With the prevalence of orthopaedic procedures that alter
load-sharing at the wrist and elbow, it was somewhat surprising
to find a relative lack of cadaveric studies that addressed the
biomechanical consequences of these operations. A better understanding
of how such procedures can affect load transmission through the
forearm can help clinicians to define surgical goals and to select
patients who are likely to have a favorable outcome.
Excision of the radial head is performed to treat comminuted
radial head fractures1-4. Long-term follow-up after radial head
excision has often revealed poor results, with the patient complaining of
loss of wrist strength and/or wrist pain2,3,5. Often these patients
have radiographically demonstrable proximal migration of the distal
part of the radius of 3 mm or more3,5-7. It is believed that this migration
may lead to ulnar-sided wrist pain. We found no cadaveric studies
in the literature that addressed the effects of radial head excision
upon radioulnar load-sharing at the wrist and upon forces in the
interosseous membrane.
The relative lengths of the radius and ulna have been altered
to treat chronic wrist pain due to Kienböck disease8-11. In an attempt
to redistribute loading at the wrist, radial shortening and ulnar lengthening
procedures have been performed on a rather empirical basis, and
not always with good results. Several authors have commented on
the complication of "overshortening" of the radius9,11,12. Ulnar-sided
wrist pain and ulnar impaction syndrome have been described with between
4 and 6 mm of radial shortening9,11,13. Biomechanical studies related
to distal radial shortening or distal ulnar lengthening are infrequent
in the literature14,15.
The objectives of the present study were: (1) to perform baseline
measurements of forces transmitted through the interosseous membrane
and the distal part of the ulna as load was applied axially to the forearm
through the wrist, (2) to repeat the measurements after incremental
levels of distal radial shortening and again after removal of the
radial head, and (3) to study the effects of varus-valgus elbow
alignment and elbow flexion on the recorded measurements in the
intact and surgically altered states.
Twenty fresh-frozen forearms were obtained from fifteen male
and five female donors; the age at the time of death ranged from
fifty-six to eighty-one years. Radiographic measurements of ulnar
variance were within normal limits in all specimens. The midpart
of the humeral shaft was sectioned and potted in a cylinder of polymethylmethacrylate
for gripping in a split clamp. The central three metacarpals were
stripped of soft tissue and potted in a cylindrical mold with polymethylmethacrylate;
the third metacarpal was aligned with the long axis of the cylinder
during potting. Load-cell attachment prongs were cemented into the
proximal part of the radius and the distal part of the ulna as described
in a previous study14 (Fig. 1Fig. 1).
All soft tissues overlying the elbow, forearm, and wrist were
left intact, with the exception of muscle tissue removed through
a dorsal approach to expose the central portion of the interosseous
membrane. With the humerus clamped, the elbow was forced into varus
alignment at 90° of flexion, and a manual axial wrist force was
applied to the potted metacarpals to load the radius and the interosseous
membrane. The most highly tensioned portion of the central band
was identified by direct palpation, and an arthroscopically implantable
force probe (MicroStrain, Burlington, Vermont) was placed into these
fibers. This device, which measures the effects of local ligament-fiber
tension, is placed into a hole punctured into the tissue. Its electrical output
is proportional to force developed within the surrounding tissue
fibers. The force probe was used to indicate when force was developing
in the tissue fibers in the central band and to record relative changes
in local fiber tension in different forearm positions.
Proximal displacement of the radial head relative to the capitellum
was measured with a linear variable differential transformer (Schaevitz
Engineering, Pennsauken, New Jersey). The cylindrical coil of the
transformer was placed into an insert threaded into the lateral
humeral condyle. An extension of the magnetic core rod passed through
the coil into a hole drilled in the lateral epicondyle and through the
subchondral plate of the capitellum such that the rod extension
contacted the center of the radial head perpendicular to its surface
(Fig. 2Fig.
2). Motion between the rod and the fixed coil represented axial
displacement of the radial head with respect to the capitellum.
