A review of records identified thirty-four children who had
undergone hemiepiphyseal stapling for late-onset tibia vara at two
institutions between January 1, 1992, and December 31, 1999. Eight of them
(with nine involved extremities) with relatively mild deformities had been
lost to follow-up less than two years after the index procedure, but, when
last seen, they had either closed or closing physes in good alignment.
Attempts to contact these patients were unsuccessful. The remaining twenty-six
patients, with forty-two involved extremities, were followed for a minimum of
two years after the surgery. In nine extremities, an osteotomy of the tibia
had been performed concomitantly with a distal femoral hemiepiphyseal
stapling, and those limbs were excluded from the study. Twenty-six patients
(thirty-three extremities) were included in the study group.
The medical records of the twenty-six patients were reviewed for
demographic information, surgical details, and complications. The result was
considered to be final at either skeletal maturity or just prior to definitive
corrective osteotomy for a residual deformity. All pertinent radiographs of
the lower extremities, including those made preoperatively and at the most
recent follow-up visit, were reviewed. We used standardized standing
long-cassette anteroposterior radiographs of the lower extremity made with the
knee facing directly
anteriorly44,45
to measure the mechanical axis deviation, the mechanical medial proximal
tibial angle, the mechanical lateral distal femoral angle, and limb-length
discrepancy44,45.
The knee joint was divided into zones for classification with a modification
of the method described by Mielke and
Stevens31
(Fig. 1). Medial (varus) zones
were labeled -1, -2, -3, and -4 depending on the severity of varus angulation.
The zone through which the mechanical axis passed was recorded.
The result was graded as normal alignment, varus undercorrection, or valgus
overcorrection on the basis of the zone through which the mechanical axis
passed on either the final radiograph or the radiograph made prior to
osteotomy. Normal alignment indicated that the mechanical axis passed through
zone 1 or -1. Mild varus undercorrection was defined as the mechanical axis
passing through zone -2, and moderate varus undercorrection was defined as the
mechanical axis passing through zone -3. Valgus overcorrection indicated that
the mechanical axis passed through zone 2, 3, or 4.
The final result was classified as successful, partially successful, or
unsuccessful. A successful result indicated that neither a proximal tibial nor
a distal femoral valgus osteotomy was necessary, either because alignment was
normal or because there was only mild residual varus deformity that was not a
concern to the patient or family. A partially successful result indicated that
a corrective osteotomy was necessary but hemiepiphyseal stapling obviated the
need to perform an osteotomy on either the tibia or the femur that would have
otherwise been necessary. An unsuccessful result indicated that an osteotomy
was required to correct the residual varus or valgus deformity.
Hemiepiphyseal stapling (Figs. 2-A, 2-B,
and 2-C) was performed at our institution when genu varum due to
late-onset tibia vara was present in a patient who was skeletally immature and
who, in the surgeon's judgment, had adequate growth remaining to allow full
correction of the deformity. The surgeon based the decision regarding adequate
remaining growth on the patient's chronological age, the widths of the physes,
and the magnitude of the deformity. Skeletal age was not routinely assessed
with the use of the Greulich and Pyle
atlas46. Patients
with a lower-extremity length discrepancy of >3 cm, laxity of the lateral
collateral ligament, closing physes with a large deformity, or substantial
pain when walking were not considered for hemiepiphyseal stapling alone.
There were twenty-one boys and five girls, with a mean age of 11.8 years
(range, 8.1 to 14.9 years) at the time of the hemiepiphyseal stapling. Fifteen
patients were white, and eleven were black. The mean body weight at the time
of the surgery was 84 kg (range, 40 to 154 kg), with the weight of
twenty-three of the twenty-six patients in excess of the 95th percentile for
their age. The patients were followed postoperatively for a mean of 3.8 years
(range, two to 6.8 years). Seven of the twenty-six patients were treated
bilaterally with the index procedure. Lateral hemiepiphyseal stapling of the
proximal part of the tibia was performed on all extremities, and concomitant
lateral hemiepiphyseal stapling of the distal part of the femur was performed
on fourteen extremities. Distal femoral stapling was indicated when >5°
of distal femoral varus (a lateral distal femoral angle of =93°) was
noted. All but three patients (three extremities) were skeletally mature at
the time of final follow-up. The results in these three patients were
considered to be final for the purpose of this study, as an osteotomy was
ultimately required to complete the correction of the deformity.
