|Home | About | Journals | Submit | Contact Us | Français|
To determine the role of the carpal ligaments during wrist flexion/extension and to understand if maintaining integrity of only the dorsal scapholunate ligament is adequate for maintaining stability of the scapholunate joint.
This study combined motion analysis and manual digitization of ligament attachment regions to generate predictions of carpal ligament length and implied strain during wrist motion and length changes after simulated ligamentous injury.
Thirteen ligaments and 22 ligament segments (sub portions) were modeled. Ligament length change with respect to wrist angle was measured. Eleven segments had minimum stretch or elongation from neutral wrist position over the entire wrist range of motion for any ligament cut condition. The remaining eleven segments had more than 10% stretch in some portion of flexion/extension. In general, ligaments had increased stretch during wrist flexion and after cutting the entire scapholunate ligament (SLL) and the dorsal intercarpal ligaments off the scaphoid.
Disruption of the membranous and palmar portions of the SLL and the dorsal intercarpal ligament off the scaphoid did not result in the development of an increased 3-dimensional scapholunate gap as measured by differences in ligament length calculations between the scaphoid and lunate. This may be indicative of a pre-dynamic instability condition (prior to clinical signs and x-ray findings) that is stabilized by the dorsal SLL preventing the increase in 3-dimensional scapholunate gap. This may also support surgical treatment recommendations, which suggest repair of the dorsal component only of the SLL will be effective. Disruption of the dorsal intercarpal ligament off the scaphoid or lunate did not result in further significant changes. Therefore the dorsal SLL has an important role in preventing scapholunate ligament instability.
These results provide insight into the abnormal kinematics as various ligaments are compromised.
Kinematics of the carpal bones is complex. Movement of the carpal bones is mostly determined by articular contact and ligamentous tethers and not by direct muscular attachments and forces. It is therefore difficult to predict motion of the bones after ligament tears using current imaging and clinical techniques. The term scapholunate instability describes a spectrum of clinical conditions making a diagnosis of mild or moderate sprains not evident on xrays difficult to diagnose . Linsheid  described scapholunate dissociation and detailed the loss of mechanical linkage between the scaphoid and lunate. Later, Mayfield  described the progressive stages of perilunate instability. It is  believed that carpal instability is more common than previously thought and that degenerative disease may be the late result of undiagnosed instability.
Scapholunate instability has been described in 3 stages: pre-dynamic, dynamic, and static [5,6] Few studies have provided insight into what ligaments contribute to these various stages of instability. Short [7,8] investigated the stabilizing role of the scaphoid and lunate ligaments. They report that the scapholunate ligament (SLL) is the primary stabilizer of the joint and the other ligaments secondary stabilizers. Meade  showed that division of the palmar portion of the SLL resulted in minimal radiographic change. Only after the radioscaphocapitate ligament (RSCL) was divided did the scapholunate gap substantially widen. Ruby  also found that dividing the dorsal capsule and entire SLL resulted in a scapholunate diastasis.
These studies indicate that the SLL and dorsal capsule are important to maintaining stability in the carpus. Then, Viegas  described more anatomical information and the role of the dorsal ligaments. He determined that these ligaments form a lateral “V” construct to afford indirect stability to the scaphoid. This research was followed up with a biomechanical evaluation by Mitsuyasu  who determined that when the dorsal intercarpal ligament (DICL) and the SLL are disrupted from the scaphoid, but the DICL was still attached to the lunate, the result was a flexed posture of the scaphoid and a widened scapholunate (SL) gap, but only when the hand was loaded (dynamic instability). When the DICL was also disrupted from the lunate, the result was a flexed posture of the scaphoid, an extended posture of the lunate, and a widened SL gap in the unloaded condition (static instability). The progression of a dynamic to a static deformity is well accepted in the clinical setting. However, before Mitsuyasu′s work , there were no detailed anatomic explanations or examples of the causes of and the differences between a dynamic and static instability.
The purpose of this study was to determine the role of the carpal ligaments during normal wrist flexion/extension motion. The second purpose was to determine if maintaining integrity of only the dorsal SLL was adequate for maintaining stability of the SL relationship.