Radial head displacement was measured at 90° of elbow flexion only;
measurements at 0° flexion were not possible.
The loading apparatus and the technique for aligning the specimen
were described in detail in a previous study14. Briefly, the potted
end of the humeral shaft was mounted on a test fixture attached
to the crosshead of an MTS materials testing machine (model 812;
Minneapolis, Minnesota). The potted metacarpals were clamped in
a cup that could be tilted in two planes to adjust wrist flexion-extension
and radioulnar deviation. This cup was attached to a load-cell mounted
on the hydraulic actuator of the test machine. The hand was fixed
in neutral wrist rotation (the plane of the metacarpals aligned
with the flexion-extension plane of the elbow) and in neutral radioulnar
deviation, and the forearm was oriented vertically, for all loading tests.
The humeral fixture was adjusted to place the humerus in the position
necessary to achieve the desired angle of elbow flexion and the
desired varus-valgus alignment of the elbow. This procedure compensated
for variations in the carrying angle of the elbow among specimens.
Prior to testing, measurements of radial head displacement relative
to the capitellum were recorded between approximately 2 Nm of varus
moment and 2 Nm of valgus moment. These were the limits used to establish
varus and valgus elbow positions for testing. Outputs from the actuator
load-cell, force probe, displacement transformer, and forearm load-cells
were recorded simultaneously as the hydraulic actuator displaced
the wrist proximally at a rate of 1 mm/sec. The wrist was maintained
in a reduced position with manual pressure during testing. As wrist-loading
was initiated, the forearm aligned itself into a stable position.
This normally occurred at a preload of approximately 5 to 10 N.
The maximum wrist force applied during these tests was 200 N. Loading
to this level was not always possible due to potential break-out
of the distal ulnar load-cell fixation prongs. All specimens were
loaded to a minimum of 133.5 N of applied wrist force. After each
group of tests (intact, after radial head excision, and after radial
shortening), repeat testing was performed at 90° of elbow flexion
with varus elbow alignment to verify repeatability.
Prior to testing, a special slotted radial plate was applied
to the distal 6 cm of the radius, approximately 1 cm proximal to
the wrist. A 10-mm section of bone was left between the second and
third screws to allow for excision of bone and shortening of the
radius (Fig. 1Fig.
1). Care was taken to protect the distal radioulnar joint and the
interosseous membrane during the application of the plate. Initial
loading tests were performed at 90° and 0° of elbow flexion, in
both varus and valgus alignment, with the wrist in neutral radioulnar
deviation. Next, radial head excision was simulated by disassembling
the proximal radial load-cell and turning the radial head fragment
anteriorly such that it could no longer contact the radial shaft
and bear load (Fig. 3Fig. 3). The annular ligament remained
intact during this process. All loading tests were repeated. The
radial head was then restored to its anatomical position, and the
prongs cemented into the radial head and radial shaft were reconnected
to the load-cell.
With the slotted plate securely fixed to the bone, an oscillating
saw was used to remove an 8-mm section of the distal part of the
radius between the second and third screws. The distance between
the resected ends of bone was recorded to serve as the reference
length (0 mm of shortening). Distal radial shortening was accomplished
by loosening the screws and adjusting the slotted plate in 2-mm increments
to the desired length. All loading tests with radial head excision
and distal radial shortening were performed at 0° and 90° of elbow
flexion, in varus and valgus elbow alignment, and with the wrist
in neutral pronation-supination and neutral radioulnar deviation.
All shortening procedures were performed while the specimen remained
in the test fixture to eliminate errors due to remounting. The original
elbow alignment was always preserved during tests with a shortened
radius. After all testing was completed, the radius was reset to
0 mm of shortening, and the 90° varus and valgus alignment tests
were repeated.