Surgical Technique
The patient was placed in a supine position on a radiolucent table
(Orthopedic Systems, Hayward, California) with adequate elevation of the
ipsilateral hip to allow access to the lateral aspect of the knee. A 3 to 5-cm
incision was made over the lateral aspect of the proximal tibial physis, just
anterior to the fibula, and extraperiosteal dissection was performed. The
physis was located with a Keith needle with the aid of fluoroscopy. Two or
three cobalt-chromium-alloy Vitallium Blount staples (Zimmer, Warsaw, Indiana)
were placed extraperiosteally across the lateral aspect of the proximal tibial
physis. Distal femoral hemiepiphyseal stapling, when indicated, was performed
through a second 3 to 5-cm incision over the lateral aspect of the distal
femoral physis. These staples were also placed after an extraperiosteal
dissection. Staple position was confirmed intraoperatively with use of
anteroposterior and lateral fluoroscopy.
Postoperatively, patients were allowed to bear weight as tolerated without
immobilization. The strategy for staple removal varied, depending on both the
correction that had been obtained and the patient's age. Ideally, anatomic
correction occurs concomitant with spontaneous physeal closure at skeletal
maturity and staple removal is optional. For patients who are nearly but not
yet skeletally mature, have minimal lower-extremity length discrepancy, and
have a mechanical axis passing through the center of the knee, percutaneous
drilling
epiphysiodesis47-49
through the lateral incision can be performed concomitant with staple removal.
In patients with substantial remaining growth, as determined by chronological
age and the width of the physis on radiographs, overcorrection well into
valgus zone 1 was allowed with the anticipation that rebound growth would
result in the mechanical axis of the lower extremity passing through the
center of the knee at skeletal maturity.
Preoperatively, the mechanical axis passed a mean of 58 mm (range,
27 to 157 mm) medial to the center of the knee. The mean mechanical axis
deviation improved to 22 mm (range, -33 to 117 mm) at the time of the last
follow-up. The mean medial proximal tibial angle (normal, 87° with a range
of 85° to 89°) improved from 77° (range, 50° to 85°)
preoperatively to 85° (range, 48° to 95°) at the time of the last
follow-up. Preoperatively, the mean lateral distal femoral angle (normal,
88° with a range of 86° to 89°) was 92° (range, 87° to
100°) in the series as a whole and 96° (range, 92° to 100°) in
the fourteen lower extremities that were treated with supplemental lateral
distal femoral hemiepiphyseal stapling. The mean lateral distal femoral angle
in these fourteen limbs improved to 86° (range, 79° to 97°).
Preoperatively, the mechanical axis passed through zone -2 in twelve knees,
zone -3 in seventeen, and zone -4 in four. At the time of the last follow-up,
the mechanical axis passed through the central portion of the knee (zone -1 or
1) in twenty (61%) of the thirty-three lower extremities. Of these twenty
extremities, which were classified as having normal alignment, eleven had had
a preoperative mechanical axis that passed through zone -2; eight, zone -3;
and only one, zone -4 (Table
I). In eleven extremities (33%), the final mechanical axis was
medial to the central zones of the knee, which was considered to be varus
undercorrection. Of these eleven extremities, four had mild varus and seven
had moderate varus. One had been corrected so that the axis passed through
zone -1, but rebound varus occurred after staple removal and the extremity
ultimately was classified as having mild varus, with the mechanical axis
passing through zone -2. A repeat hemiepiphyseal stapling was performed, and
ultimately the alignment was in zone 1, but the result was considered to be a
varus undercorrection because of the repeat stapling. The patient and his or
her family were satisfied with the alignment of all of the four extremities
that were classified as being in mild varus, and they declined additional
surgery although a tibial osteotomy to correct the remaining deformity was
offered.
At the time of the final follow-up, normal alignment was noted in fourteen
(61%) of the twenty-three extremities with juvenile tibia vara and six of the
ten with adolescent tibia vara. With the numbers available, there was no
significant difference between these two groups with regard to final alignment
(Table II).
