Multiple osteochondromas frequently affect the distal aspects of the radius
and ulna and can result in severe deformity of the wrist and forearm. Such
deformity is caused by a combination of shortening of the ulna, bowing of the
radius, and ulnar deviation of the wrist, and, occasionally, radial head
dislocation1. Radial
articular deformity in association with multiple osteochondromas (an
increasing radial articular angle) may be induced by tethering of the
shortened distal aspect of the ulna, as first proposed by Solomon in
19612, and
subsequently supported by several other
studies3-5.
Solomon also theorized that relative ulnar shortening was accommodated by
either bowing of the radius or dislocation of the radial head proximally.
However, more recently Burgess and Cates performed a radiographic evaluation
of a forearm deformity in thirty-five patients (sixty-five forearms) with
multiple osteochondromas and showed that ulnar shortening was not related to
subsequent shortening of the radius or to distal radial deformity, providing
evidence against the ulnar tether
theory6.
Alternatively, their results suggest that both the radius and the ulna exert
proportional suppression of growth, and the loss of radial length is caused by
involvement of the radius in the pathological process, and not by an ulnar
tether. This speculative argument was supported by Ogden on the basis of a
histological study of an osteochondroma in the fibula of a three-year-old
girl, which showed that the osteochondroma invaded the epiphysis and had the
potential to further invade the growth
plate7. Masada et
al. classified forearm deformities that were due to multiple osteochondromas
and reported the outcome of several surgical procedures; these data showed
that the bulk of osteochondromas themselves plays an important role in
inducing radial
bowing8. Therefore,
forearm deformity may be the result of various etiologic factors associated
with multiple osteochondromas (ulnar shortening, growth disturbance, and/or
compression by the tumor itself).
Tumor excision is the most common surgical treatment, and this procedure
can be combined with other surgical methods, including ulnar lengthening,
radial osteotomy, and hemiepiphysiodesis, depending on the severity and
pattern of the
deformity3-5,8-13.
However, the variety of combinations of surgical procedures has made it
difficult to assess the effectiveness of each procedure for the correction of
the deformity, leading to a long-standing controversy regarding the overall
effectiveness of the excision of osteochondromas in controlling the
progression of a deformity. Masada et al. suggested that a simple excision can
prevent the progression of the disease and is effective in controlling radial
bowing8; however,
Fogel et al., in a study of simple excision in ten patients, found that early
excision of the osteochondroma alone did not improve the wrist deformity,
although the procedure did reduce the rate of progression of ulnar shortening
compared with the preoperative
rate10.
The purpose of the current study was to investigate the effect of simple
excision of osteochondroma(s) on the correction of the deformity, on the basis
of the changes in radiographic parameters. Various patterns of deformity were
classified radiographically according to the location of the tumor, and these
were related to the etiologies of the deformity. We hypothesized that a
forearm deformity can be defined by the tumor location and that the type of
deformity influences the effectiveness of simple excision for the correction
of the deformity.
Fifty-seven surgical procedures in twenty-five patients (thirty forearms)
with multiple osteochondromas were performed in our department between 1983
and 2005. Excision of osteochondromas was performed in all patients.
Accompanying procedures included ulnar lengthening (eighteen forearms), radial
osteotomy (two forearms), and radial hemiepiphysiodesis (seven forearms). The
ulnar lengthening was usually performed when the radial head was dislocated
proximally in association with severe ulnar shortening, and radial osteotomy
or radial hemiepiphysiodesis was performed when the lunate had slipped beyond
the ulnar edge of the distal part of the radius with increased radial
inclination of the distal end of the radius. The purpose of this study was to
assess the effectiveness of tumor excision for the correction of a deformity;
therefore, fourteen forearms in thirteen patients who had a simple tumor
excision as an isolated procedure and were followed for more than twenty-four
months were included in the study (Table
I). Informed consent was obtained preoperatively from the patients
and parents on the basis of the surgical indication as determined at our
institution, and the patients returned for follow-up evaluation
postoperatively. The patients included six boys and seven girls. In one
patient, tumors were excised from both forearms. The average age at the time
of surgery was 7.9 years (range, four to fourteen years), and the operation
was performed when the patients were less than ten years old, except for two
children who had surgery when they were twelve and fourteen years old. The
average follow-up period was fifty-three months (range, twenty-four to
ninety-seven months).