Statistical Analysis
For calculations of load-sharing, it was assumed that forces
transmitted between the radius and ulna through the capsular ligaments
proximally and distally were negligible. Thus, the sum of the ulnar force
and radial force equaled the applied axial force (both proximally
and distally). The difference between the calculated proximal ulnar
force and the measured distal ulnar force was assumed to be the load
transmitted from the radius to the ulna through the interosseous
membrane. All mean load-sharing values were calculated as a percentage
of the applied wrist force, which was 133.5 N. Mean radial displacements
and force-probe outputs were also calculated at this applied force
level. A two-way analysis-of-variance model with repeated measures
was used to determine the significance of differences in mean load-sharing
percentages between elbow flexion angles, between incremental levels
of distal radial shortening in varus and valgus elbow alignment,
and between status conditions of the radial head (intact compared
with excised) at 90° of flexion. Pairwise comparisons between means
were made with the Student-Newman-Keuls test. The level of significance
was p < 0.05.
Radial Head Excision
In the intact state (load-cells installed and the elbow in 90°
of flexion), the average gap between the radial head and the capitellum
produced by an applied 2-Nm varus moment ranged from 0.75 to 4.77
mm. The mean gap was 1.97 mm (standard deviation, 0.97 mm).
With the intact forearm in 90° of flexion (with the wrist in
neutral position), application of 133.5 N of wrist load resulted
in a mean distal ulnar force of 7.1% of the applied wrist force
with the elbow in valgus alignment and 27.9% of the applied wrist force
with the elbow in varus alignment (Fig. 4Fig. 4). Corresponding forces in
the interosseous membrane averaged 4.0% and 51.2%, respectively.
With removal of the radial head, the mean distal ulnar force increased
to 42.4% of the applied wrist force; this was significantly higher
than the corresponding mean in the intact state, in both varus (p < 0.008) and
valgus (p < 0.001) elbow alignment (Fig. 4Fig. 4). The mean force in the interosseous
membrane after radial head excision was 58.8% of the applied wrist
force. This was significantly higher than the mean in the intact
state with the elbow in valgus alignment (p < 0.0001), but it
was not significantly different from the mean in the intact state
with the elbow in varus alignment (p < 0.08).
The output of the force probe in the intact state was negligible
with the elbow in valgus alignment. The mean force-probe output
following radial head excision was significantly higher than the
mean in the intact state when the elbow was tested in valgus alignment
(p < 0.0001), but it was not significantly higher when the elbow
was tested in varus alignment (p < 0.14, power = 0.78).
Radial Shortening
Since there were no significant differences between the effects
of radial shortening during tests conducted at 0° of elbow flexion
and the effects during tests conducted at 90° of flexion, all results reported
were derived from the tests performed at 90° of flexion (with the
wrist in neutral rotation and neutral radioulnar deviation).
The key factor related to changes observed with radial shortening
was the presence of a gap between the radial head and the capitellum.
This gap was created in two ways during these experiments: by placing
the elbow in varus alignment and by shortening the distal part of
the radius (which acted to pull the radial head away from the capitellum).
The effects of distal radial shortening upon measurements recorded
during the loading tests are illustrated by data for a single specimen
(Figs. 5-AFigs.
5-A and 5-B5-B).
As this specimen was loaded in valgus elbow alignment with 0 mm
of radial shortening, proximal displacement of the radial head was
negligible (since it was contacting the capitellum), force in the distal
part of the ulna was low, proximal radial force increased linearly
from 0 (indicating direct transfer of load through the radius to
the elbow), and force-probe output was 0 (indicating no transfer
of load through the interosseous membrane) (Fig. 5-AFig. 5-A).
Shortening the distal part of the radius by 6 mm decreased the separation
distance between the proximal and distal ends of the radius, thereby
slackening the interosseous membrane and making the wrist ulnar-positive
(the ulna distal to the radius). After 6 mm of shortening, virtually
all of the applied wrist force was transferred to the distal part of
the ulna. The low level of distal radial force displaced the radius
approximately 2 mm, until radial head contact occurred at approximately
130 N of applied wrist force, as indicated by a change in the slope
of the proximal radial load-cell-output curve and by a leveling
off of the radial head displacement curve (Fig. 5-AFig. 5-A).
