Patients were included in this study if they met the following
criteria: (1) skeletal immaturity, (2) established lateral condylar nonunion
of two years or more, (3) a cubitus valgus deformity of =20°, and (4)
postoperative follow-up of three years or more. Eight patients who met these
criteria were treated with this new technique between 1994 and 2000 (see
Appendix). The mean age at the time of the lateral condylar fracture was 3.7
years (range, two to six years). Six children had been treated with
immobilization in a cast without closed reduction for three to four weeks. The
other two children had been treated by bone setters with only traditional
manipulation and local herb paste application. The mean age at the time of our
procedure was 8.6 years (range, 4.4 to 12.2 years). The mean interval between
the lateral condylar fracture and our definitive procedure was 4.9 years
(range, 2.1 to 9.0 years).
Preoperative Evaluation
The medical records were reviewed with regard to preoperative complaints
about cosmetic appearance, pain, joint stability, and ulnar nerve
symptoms.
Preoperative physical examination included determination of the range of
motion of the elbow, carrying angle, and joint stability and a neurological
evaluation. The clinical carrying angle was measured with a goniometer with
the elbow in maximum extension and supination.
Preoperative radiographic assessment included anteroposterior and lateral
radiographs of both elbows. Only nonrotated or minimally rotated lateral
condylar nonunions as demonstrated on radiographs were treated with the new
technique. The anteroposterior radiographs were made with the arm in maximum
extension and supination. Elbow alignment was assessed by measuring the
humerus-ulna angle on the anteroposterior radiographs prior to surgery.
Preoperative planning included a tracing of the injured and uninjured elbows
on the anteroposterior radiographs. The tracing of the injured side was then
reversed and superimposed on the tracing of the normal side to determine the
desired corrective angle (a).
Surgical Technique
The patient was placed in the prone position with the affected extremity
resting on an arm support. After inflation of a pneumatic tourniquet, the
surgery was performed on the elbow through a midline posterior skin incision
extending from 6 cm proximal to 2 cm distal to the olecranon. The lateral
border of the triceps muscle was mobilized and retracted medially to expose
the lateral condylar nonunion. To avoid postoperative joint adhesions and/or
osteonecrosis of the lateral condyle, no attempt was made to mobilize the
lateral condyle for the purposes of reduction and only fibrous tissue in the
fracture gap was débrided. The lateral condyle was stabilized in situ
with two small cancellous compression screws inserted across the lateral
condyle into the metaphysis of the distal aspect of the humerus. No bone graft
was used in any patient.
The triceps muscle was then split along the midline in order to expose the
distal aspect of the humerus for the domeshaped osteotomy
(Fig. 1-A). In order to avoid
injury to the ulnar nerve, it was routinely identified, released from the
cubital tunnel, transposed anteriorly, and protected with a Penrose drain
during the procedure. The periosteum was incised longitudinally, and
subperiosteal dissection was performed to expose the distal humeral
metaphysis. The periosteum of the distal aspect of the humerus is typically
very thick and tends to blend rather strongly into the perichondrium. This
thick portion needs to be detached by sharp dissection along the junction of
the perichondrium. This avoids injury to the physes and articular cartilage.
The junction between the periosteum and the perichondrium at the medial border
of the distal humeral metaphysis was defined as point A
(Fig. 1-B). The intersection of
the midline axis and the upper margin of the olecranon fossa (point O) was
designated as the center of the dome. With the OA segment as the base, a
second line was drawn from point O to form an angle equal to the planned
correction angle (a). Furthermore, the intersection of this line and the
medial border of the distal aspect of the humerus was designated as point B.
The length of segment OB was designated as the radius of the dome. The arc of
the dome osteotomy was thus defined by these parameters.
During the osteotomy, two small Chandler retractors were placed alongside
the anterior humeral cortex in order to protect the neurovascular structures
in the antecubital fossa. Interrupted drill-holes were made along the proposed
osteotomy arc with use of a 1.8-mm Kirschner wire. The osteotomy was then
completed with a small osteotome. Subsequent to the osteotomy, point B was
split into two points, point B' on the proximal fragment and point B" on
the distal fragment. Next, the distal fragment of the osteotomized humerus was
rotated along the osteotomy arc until point A on the distal fragment met point
B' on the proximal fragment. The distal humerus was thus realigned as planned
(Figs. 2-A and 2-B).