The forearm and wrist deformity were evaluated with use of plain
anteroposterior radiographs according to the method reported by Burgess and
Cates (Fig.
1)6. In
this method, a linear axis is determined by a line connecting the ulnar
borders of the distal and proximal physeal plates of the radius. Ulnar
shortening is measured as the distance between the intersection of a line
drawn perpendicularly at the level of the distal end of the ulna to the linear
axis and the ulnar border of the distal physeal plate of the radius; the
percentage of ulnar shortening is then calculated by dividing ulnar shortening
by the length of the linear axis of the ulna. The radial articular angle is
the angle between the linear axis of the radius and a line drawn along the
distal articular surface of the radius. Radial bowing is defined as the
greatest distance from the radial diaphysis to the linear axis of the radius
and is expressed as the percentage of radial bowing, which is determined by
dividing the radial bowing by the length of the linear axis. Carpal slip is
determined by measuring the distance between the extreme ulnar side of the
lunate and a continuation of the linear axis of the ulna, with this distance
expressed as a percentage of the total length of the lunate. The extreme ulnar
side of the lunate is defined as the contact point of a line drawn parallel to
the linear axis of the lunate. In two four-year-old patients (Cases 4 and 6),
the lunate had not yet ossified and carpal slip was excluded from the
evaluation.
Two patterns of deformity were defined on the basis of the location of the
tumor. In the six forearms in Group 1, the osteochondroma was present only at
the junction of the metaphysis and diaphysis of the distal aspect of the ulna,
and it resulted in compression of the radius, which was mainly reflected in
altered radial bowing (Figs. 2-A and
2-B). The deformity of the distal end of the radius in this group
was usually minimal. In the eight forearms in Group 2, the tumors were present
at both the distal end of the ulna and the ulnar side of the radius, and they
were in contact with each other (Figs. 3-A
and 3-B). The percentage of ulnar shortening and the percentage of
radial bowing were less severe in Group 2 compared with Group 1. No patient in
either group had dislocation of the radial head; however, one patient (Case 4)
in Group 1 showed lateral subluxation of the radial head, with severe ulnar
shortening.
Osteochondromas were excised completely under radiographic image control
because there was no clearly identifiable border between the base of the tumor
and the normal bone. Usually the normal bone was partly excised at the base of
the tumor with great care taken to not injure the growth plate. In Group 1,
the osteochondroma of the distal part of the ulna was approached between the
extensor carpi ulnaris and the flexor carpi ulnaris. In Group 2, the
osteochondroma of each bone was excised through two separate skin incisions in
order to reduce the possibility of creating a synostosis between the distal
end of the radius and the ulna after the
excision12. The
osteochondroma on the ulnar side of the distal part of the radius was exposed
through a radiopalmar skin incision with detachment of the pronator quadratus
from the radius.
We compared the groups in terms of the preoperative radiographic parameters
using the Mann-Whitney test. To evaluate the effect of simple tumor excision
on the correction of the deformity, the radiographic parameters before surgery
and at the time of the final follow-up were compared with use of the paired t
test and the groups were compared, with use of the Mann-Whitney test, with
respect to the changes in each parameter after surgery. A p value of 0.05 was
considered to be significant.
Preoperative Findings
Preoperative and postoperative radiographic parameters are shown in
Table I. Groups 1 and 2 were
compared with respect to these parameters. The average percentage of ulnar
shortening, percentage of radial bowing, radial articular angle, and
percentage of carpal slip were 8.3%, 10.2%, 30.5°, and 70.3%,
respectively, in Group 1 and 5.4%, 6.9%, 34.5°, and 71.6%, respectively,
in Group 2. Before surgery, the percentage of ulnar shortening and percentage
of radial bowing were more severe in Group 1. A significant difference was
recognized in the percentage of radial bowing (p = 0.010). There were no
apparent differences in radial articular angle or carpal slip between Groups 1
and 2.
Postoperative Findings
There were no significant differences between Groups 1 and 2 in terms of
age at the time of surgery (average, 6.9 and 8.7 years, respectively) or
length of follow-up (fifty and fifty-six months, respectively).