Six fresh-frozen cadaver upper extremities free from visible or radiographically identifiable deformities and/or degenerative changes were studied. There were 3 right and 3 left wrists, 5 were male and one was female with an age range of 28–59 years. Carpal kinematics were measured using previously reported methods [12,13]. Briefly, titanium screws were placed into the dorsal surface of the radius, ulna, scaphoid, lunate, triqetrum, and capitate under fluoroscopic guidance. Two millimeter diameter graphite rods were glued to the head of each screw. At the end of each rod was a triad pin that had three 5-mm diameter spheres placed in a cruciform arrangement on top. The plastic spheres were coated with 3M reflective tape (St Paul, MN). The sphere and rods were oriented to ensure that they would not come into contact with each other during full passive flexion/extension of the wrist. The rods were rigidly attached in the bones. The rod weight was approximately 0.8g and did not flex or vibrate independently (visual observation). They were positioned to avoid tendons and avoid or minimize tethering of the capsule and/or ligaments. The flexor carpi ulnaris and extensor carpi ulnaris tendons were connected to each other by wires that looped around a freely moving pulley. The flexor carpi radialis and extensor carpi radialis brevis tendons were treated identically. Each loop was weighted with 11.1N (total across the wrist 22.2N or 10 lbs.) of weight. One pulley attached to the radial extensor/flexor pair and the other to the ulnar extensor/flexor pair. These floating pulleys allowed free movement of the tendon pairs during motion and allowed them to move synergistically and to approximate normal muscle tone. A 3 mm diameter metal pin was placed into the medullary canal of the third metacarpal and was used as a guide to move the wrist passively through a flexion/extension motion (minimizing out of plane motion). The elbow was fixed in 90 degrees of flexion, the forearm in neutral, and the arm placed on the table for testing (humerus parallel to the table). Radial and ulnar deviation were not restricted.
A 6-camera optical motion analysis system (Motion Analysis Corp. Santa Rosa, CA) was used to track the 3 reflective markers for each carpal bone as the wrist was manually driven through 2 cycles of wrist flexion-extension. Then each wrist was tested in a series of progressive and cumulative disruptions of the intercarpal ligaments. Each sequence simulated our hypothesized series of increasing stages of carpal instability of the wrist. The ligaments were cut, through a small incision of the wrist capsule to simulate a total tear of each ligament. The order of ligament cuts were as follows: 1) membranous portion of the SLL (membranous SLL) and palmar portion of the SLL (palmar SLL), 2) DICL off the scaphoid (scaphoid DICL), 3) dorsal portion of the SLL (dorsal SLL), 4) DICL off the lunate lunate (DICL), and 5) the entire lunotriquetral ligament (LTL). Table 1 lists the ligament cut sequence and hypothesized clinical implication. Table 2 lists definitions of all ligaments that were modeled. .
Computed tomographic (CT) scans of the specimens with attached markers were performed (General Dynamics CT scanner, Waukesha, WI) generating .625 mm thick transverse contiguous slices of the wrist after collecting the first sequence. A 3-dimensional surface model of each bone and associated markers was subsequently reconstructed from the CT scans of the wrists in neutral wrist position using Mimics (Materialize, Ann Arbor, MI). Each specimen was subsequently disarticulated to confirm that all ligaments were cut. Finally, ligament attachment and articulating surface regions were identified using a 3-dimensional digitizer (Microscribe MX, Solutions Technologies Inc., Odella, MD) [14,15]. The manually digitized bones were aligned with their corresponding CT models. The 3-dimensional carpal bone models were then animated using the motion measured by the 6 camera system. The angle between the radius and capitate was calculated and determined the global wrist angle for each motion sequence investigated.
Ligament lengths during the animated motions for each sequence were determined as the straight line between the center of the attachment region determined by previous methods . Ligament wrapping was modeled as an approximation of the 3-dimensional path of the ligament over the surface of the bone or bones and was calculated as an arc of a convex hull about a polygonal plane intersection with the triangulated bone surface.
