Safety of Releasing the Volar Capsule During Open Treatment of Distal Radius Fractures: An Analysis of the Extrinsic Radiocarpal Ligaments’ Contribution to Radiocarpal Stability

Abstract

Purpose

The contribution of the extrinsic radiocarpal ligaments to carpal stability continues to be studied. Clinically, there is a concern for carpal instability from release of the volar extrinsic ligaments during volar plating of distal radius fractures in which the integrity of the dorsal ligaments may be unknown. The primary hypothesis of this study was that serial sectioning of radiocarpal ligaments would lead to progressive ulnar translation of the carpus.

Methods

We studied the stabilizing roles of the radioscaphocapitate (RSC), short radiolunate (SRL), long radiolunate (LRL), and dorsal radiocarpal (DRC) ligaments. We sequentially sectioned these ligaments in 2 groups of 5 matched pairs and measured the motion of the scaphoid and lunate with the wrist in passive neutral alignment, radial deviation, ulnar deviation, and simulated grip. Displacement of the lunate in the radioulnar plane was used as a surrogate for carpal translation. The groups differed only by the order in which the ligaments were sectioned.

Results

In the intact state, the lunate translated ulnarly during simulated grip and radial deviation, whereas radial translation, relative to its position under resting tension, was observed during ulnar deviation. With serial sectioning, the lunate displayed increased ulnar translation in all wrist positions for both groups 1 and 2. The magnitude of ulnar translation exceeded 1 mm after sectioning the LRL plus RSC along with either the DRC or the SRL.

Conclusions

Sectioning of either the DRC or SRL ligaments along with release of the RSC and LRL ligaments leads to notable although minimal (<2 mm) ulnar lunate translation.

Clinical relevance

Isolated sectioning of individual radiocarpal ligaments, such as for visualization of the articular surface of the distal radius, leads to minimal ulnar translation. Because prior clinical work found no clinical complications after volar capsule release, it is posited that translation less than 2 mm creates subclinical changes in carpal mechanics.

FRACTURES OF THE DISTAL RADIUS are common injuries, with an incidence of approximately 16.2 fractures per 10,000 person-years in the United States.¹ Although many fractures are treated with nonsurgical measures such as closed reduction and immobilization, others require surgical intervention.² When open reduction and internal fixation is indicated, volar or dorsal approaches can be employed. Currently, surgeons less commonly use the dorsal approach for concerns of tendon irritation, although this approach allows for visualization of the distal radius articular surface. Visualization of the articular surface through the volar approach to the distal radius is avoided owing to concerns of carpal instability from release of the volar extrinsic ligaments.^3,4 For example, Siegel and Gelberman⁵ found that horizontal and vertical oblique radial styloidectomies consistently disrupted the palmar radiocarpal ligaments, which Fisk⁶ and Blevens et al⁷ showed to be central to preventing ulnar slide of the carpus. Moreover, cases have been described in which rupture of the palmar radiocarpal ligaments after traumatic injury resulted in ulnar translation of the entire carpus.^8,9

The roles of the extrinsic ligaments in normal carpal kinematics are complex and remain areas of ongoing investigation.^10,11 For instance, the extent to which the extrinsic ligaments can be safely released, and the subsequent effects on carpal translation, have not been described. Gaining a more complete understanding of the roles of the extrinsic radiocarpal ligaments in maintaining carpal stability may inform surgical techniques for visualizing the articular surface during volar plating of fractures of the distal radius. Although they are not encountered during the volar approach, the dorsal structures were included in our study because the integrity of the dorsal capsule is often unknown at the time of distal radius fracture fixation.

The purpose of this study was to evaluate the contributions of the dorsal and volar radiocarpal ligaments to carpal translation relative to the radius. Displacement of the lunate in the radioulnar plane was used as a surrogate for carpal translation. The primary hypothesis was that serial sectioning of the extrinsic radiocarpal ligaments would lead to progressive ulnar translation of the lunate. The secondary hypothesis was that isolated sectioning of individual radiocarpal ligaments would not lead to gross radiocarpal instability (>2 mm).^12–15

MATERIALS AND METHODS

We used 5 fresh-frozen matched pairs of cadaver upper extremities in this study (mean age, 46 ± 16 years; range, 22–66 years; 2 female and 3 male). Specimens were excluded if they showed evidence of previous fracture, wrist ligamentous injury, or bone lesion. Specimens were stored in a −20°C freezer before dissection and testing. Each specimen underwent 2 freeze–thaw cycles: the first for dissection and preparation and the second for kinematic evaluation.