One medial and one or two lateral crossed percutaneous 1.8-mm Steinmann
pins were used to stabilize the osteotomy. During insertion of the pins, the
entry and exit points should be directly visualized, thereby minimizing the
potential for neurovascular injury and eliminating the need for fluoroscopy.
After correction of the cubitus valgus deformity, any initial tension on the
ulnar nerve had resolved. The nerve was transposed subcutaneously prior to
wound closure. The same surgeon performed all of the procedures in this
study.
Postoperatively, the upper extremity was immobilized in a long-arm
posterior splint with the elbow in 90° of flexion for three weeks. The
affected elbow was then placed in a sling, and the parents were instructed to
have the child perform active-assisted extension and flexion exercises for
thirty minutes twice a day. However, if the compliance of the patient and
family were questionable, immobilization in the splint was continued for
another week. The pins were removed at the end of the fourth week in an
outpatient clinic.
Postoperative Evaluation
The operative time, estimated blood loss, neurovascular complications,
wound healing, and pin-track condition were recorded.
The clinical carrying angle, range of motion, muscle power, joint
stability, and the postoperative scar were assessed by the same examiner at
the final follow-up evaluation. To avoid bias, the follow-up examinations were
not performed by the operating surgeon. Subjective opinions with regard to the
appearance of the elbow and the resultant postoperative scar were recorded.
Anteroposterior and lateral radiographs of the elbow were made, and the
humerus-ulna angle was measured on the anteroposterior radiograph.
A clinical score was determined for each patient with use of our
modification to the scoring system described by Dhillon et
al.7
(Table I). This system is based
on pain, function, and cosmetic appearance. Function was assessed with use of
pain or weakness on activity and the measured range of motion of the elbow.
Cosmetic appearance was based on the postoperative humerus-ulna angle.
However, the medial shift in the humerus-ulna axis following the osteotomy was
added as a fourth parameter to evaluate whether the dome-shaped osteotomy
induced a secondary prominence of the medial epicondyle. The medial shift in
the humerus-ulna axis was defined as (CD/AB) × 100%, with CD indicating
the distance from the olecranon tip to the midline of the humerus and AB
indicating the diaphyseal radius at the narrowest portion of the humeral
shaft. The medial shift in the humerus-ulna axis was given a score of 3 points
when it was =100%, 2 points when it was between 101% and 200%, 1 point when
it was between 201% and 300%, and 0 points when it was >300%
(Table I). The maximum score
for this modified Dhillon scoring system was 12 points. A score of 12 points
was classified as an excellent result; 10 to 11 points, a good result; 9
points, a fair result; and =8 points, a poor result.
Preoperatively, the cubitus valgus deformity and its progression
were the major concerns of all eight patients. However, four patients also had
pain in the involved elbow during sports activities. Three children who had
elbow pain also complained of a "giving-way" sensation when
conducting a pushing-up motion. No patient reported weakness. No muscle
weakness in the ulnar nerve distribution was observed on the preoperative
physical examination. The preoperative elbow range of motion, compared with
that of the contralateral unaffected elbow, was normal in four patients (Cases
1, 5, 6, and 8). Three patients had a flexion contracture that ranged from
10° to 15°, and two patients had lost 10° and 15° of full
flexion. The mean preoperative humerus-ulna angle was 31° (range, 24°
to 36°). The mean duration of the operation was eighty-one minutes (range,
seventy-two to eighty-eight minutes).
No intraoperative complications occurred during any of the procedures.
Postoperatively, no patient had a wound or pin-track infection. No
percutaneous pins had to be removed prior to the union of the osteotomy. For
all eight patients, correction of the deformity was well maintained throughout
the immediate postoperative period. No patient had loosening of fixation or
loss of correction, and no patient had revision surgery.
All supracondylar osteotomies of the distal aspect of the humerus healed
uneventfully by the end of the fourth week, at which time the percutaneous
pins were removed. Postoperative radiographs of the elbow demonstrated that
the lateral condylar nonuion united within two months in four patients and
within three months in the remaining four patients. Elbow range of motion
improved progressively until the sixth month postoperatively and then did not
improve further.
