Factors Affecting the Sagittal Alignment of the Lunate

MATERIALS AND METHODS

The database from our earlier studies provided material for the present study and included CT scans of the wrists of 70 healthy volunteers who had never had any symptoms pertaining to their wrist joint. The mean age of the volunteers was 34 (range 19–55) years. There were 49 men and 21 women. A standard posteroanterior and a true lateral radiograph of wrist were taken to exclude any radiological abnormality.

All CT wrists were done with the subjects lying prone, keeping the ipselateral shoulder abducted overhead with the elbow at 90° and with the palm facing the table. The forearm and hand were tied to a custom designed positioning device to keep them parallel to the CT table while maintaining the wrist in neutral alignment (Fig 1). A posteroanterior scout view was used, not only to outline the exact area to be scanned, but, also, to confirm whether the wrist was placed in neutral radio-ulnar deviation, maintaining the axis of the third metacarpal collinear with the axis of the radius. The proximal carpal row rotates dorsally during ulnar deviation and volarly during radial deviation of the wrist. It is necessary, therefore, to position the wrist precisely, particularly in terms of its ulnar or radial deviation, for accurate measurement of the various sagittal alignments of the proximal carpal row. One mm thick sections were taken at an interval of 1.5 mm from the distal radioulnar joint to the proximal end of the metacarpals.

Sagittal sections through the middle of the capitate head were used to measure the volar tilt of the distal radial articular surface and the sagittal axis of the lunate. Sagittal sections through the triquetrohamate joint demonstrated that the triquetral articular surface has variable shape and inclinations, except for the first few radial sections, which demonstrated the triquetral articular surface to form a straight line with consistent shape and inclination. The radial-most section through the triquetrohamate joint, therefore, was used to draw the sagittal axis of the triquetrum, which is a line perpendicular to the straight articular surface of the triquetrum (Fig 2A). A line perpendicular to the line tangential to the two poles of the lunate formed the sagittal axis of the lunate (Fig 2B). The angles which the two axes measured with the line along the long axis of the radius were defined as the “Sagittal Angle of the Triquetrum” and the “Sagittal Angle of the Lunate”. Measurements of the height of the lunate were drawn according to Watson et al. (1996) (Fig 2B) on the sagittal section through the midpoint of the lunate and sections at 3 mm on either side of it and were defined as the “Wedge Ratio-Sagittal (middle)”, “Wedge Ratio-Sagittal (lateral)” and “Wedge Ratio-Sagittal (medial)”.

The largest circle which fitted congruently with the articular surface of the head of the capitate was drawn and the centre of the circle marked. The same procedure was repeated for the articular surface of the distal radius. Longitudinal axes through the centres of the two circles were drawn parallel to the axis of the radius and were defined as the “Midcarpal Axis” and the “Radio-carpal Axis”. The distance between the two axes was measured and taken as a positive value if the midcarpal axis was dorsal to the radiocarpal axis and as a negative value if the midcarpal axis was volar to the radiocarpal axis (Fig 3).

Axial sections of 1 mm thickness and interval were reconstructed from the sagittal images. Axial sections through the distal one fourth of the scapholunate joint provided a consistent, straight, lateral border and were used to draw the lunate axis. The angles which the lunate axis and the line along the luno-triquetral joint formed with the zero rotation sagittal line measured the axial angle of the lunate and the luno-triquetral angle, respectively (Fig 2C) (Gupta and Moosawi, 2002).

Correlation among various parameters was derived using the bivariate correlation procedure and Pearson’s coefficient of correlation. The level of significance was set at P < 0.05. All statistical analyses were performed using SPSS for Window.

RESULTS

The lunate was seen to be of dorsal type with a narrow dorsal segment in 34 subjects (50%), of volar type with a narrow volar segment in three subjects (4%) and of neutral type with equal dorsal and volar segments in two subjects (3%), in all three sagittal images. Thirty-one subjects (43%) had a lunate with mixed shape, with two, or three, shapes seen in the same lunate on the various chosen sagittal sections. The mean wedge ratio-sagittal (medial), wedge ratio-sagittal (middle) and wedge ratio-sagittal (lateral) were 1.09 (SD 0.12), 1.12 (SD 0.14) and 1.20 (SD 0.20), respectively.

The midcarpal axis was found to be dorsal to the radiocarpal axis in 21 subjects (30%) with the mean distance between the two axes being 1.6 mm (SD 0.87), ranging from 1 to 4 mm. The midcarpal axis was found to be volar to the radiocarpal axis in 29 subjects (41%). The mean distance between the two axes in this group was −1.3 mm (SD 0.71), ranging from −1 to −4 mm. The two axes were nearly collinear in 20 subjects (29%). The distance between the two axes showed a significant correlation with the sagittal alignment of the lunate (r = −0.8).

The mean volar tilt of the distal radial articular surface and the sagittal angle of the lunate were 10° (SD 8.01) and 7° (SD 11.88), respectively. Most lunates did not demonstrate a shape of a similar type on the three chosen sagittal CT sections but showed a mixed pattern with two, or even all, the three types of shape in the same lunate. Therefore, it was not possible to classify all lunates based on their shape on the CT measurements. No significant correlation was seen between the sagittal alignment of the lunate and the three wedge ratio measurements of the height of the lunate or with the volar tilt of the distal radial articular surface.

The mean sagittal angle of the triquetrum, the axial angle of the lunate and the luno-triquetral angle were 12° (SD 10.02), 13° (SD 9.34) and 1° (SD 11.36), respectively (Table 1). The sagittal angle of the lunate demonstrated significant correlation with the alignment of the triquetrum in both planes, i.e. the sagittal angle of the triquetrum (r = 0.6) and the luno-triquetral angle (r = 0.5).

