CLINICAL ORTHOPAEDICS AND RELATED RESEARCH
Number 0, pp. 000000
Coxa Vara
A Novel Measurement Technique in Skeletal Dysplasias
Chang-Wug Oh, MD*; Mihir M. Thacker, MD; William G. Mackenzie, MD, FRCSC;
and Eric C. Riddle, MHS, PA-C
Coxa vara can be a progressive deformity in children with
skeletal dysplasia. Preoperative anteroposterior pelvic radio-
graphs of 30 children with spondyloepiphyseal dysplasia con-
genita and spondyloepimetaphyseal dysplasia were used to
test the reliability of a new radiographic measure of coxa
vara, the Hilgenreiner-trochanteric angle. An additional 10
patients (20 hips) with coxa vara deformities needing valgus-
producing proximal femoral osteotomies also were reviewed.
Interobserver reliability with plain radiographs was 0.929
for the left side and 0.914 for the right side using interclass
correlation coefficients. Intraobserver reliability also was
high, with an interclass correlation coefficient of 0.875.
Twelve hips corrected by osteotomy had adequate ossifica-
tion to measure the Hilgenreiner-epiphyseal angle, head-
shaft angle, and Hilgenreiner-trochanteric angle. Only one of
these hips had a recurrence. The results were good in all of
the other ossified hips. Eight hips had limited ossification;
only two of these hips maintained acceptable alignment. Six
hips had less postoperative correction and progressive defor-
mity at the final followup. We present a novel measurement
technique to determine the degree of coxa vara deformity in
children with delayed or absent ossification of the capital
femoral epiphysis.
Received: June 3, 2003
Revised: December 22, 2003; February 8, 2005; August 29, 2005; November
17, 2005
Accepted: December 12, 2005
From the *Department of Orthopedic Surgery, Kyungpook National Univer-
sity Hospital, Samdok, Chung-gu, Daegu, Korea; and the Department of
Orthopaedic Surgery, Alfred I. duPont Hospital for Children, Wilmington,
DE.
Each author certifies that he or she has no commercial associations (eg,
consultancies, stock ownership, equity interest, patent/licensing arrange-
ments, etc.) that might pose a conflict of interest in connection with the
submitted article.
Each author certifies that his or her institution has approved the human
protocol for this investigation and that all investigations were conducted in
conformity with ethical principles of research, and that informed consent was
obtained.
Correspondence to: Eric C. Riddle, MHS, PA-C, Department of Orthopaedic
Surgery, Alfred I. duPont Hospital for Children, 1600 Rockland Road, Wil-
mington, DE. Phone: 302-651-5890; Fax: 302-651-5951; E-mail:
eriddle@nemours.org.
DOI: 10.1097/01.blo.0000203476.81302.24
Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
2006 Lippincott Williams & Wilkins
Level of Evidence: Diagnostic studies- Level III-1 (Study of
nonconsecutive patients; without consistently applied refer-
ence gold standard. See the Guidelines for Authors for a
complete description of levels of evidence.
Coxa vara is defined as a neck-shaft angle less than 110.
It has been classified as dysplastic, congenital, develop-
mental, or traumatic depending on its etiology.1 It is char-
acterized radiographically by a decrease in the neck-shaft
angle, and clinically by a waddling gait or limb-length
discrepancy. Spondyloepiphyseal dysplasia congenita and
spondyloepimetaphyseal dysplasia are two rare forms of
osteochondrodysplasia. They are associated with progres-
sive coxa vara deformity and delay in ossification of the
capital femoral epiphysis, which often is not ossified until
the second decade. Measurements used to define the de-
gree of coxa vara are the head-shaft angle, neck-shaft
angle, and Hilgenreiner-epiphyseal (H-E) angle.
The head-shaft angle is a more reproducible measure-
ment of varus deformity than the neck-shaft angle for se-
verely deformed hips.24 Carroll et al found the head-shaft
angle to be an unreliable indicator of appropriate correc-
tion in children with coxa vara deformity having a valgus-
producing osteotomy.4 The H-E angle is defined as the
angle between the physis and the horizontal line of Hilgen-
reiner (Fig 1A). This measurement has been used to de-
termine the surgical correction necessary to avoid recur-
rence of deformity, and for surgical indications.4 An ossi-
fied femoral epiphysis is essential for using all of these
measurement techniques. In patients with spondyloepiphy-
seal dysplasia and other osteochondrodysplasias, ossifica-
tion of the femoral head and neck often is delayed or even
absent in severe occurrences. In this situation it becomes
difficult to use traditional measurements to assess the se-
verity of the proximal femoral deformity.
We devised a new radiographic measurement to deter-
mine the degree of coxa vara deformity in children. This
technique does not require epiphyseal ossification. The
purpose of our study was to test the reliability of this new
 Clinical Orthopaedics
2 Oh et al and Related Research