Of the seven patients (seven extremities) with moderate varus
undercorrection, one was later found to have a medial proximal tibial physeal
bar that had not been apparent at the time of stapling. Two patients had early
staple migration or breakage, which was not revised. Early in the series, one
patient had distal femoral varus that had not been appreciated at the time of
the proximal tibial stapling. The proximal tibial deformity was fully
corrected, but the patient was left with residual femoral varus, which was
later corrected with a distal femoral osteotomy. One patient was fifteen years
of age at the time of the hemiepiphyseal stapling and had insufficient
remaining growth to correct the deformity. The two remaining extremities with
moderate undercorrection had had severe varus preoperatively, with the
mechanical axis passing through zone -4 in both and a mechanical axis
deviation of 118 mm in one and 157 mm in the other. Although the patients were
young when the hemiepiphyseal stapling was performed, the deformities
corrected very slowly, and an osteotomy was later performed because of
increasing knee pain.
The mechanical axis of two extremities passed lateral to the central zones
of the knee (through zone 2) at the time of follow-up, and these limbs were
considered to have valgus overcorrection. Both of these patients had failed to
return for follow-up at the requested time; they returned only after the
valgus overcorrection had developed.
The results were also examined with regard to whether the procedure was
successful in correcting the limb adequately to obviate the need for
additional surgery. Of the twelve extremities (eleven patients) with a mild
preoperative deformity (zone -2), eleven had normal final alignment and one
had worsening of the varus in spite of the hemiepiphyseal stapling
(Table III).
Of the seventeen extremities (sixteen patients) with a moderate
preoperative deformity (zone -3), fourteen had alignment that was satisfactory
to the patient at the time of follow-up. Of these fourteen extremities, eight
(eight patients) had normal alignment at the time of follow-up, four (four
patients) had residual mild varus undercorrection, and two (two patients) had
valgus overcorrection. The hemiepiphyseal stapling in these fourteen
extremities was considered to be successful in obviating the need for
additional surgery. The remaining three extremities with moderate preoperative
deformity had moderate varus undercorrection. In one of them, the distal part
of the femur was adequately corrected and, although a tibial osteotomy was
ultimately required, a femoral osteotomy was not. In the second extremity with
moderate varus undercorrection, the tibia was corrected but there was residual
varus of the distal part of the femur that required a distal femoral
osteotomy. The hemiepiphyseal stapling in these two extremities was considered
to be partially successful because, although an osteotomy was required in
either the femur or the tibia, it was not needed in both, as would have been
necessary without stapling. The final patient with residual moderate varus
required a tibial osteotomy despite hemiepiphyseal stapling and was considered
to have an unsuccessful result.
Of the four patients (four extremities) with a severe preoperative
deformity (zone -4), one obtained complete correction and was considered to
have a successful result. The other three patients subsequently required both
femoral and tibial osteotomies for correction of the deformity and were
considered to have an unsuccessful result. Thus, according to our criteria for
outcome assessment, hemiepiphyseal stapling was successful in twenty-six
extremities, partially successful in two, and unsuccessful in five. Of the
twenty-nine extremities with mild or moderate preoperative varus, twenty-five
had a successful result; two, a partially successful result; and two, an
unsuccessful result.
Four patients (four extremities) had a lower-extremity length discrepancy
of >2 cm (mean, 2.7 cm; range, 2.5 to 3.0 cm) at the time of the final
follow-up. The result was classified as unsuccessful for three of these four
extremities, and the residual discrepancy was corrected at the time of a
definitive osteotomy. The remaining patient was asymptomatic with a 3.0-cm
discrepancy and declined further treatment. No patient was found to have
limited knee motion or to have knee hyperextension suggestive of recurvatum at
the time of the final follow-up.
The staples were ultimately removed from ten extremities. Four of these
extremities were considered failures and had the staples removed near or after
skeletal maturity or at the time of an osteotomy. Another extremity had
recurrence of the deformity after staple removal, as discussed previously, and
was treated with repeat stapling. Three other extremities (two with a
mechanical axis passing through zone -1 and one with a mechanical axis passing
through zone -2) had the staples removed, after skeletal maturity, at the
family's request. The final two extremities had the staples removed prior to
skeletal maturity. At the time of the final follow-up (two and three years
after the staple removal), which was after skeletal maturity, those two
extremities had a mechanical axis passing through either zone 1 or zone
-1.
One patient had a common peroneal nerve neurapraxia, probably due to
excessive retraction at surgery. It resolved spontaneously three months
postoperatively. No wound infections or other early complications developed.
The staple backed out of the bone in five extremities. A reoperation was
performed to reinsert staples in three of these extremities and ultimately led
to a successful result. The other two patients with staple migration did not
return for follow-up until after skeletal maturity. Both had residual moderate
varus at that time.