The average changes in the radiographic parameters for all thirteen
patients from before surgery to the final follow-up were 6.7% to 6.1% of ulnar
shortening, 32.8° to 37.2° in the radial articular angle, 8.3% to 7.7%
of radial bowing, and 71.1% to 64% of carpal slip. With this small number of
patients, none of these changes were significant.
Changes in Radiographic Parameters After Surgery
Ulnar Shortening (Fig.
4)
The average percentage of ulnar shortening improved from 8.3% before
surgery to 6.5% at the time of the final follow-up in Group 1 (p = 0.047),
whereas no change was detected, with the numbers studied, in Group 2 (5.4% to
5.8%, respectively). The difference between the two groups with respect to the
average change in the percentage of ulnar shortening (—1.8% in Group 1
and +0.4% in Group 2) after surgery was significant (p = 0.039). In Group 1,
improvement of >2% was obtained in three patients (Cases 2, 3, and 4)
(Figs. 2-A through 2-D) and two
patients (Cases 1 and 5) showed almost no change in the percentage of ulnar
shortening. In contrast, one patient in Group 2 (Case 13) showed a
deterioration in the percentage of ulnar shortening of >2%.
Radial Bowing (Fig.
5)
The mean percentage of radial bowing in Group 1 improved significantly from
10.2% before surgery to 7.3% at the time of the final follow-up (p = 0.010).
Conversely, the mean percentage of radial bowing in Group 2 showed significant
deterioration from 6.9% before surgery to 8.0% at the time of the final
follow-up (p = 0.017). The difference between the groups with respect to the
average change after surgery (—2.9% in Group 1 and +1.1% in Group 2) was
also significant (p = 0.002).
Radial Articular Angle
(Fig. 6)
At the time of the final follow-up, the mean radial articular angle had not
changed in Group 1 (30.5° before surgery to 30.0° at the time of the
final follow-up), but it had significantly deteriorated from 34.5° to
42.6° in Group 2 (p = 0.049). The difference between Groups 1 and 2 with
respect to the mean change after surgery (—0.5° and +8.1°,
respectively) was not significant (p = 0.106). In four patients (Cases 7, 8,
11, and 12) in Group 2, the radial articular angle had increased by
=10° at the time of final follow-up compared with the preoperative
value.
Carpal Slip (Fig.
7)
The average carpal slip in Groups 1 and 2 was 70.3% and 71.6%,
respectively, before surgery and 61.1% and 65.4%, respectively, at the time of
the final follow-up. In both groups, the average carpal slip showed slight
improvement; however, in two patients (Cases 7 and 8) in Group 2, carpal slip
increased severely and reached values of >100% with accompanying increases
in the radial articular angle at the time of the final follow-up
(Fig. 3-B).
Recurrence of the Tumor
Some degree of tumor recurrence was recognized on the distal aspect of both
the ulna and the radius in five patients (Cases 7 and 8 and, on the left side,
in Cases 10, 12, and 13) in Group 2 and on the distal aspect of the ulna in
two patients (Cases 2 and 6) in Group 1 at the time of the final follow-up.
Recurrence was more frequent in the patients with the longer follow-up period.
However, there were no clear differences in the radiographic parameters after
surgery between the patients with and those without tumor recurrence in either
group.
In our patients with osteochondromas of the distal aspect of the forearm,
the pattern of deformity differed according to the location of the tumor. In
Group 1 (with the tumor in the metaphyseal-diaphyseal area and located far
from the growth plate of the ulna), the percentage of ulnar shortening and the
percentage of radial bowing were worse than those in Group 2, leading to
compression of the radius. Radial bowing in the patients in Group 1 was,
therefore, thought to be partly due to compression of the radius caused by the
tumor, as well as to relative ulnar
shortening8. The
radial articular angle was usually minimal in the patients in Group 1,
although ulnar shortening was severe; this is inconsistent with the ulnar
tether theory and in agreement with the conclusions of Burgess and
Cates6. Simple
excision of the tumor resulted in improvement in the relative amounts of ulnar
shortening and radial bowing in Group 1, indicating that radial bowing can be
expected to correct by normal remodeling of the bowed radius, as well as
through recovery of the ulnar growth rate after removal of the tumor,
especially in younger patients with a greater potential for remodeling. In
Group 1, there appeared to be little tethering effect between the distal
aspects of the ulna and the radius; therefore, simple tumor excision should be
enough to correct the forearm deformity when there is no dislocation of the
radial head proximally.