We followed a previously described protocol for loading and testing carpal kinematics in cadavers.^16,17 Specimens were transected at the elbow joint and dissected, preserving the extensor carpi radialis longus (ECRL), extensor carpi radialis brevis (ECRB), abductor pollicis longus (APL), extensor carpi ulnaris (ECU), flexor carpi ulnaris (FCU), and flexor carpi radialis (FCR) tendons. The tendons were transected at the myotendinous junction and connected to weights by heavy string. The fingers were disarticulated at the meta-carpophalangeal joints and all other flexors and extensors were removed. Care was taken to preserve the extensor and flexor retinacula. A 3.2-mm-diameter rod was inserted retrograde into the third metacarpal intramedullary canal. For motion tracking, threaded K-wires (2.0 mm diameter) were inserted into the scaphoid, lunate, radius, and third metacarpal from the volar aspect of the wrist. Appropriate wire position was confirmed fluoroscopically and visually to ensure that the wires did not impinge on carpal motion. These K-wires were used as attachment points for custom motion-tracking system rigid body trackers (3D Creator, Boulder Innovation Group, Boulder, CO). Previous testing showed that the relative motion between the rigid body trackers is accurate to 0.04 mm and 0.1° (root-mean-square error). The rigid body trackers weigh 12 g. To minimize the influence of tracker weight on carpal motion, the rigid body trackers were placed as close to the bones as possible to minimize the moment arms and torques placed on the bones.

Forearm rotation was locked in neutral alignment with two 2.0-mm diameter transosseous K-wires through the shafts of the radius and ulna. The olecranon process of the ulna was then removed to facilitate attachment to a custom testing device. The third metacarpal intramedullary rod was connected to a guide rail and used to control radioulnar deviation (Fig. 1) The guide did not restrict rotation or axial translation of the third metacarpal.

FIGURE 1:

Experimental setup. A specimen secured in the custom testing device with rigid body motion trackers attached and transosseous K-wires inserted.

During passive testing, the operator translated the third metacarpal rod along a guide rail to bring the wrist joint into neutral alignment, 10° radial deviation, and 20° ulnar deviation. The ECRL, ECRB, APL, ECU, FCU, and FCR tendons were tensioned with 1.5-N weights to simulate resting muscle tension.^16,17 Wrist positions were verified using a digital level with a built-in angle sensor. This was done by aligning the digital level with the third metacarpal rod and determining the angle with respect to the vertical axis. Kinematic data were additionally acquired during simulated grip with the wrist in neutral alignment. The tendon weights during simulated grip were proportional to the physiologic cross-sectional area of the muscles and electromyographic activity for a total simulated grip force of 98 N (ECRL = 24.5 N; ECRB = 13.7 N; APL = 9.8 N; ECU = 14.7 N; FCU = 21.5 N; and FCR = 13.7 N).¹⁶ Three trials of kinematic data were acquired for each wrist position. All kinematic data in all positions were acquired with the wrist in static equilibrium.

The matched pairs were divided into 2 groups of 5 specimens each with right and left sides arbitrarily assigned to each group. We chose this sample size based on a previous study that showed a power of 89% to detect a 2 mm difference in these specimens.¹⁸ The specimens were then tested in the intact state. The first group of specimens underwent isolated sequential sectioning of the dorsal radiocarpal (DRC) ligament, followed by the radioscaphocapitate (RSC) and long radiolunate (LRL), and finally the short radiolunate (SRL) ligaments (group 1). The second group of specimens (group 2) underwent sectioning of the same ligaments in reversed order: isolated sectioning of the SRL, followed by the LRL and RSC, and finally the DRC. For clarity, kinematic data were gathered after sectioning of one ligament, then again after sectioning of 3 total ligaments, and a final time with all 4 ligaments sectioned. This was true for specimens in both groups 1 and 2, with only the order of sectioning reversed.

We released the extrinsic ligaments off the radius based on the data and anatomical landmarks described by Zumstein et al.¹⁹ In terms of percentiles of the total bony width of the radius, in which 0 represents the most ulnar aspect of the radius and 100 represents the most radial aspect of the radius, the DRC attached from the 16th to the 52nd percentile, the RSC and LRL attached from the 87th percentile dorsally to the 59th percentile volarly, and the SRL attached from the 14th to the 41st percentile. With respect to anatomic landmarks, the DRC attached from the dorsal ulnar corner to just adjacent to the ulnar border of Lister tubercle. The RSC and LRL attached from the tip of the radial styloid process to approximately the middle of the scaphoid fossa. The SRL attached from the radial edge of the lunate to the volar ulnar corner.