The goal of surgical treatment of late-onset tibia vara is
restoration of normal knee and limb alignment as well as normal joint
orientation in order to minimize the risk of degenerative arthritis of the
knee. Historically, the mainstay of treatment of late-onset tibia vara has
been valgus osteotomy of the proximal part of the tibia. Alternatives to
osteotomy include
hemiepiphysiodesis8,50
and hemiepiphyseal stapling. Hemiepiphysiodesis, however, has the disadvantage
of being permanent, as it irreversibly stops growth capacity, and thus
requires accurate estimation of remaining growth, which is
difficult51-53.
Hemiepiphyseal stapling, as introduced by
Blount30,43,
is relatively easy to perform, perioperative morbidity is minimal, correction
occurs at the site of the deformity, and longitudinal growth can resume
following staple removal.
The primary criterion that we used to evaluate the preoperative severity of
the deformity and the outcome of treatment was the mechanical alignment of the
lower extremity. Once hemiepiphyseal stapling is performed, it is critical to
follow the patient closely both clinically and radiographically to monitor
correction. Patients and their families often note improvement—indeed,
full correction of the deformity—and tend not to return for follow-up,
despite preoperative warnings about the potential for overcorrection. Two of
our patients had valgus overcorrection requiring operative treatment. We
recommend achieving slight overcorrection of the varus deformity in patients
with substantial remaining growth and allowing the mechanical axis to pass 1
to 2 cm lateral to the center of the knee prior to staple removal. Following
staple removal, rebound varus will occur, usually bringing the mechanical axis
back to the center of the knee. Again, clinical and radiographic follow-up is
essential. Varus did recur in one of our patients. In some older adolescent
patients with a minimal limb-length discrepancy, we have removed the staples
when the mechanical axis passed through the center of the knee and have
performed a proximal tibial epiphysiodesis to prevent rebound varus.
In retrospect, it was realized that three patients who were treated with
hemiepiphyseal stapling in our series had had insufficient remaining growth to
allow full correction of the deformity. Nevertheless, two of these patients,
who did not obtain full correction of the deformity, did obtain complete
correction of the femoral portion of the deformity, which otherwise would have
required a corrective osteotomy. Three patients with a severe preoperative
deformity obtained little correction despite the stapling and the seemingly
adequate growth remaining; these patients were morbidly obese, and the medial
aspect of the physis did not grow sufficiently to correct the deformity
following the lateral hemiepiphyseal stapling. It is difficult to determine
the amount of remaining growth of a pathologically altered medial-proximal
aspect of a tibia, and determining bone age preoperatively may help one to
ascertain which patients may not obtain full correction. We have not utilized
this approach because we believe that the growth in this area does not follow
age-dependent norms.
It is of crucial importance to adequately assess the deformity
preoperatively, as distal femoral varus is frequently a major component of the
overall varus
deformity10. One of
our patients ultimately required a distal femoral osteotomy after
hemiepiphyseal stapling of the proximal part of the tibia had completely
corrected the tibial deformity. Because of the relative abduction of the hip
associated with the genu varum deformity, in spite of distal femoral varus the
knee joint can appear parallel to the floor on a standing anteroposterior
radiograph of the lower extremity, leading to the impression that the
deformity is completely within the proximal part of the tibia. Careful
preoperative planning and assessment of the deformities are essential to the
success of operative intervention.
Historically, hemiepiphyseal stapling has been reserved for adolescents.
More recently, Mielke and Stevens showed that hemiepiphyseal stapling, with
careful insertion and removal of the staples without injury to the periosteum,
is a safe and effective treatment for younger children who have an angular
deformity of the
knee31. In our
series, stapling proved to be safe and effective for patients who were ten
years of age or younger. We obtained complete correction in all five patients
of that age.
In conclusion, hemiepiphyseal stapling is indicated in patients with
late-onset tibia vara if the proximal tibial physis is still open and the
varus deformity is mild or moderate. When distal femoral varus is present,
lateral hemiepiphyseal stapling of the distal part of the femur should be
performed concomitantly. Hemiepiphyseal stapling can be effective in selected
patients with severe varus deformity, although osteotomy may be indicated in
the majority of those patients. Frequent clinical and radiographic follow-up
is mandatory to monitor the correction. ?