Keywords
arthrokinematics - contact - distal radioulnar joint - intercartilage distance - triangular
fibrocartilage complex
Distal radius fractures are the most common type of upper extremity fracture in the
United States.[1] Factors such as osteopenia, comminution, age over 60 years, and a high degree of
initial displacement may predispose these to malunion.[2] Residual dorsal angulation is the most common deformity, and the consequences of
this have been the most widely studied. Specifically, residual dorsal angulation can
alter forearm mechanics, with effects on both the range/axis of forearm rotation[3]
[4]
[5] and torque required for prosupination.[6]
[7] Moreover, increased dorsal angulation may cause dorsal intercalated segmental instability[8] and change the excursion and moment arms of the wrist muscles.[9]
[10]
The consequences of distal radius malunion on the distal radioulnar joint (DRUJ) have
been the subject of further enquiry. Dorsal angulation of the distal radius has a
significant effect on the DRUJ, causing incongruity,[11] instability,[12] and abnormal load transfer across the joint.[13] Persistent disability from malunion has been observed clinically, with symptoms
including ulnar-sided wrist pain, deformity, restricted forearm rotation, and limitations
in grip strength.[14]
[15] Dysfunction related to these may be exacerbated in the setting of associated triangular
fibrocartilage complex (TFCC) rupture and DRUJ instability.[16] These symptoms may, in part, relate to the biomechanical effects of distal radius
malunion on the DRUJ.
Arthrokinematics, or the specific movement of joint surfaces,[17] are not well understood for the DRUJ in the setting of distal radius malunion. Using
in vivo methods, previous authors have documented a reduction in the contact area
between the ulnar head and sigmoid notch with malunion.[18]
[19] In vivo methods use live subjects with multiplanar distal radius deformities of
variable severity. In vitro techniques use cadaveric specimens and allow for individual
deformities to be isolated and different conditions to be simulated, such as TFCC
rupture. This permits a categorical analysis of the effects of each parameter on the
arthrokinematics of the DRUJ.
Accurate indirect measurement of joint contact can be achieved using in vitro techniques.
Intercartilage Distance (ICD) is one such technique, which utilizes computed tomography
(CT)-based bone and cartilage models, fiducial-based registration, and optical tracking
motion capture data.[20] It has been used previously to characterize DRUJ contact in the intact state.[21]
The purpose of this in vitro study was to utilize ICD to examine the effects of dorsal
angulation deformity on DRUJ contact patterns throughout simulated active forearm
rotation. Our hypothesis was that the contact area would decrease with progressive
dorsal angulation, and that the centroid of contact would become more volar and distal
in the sigmoid notch with increasing deformity. We also hypothesized that simulated
TFCC rupture would decrease contact area at the sigmoid notch and increase the variability
of the contact path of the centroid.
Methods
Specimen Preparation
Testing was performed on eight cadaveric forearm specimens (mean age 60 years; range
29–75 years; six men and two women) with no CT evidence of osteoarthritis. We established
that this sample size would provide a power of 80% to detect changes of 0.8 mm in
centroid position and 20 mm2 in contact area at the 0.05 confidence level. The distal tendons of the wrist extensors
(extensor carpi radialis longus [ECRL], extensor carpi ulnaris [ECU]), wrist flexors
(flexor carpi radialis [FCR], flexor carpi ulnaris [FCU]), pronator teres [PT], and
biceps [BIC] were then sutured using #2 Ethibond (Ethibond Excel, Ethicon Inc., Piscataway,
NJ).
Sutures were passed through alignment guides that reproduced the physiologic line
of action of each muscle. ECRL and ECU were routed through a lateral epicondyle sleeve,
while PT, FCR, and FCU were routed through a medial epicondyle sleeve. The supinator
[SUP] was modeled by placing a suture anchor in the radial tuberosity and routing
the attached suture through a Delrin sleeve which traversed the supinator crest to
the posterolateral aspect of the ulna. The sutures of ECRL, ECU, FCR, FCU, and SUP
were attached to individual pneumatic actuators (Airpot Corporation, Norwalk, CT).
Simulation of Motion
A servo motor (SM2315D; Animatic, Santa Clara, CA) was used to simulate active motion,
with a resistive counterforce provided by a pneumatic actuator. Active supination
was initiated by attaching BIC to the servo motor set to motion control at a constant
tendon velocity of 5 mm/s. The muscles were loaded using ratios based on a previous
investigation of forearm muscle EMG and cross-sectional area.[22] Constant tone loads of 10 N were applied to the FCU, FCR, ECU, and ECRL. Simultaneous
pneumatic actuator loads were regulated by proportional pressure controllers (MAC
Valves, Wixom, MI) under computer control using custom programmed software (LabVIEW,
National Instruments, TX).