A 1-way analysis of variance was used to test for differences in ligament segment length due to ligament instability. Unless explicitly stated otherwise, a P value of less than 0.01 was considered to be statistically significant.
The percent stretch at each wrist position from the normal condition was calculated as: [(ligament length for cut condition – ligament length for no ligament cut condition)/ligament length for no ligament cut condition]*100. To estimate the precision of the ligament length measurements, a subset of 11 ligaments were identified which exhibited no significant change in length during wrist motion, selected from a random selection of 4 specimens at the trial condition. The average standard deviation in the estimate of ligament length for these ligaments with no change in length was 0.5 millimeter.
Data were analyzed for changes in each degree of global wrist angle, however this continuous data was subsampled at 5 wrist intervals (between 24 degrees flexion and 28 degrees extension) to simplify the analysis. Figures 1–3 show ligament lengths at 5 wrist positions for all 22 ligament segments measured in the normal condition. During normal wrist flexion/extension only 8 ligaments had greater than 10% stretch or elongation from neutral wrist position (Table 3).
The overlapping (common) wrist range of motion among the samples (35 degrees extension to 29 degrees flexion) was analyzed for the ligament cut conditions. Eleven ligament segments had less than 10% stretch over the entire wrist range of motion (average of 0.9, maximum of 8%) for any ligament cut condition. The remaining 11 ligaments had over 10% stretch in some portion of the wrist range of motion during 1 or more of any ligament condition (bolded segments in Table 2, segments 4–10, 12–13, and 20). Figure 4 shows the dorsal SLL raw lengths for each ligament cut sequence. Figures 5–7 show the maximum % stretch for ligaments after each sequential ligament cut. In general ligaments had increased stretch during wrist flexion and after the dorsal SLL was cut (condition #4). It should be noted that the term ligament stretch is used even in conditions for which the ligament was cut. Dashed lines in the figures indicate measurements when the ligament is not intact and therefore represents the distance between bony attachments sites.
Previous studies have investigated the dynamic and static aspects of scapholunate instability (defined radiographically as an increase gap between the scaphoid and lunate) and the roles of the DICL . They found that disruption of the entire SLL and the scaphoid DICL resulted in a flexed scaphoid and widened scapholunate gap only when the wrist was loaded (clinical dynamic scapholunate instability). However, when the lunate DICL was disrupted the resulting instability showed a flexed scaphoid, extended lunate, and widened scapholunate gap in both loaded and unloaded conditions (clinical static scapholunate instability). This previous study  only tested static wrist positions; therefore the purpose of the current study was to test the kinematics of scapholunate instability over a range of flexion/extension wrist positions.
Our results found a significant shift in ligament lengths (up to 60%) after the dorsal SLL was cut. Little length changes were observed after the DICL was cut from the lunate or after detachment of the lunotriquetral ligament from the lunate. This is in contrast with previous studies in the 1970 and 1980's [3,17] that reported that the primary stabilizer of the scapholunate joint is the radioscaphoid ligament and not the SLL. Since then others have reported different kinematic results, and our results agree with those who have modeled ligament injury. Meade  investigated the changes seen on radiographs during sequential sectioning of selected ligaments. They reported the scapholunate space averaged 1.6 mm, smaller than our measurements of 7.0 mm, 4.2 mm, and 5.5 mm in the palmar, membranous, and dorsal portions of the SLL. Differences are most likely due to the 2-dimensional nature of Meade's study and their estimation of the ligament insertion sites.
Short [7,8] investigated SLL instability and reported scapholunate gaps in a slightly different order of ligament cuts. Our membranous SLL results show a 5 to 8 mm increase after the DICL is cut off the lunate and is in agreement with their findings of 3.1 to 7.1mm. Our report supports their finding that as long as the SLL is intact the dorsal radiocarpal ligament and the DICL can be cut without altering scapholunate kinematics.