The 2 mm of proximal radial displacement indicated that the radial
head had been pulled distally away from the capitellum when the
distal part of the radius was shortened by 6 mm. Since the interosseous
membrane had been preslackened by the radial shortening procedure,
force-probe output remained 0 even though the radial head displaced proximally
(Fig. 5-AFig.
5-A).
When this specimen was loaded in varus elbow alignment with 0
mm of shortening, the radial head was not in contact with the capitellum
initially; distal ulnar force remained low, and most of the applied
wrist force was transferred to the radius, which displaced proximally
until radial head contact occurred at approximately 115 N of applied wrist
force, thereby generating force-probe output from the central band
of the interosseous membrane (Fig. 5-BFig. 5-B). Application of additional
wrist force after the point of radial head contact caused proximal
radial force to increase in a linear fashion, with little additional
proximal radial displacement (Fig. 5-BFig. 5-B). The radial head gap due
to varus elbow alignment and 6 mm of radial shortening was so great
that radial head contact never occurred during the loading test.
Nearly all of the applied wrist force was transferred to the distal
part of the ulna, while the remaining portion applied to the distal
part of the radius continued to displace the radius proximally without
generating force-probe output, as a result of the preslackened condition
of the interosseous membrane (Fig. 5-BFig. 5-B).
With valgus elbow alignment, distal radial shortening increased
the mean displacement of the radius and the mean wrist force required
for radial head contact (Table ITable I). Both of these effects
were a direct result of the radial head being pulled distally during
the shortening procedure. With varus elbow alignment, 133.5 N of
applied wrist force was not always sufficient to achieve radial
head contact in the shortened condition. There was radial head contact
in all specimens with 0 mm of shortening, in seventeen of twenty
with 2 mm (p < 0.25 for comparison between 0 and 2 mm), in eight
of twenty with 4 mm (p < 0.04 for comparison between 2 and 4
mm), and in one of twenty with 6 mm (p < 0.02 for comparison
between 4 and 6 mm).
The mean distal ulnar force increased with progressive radial
shortening in both varus and valgus elbow alignment (Fig. 6Fig. 6). After
6 mm of radial shortening, the mean distal ulnar force increased
to 92.4% (in varus alignment) and 60.9% (in valgus alignment). A
linear regression of the mean values yielded a slope of 9.6% increase
in distal ulnar force per millimeter of radial shortening with the
elbow in varus alignment and 9.1% with the elbow in valgus alignment.
The average of the r2 values for linear curve fits for individual
specimens was 0.86 in varus elbow alignment and 0.94 in valgus elbow
alignment. Equal load-sharing between the radius and the ulna at
the wrist occurred with approximately 5 mm of radial shortening
in valgus elbow alignment and with approximately 2 mm in varus elbow
alignment (Fig. 6Fig.
6).
With valgus alignment, forces in the interosseous membrane were
not significantly different from 0 (Fig. 7Fig. 7); no measurable force-probe
output was recorded during any radial shortening test with the elbow
in valgus alignment (p values were greater than 0.16, power = 0.74).
With the elbow in varus alignment, 133.5 N of applied wrist force
was always sufficient to close the initial gap between the radial
head and the capitellum in the intact condition (0 mm of shortening),
thereby loading the interosseous membrane. With progressive radial
shortening, more applied wrist force was shifted to the distal part
of the ulna (Fig. 6Fig.
6) and less force was carried by the progressively slackened interosseous
membrane (Fig. 7Fig.
7). The mean force in the interosseous membrane decreased from 51.2%
(0 mm of distal radial shortening) to 0% (6 mm of distal radial
shortening). The force in the interosseous membrane decreased significantly
with each 2-mm increment of radial shortening (0 compared with 2
mm, p < 0.0001; 2 compared with 4 mm, p < 0.002; and 4 compared with
6 mm, p < 0.001).