The clinical and radiographic postoperative follow-up period was a mean of
4.5 years (range, 3.1 to 7.7 years). At the last follow-up evaluation, the
total available arc of elbow motion in the flexion-extension plane improved in
three patients (Cases 3, 4, and 7), remained unchanged in four patients (Cases
1, 5, 6, and 8), and decreased in one patient (Case 2). The mean total arc of
elbow motion was 131° (range, 120° to 145°) preoperatively and
131° (range, 110° to 145°) postoperatively. No loss of elbow
motion resulted from the surgical procedure, except in one boy who lost
10° of motion. Three of the four patients who had preoperative elbow pain
had complete relief of pain postoperatively. The remaining patient reported
that the preoperative pain had decreased. The three patients who had elbow
instability preoperatively had no such symptoms at the time of the latest
follow-up.
The mean postoperative humerus-ulna angle at the last follow-up examination
measured 5.5° (range, 3° to 9°) of valgus. All patients were
satisfied with the overall cosmetic outcome of surgery. Muscle weakness was
not reported by any patient or observed in any postoperative assessment.
According to our modification of the Dhillon scoring system, the result was
rated as excellent for two patients (Fig.
3-A, 3-B,
3-C, 3-D), good for four
patients, and fair for two patients. No patient subsequently had a recurrent
cubitus valgus deformity or had a medial shift of the humerus-ulna axis.
Nonunion of a fracture of the lateral condyle can result in a
progressive cubitus valgus
deformity5,8
and late ulnarnerve
neuropathy5,9,10.
Currently, there are several treatment options. For ulnar-nerve neuropathy,
anterior transposition of the ulnar nerve is a common
procedure11,12.
Supracondylar osteotomy of the distal aspect of the humerus is also generally
believed to be necessary to relieve the tension on the ulnar nerve when the
cubitus valgus deformity is
=20°11,12.
However, treatment of the lateral condylar nonunion is controversial because
of the reported high rates of
complications1-6.
Nonetheless, the advantages of treating the nonunion itself that have been
reported include improvement in elbow
stability11,12,
less elbow pain during
sports11,12,
and the prevention of further progression of the cubitus valgus deformity and
possible tardy ulnar
palsy12.
In this study, the main indication for treatment was the progression of the
cubitus valgus deformity. We used a domeshaped supracondylar osteotomy of the
distal aspect of the humerus to correct the deformity. However, we believe
that the deformity would have recurred if the nonunion was not repaired
because our patients were all skeletally immature. Therefore, the lateral
condylar nonunion was stabilized to prevent recurrence of the deformity. The
postoperative cubitus valgus deformity was well corrected and was maintained
at a mean follow-up interval of 4.5 years. Additionally, the elbow pain that
occurred in four patients was relieved in three patients and was decreased in
one. The preoperative elbow instability in three patients was also eliminated.
However, it should be stressed that only nonunions without substantial
rotation should be treated with this new technique.
Attempted anatomical reduction and rigid fixation of the lateral condyle
has frequently resulted in complications including loss of
motion3-6,
osteonecrosis of the
fragment1,3,4,
and persistent
nonunion6,7.
It appears that extensive soft-tissue stripping, undertaken while attempting
to mobilize and reduce the fracture fragments anatomically, constitutes the
main cause of these
complications2,11.
As a consequence, several authors have recommended in situ arthrodesis, rather
than anatomic reduction, of the nonunited lateral condylar
fragment2,11,13.
To improve the outcome of cubitus valgus secondary to a lateral condylar
nonunion, we developed this new technique. In comparing the new method with
traditional surgical procedures, three advantages can be readily identified:
the need for only a single posterior midline skin incision, a posterior
approach to the lateral condylar nonunion, and a dome-shaped supracondylar
osteotomy of the distal aspect of the humerus.
The traditional techniques used to treat a lateral condylar nonunion
usually require a medial skin incision for anterior transposition of the ulnar
nerve and a separate lateral skin incision for bone-grafting and internal
fixation of the lateral condyle and a possible concomitant supracondylar
corrective
osteotomy12. With
two scars, the cosmetic appearance can be compromised particularly because the
lateral scar crosses the Langer lines in this area and therefore tends to
hypertrophy14,15.
We have found that the posterior scar is more acceptable cosmetically. Fewer
of these scars become hypertrophic, and the posterior location has led to
better patient
acceptance16.