Many [18–20] have used CT or magnetic resonance imaging to investigate changes in lengths of carpal ligaments in-vivo in non-injured wrists. Upal  reported lengths of 5.1 mm to 6.0 mm for the palmar SLL and 3.5 mm to 4.3 mm for the dorsal SLL. Our study agrees and found lengths of 5.5 mm and 7.0 mm for the palmar SLL and dorsal SLL respectively. They also report that the palmar SLL shortens 29% and elongates 27% and the dorsal SLL shortens 26% and 4% in extension and flexion respectively. Our results did not find similar results (palmar SLL elongated from 18.8 to 13.0 percent and the dorsal SLL elongated from 12.4 and 5.7 percent). Differences are most likely due to their calculation of length change between 60 degrees of extension and 60 degrees of flexion as compared to our calculation from neutral wrist position.
Xu  investigated length changes of 5 ligaments during radioulnar deviation. Our results agree with their calculations in neutral wrist position of 17, 11, 25, and 17 mm for the radioscaphocapitate ligament, long radiolunate ligament, DICL, and dorsal radiocarpal ligament respectively, our results were 26, 17, 28, and 12 mm. Our radioscaphocapitate ligament is slightly longer but may be due to how the paths were calculated.
Lee  reported ligament lengths of the three sub regions of the SLL to be 4.6, 5.7 and 3.0 mm. Our findings were 7.0, 5.5 and 4.2 mm respectively differing only in the dorsal region. Though this 2.5 mm difference is large, it may be in the way that the insertion sites were calculated because our definition of the insertion sites was measured directly for each specimen. Moritomo  investigated the ulnocarpal ligaments and found changes in length from neutral to extension and neutral to flexion for the ulnotriquetral (1.4 and 2.2 mm) and ulnolunate (5.2 and 0.7 mm) ligaments. This is larger than our study (ulnotriquetral ligament; −.49 and −1.7 mm; ulnolunate ligament: −0.1 and 0.4 mm) and may be due to a smaller wrist range of motion tested.
Overall our study supports data reported in recent publications. Limitations of our study are similar to any other biomechanical study that uses cadaver material in that these models may or may not reproduce in vivo kinematics. However, using these techniques, we are able to analyze dynamic carpal motion. In addition, each wrist was compared to itself, reporting a change from the normal condition, providing results only on the effect of the ligament sectioning and reducing any individual specimen differences. We do not report data in extreme wrist flexion and extension. It should be noted that our ligament length measurements are 3-dimensional distances that may not be directly comparable to 2-dimensional radiographic measurements. Finally, we tested ligament section of the scaphoid DICL first then the SLL. It is the clinical author's (SFV) opinion that many pre-dynamic injuries (prior to clinical signs and x-ray findings) are not complete DICL tears and that partial fibers remain resulting in a dynamic clinical presentation. This study cannot provide proof that the scaphoid DICL is a significant stabilizer. However, it is felt that surgical treatment of these types of injury should attempt to repair both the dorsal SLL and the DICL.
Disruption of the membranous SLL, palmar SLL, and the scaphoid DICL did not result in the development of a static collapse as measured by significant differences in ligament length between the scaphoid and lunate. This weakened ligament condition, stabilized by the dorsal SLL, may be indicative of a pre-dynamic instability. However, complete disruption of the SLL did result in the development of an increased scapholunate gap. Further disruption of the lunate DICL did not result in additional gap changes, indicating a possible dynamic and finally a static clinical collapse. Dynamic instability may be complete disruption of all components of the SLL with an intact or partially intact DICL (observations by the clinical author, SFV, intraoperatively). However, this was not tested since the primary goal was to evaluate repair of the SLL as a reasonable surgical goal for repair/reconstruction of SL instability. This study suggests that the SLL and DICL have important roles in stabilizing the scaphoid and lunate, the most important component of which is the dorsal component of the SLL. Surgical treatment of scapholunate instability should include at least the dorsal component of the SLL and perhaps also the DICL at its scaphoid and lunate attachments.
The project was supported by Grant Number 5R01AR049354 from the National Institute of Arthritis and Musculoskeletal and Skin Diseases at NIH.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.