Fig 1AB. The illustrations show the (A) Hilgen-

reiner-epiphyseal (H-E) angle and (B) Hilgenreiner-
trochanteric physis (H-T) angle, which are used to
measure coxa vara.

measurement and determine the amount of correction

needed to obtain a satisfactory outcome in a group of
children with spondyloepiphyseal dysplasia congenita and
spondyloepimetaphyseal dysplasia.

MATERIALS AND METHODS

From 1985 to 2002, 46 patients with spondyloepiphyseal dys-

plasia congenita or spondyloepimetaphyseal dysplasia were
treated at our institution. Demographic information and infor-
mation regarding any complications were obtained from hospital
records, radiographs, or the patient during the clinical ex-
amination. Our study was given institutional review board ap-
proval.
In these skeletal dysplasias, the capital femoral epiphysis, the
physis, and metaphyseal fragment can slide down the medial
femoral shaft. In progressive coxa vara deformity, the greater
trochanter moves proximally in relation to the acetabulum. De-
layed or poor ossification of the capital femoral epiphysis made
it difficult to see the proximal femoral physis and measure the
H-E angle. The new measurement technique was developed to
assess the deformity accurately in the presence of limited epiph-
yseal ossification. A line was drawn connecting the medial edge
of the physeal line of the greater trochanter to the outer point of
the triradiate cartilage. The angle was measured between this line
and Hilgenreiners line (Fig 1B). This was named the H-T angle.
The H-T angle was considered negative if the medial edge of the
physeal line of the greater trochanter was proximal to the Hilgen-
reiners line, indicating a greater degree of coxa vara.
The H-T angles of 30 patients (60 hips) with spondylo-
epiphyseal dysplasia congenita or spondyloepimetaphyseal dys-
plasia were recorded on two separate occasions 6 weeks apart.
These were measured before any treatment including valgus-
producing osteotomy to determine reliability of the new tech-
nique. The radiographs were reviewed by three independent ob- Fig 2AB. The radiographs show progression of coxa vara in
servers with varying degrees of experience: a senior pediatric Patient 6 with spondyloepiphyseal dysplasia congenita at (A) 5
orthopaedic surgeon (WGM), a pediatric orthopaedic fellow years and (B) 11 years. The femoral epiphysis is ossified par-
(MMT), and a physician assistant (ECR). These measurements tially with a distinct triangular metaphyseal fragment.

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Number 0
Month 2006 Measuring Coxa Vara in Skeletal Dysplasias 3