After kinematic testing, a handheld probe was used to acquire landmark points on the radius, third metacarpal, scaphoid, and lunate. The volumetric centroids of the scaphoid and lunate were determined from points acquired on the surface of each bone after resection from the wrist. Motion for each wrist position was calculated with respect to the intact, neutral, and resting tension postures (reference state). Translations of the scaphoid and lunate were calculated from the motion of the volumetric centroids with respect to the radial coordinate system.²⁰ The positions of the scaphoid and lunate in the reference state were defined as zero translation. Ulnar translation was defined as positive, and radial translation was defined as negative. Displacement of the lunate in the radioulnar plane was used as a surrogate for carpal translation. Rotations were calculated from helical axis rotation projections onto the radius-based coordinate system.²¹ Rotations of the third metacarpal were used to verify that proper alignment of the specimens was achieved during kinematic testing. A mixed-effects linear model with a random effect for specimen was used to evaluate the effects of sectioning on scaphoid and lunate kinematics. Significance was set at α < 0.05.

The threshold of 2 mm was informed by studies conducted by Gilula and Weeks,¹² Pirela-Cruz et al,¹³ Wollstein et al,¹⁴ and Gupta and Al-Moosawi.¹⁵ Gilula and Weeks proposed that ulnar translation of the carpus can be determined to be present radiographically when 50% or more of the lunate lies medial to the distal radius. Quantitatively, this can be expressed by the ratio of the lunate overhang (Lo) (the amount medial to the radius) to the lunate width (Lw). If Lo / Lw is greater than 0.5, ulnar translocation of the carpus is present. Wollstein et al showed that the mean value of the Lo/Lw ratio in healthy, intact specimens is 0.4, with a 95% confidence interval from 0.37 to 0.44. Gupta and Al-Moosawi found that the mean mediolateral diameter of the lunate was 12.80 mm with a standard deviation of 1.37 mm. Thus, it follows that a 10% translation of the lunate relative to the radius (0.1 × Lw = 1.28 mm) in the ulnar direction indicates the presence of ulnar translation of the carpus. Therefore, our threshold was set at 2 mm to consider the standard errors of the preceding measurements and be specific for conditions in which the carpus is grossly unstable.

RESULTS

Table E1 (available on the Journal’s Web site at www.jhandsurg.org) presents lunate and scaphoid translations and rotations for all wrist positions and states. P values for all comparisons are presented in Table E2 (available on the Journal’s Web site at www.jhandsurg.org). A P value threshold of less than .05 was used to determine statistical significance.

In the intact state, the lunate translated ulnarly during simulated grip and radial deviation, whereas radial translation was observed during ulnar deviation, relative to the reference state (Fig. 2). With serial sectioning, the lunate displayed increased ulnar translation in all wrist positions for both group 1 and group 2. Ulnar translation exceeded 1 mm after sectioning the LRL plus RSC along with either the DRC or the SRL. Similar trends in ulnar translation with serial sectioning were observed in the scaphoid for both group 1 and group 2 (Fig. 3).

FIGURE 2:

Radioulnar translation of the lunate with serial sectioning of the capsular ligaments for group 1 (A) and group 2 (B). Subtitles of individual charts indicate the angle of the third metacarpal as well as the amount of tension applied to tendons. The x axis denotes the ligaments severed. The y axis shows translation of the volumetric centroid relative to its position in the reference state. Positive translations are directed ulnarly, and negative translations radially. The intact neutral–resting tension wrist position served as the reference state for all motion. Horizontal dashed lines represent 2 mm ulnar translation from the median value of the measurements under resting tension, the threshold for gross instability. Shared subscripts denote no significant difference between test configurations (P > .05). P values for all comparisons are presented in Table E2. Deg, degree.

FIGURE 3:

Radioulnar translation of the scaphoid with serial sectioning of the capsular ligaments for group 1 (A) and group 2 (B). Positive translations are directed ulnarly. Shared subscripts denote no significant difference between test configurations (P > .05). P values for all comparisons are presented in Table E2. Deg, degree.

With the wrist in radial deviation and simulated grip, the scaphoid translated significantly proximally after sectioning of the SRL plus LRL plus RSC ligaments (group 2). The scaphoid translated significantly volarly after sectioning of the DRC plus LRL plus RSC ligaments with the wrist in both radial and ulnar deviation (group 1). With the wrist in radial deviation, significant volar translation occurred after sectioning of the SRL plus LRL plus RSC ligaments (group 2). No clear trends were observed in proximodistal or volar-dorsal translation of the lunate.

With the wrist in 20° ulnar deviation, the lunate and scaphoid pronated significantly after sectioning of DRC plus LRL plus RSC ligaments (group 1). In group 2, after sectioning of the SRL plus LRL plus RSC ligaments, the lunate pronated significantly for all wrist positions except 20° ulnar deviation, whereas the scaphoid pronated significantly during simulated neutral grip and radial deviation.

