experiment "NLL" > 100%, or 16 degrees). Loading reduces the FEM component of the scaphoid regardless if the STT ligaments are intact or cut, though this not statistically significant. Figure 10 illustrates scaphoid rotation around the Zaxis which describes the RUD of the scaphoid during the RUD of the wrist. With the STT ligaments intact (experiment "NLL") the scaphoid motion occurs close to the neutral position, as this curve crosses the abscissa. When the STT ligaments are sectioned (experiment "2LL") or the wrist is loaded (experiment "NHL" or "2HL") the scaphoid is deviated ulnarly, illustrated by the parallel shift of the curves to a positive value on the ordinate. Using the ANOVA method for a statistical comparison, there is a significant difference (p < 0.001) between the kinematic behavior of the scaphoid RUD motion component, during the RUD of the wrist, with a high loading (experiment "NHL", "2HL") compared to the experimental setup with a low load (experiment "NHL", "2HL"). However, no statistic significant difference could be found between the motion pattems of the scaphoid in the low load conditions with intact or sectioned STT ligaments (experiment "NLL" compared to experiment "2LL") . Figure 11 shows the rotation of the scaphoid around the Xaxis which represents the pronation and supination motion component of the scaphoid during RUD of the wrist. With intact STT ligaments (experiment "NLL") the range of rotation is 14 degrees, which is considerably reduced (p < 0.001) after sectioning the STT ligaments in experiment "2LL". This difference is significant only for the motion arc from ulnar deviation to the neutral position of the wrist. From neutral position to maximal radial deviation, the scaphoid pronates 6 to 7 respectively degrees for the experimental condition "NLL" and "2LL". For this second part of the motion no statistical difference was found. Another interesting pattem is shown in figure 11: when the wrist is loaded pronation and supination of the scaphoid is reduced to 50% in the experiments "NHL" and "2HL", as compared to experiment "NLL" and "2LL". This difference is also statistically significant (p < 0.001). There is no statistical significant difference concerning the motion pattern of the scaphoid in the high load conditions whether the STT ligaments are intact or not (experiment "NHL", "2HL").
In summary the pronation and supination motion component of the scaphoid is the most affected motion component when sectioning the STT ligaments or loading is performed during the RUD motion of the wrist.
The role of the scaphoid in understanding carpal instabilities has been challenging. Interest has focused on the scapholunate ligament as being responsible for the rotatory instability of the scaphoid (RlS).'7'' However a lesion of the STT ligaments may occur in traumatic conditions which are clinically seen as a luxation of the trapezium or of the scaphoid itself. Concurrent attenuations of the ligaments at either end of the scaphoid may also occur. The aim of this study was to assess the importance of the STT ligaments for the kinematics of the scaphoid during standardized wrist motion. To our knowledge there are no previous studies of the kinematic behavior of the scaphoid with special regard to the STT ligaments using a magnetic field tracking system and measurement method (3Space Isotrack).
Our previous experience on kinematic data analysis was based on the roentgen-stereophoto-grammetric method applied in different experimental protocols related to carpal kinematics.8'2'823 The disadvantages of this method are that an incision of the joint capsule has to be made for implantation of 4 metal markers in each carpal bone and a continuous registration of sensors motion cannot be done. Xray exposure has to be performed in a static condition. The analysis of kinematic measurements of bones in biomechanical and kinematic studies (3) has reached a high technological standard. The most accurate results are obtained with the roentgenstereophoto-grammetric system, The sonic digitizer method, 3 33 34 and the sixdegreeoffreedom spatial linkage. 24 Photographicalstereometric methods using optical signals from a light emitting diode (L.E.D.) such as in gaitanalysis are not suitable for an analysis of carpal kinematics. All these methods require a fixed position of one of the rigid bodies when measurement is performed. With the sonic digitizer, data collection is performed in 25 intervals of the wrist motion.3 Using the roentgenstereogrammetric method, measurements of wrist motion in increments of 10 degrees were chosen. The linkage method has not been applied for measurements of carpal kinematics. The motion between the different positions of the bone has to be extrapolated using different algorithms.
Most kinematic analyses performed in human cadaver hands were obtained from a population aged between 6080 years, as this corresponds to natural mortality in an average population. Some authors report on carpal kinematics in hands of younger donors: 57 and 44 years in the study of Woltring et al. however this particular study was done in only 2 hands. In most papers specification of sex and age of the hands used for kinematic data analysis is not found. As kinematic data analysis is very time consuming, most papers only report on experiments performed with 2 to maximal 6 specimens. 8'22330 In the present study the data of nine wrists with no pathologic findings detected in preexperimental examination is reported.
This study introduces for the first time the magnetic field technology (Isotrack 3 Space) for an experimental analysis of the kinematic behavior of the scaphoid. Advantages of this technology are the capability for continuous data acquisition and incremental motion analysis. This allows a graphic display of the continuous motion, rather than the static defined end positions of the scaphoid, such as the roentgen-stereophoto-grammetric method. This technology allows comparison of the scaphoid motion in intermediate positions during motion under different conditions, with continous data acquisition.
Although described in more technological biomechanical papers, the concept of motion analysis presented in this paper is based on the kinematics of rigid bodies in 6 DOF. The methodologic approach is briefly summarized: a rigid body is defined in classical mechanics as "a system of mass points subject to the holonomic constraints that the distances between all pairs of points remain constant throughout the motion".9 This can be assumed for the scaphoid as no change of the shape of the bone is to be expected, even when the hand is loaded. For a rigid body 3 reference points are required to determine the position and orientation of the object in space. A kinematic analvsis with a complete and accurate quantitative description of motion requires 15 data variables.'29 These include the position vectors, linear velocity and acceleration of the rigid bodies segment's center of mass, angular orientation, angular velocity and angular acceleration. All these criteria are available with magnetic field technology (Isotrack 3Space System).