This study was performed with fresh-frozen cadaveric specimens
without simulated muscle forces. Our model assumed that the muscle
forces applied to tendons spanning the wrist during grip place the joint
in a state of equilibrium that produces anatomical alignment of
the carpal bones. This anatomical alignment was maintained with
manual pressure during all wrist-loading tests. This methodology helped
to ensure that the wrist was loaded in a reproducible manner. Palmer
and Werner16 loaded the wrist by applying weights to cords sutured
to muscle tendons spanning the wrist. Although this configuration
more closely simulated physiological wrist-loading, the combination
of weights selected to produce standardized wrist-loading was somewhat
arbitrary, and reduction of the wrist during loading was not controlled.
To our knowledge, the present study is the first in which an
implantable force probe was used to measure the tension that developed
within the central band of the interosseous membrane. A post
hoc calibration of force-probe output versus central-band
tension was not performed since it would have required dissociation
of the radius and ulna at the wrist and elbow joints. This would
have destabilized the specimen and necessitated constraining the
two bones in rigid test fixtures, thereby creating an artificial
relative motion pathway that would not have strained the central
band in a physiological manner. Although the direct relationship
between force-probe output and central-band tension was unknown
for this specific tissue, prior calibration studies involving the
anteromedial band of the anterior cruciate ligament have shown that
force-probe output is directly proportional to resultant force in the
tissue into which it is inserted17. Force-probe output in the present
study was most useful for indicating when tension was developing
in the central band and for comparing changes in output when variables
such as elbow alignment and flexion angle were altered.
The custom-designed load-cells used in these experiments were
made as small as possible so that the soft-tissue resection necessary
for insertion was minimal and, more importantly, so that the load-cells
could be placed as close as possible to the sites of measurement
(the distal part of the ulna and the proximal part of the radius).
Under 133.5 N of applied force, the initial gap between the load-cell prongs
closed approximately 0.4 mm. The measured load-sharing percentages
could have been affected slightly by alteration in relative radioulnar lengths
as the load-cells were compressed. Fixation of the load-cell prongs
into the distal part of the ulna and the proximal part of the radius
was another source of axial compliance, especially in specimens with
poor bone quality. Deformation of the proximal radial load-cell
would help to explain why force-probe output with varus elbow alignment continued
to increase even after radial head contact had occurred. Load-cell
compression as well as articular cartilage deformation would both
allow further tensioning of the interosseous membrane with continued
wrist-loading after the point of radial head contact.
Since we used custom-built load-cells for this study, a discussion
of possible sources of error related to their design is appropriate.
The primary output from the implanted load-cell is produced by axial
compression force, which, by virtue of its offset from the strain-gauged
section of the beam element, generates an output voltage from the
simple bending bridge circuit. A second important source of output
voltage from the bending bridge circuit is the bending moment at
the implant site from compression loading of a curved column.
Two types of bending moments at the load-cell installation site
merit consideration: in-plane bending moments, which are of major
concern, and transverse-plane bending moments, which are of minor
concern because of the very low output signal. Although the transformer
output voltage reflects both axial compression and curved beam bending,
our calibration scheme makes it unnecessary to separate these two
modes of loading. Since each bone is loaded during calibration as
it is during an actual test, any bending effects are automatically
incorporated into the linear scale factor that relates load-cell
output signal to applied compression force. If only a bench load-cell
calibration under axial compression were performed (as we have done
to verify load-cell linearity), then in situ load-cell
outputs would be in error as a result of in-plane bending moments
at the installation site. In any case, we believe that bending moments
at the slit cross section were small because (1) the slit cross
sections of the long bones were close to the radial and ulnar heads,
where offset distances to the lines of action of the resultant force
vectors are small (typically less than 0.2 in [5.1 mm]), and (2) the
slit cross sections of both bones were located in relatively straight
regions of the bones, where little curvature is present.