Motion Tracking and Kinematic Data Acquisition
Infrared marker triads were rigidly affixed to the proximal radius and ulna using
custom Delrin pedestals and the arc of simulated active supination was tracked using
an Optotrak Certus (Northern Digital Inc, Waterloo, Ontario, Canada) optical motion
capture system[23] ([Fig. 1]).
Fig. 1 Depicting a cadaveric specimen mounted in a custom forearm motion simulator. The
humerus and ulna are rigidly secured. The outrigger stabilizes a third metacarpal
pin holding the radiocarpal joint in a neutral position. Optical tracking markers
are mounted on Delrin posts affixed to the radius and ulna. Pneumatic actuators and
the servo motor are attached to a Delrin base.
Simulation of Distal Radius Deformity
A previously described, custom-engineered adjustable implant was applied to the volar
aspect of the distal radius for each specimen.[6] This permitted the creation of simulated dorsal angulation deformities. The central
appliance of the implant was removable and exchanged for each deformity condition.
To install the device, a 20-mm corticocancellous segment of volar distal radius was
removed 2 mm proximal to the DRUJ using an oscillating saw. The dorsal cortex was
initially left intact as a bone bridge. Medullary bone from the distal radius metaphysis
and shaft was curetted away and cavities were filled with polymethylmethacrylate cement.
The adjustable implant was then fixated using bone screws in a neutral position ([Fig. 2]).
Fig. 2 The custom adjustable implant is inset into the distal radius osteotomy with a dorsal
intact bone bridge. Depicted is a schematic and clinical photo, with the implants
fixation augmented by intramedullary cement.
Four different deformity conditions were tested: no deformity (Straight Wedge [SW]),
dorsal angulation of 10 degrees (DA10), 20 degrees (DA20), and 30 degrees (DA30).
The SW configuration of the adjustable implant kept the proximal and distal radius
fragments in their original anatomic alignment, while the dorsal angulation configurations
introduced progressive dorsal tilt of the articular surface relative to the original
anatomy ([Fig. 3]).
Fig. 3 Depicting the four different deformity conditions including the straight wedge, dorsal
angulation of 10 degrees (DA10), 20 degrees (DA20), and 30 degrees (DA30). Note that
the deformities are angulated relative to the original anatomy and do not represent
the absolute dorsal angulation value as would be measured on a conventional lateral
radiograph.
Testing Procedure
The specimens were kept hydrated throughout testing using 0.9% normal saline and closure
of the skin envelope between implant exchanges. Kinematic data were gathered with
the implant in the neutral (SW) position and for the dorsal angulation deformities
with the TFCC intact. Once testing of the intact state had concluded, the TFCC was
divided off its ulnar insertion ([Fig. 4]).
Fig. 4 A photo of the sectioned TFCC, with no residual fibers inserting on the ulnar styloid
or fovea. TFCC, triangular fibrocartilage complex.
Subsequently, all deformity testing was repeated for the TFCC insufficient state ([Fig. 4]). At the conclusion of the testing protocol, the forearm was dissected and the bones
were denuded of soft tissue. Landmarks on the distal radius, implant, proximal radius,
and ulna were digitized relative to the attached motion trackers. This permitted the
creation of a three-dimensional anatomic coordinate system; therefore, the kinematic
data could be transformed to describe the position of the radius relative to the ulna.
Intercartilage Distance Measurement Technique
[Fig. 5] provides a flowchart summarizing the stages of data processing which follow kinematic
data acquisition to create an ICD measurement. Steps are included from both the experimental
phase and from the volumetric data acquisition phase using the denuded specimens.
Fig. 5 A flowchart detailing the stages of postexperiment data processing for application
of the Intercartilage Distance algorithm.
Data Analysis
All eight specimens were used for ICD contact analysis. The ICD algorithm was used
to generate a contact patch and contact centroid for every 10 degrees interval of
forearm rotation. The optical tracking system was unable to capture the extremes of
forearm rotation due to loss of tracker visualization; therefore, an arc from −60
(60 degrees of supination) to +40 (40 degrees of pronation) was analyzed.
Centroid coordinate data from eight specimens was also evaluated. An anatomical coordinate
system was assigned to the sigmoid notch of the distal radius, with a point designated
as its center. Contact centroid position relative to the sigmoid notch center was
then calculated in mm, for both the proximal–distal (X) and volar–dorsal (Y) axes.