The percentage of ulnar shortening in the patients in Group 2 was less
severe than that in Group 1, which may reflect a growth disturbance of the
distal aspect of the radius due to involvement by the tumor. Burgess and Cates
described two patients with an increased radial articular angle in conjunction
with positive ulnar
variance6, in whom
the radiographic findings showed the tumor to be on the ulnar side of the
distal part of the radius. Masada et al. classified forearm deformity
according to the location of the
osteochondromas8. In
that classification system, Type-III deformities are those in which the tumor
is in the distal aspect of the radius and is accompanied by relative radial
shortening.
The effectiveness of different surgical approaches to correct a forearm
deformity due to multiple osteochondromas was reported by Fogel et
al.10 and included
the interesting case of a patient who had the same type of deformity as those
in Group 2 in our study; that is, the tumors were on both the ulnar side of
the distal aspect of the radius and the distal part of the ulna and were in
contact with each other. After observation of that patient for eight years,
the deformity of the wrist had progressed and radial hemiepiphysiodesis was
performed to correct the deformity by retarding growth on the radial side of
the distal part of the
radius10. However,
this procedure also resulted in a difference in forearm length of 2.3 cm
compared with the contralateral forearm. In contrast to the findings in that
patient, Wood et al. stated that if an osteochondroma is located on the
radius, or if osteochondromas from both the radius and the ulna push against
each other, only minimal deformity
occurs5.
In the current study, the percentage of ulnar shortening and the percentage
of radial bowing were less severe in the patients in Group 2 than in those in
Group 1, which is consistent with the findings of Wood et
al.5. However, we
suspect that the ulnar aspect of the distal part of the radius is more
susceptible to growth disturbance compared with other parts of the radius.
When tumors are on both the ulnar side of the distal part of the radius and
the distal aspect of the shortened ulna and are in contact with each other, a
tethering effect between the tumors might occur in a longitudinal direction
and result in an increasing radial articular angle and radial bowing with the
passage of time because of a growth disturbance of the ulnar side of the
distal aspect of the radius. Solomon described a reason for the valgus
deformity of the distal part of the tibia when osteochondromas develop in both
the distal part of the tibia and the distal aspect of the
fibula2. He
concluded that the valgus deformity was caused by a tethering effect of the
shortened fibula and not by uneven growth of the tibial physis because the
distal tibial physis remained in its normal horizontal orientation. However,
in our patients in Group 2, the distal radial physis showed ulnar tilting
(Fig. 3-B), which suggests a
growth disturbance of the ulnar side of the distal radial physis.
In our study, simple excision of the tumor was unable to correct the
deformity in the patients in Group 2, and the radial articular angle and the
percentage of radial bowing showed significant increases. It is difficult to
conclude whether simple tumor excision in the patients in Group 2 was
effective for the correction of the deformity because of the high recurrence
rate of the tumor in five of the eight forearms. Such recurrence of the tumor
after excision has been reported to be frequent. Shin et al., in a study
involving twenty-two patients managed with simple excision of osteochondromas,
noted that the recurrence rate was 53.8% in patients who were less than ten
years old13. We
think that the tethering effect and the growth disturbance of the affected
portions persisted even after tumor excision irrespective of tumor recurrence
in Group 2. This resulted in the deterioration of the radial articular angle
and radial bowing. Actually, the percentage of radial bowing in three patients
(Cases 8 [right side], 9, and 11) and the radial articular angle in one
patient (Case 11) deteriorated, although there was no recurrence of the tumor.
We recommend ulnar lengthening to completely release the tethering and to
support the ulnar carpus when the radial articular angle and carpal slip are
severe in patients with a Group-2 type of deformity.
In conclusion, the pattern of forearm deformity can be classified on the
basis of the location of the tumor. This pattern is determined by the absence
(Group 1) or presence (Group 2) of the tethering effect between the ulnar side
of the distal aspect of the radius and the distal end of the ulna, and the
ability of simple surgical excision of the tumor to correct deformity is
influenced by this pattern of deformity. We note that the number of patients
in the current study was small, and a more extensive study is necessary to
confirm these conclusions. ?