A second source of in-plane bending at the load-cell site could
occur from distributed oblique loading along the midlength portions
of the bones produced by tension generated within the interosseous membrane.
Our results indicate that tension within the interosseous membrane
with the elbow in valgus alignment is low and that wrist load is
essentially transmitted in a straight line up the radius to the
elbow. With varus elbow alignment, however, there is an initial
gap between the radial head and the capitellum, which allows the
radius to displace proximally with respect to the ulna upon application
of wrist load, thereby generating tension within the interosseous
membrane. This produces distributed tensile loading in the midshaft
region of the radius, where the resultant force vector in the interosseous
membrane acts in a direction opposite to that of the applied wrist
load at a slight angle from the long axis of the radius. The line
of action of this resultant force vector in the interosseous membrane
may not pass through the neutral (bending) axis at the installation
site of the proximal radial load cell, thereby generating a bending moment
in the bone at this point.
Using a sample forearm specimen, we estimated the offset distance
of the interosseous membrane load vector from the center of the
slit cross-sectional area. Assuming that all 133.5 N of wrist force is
carried by the interosseous membrane and the interosseous membrane
force vector acts at an offset distance of 0.3 in (7.6 mm) from
the centroid of the slit area, we calculated the load-cell output
voltage produced by an interosseous membrane-generated bending moment
at this point to be 12.7% of the output voltage generated by the
133.5-N axial load. This estimated error probably represents a maximum
because (1) all of the interosseous membrane force was assumed to
be acting in the most sensitive bending plane (which it does not),
and (2) it was assumed that all applied wrist force was acting through
the interosseous membrane (which typically is not the case). We
estimate that a similar maximum error could exist with respect to
the effects of interosseous membrane tension on the distal ulnar
load-cell output. It should be noted that both load-cell installation
sites lie "outside" of the thickened central interosseous membrane
insertion sites on both bones.
To our knowledge, this is the first study in which changes in
load distribution at the wrist after radial head excision were measured
and the first in which calculated changes in interosseous membrane
force after radial head excision were compared with direct measurements
of central-band activity. Load-sharing in the forearm after radial
head excision was similar to that recorded in the varus test condition
(a gap between the radial head and the capitellum). In fact, radial
head excision can be viewed as an extreme type of the varus loading
condition at the elbow (that is, a constant gap at the proximal part
of the radius), in which all distal radial force is transferred
to the proximal part of the ulna through the interosseous membrane.
Severely comminuted fractures of the radial head pose an unsolved
clinical problem, as numerous small fragments often make internal
fixation difficult. Radial head excision and arthroplasty with a silicone
spacer, allograft, or metal prosthesis have had varied results1,5,18.
Even with radial head replacement, patients can have proximal radial migration
on the order of 2 to 3 mm, ulnar-based wrist pain, and weakness1,5,6.
Sowa et al.5 suggested that the success of radial head excision
and joint-leveling procedures is dependent on the integrity of the
central band of the interosseous membrane. Our results support this
view. We found that, when the interosseous membrane was intact,
distal ulnar loads were limited to less than half of the applied
wrist force. If the interosseous membrane were ruptured in conjunction
with an excised radial head, we would expect the entire applied
wrist force to be shifted to the distal part of the ulna, increasing
the likelihood of ulnar-sided wrist pain. Some authors have suggested
that if a patient is found to have a substantial injury of the interosseous
membrane and requires radial head excision, the clinician should
consider immobilization in a long arm cast or pinning of the distal
radioulnar joint in order to promote healing of the interosseous
membrane5,19.
Despite the frequency of surgical procedures for treatment of
Kienböck disease, we found only two cadaveric studies that addressed
changes in forearm load transmission after radial shortening or
ulnar lengthening. Werner et al.15 reported that lengthening of
the ulna by 2.5 mm increased distal ulnar load from 18% to 42% (p < 0.05).