In group 1, after sectioning of the DRC plus LRL plus RSC ligaments, the scaphoid radially deviated significantly for all wrist positions except 20° ulnar deviation. With the wrist in radial deviation and simulated grip, the scaphoid radially deviated significantly after sectioning of the SRL plus LRL plus RSC ligaments (group 2).

No clear trends were observed in flexion-extension rotation of the scaphoid or flexion-extension and radioulnar deviation of the lunate.

DISCUSSION

We found progressive ulnar translation of the lunate after sequential sectioning of the extrinsic radiocarpal ligaments. However, isolated sectioning of individual radiocarpal ligaments did not lead to radiocarpal instability greater than 2 mm. After sectioning of either the DRC or the SRL with the other 3 ligaments still intact, the carpus translated less than 1 mm. Ulnar translation of more than 2 mm did not occur until complete sectioning of all 4 extrinsic radiocarpal ligaments (SRL, LRL, RSC, and DRC). This result was independent of the group to which the specimens were assigned. Maximum translation of 2.7 mm was observed with all ligaments severed while under simulated grip. With 3 of 4 ligaments sectioned, the condition that most replicates sectioning of the volar ligaments with intact dorsal structures during a distal radius fracture repair, maximum ulnar translation of 1.5 mm, was observed with simulated grip.

These findings are consistent with previous cadaveric studies that demonstrated that ulnar translation of the carpus (>2 mm) occurred only after complete disruption of all radiocarpal ligaments.^4,18,22 Viegas et al¹⁸ found that sectioning of the RSC and the radiolunate ligaments did not produce notable ulnar translation of the carpus on radiographs (as defined by a lunate overhang to lunate width ratio greater than 0.5), contact pressure-sensitive paper (as defined by x, y coordinates relative to the intact state), or clinical examination (as defined by manual displacement of the scaphoid onto or ulnar to the interfossal ridge of the radius). They noted the occurrence of palmar translation of the carpus after additional sectioning of the ulnocarpal ligaments, and ulnar translation of the carpus after severe and global disruption of the ligaments. Similarly, using a single fixed specimen, Rayhack et al²² noted that ulnar translation did not occur by manual displacement until both radiolunate and RSC origins were entirely sectioned. However, when comparing these studies, it should be noted that Viegas et al and Rayhack et al sectioned the ulnar extrinsic ligaments in addition to the volar and DRC ligaments sectioned in this study. Moreover, our study incorporated 3-dimensional motion capture technology and simulated grip and wrist positioning to address some of the limitations of the earlier studies. Evaluation of the specimens using conventional radiography was avoided owing to its low sensitivity (60% to 80%) for diagnosing carpal instability.²³

A previous study²⁴ found no radiographic or clinical signs of ulnocarpal translation in 11 patients who underwent an open volar capsular release of the radial collateral ligament, the RSC, the LRL, and part of the SRL for treatment of wrist stiffness, a sequela of previously treated distal radius fractures with volar locking plates. However, that analysis was completed using radiographs, whereas our cadaveric analysis recorded volumetric data with the wrist in various positions and under variable amounts of load. Considering that the greatest translations occurred with simulated grip, it is possible that carpal instability was missed by static radiography. However, the results of our study show that isolated sectioning of radiocarpal ligaments in the absence of additional ligamentous or bony trauma results in minimal (<1-mm) radiocarpal translation. This can occur from release of even the dorsal extrinsic ligament (ie, DRC), which is commonly done during procedures addressing the carpus. Because prior clinical work found no complications after volar capsule release, it is posited that translation less than 2 mm creates subclinical changes in mechanics. The long-term effect of small changes in radiocarpal stability is unknown.

A limitation of this study was that its analysis was based on cadaveric specimens, which may not reflect in vivo motion. Prior work, however, demonstrated similar behavior between cadavers and in vivo analyses.²⁵ The experimental study does not replicate the full range of wrist motion or loads, or the dynamics of wrist motion in an active state. Although we inferred kinematics from static positions, we did not acquire time series data during changes of position. The ulnocarpal and the intrinsic wrist ligaments were also not tested. Changes in radiocarpal and ulnocarpal contact pressures were not measured. These limitations do not detract from the validity of the conclusions drawn from this study because our focus was on the role of the volar and dorsal radiocarpal ligaments in maintaining carpal stability under 4 clinically important sets of conditions. Future investigations may contribute to knowledge in this field by studying the role of the ulnocarpal and intrinsic wrist ligaments in maintaining carpal stability or by studying the motion of the carpals using a dynamic model under various loads.