For nonrotated and minimally rotated lateral condylar nonunions, the
fracture can be readily identified on the anteroposterior radiograph.
Therefore, the fracture surfaces can be exposed directly through the posterior
approach. If no attempt at anatomical reduction of the lateral condylar
fragment is made, extensive soft-tissue stripping can be avoided. Under direct
vision, the fibrous tissue embedded in the gap between the fracture fragments
can be removed with relative
ease17. In our
experience, only the fracture surfaces over the metaphysis need to be
refreshed and the portion in the physis and epiphysis can essentially remain
untouched. As part of this technique, the lateral condyle only needs to be
fixed in situ by means of two small cancellous compression screws. In this
study, all of the nonunions could be easily compressed and no bone graft was
necessary. Because the posterior approach allows direct access to the nonunion
site and reduces the amount of soft-tissue dissection necessary to achieve
compression, the risk of jeopardizing vascularity of the lateral condyle can
be avoided. No osteonecrosis of the lateral condyle occurred in this
series.
A closing-wedge osteotomy has been used traditionally to correct the
cubitus valgus deformity, although there have been a few associated
disadvantages4,12.
The most frequently reported disadvantage is the loss of
correction4. With
lateral condylar nonunion, the lateral side of the distal humeral metaphysis
usually is somewhat hypoplastic and there may be lateral instability of the
elbow12. In order
to stabilize the lateral condyle, a layer of heavy fibrous tissue often
develops around the nonunion site. In a varus closing-wedge osteotomy, the
lateral fibrous tissue is stretched and creates a large valgus moment on the
lateral side of the elbow. This valgus moment renders the osteotomy site
unstable and makes it difficult to appropriately fix the site to maintain the
achieved correction. The second disadvantage to a closing-wedge osteotomy is
the resultant decrease in postoperative elbow range of
motion12. With a
closing-wedge osteotomy, the soft tissues in the lateral aspect of the elbow
tend to be overstretched, thus reducing the range of motion of the elbow.
Therefore, posterior capsulotomy and partial lateral triceps tenotomy are
frequently necessary in order to improve the postoperative range of
motion12. The third
disadvantage is the possibility of a secondary elbow deformity developing
subsequent to surgery. For the closing-wedge osteotomy, a large triangular
bone block usually needs to be removed. This typically results in a large
difference in surface areas between proximal and distal fragments at the
osteotomy site. After positioning the larger distal fragment against the much
smaller proximal fragment, a medial epicondylar prominence often results
because of a medial translation of the humerus-ulna axis. The cosmetic outcome
of the procedure is compromised because of these anatomical changes. When a
dome-shaped osteotomy is used, the rotation center of the distal humeral
fragment tends to remain at the midline of the humerus, such that only a small
rotation arc is necessary to correct the deformity. A much smaller valgus
moment is created, and the round distal fragment is well contained within the
dome of the proximal fragment. Therefore, with a dome-shaped osteotomy,
mechanical stability at the osteotomy site is enhanced and, as evidenced by
the results of our study, loss of correction is unlikely to occur. In
addition, rotating the distal humeral fragment with its rotation center
located at the midline of the humerus decreases the likelihood that it will
shift medially to create a medial epicondylar prominence.
We concluded that, with better exposure of the lateral condylar nonunion
through a posterior approach, the nonunion can be stabilized effectively and
postoperative loss of motion and osteonecrosis of the lateral condyle can be
avoided. With a dome-shaped supracondylar osteotomy, we can correct the
cubitus valgus deformity and avoid the development of a medial epicondylar
prominence. In carefully selected patients, this new technique can be an
effective method for the treatment of this clinically challenging problem.
A table presenting clinical and radiographic details on all patients is
available with the electronic versions of this article, on our web site at
(go to
the article citation and click on "Supplementary Material") and on
our quarterly CD-ROM (call our subscription department, at 781-449-9780, to
order the CD-ROM).
The authors did not receive grants or outside funding in support of their
research or preparation of this manuscript. They did not receive payments or
other benefits or a commitment or agreement to provide such benefits from a
commercial entity. No commercial entity paid or directed, or agreed to pay or
direct, any benefits to any research fund, foundation, educational
institution, or other charitable or nonprofit organization with which the
authors are affiliated or associated.