were recorded in an identical manner, except that all radiographs
were presented in a different order with a new identification
number at the second session. The radiographs were not avail-
able to any of the observers between sessions. Data were on the
interval scale of measurement. Therefore, interobserver reliabil-
ity was assessed using interclass correlation coefficients adjusted
for complete agreement. The interobserver reliability was as-
sessed between the ratings at the two sessions. Intraobserver
reliability for the observations by the same observer at the two
sessions also was assessed using interclass correlation coeffi-
cients adjusted for complete agreement.18 Data were analyzed
using SPSS software (SPSS Inc, Version 11.0.1, Chicago, IL).
Ten patients (20 hips) who had coxa vara deformities and
valgus proximal femoral osteotomies also were retrospectively
reviewed. The medical genetics department confirmed the diag-
nosis of spondyloepiphyseal dysplasia congenita in nine patients
and spondyloepimetaphyseal dysplasia in one patient. The de-
formities were measured preoperatively and postoperatively us-
ing the Hilgenreiner-epiphyseal angle, head-shaft angle, and
Hilgenreiner-trochanteric angle to determine the degree of sur-
gical realignment necessary to obtain a satisfactory result. Indi-
cations for surgery included radiographic progression of the
varus deformity and progressive clinical complaints. Two ortho-
paedic surgeons (WGM, JRB ) performed the osteotomies on the
10 patients (five boys and five girls). Valgus osteotomy was
done at an average of 8.7 years (range, 4.313.5 years), and
followup averaged 4.9 years (range, 38 years). All patients were
older than 10 years at the final followup (mean, 13.8 years).
There were no complications from the operative procedures,
such as premature physeal closure, infection, or hardware-related
issues. These patients had bilateral procedures. A bilateral hip
spica cast was not used consistently. Standing anteroposterior
(AP) radiographs of the lower extremities with the patella facing
forward were taken preoperatively and postoperatively. Supine
views also were taken because the severe lordosis and pelvic
obliquity when standing made radiographic evaluation difficult.
Pillows placed under the knees were used to help reduce lumbar
lordosis. We assessed the amount of epiphyseal ossification (full
or partial) and measured the head-shaft angle if possible, the H-E
angle, and the Hilgenreiner-trochanteric (H-T) angle. In addition,
we prospectively studied a control group of 20 normal hips (10
boys and 10 girls; range, 113 years) that had serial radiographs
of the hip at 1-year or 2-year intervals to determine normal
values. One hip from each AP radiograph of the pelvis was
evaluated from each child.
RESULTS
In the 30 patients with spondyloepiphyseal dysplasia
congenita or spondyloepimetaphyseal dysplasia, the
interclass correlation coefficients ratings for the H-T angle
was 0.929 for the left hips and 0.914 for the right hips.
The intraobserver reliability analysis showed interclass
correlation coefficients of 0.875, indicating substantial
reliability. In the control group, the H-T angle averaged

TABLE 1. Changes in Hilgenreiner-Epiphyseal (H-E) and Hilgenreiner-Trochanteric

Physis (H-T) Angles
H-E Angle H-T Angle
Patient Epiphyseal Before After Final Before After Final
Number Side Ossification Surgery Surgery Followup Surgery Surgery Followup
1 L No NA NA NA 29 8 20
R No NA NA NA 30 5 21
2 L No NA NA NA 0 11 3
R No NA NA NA 0 10 3
3 L No NA NA NA 14 7 26
R No NA NA NA 0 8 2
4 L No NA NA NA 16 25 22
R No NA NA NA 10 30 29
5 L Yes 63 31 23 21 13 12
R Yes 72 25 22 15 14 13
6 L Yes 54 23 21 0 23 23
R Yes 60 33 24 6 20 20
7 L Yes 48 5 5 8 15 15
R Yes 60 5 5 0 20 20
8 L Yes 56 27 32 19 29 25
R Yes 60 23 27 18 30 26
9 L Yes 60 30 56 5 27 5
R Yes 59 25 29 16 25 16
10 L Yes 90 23 20 6 16 17
R Yes 67 21 25 5 18 19

Mean 62.4 22.6 24.1 4.3 17.7 9.35

Patients 19 had spondyloepiphyseal dysplasia; Patient 10 had spondyloepimetaphyseal dysplasia; NA = Not available

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Clinical Orthopaedics
4 Oh et al and Related Research

32.8 (range, 26.834.8) and varied from 3035 af-

ter 2 years. The H-E angle averaged 13.8 (range, 10.9
17).
Twelve hips (six patients) corrected by proximal femo-
ral osteotomies were ossified. Ossification of the capital
femoral epiphysis was adequate to see the physeal line in
these patients (Fig 2). The changes in H-E and H-T angles
are listed in Table 1. After surgical realignment, all but one
ossified or partially ossified hips had good results as mea-
sured by the three angles (Fig 3). One hip had a recurrence
(H-E angle 56, head-shaft angle 92, and H-T angle 5).
Ossification of the capital femoral epiphysis was limited in
eight hips (four patients) (Fig 4), and the H-E and head-
shaft angles could not be measured (Table 1). Immediate
postoperative alignment was satisfactory in two of eight
hips, (Patient 4) and these hips maintained good correction
at followup. The remaining six hips had less immediate
postoperative correction and progressed at final