In the experimental setup the physiologic motion pattern of the scaphoid was simulated with 1.4 kg crossing the wrist. In the second experimental condition 23.5 kg were applied. Under clinical circumstances this is a realistic force transmission through the wrist.
The different motion patterns of the scaphoid with intact STT ligaments it a low load condition (experiment "NLL") are:  when the wrist is moved from extension to flexion in a 60 degrees arc, the scaphoid performs a flexion arc of 50 degrees; pronation and supination amount to 8 degrees and the RUD of the scaphoid has a range of 2 degrees (Figure 4);  during the RUD of the wrist in an arc of 35 degrees the main motion component for the scaphoid is FEM of 18 degrees. The amount of pronation and supination motion is 14 degrees. The scaphoid rotates around the Zaxis only 2.5 degrees during RUD, of the wrist (Figure 5). After sectioning the STT ligaments all three components of scaphoid motion are reduced. This could be found for FEM as well as for RUD of the wrist when a low load condition was simulated (experiments "NLL" and "2LL"). These experiments underline the importance of the STT ligamentous complex for the stability of the scaphoid. Our data for the normal motion of the scaphoid with the STT ligaments intact can be compared with other kinematic studies where the scaphoid motion was analyzed using the screw axis concept. 30 No description of a similar experimental protocol on the effect of loading was found in the literature. Loading led the scaphoid from its normal position to a flexed, ulnar deviated and pronated position when RUD motion of the wrist was performed (figure 9, figure 10, figure 11). When FEM of the wrist was done the scaphoid moved under loading in a flexed, radialdeviated and pronated position (figure 6, figure 7, figure 8).
Sectioning of the STT ligaments leads to a slight decrease of the 3 motion components (rotation of the scaphoid around the X, Y, and Zaxes) of the scaphoid during FEM of the wrist. During FEM of the wrist, the scaphoid is "stabilized" against rotation by loading which reduces the pronation supination considerably (p < 0.001). This happens whether the STT ligaments are intact or not. In the high load condition there is no difference in the motion pattem of the scaphoid between the experiments "NHL" and "2LL".
The pronation supination motion component of the scaphoid is altered most significantly when the STT ligaments are sectioned. The other two motion components of the scaphoid (FEM and RUD) are not changed in their characteristics after sectioning of the STT ligaments was performed. Loading reduces the range of motion for all three motion components of the scaphoid during RUD of the wrist. Under the high load condition (experiment "NHL" and "2HL") there is no difference in the kinematic behavior of the scaphoid whether the STT ligaments are intact or cut. Scaphoid motion decreases after sectioning the STT ligaments. The motion characteristics for the 3 motion components of the scaphoid remain however the same without regard to the STT ligaments intact or sectioned. Loading of the hand results in an overall decrease of the spatial motion of the scaphoid during FEM and RUD of the wrist. This phenomenon is due to the close attachment of the tendon sheath of the FCR to the palmar aspect of the scaphoid. Anatomic dissection of the specimens used in this study showed that the tendon sheath of the flexor carpi radialis is in direct contact to the palmar side of the trapezium and scaphoid. The FCR tendon inserts at the base of the second metacarpal. The palmar STT ligaments cannot be isolated from the tendon sheath of the FCR as shown in figure 12.
This finding and the kinematic data suggest that the FCR has to be considered as a "dynamic" stabilizer of the scaphoid during the wrist motion. The effect of the FCR tendon on the kinematics of the scaphoid is explained by the "bow stringing" effect which causes an extension moment on the distal pole of the scaphoid when this tendon is loaded (experiment "NHL" and "2HL" figure 13).
The present concept of classification of carpal instabilities describes the position of the lunate with regard to its angular inclination to the radius as flexed or extended, as an intercalated segment (VISI and DISI deformity). A similar description of the position of the scaphoid conceming the different instability pattem is not available. Based on these experiments, this study suggests a new classification for the RIS which may be described in a similar fashion as for the lunate:  for the rotatory instability of the scaphoid due to the scapholunate dissociation alone, the abbreviation RISA for this type can be used;  if dissociation between the scaphoid, trapezium and trapezoid (STTD) is responsible for the collapse deformity of the scaphoid, the abbreviation RISB can be used;  if both ligamentous structures of the scaphoid, the distal intrinsic STT ligaments and the proximal intrinsic scapholunate and/or extrinsic scaphoradial ligaments are ruptured or attenuated, this should be described by the abbreviation RISC.
Furthermore our clinical experience may suggest that a secondary attenuation of the STT ligaments occurs as a result of chronic scapholunate dissociation. Under clinical circumstances the evaluation of the instability pattern of the scaphoid should include all assessment of the STT ligaments according to the previous suggested classification for the rotatory instability of the scaphoid (RIS).
This may be helpful for a further classification of carpal instabilities as treatment options for the rotatory instability of the RIS will vary among the different conditions in the 3 different RISA, RISB and RISC types of scaphoid collapse due to ligamentous carpal instabilities.
This study was supported in part by the Deutsche Forschungsgemeinschaft
D.F.G. (German Research Society). This experimental study was
done by the first author at the Orthopedic Biomechanics Laboratory
of the Mayo Clinic, Rochester, Minnesota, USA.