The effects of forearm rotation angle, distal radius deformity, and TFCC sectioning
on DRUJ contact area and contact centroid position were evaluated. A three-way repeated
measures analysis of variance (ANOVA) was performed, with independent variables of
forearm rotation angle, distal radius deformity, and TFCC condition.
To determine if the centroid pathways were more variable after TFCC sectioning, the
standard deviation values for each 10 degrees interval of forearm rotation were compared
using a one-way repeated measures ANOVA for matched deformities. Both the proximal–distal
and dorsal–volar axes were assessed.
Data imputation using a linear regression model was used to reconstitute missing contact
area and centroid coordinate values. A Greenhouse–Geisser's correction was applied.
Statistical significance was set at p < 0.05. Data presented are the mean DRUJ contact area ± standard deviation unless
otherwise specified. We used a Bonferroni correction for multiple comparisons to compare
main effects.
Results
There was no significant effect from deformity on contact area in the DRUJ (p = 0.30). Forearm rotation angle had a significant effect on contact area (p = 0.004), with measurements being highest between 10 and 30 degrees of supination.
TFCC sectioning caused a significant decrease in contact area in the DRUJ (p = 0.030), with a mean reduction of 11 ± 7 mm2 between the TFCC intact and sectioned conditions across all variables ([Figs. 6] and [7]).
Fig. 6 Depicting the mean +1 SD of DRUJ contact area for the normal condition (SW) and with
an increasing degree of dorsal angulation deformity (DA10/20/30). Measurements were
made at 10 degrees intervals of forearm rotation, from 60 degrees of supination to
40 degrees of pronation. DRUJ, distal radioulnar joint; SD, standard deviation; SW,
straight wedge.
Fig. 7 Depicting the mean + 1 SD of DRUJ contact area after TFCC sectioning, for the normal
condition (SW) and with an increasing degree of dorsal angulation deformity (DA10/20/30).
Measurements were made at 10 degrees intervals of forearm rotation, from 60 degrees
of supination to 40 degrees of pronation. DRUJ, distal radioulnar joint; SD, standard
deviation; TFCC, triangular fibrocartilage complex.
The position of the contact centroid along the volar–dorsal axis moved volarly with
supination for all variables (p < 0.001). Deformity had a significant effect on the location of the contact centroid
along this plane (p = 0.043). Relative to the SW position, the mean centroid position moved 0.3 ± 1 mm
volar in 10 degrees of dorsal angulation, 0.1 ± 0.9 mm volar in 20 degrees of dorsal
angulation, and 0.6 ± 0.9 mm volar in 30 degrees of dorsal angulation. There was no
effect from sectioning the TFCC on the volar–dorsal position of the centroid (p = 0.24). Variability of the centroid pathway was significantly increased along the
volar–dorsal axis after TFCC sectioning (p < 0.001), with a 16% increase in the magnitude of standard deviation values for each
angle of forearm rotation across deformities.
The position of the contact centroid along the proximal–distal axis moved proximally
with supination for all variables (p = 0.043). Deformity did not have a significant effect on the location of the contact
centroid along this plane (p = 0.17). There was no effect from sectioning the TFCC on the proximal–distal position
of the centroid (p = 0.21). Variability of the centroid pathway was significantly increased along the
proximal–distal axis after TFCC sectioning (p = 0.004), with a 50% increase in the magnitude of standard deviation values for each
angle of forearm rotation across deformities ([Figs. 8] and [9]).
Fig. 8 The position of the contact centroid on the face of the sigmoid notch during forearm
rotation. Mean centroid position is displayed for TFCC intact specimens. TFCC, triangular
fibrocartilage complex.
Fig. 9 The position of the contact centroid on the face of the sigmoid notch during forearm
rotation. Mean centroid position is displayed for TFCC sectioned specimens. TFCC,
triangular fibrocartilage complex.
Discussion
This study demonstrated that contact area in the DRUJ is variable and dependent on
the angle of forearm rotation. Contact area was maximal between 10 and 30 degrees
of supination during the conditions tested. These findings are consistent with the
literature, with reports indicating that the highest DRUJ contact area values occur
across 10 to 30 degrees of supination.[24]
[25]
[26] We noted that the contact centroid on the sigmoid notch moved volarly and proximally
with progressive supination. This was also expected and is in agreement with the published
literature on DRUJ kinematics[27]
[28]
[29]
[30] and contact.[21]
[31]
Simulated malunion with dorsally angulated distal radius deformities influenced DRUJ
contact. Increasing dorsal angulation caused the contact centroid to move progressively
more volar in the sigmoid notch. This was in keeping with our hypothesis, and relates
to the distal radius being dorsally displaced relative to the ulnar head during forearm
rotation.[3]
[32]
We found no correlation between the amount of simulated distal radius deformity and
contact area in the DRUJ. This finding was unexpected, given the sensitivity of this
technique for subtle contact area changes[20] and the known effects of dorsal angulation deformity on DRUJ biomechanics.[6]
[7]
[12]
[18]
[32]
[33]
[34] It is possible that DRUJ contact area does not change with progressive dorsal angulation
of the distal radius. Alternatively, the lack of difference in our study may relate
to the arc of motion studied (60 degrees of pronation to 40 degrees of supination).