Varus-valgus alignment was not controlled in their tests, and forces
in the interosseous membrane were not calculated. Nevertheless,
the 24% increase in distal ulnar load per 2.5 mm of ulnar lengthening
in the study by Werner et al. roughly corresponds to the 20% increase
in distal ulnar load per 2 mm of radial shortening measured in our
study. Markolf et al.14 found an approximately 60% increase in distal ulnar
force after 6 mm of distal radial shortening (p < 0.05), but
incremental changes in distal ulnar force were not equal for 2-mm
increments of radial shortening. The ratio of an approximately 10% increase
in distal ulnar load per 1 mm of radial shortening found in the
present study held true for both the varus and the valgus condition
and would likely apply to ulnar lengthening as well.
Markolf et al.14 studied load-sharing with the elbow in 45° of
flexion and in valgus alignment and showed that approximately equal
load-sharing at the wrist occurred after 4 mm of radial shortening. This
value closely corresponds to the 5 mm of shortening that produced
equal load-sharing with the elbow in valgus alignment in the present
study. However, Markolf et al. did not assess the effect of shortening
with the elbow in varus alignment, which is when the interosseous
membrane is most active.
One of the most interesting findings of the present study was
the effect of radial shortening upon the function of the interosseous
membrane. In some specimens, 6 mm of radial shortening was difficult to
achieve because of soft-tissue resistance. With valgus elbow alignment,
the interosseous membrane remained unloaded after all amounts of
radial shortening, a finding that is consistent with constant radial
head contact and direct load transfer through the radius. With varus
elbow alignment, the interosseous membrane was loaded with 0 mm
of shortening because of the presence of an initial radial head
gap. An interesting observation was made as the radius was progressively
shortened. The force transmitted through the interosseous membrane
progressively decreased. When the wrist was loaded following 6 mm
of distal radial shortening, the radius displaced proximally (indicating
that some distal radial force was present), but no load was being
carried by the interosseous membrane (a finding confirmed both by
calculation of the force in the interosseous membrane and by direct
force-probe measurements). The only way for this to occur would
be if the interosseous membrane had been preslackened by the radial
shortening procedure. We believe that most of the change in relative length
(and hence slackening of the interosseous membrane) produced by
the shortening procedure was achieved by pulling the distal part
of the radius proximally away from the carpus, not by pulling the radial
head away from the capitellum. This conclusion is based upon the
observation that 6 mm of distal radial shortening in valgus alignment
increased mean proximal radial displacement by only 0.5 mm. The
remaining mean 5.5 mm of change in the relative length must have
occurred at the wrist. Although radial shortening and ulnar lengthening may
have similar clinical outcomes, they have markedly different effects
in terms of the load-transmission function of the interosseous membrane
when the forearm is in varus alignment. It is unknown whether these
differences translate into clinical superiority of one procedure
compared with the other.
Although the precise degree of distal radial off-loading is not
always clear in the surgical treatment of Kienböck disease, equal
load-sharing at the wrist might be a reasonable goal. If equal load-sharing
is to be achieved, should it be equal in varus or valgus elbow alignment?
Nakamura et al.9,13 and Weiss et al.11 both reported the development
of ulnar-sided wrist pain necessitating additional surgery in forearms
that had been ulnar-neutral. Our results suggest that with the elbow
in varus alignment, 5 mm of shortening would result in distal ulnar
loading of approximately 80%, which could produce ulnar-sided wrist
pain. With the elbow in valgus alignment, 2 mm of shortening would
result in a distal ulnar load of approximately 20% and perhaps not off-load
the radius enough. Weiss et al. suggested that 2 mm of shortening,
regardless of the ulnar variance, should create enough relative
unloading to alleviate symptoms. When planning a shortening procedure,
the surgeon should measure the initial ulnar variance and consider
that 1 mm of radial shortening will increase distal ulnar load by approximately
10%, regardless of elbow alignment.