Fig 4AB. The radiograph shows (A) bilateral coxa vara in

Patient 1 with spondyloepiphyseal dysplasia congenita with no
ossification of the femoral epiphysis. (B) At 10 years, a radio-
graph shows progression to severe deformity with upward dis-
placement of the greater trochanter.

followup (Fig 5). A metaphyseal fragment (triangular frag-

ment in the inferomedial neck) of variable sizes was seen
in 18 hips (nine patients). The metaphyseal fragment
healed in hips that maintained satisfactory correction at
final followup (Table 2).

DISCUSSION

Measuring the degree of coxa vara and correcting the de-

Fig 3AB. The radiographs show (A) a valgus osteotomy at
12 years, and (B) correction that has been maintained at 16
years, and the triangular metaphyseal fragment is healed.
formity in children with spondyloepiphyseal dysplasia
congenita and spondyloepimetaphyseal dysplasia can be
challenging because of lack of epiphyseal ossification at
an early age. Additionally, the amount of surgical correc-
tion needed to obtain a satisfactory result without recur-
rence has not been reported in a large series of patients

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Number 0
Month 2006
Fig 5AB. The radiograph shows (A) a subtrochanteric val-
gus osteotomy with internal fixation in a 12-year-old patient
(Patient 1). (B) Recurrent deformity was evident in the patient
at 16 years. The femoral epiphysis has limited ossification.
with spondyloepiphyseal dysplasia congenital or spondy-
loepimetaphyseal dysplasia. Coxa vara is defined as a de-
crease in the angle between the head or neck and shaft of
the femur, often in the presence of a metaphyseal defect.
The cause is not clear, but is thought to be related to a
physeal abnormality, which results in an inability to with-
stand the applied mechanical shear forces.14
The tip of the greater trochanter typically is at the same
level as the femoral head center in adults, and usually is
inferior to this level in children. The physeal line of the
greater trochanter usually is visible, and is more distal to
the center of the femoral head than the tip of the greater
trochanter.13 Hilgenreiners line12 is the horizontal line
connecting the triradiate cartilages and typically is above
the level of femoral head center. We drew a line connect-
ing the medial edge of the physeal line of the greater
trochanter to the outer point of the triradiate cartilage. The
angle between this line and Hilgenreiners line was mea-
Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Measuring Coxa Vara in Skeletal Dysplasias 5
sured as the H-T angle. Measuring the H-E angle or head-
shaft angle is difficult with delayed or absent ossification
of the capital femoral epiphysis. The H-E angle is useful
for surgical indications and outcome.3,4,20 We could not
measure this angle accurately in our patients. Ultrasound
can be used to evaluate coxa vara in children.9 However,
this technique is operator dependent, and the expertise
may not be available in some institutions. A limitation of
this study was the use of only conventional radiographs to
measure the H-T angle. Using coronal plane cuts of mag-
netic resonance images would have enabled us to assess
the degree of coxa vara in the presence of limited ossifi-
cation. However, these patients did not have this expensive
test. All landmarks needed for the H-T angle were easily
identifiable even in patients with delayed or absent ossi-
fication of the capital femoral epiphysis.
Patients with defective COL2A1 protein, as seen in
spondyloepiphyseal dysplasia congenita and spondylo-
epimetaphyseal dysplasia,6,16,19,23 often have coxa vara
develop. Our patients had the severe congenital form of
spondyloepiphyseal dysplasia congenita according to the
classification of Wynne-Davies and Hall.22 In spondylo-
epiphyseal dysplasia congenita and spondyloepimetaphy-
seal dysplasia, delayed ossification of femoral head is
characteristic, and the head may remain unossified until
later in life. In our study, 40% of the hips did not show
ossification by the time the patients were 10 years.
The H-T angle was greater than 30 in the control
group. An H-T angle less than 0 occurs in coxa vara and
congenital hip dislocation. An H-E angle greater than 60
has been used as an indication for surgical correction.15,21
We suggest an H-T angle less than 0 to be an indication
for surgical correction of coxa vara. Carroll et al4 recom-
mended correction of the H-E angle to less than 38 to
avoid recurrence of coxa vara. In our study, an H-T angle
of 13o or greater was not associated with recurrent varus.
The origin of the triangular metaphyseal fragment in the
inferomedial neck in coxa vara has been discussed.5,8,11,17
It is assumed that abnormal shearing forces produce
a defect of the femoral neck (triangular fragment),
which represents an insufficiency fracture.14 The size of
the triangular fragments in our series varied greatly. The
children typically had good results when this fragment
healed.
Surgical techniques using blade plates and screw side
plate devices have improved bone fixation. Investigators
using these fixation methods3,4,7 reported better results
than with previous methods including pin fixation and
casting.1,2,10,13 However, these earlier studies included
only a few children with spondyloepiphyseal dysplasia
congenita. In a study of five hips in children with spon-
dyloepiphyseal dysplasia, Widmann et al21 reported that
 Clinical Orthopaedics
6 Oh et al and Related Research