DISCUSSION

The lunate is known to have a tendency to extend under compressive axial loading (Kauer, 1986). The main factor responsible for this is the lunate’s wedge shape, with a narrower dorsal part. The axis of the midcarpal joint, classically described as lying dorsal and parallel to that of the radiocarpal joint, also converts the axial load into a force couple rotating the lunate dorsally (Kauer, 1986; Klienman, 2000). This latter force, although as important in moving the lunate to rotate dorsally as is the shape of the bone itself, has, somehow, not attracted much attention in the literature. This is the first study to identify three different patterns of sagittal alignment between the midcarpal axis and the radiocarpal axis and recognise that the classical description of the midcarpal axis lying dorsal to the radiocarpal axis, thereby favouring extension of the lunate, is found in only 30% of wrist joints.

Sagittal alignment of the lunate is known to be very varied and this study confirmed this to range from −19° to 44°. The relationship between the midcarpal axis and radiocarpal axis in terms of dorsovolar displacement has a key role in determining the sagittal alignment of the lunate. A positively measured distance between the two axes, i.e. a dorsally placed midcarpal axis, favours dorsal inclination of the lunate while a negatively measured distance, or volarly placed midcarpal axis, favours volar inclination of the lunate (Fig 3).

In a study on plain lateral radiographs, Watson et al. (1996) described the lunate as having three different shapes but found no correlation between the shape of the lunate and its sagittal alignment. The shape pattern of the lunate and its prevalence could not be substantiated in a CT study (Gupta and Moosawi, 2002). In the present study, the lunate has been shown to have a shape of the mixed pattern in 43% subjects, in which two, or even all three, shapes of the lunate were found in the same lunate on the various chosen sagittal sections. It is, therefore, not surprising that the alignment of the lunate failed to conform to the pattern of its shapes (Fig 4).

The dorsal placement of the luno-triquetral articulation provides extension torque to the lunate, balancing the flexion torque on it exerted through the scaphoid and, thus, maintaining equilibrium. The distal pole of the scaphoid forms the first structure of the proximal carpal row to offer resistance to any axial force acting from distal to proximal. It is possible, therefore, that most of the axial force is exerted upon the scaphoid itself, producing a flexion torque due to the volarly placed distal pole of the scaphoid. Relative supination of the scaphoid, relative positioning of the trapezium-trapezoid hood over the distal pole of the scaphoid and the length of the scaphoid are among a few possible factors to affect the scaphoid flexion torque. The role of these factors need further study.

The axial load across the wrist is said to rotate the scaphoid and lunate in opposite directions, creating continuous tension in the scapholunate interosseous ligament (Zdravkovic et al., 1995). As a consequence, patients with injury to the scapholunate interosseous ligament develop a dorsal intercalated segmental instability (DISI) pattern with a dorsally rotated lunate and a flexed scaphoid with rotary subluxation. Patients with arthroscopically proven scapholunate ligament injuries do not always take up the DISI pattern (Kelly and Stanley, 1990). It is, therefore, conceivable that the resultant torque on the lunate and the triquetrum combined may not always be that of extension.

Axial loading on the triquetrum can produce torque in the dorsal direction only because of its articulation on the lunate dorsally and its articulation with the hamate facing dorsally. The factors which affect the torque on the triquetrum include the sagittal angle of the triquetrum and the luno-triquetral angle, which, in this study, showed significant correlation with the sagittal alignment of the lunate. However, torque due to axial loading of the lunate, per se, is not constant because of the varying shape of the lunate and the varying patterns of sagittal alignment between the midcarpal axis and the radiocarpal axis. This may explain the clinical situation in which patients with arthroscopically proven scapholunate ligament injuries did not take up the DISI pattern.

It is evident that the proximal carpal row under axial loading is subjected to differential torque as a result of the varied shapes and the anatomical configuration of the three carpal bones. Axial loading of an independent scaphoid would cause it to flex. Axial loading of the triquetrum would extend it, while loading the lunate could rotate it to either side, depending upon the pattern of its shape or the axes of the midcarpal or the radiocarpal joint. However, it is the combined torque of the three carpal bones which ultimately determines that the whole proximal carpal row rotate in a particular direction under axial loading, in an intact carpus. The whole of the proximal carpal row, including the lunate, is reported to undergo flexion under experimental loading of the wrist joint with a uniformly applied compression force (Kobayashi et al., 1997). Elimination of the physiological axial load in aesthetised patients with complete muscle relaxation produced extension of the lunate and the scaphoid, suggesting that axial loading as a result of the normal tone of the forearm muscles tends to flex both the scaphoid and the lunate (Gupta, 2002). It may, therefore, be concluded that the combined torque from axial loading of the three carpal bones always remains that of flexion.

There are healthy individuals whose lunate maintains an attitude of dorsal rotation. With axial loading producing a flexion torque in the intact carpus, it may be difficult to explain occurrence of such an attitude of dorsal rotation of the lunate, unless there exists another mechanism to rotate the carpus. It is concluded that the relationship between the midcarpal axis and the radiocarpal axis, in terms of dorsovolar displacement, has a key role in determining the sagittal alignment of the lunate, while the shape of the lunate may be only a secondary consideration. The relative length of the volar and the dorsal wrist capsular ligaments and the balance of equilibrium between the two could possibly be one of the factors to determine the pattern of the two axes. It may be a wise step to incorporate proper tissue balancing between the volar and the dorsal wrist capsular ligaments while conceptualising the development of any future wrist prostheses.