Other authors have noted the greatest effect of deformity at the extremes of forearm
rotation, with limitations in pronation[3] and supination[7] beyond 50 degrees of rotation. It is also possible that no difference from deformity
was observed because of the type of deformity tested. Previous authors have noted
more significant kinematic changes from combined deformities[6] or shortening compared with dorsal angulation.[33] Finally, we may have been underpowered with a small sample size to show a statistically
significant difference on contact area between deformity groups.
Our findings are interesting to contrast to in vivo studies of the DRUJ in the setting
of distal radius malunion.[18]
[34] Crisco et al[18] noted that deformity had a significant effect on interbone joint spacing area (their
proxy for joint contact area), but that forearm rotation angle had no effect. They
demonstrated less contact in malunited wrists with a contact centroid which moved
more proximally. This was in contrast to our findings, which showed a significant
effect of forearm rotation angle on contact area, and did not demonstrate a change
in DRUJ contact area with deformity. Moreover, unlike Crisco et al,[18] we noted no change in the position of the contact centroid along the proximal–distal
axis with deformity, but did find that it displaced volarly with progressive dorsal
angulation. Their values for absolute contact area in normal were also significantly
higher than in our study and those documented in the DRUJ by other authors using Tekscan.[24]
[25]
[26] There are multiple reasons that could explain the discrepancies: (1) ICD is more
accurate than interbone distance as the true cartilage thickness is accounted for
in the bone–cartilage model, compared with interbone distance where an arbitrary number
is used to create the proximity map. (2) Their technique uses multiple static positions
to extrapolate kinematics; thus, pathways depicting forearm motion may not be entirely
accurate. (3) They were evaluating multiplanar deformities which included shortening,
as opposed to isolated dorsal angulation deformities as in our study. (4) Their measurements
are based on a live population who has an almost complete active range of motion despite
their chronic deformity. In vitro specimens are unable to compensate their soft tissue
compliance for increasing levels of deformity.
Our study also examined the effect of the TFCC on contact area in the DRUJ. We demonstrated
a significant effect of simulated TFCC rupture on contact area in the DRUJ, with a
mean contact reduction of 11 ± 7 mm2 after sectioning. This was to be expected, as once TFCC failure occurs, forces across
the DRUJ relax considerably.[13] It is generally believed that the TFCC complex constrains the DRUJ up to a certain
limit in the setting of distal radius deformity. Some authors have experienced that
only moderate deformities can be reproduced with an intact TFCC complex.[3]
[35]
[36] Sectioning of the TFCC allows for more extreme malpositions to be achieved[6] with decreased torques to achieve full prosupination.[7] In light of the above, it is interesting then that no difference was found from
TFCC sectioning on contact centroid position. In our study, after TFCC sectioning,
there was a 16% increase in the magnitude of standard deviation values for the contact
centroid position along the dorsal–volar axis, and a 50% increase along the proximal–distal
axis. This implies a dramatic increase in the variability of the contact centroid
pathway after sectioning of the TFCC. This variability likely explains why no significant
difference was found.
The limitations of this study include the inability for cadaveric specimens to undergo
soft tissue adaptation, unlike the in vivo condition. Moreover, the results are less
generalizable because only uniplanar deformity was tested, while malunion is usually
composed of combination of shortening, angulation, translation, and rotation. The
advantage of an in vitro method for studying contact area, compared with in vivo methods,
is that test parameters are better controlled and effects of individual deformities
can be isolated. Fewer assumptions are made for changes in kinematic pathways, as
testing occurs continuously throughout an arc of motion. Finally, the cartilage models
created from specimens CT scanned in air create excellent cartilage definition and
more accurate models.
In conclusion, increasing dorsal angulation deformity has no apparent effect on contact
area in the DRUJ, but causes the contact centroid position to displace volarly. Simulated
TFCC rupture reduces the DRUJ contact area and significantly increases the variability
of the contact centroid pathway during forearm rotation. Future directions include
testing other deformities, including dorsal translation, combined deformities, and
volar deformities to increase the generalizability of the results.