TABLE 2. Effect of Epiphyseal Ossification on Outcome

Triangular Fragment
Age at
Patient Surgery Followup Epiphyseal Before Final
Number Side (years) (years) Ossification Surgery Followup
1 L 11.3 3.7 No Yes Yes
R 11.3 3.7 No Yes Yes
2 L 4.8 6.2 No Yes Yes
R 4.8 6.2 No Yes Yes
3 L 13.5 3.5 No NA NA
R 13.5 3.5 No NA NA
4 L 9 8 No Yes Yes
R 9 8 No Yes Yes
5 L 12.9 3 Yes Yes Healed
R 12.9 3 Yes Yes Healed
6 L 12.1 4 Yes Yes Healed
R 12.1 4 Yes Yes Healed
7 L 7.1 3 Yes Yes Healed
R 7.1 3 Yes Yes Healed
8 L 6.3 5.7 Yes Yes Healed
R 6.3 5.7 Yes Yes Healed
9 L 4.3 7.8 Yes Yes Yes
R 4.3 7.8 Yes Yes Yes
10 L 5.4 4.6 Yes Yes Healed
R 5.4 4.6 Yes Yes Healed

Mean 8.7 4.9

Patients 19 had spondyloepiphyseal dysplasia congenita; Patient 10 had spondyloepimetaphyseal dysplasia; NA = Not
available

one hip required revision surgery, but the cause of recur-
rence was not well described. In our study, hips with an
ossified capital femoral epiphysis maintained good long-
term correction. Most hips in the delayed ossification
group had recurrence of coxa vara, although the average
correction in this group was greater than in the children
with epiphyseal ossification. Six of eight patients had their
hips corrected to an H-T angle of 11 or less. This was
likely a factor in the progression of the deformity postop-
eratively. However, lack of epiphyseal ossification may
not be the only contributing factor. The age of the patient
at the time of surgery and the presence of acetabular dys-
plasia or limb-length discrepancy may contribute to recur-
rence of coxa vara postoperatively.4
The outcome of surgical realignment of coxa vara de-
formity in spondyloepiphyseal dysplasia congenita or
spondyloepimetaphyseal dysplasia seems to be influenced
by the degree of ossification of the capital femoral physis
and the amount of surgical correction. The development of
a reliable measure of coxa vara may aid in determining the
appropriate amount of correction necessary to achieve a
satisfactory outcome.
Acknowledgment
We thank Dr. J. Richard Bowen (JRB) for contributing his pa-
tients to our study.
References
1. Amstutz H, Wilson P Jr. Dysgenesis of the proximal femur (coxa
vara) and its surgical management. J Bone Joint Surg. 1962;44:
124.
2. Borden J, Spencer G Jr, Herndon C. Treatment of coxa vara in
children by means of a modified osteotomy. J Bone Joint Surg.
1966;48:11061110.
3. Burns KA, Stevens PM. Coxa vara: another option for fixation. J
Pediatr Orthop B. 2001;10:304310.
4. Carroll K, Coleman S, Stevens PM. Coxa vara: surgical outcomes of
valgus osteotomies. J Pediatr Orthop. 1997;17:220224.
5. Chung SMK, Riser W. The histological characteristics of congenital
coxa vara: a case report of a five year old boy. Clin Orthop Relat
Res. 1978;132:7181.
6. Cole WG, Hall RK, Rogers JG. The clinical features of spondylo-
epiphyseal dysplasia congenita resulting from the substitution of
glycine 997 by serine in the alpha 1(II) chain of type II collagen. J
Med Genet. 1993;30:2735.
7. Cordes S, Dickens DR, Cole WG. Correction of coxa vara in child-
hood. J Bone Joint Surg. 1991;73:36.
8. Currarino G, Birch JG, Herring JA. Developmental coxa vara as-
sociated with spondylometaphyseal dysplasia (DCV/SMD): SMD-
corner fracture type (DCV/SMD-CF) demonstrated in most re-
ported cases. Pediatr Radiol. 2000;30:1424.
9. De Pellegrin MP, Mackenzie WG, Harcke HT. Ultrasonographic
evaluation of hip morphology in osteochondrodysplasias. J Pediatr
Orthop. 2000;20:588593.
10. Desai S, Johnson L. Long-term results of valgus osteotomy for
congenital coxa vara. Clin Orthop Relat Res. 1993;294:204210.
11. Fairbank HAT. Coxa vara due to congenital defect of the neck of the
femur. J Anat. 1928;62:232237.
12. Hilgenreiner H. Early diagnosis and early treatment of congenital
dislocation of the hip. Med Klin. 1925;21:13851389,14251429.

Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Number 0
Month 2006
13. Kelikian AS, Tachdjian MO, Askew MJ, Muvali J. Greater trochan-
teric advancement of the proximal femur: a clinical and biomechan-
ical study. Proceedings of the 11th Open Scientific Meeting of the
Hip Society. St Louis, MO: CV Mosby; 1983:77103.
14. Kim HT, Chambers HG, Mubarak S Jr, Wenger DR. Congenital
coxa vara: computed tomographic analysis of femoral retroversion
and the triangular metaphyseal fragment. J Pediatr Orthop. 2000;
20:551556.
15. Lahdenranta U, Pylkkanen P. Early and late results of Bracketts
operation for pseudarthrosis of the neck of the femur in infantile
coxa vara: a review of 30 operations. Acta Orthop Scand. 1977;48:
7479.
16. Langer LO Jr, Brill PW, Ozonoff MB, Pauli RM, Wilson WG,
Alford BA, Pavlov H, Drake DG. Spondylometaphyseal dysplasia,
corner fracture type: a heritable condition associated with coxa vara.
Radiology. 1990;175:761766.
17. Pauwels F. Biomechanics of the Normal and Diseased Hip. New
York, NY: Springer-Verlag; 1976:2429.
Copyright Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Measuring Coxa Vara in Skeletal Dysplasias 7
18. Rosner B. Fundamentals of Biostatistics. 4th ed. Belmont, NY: Dux-
bury Press; 1995.
19. Slaney SF, Hall CM, Atherton DJ, Winter RM. A new syndrome of
spondyloepimetaphyseal dysplasia, eczema and hypogammaglobu-
linaemia. Clin Dysmorphol. 1999;8:7985.
20. Weinstein J, Kuo K, Millar E. Congenital coxa vara: a retrospective
review. J Pediatr Orthop. 1984;4:7077.
21. Widmann RF, Hresko MT, Kasser JR, Millis MB. Wagner multiple
K-wire osteosynthesis to correct coxa vara in the young child: ex-
perience with a versatile tailor-made high angle blade plate
equivalent. J Pediatr Orthop B. 2001;10:4350.
22. Wynne-Davies R, Hall C. Two clinical variants of spondyloepi-
physeal dysplasia congenita. J Bone Joint Surg. 1982;64:435
441.
23. Zabel B, Hilbert K, Stoss H, Superti-Furga A, Spranger J, Winter-
pacht A. A specific collagen type II gene (COL2A1) mutation pre-
senting as spondyloperipheral dysplasia. Am J Med Genet. 1996;
63:123128.