The first seventy-five periacetabular osteotomies were
performed in sixty-three patients between April 1984 and December
1987. The male:female ratio was 1:3.4, and the average age was 29.3 years
(range, thirteen to fifty-six years). An underlying neurological
disease was found in six hip joints, a posttraumatic acetabular deficiency
was found in two, and a proximal femoral focal dysplasia was seen
in another two. Twenty-three hip joints (31 percent) had
undergone previous surgery because of dysplasia (Table I). Thirty-seven
hips (49 percent) were classified as having group-III dysplasia
(dysplasia without subluxation of the femoral head) according to
the Severin system3. Thirty-three
hips (44 percent) were classified as group IV (with femoral head
subluxation); four hips (5 percent), as group V (a secondary acetabulum);
and one hip (1 percent), as group II. All patients presented with
pain at the time of surgery, and fifty-five (73 percent) of the
hips restricted the patients’ ability to walk.
Radiographic evaluations were performed on a standard anteroposterior
pelvic radiograph and a false-profile radiograph according to the
technique described by Lequesne and de Sèze4. The dysplasia was measured
on the basis of the lateral center-edge angle of Wiberg5, the anterior center-edge angle of
Lequesne and de Sèze4,
and the acetabular index described by Tönnis6 for the obliquity of the acetabular
roof. Lateralization of the femoral head was measured as the distance
between the medial edge of the femoral head and the ilioischial
line. The distance was measured preoperatively and postoperatively
and was compared, when possible, with that of the normal, contralateral
hip. The integrity of the Shenton line was recorded preoperatively
and postoperatively. The osteoarthritis at the time of surgery was graded
on radiographs according to the criteria of Tönnis6. Grade 1 indicates a widened sclerotic
zone and minimal osteophytes; grade 2, a moderate loss of joint
width and cysts; and grade 3, degenerative findings with a joint
width of less than one millimeter.
The clinical evaluation included scoring of the overall result
as well as the pain level, range of motion, and walking ability,
with a maximum of 6 points for each, with the system of Merle d’Aubigné and
Postel7 (Table II). The acetabular
rim was assessed for lesions with the impingement test8.
The surgery was performed with the technique described earlier by
the senior author (R. G.) and colleagues1.
After reorientation of the acetabulum, intraoperative radiographs
were made routinely. As a result, an additional intertrochanteric
osteotomy was judged to be necessary to improve joint congruency
in sixteen hips. An abduction osteotomy was done in thirteen of
these hips; an adduction osteotomy, in two; and an extension
osteotomy, in one. The operating time averaged 3.5 hours
(range, two to five hours), blood loss averaged 2000 milliliters
(range, 750 to 4500 milliliters), and an average of four
red blood-cell units (range, one to eleven units) was required.
Seventy-one hips (95 percent) had adequate follow-up,
at an average of 11.3 years (range, ten to 13.8 years). One patient
with two involved hips (3 percent) died during the observation period,
six years after the periacetabular osteotomy. Two patients (3 percent)
were lost to follow-up early after the osteotomy.
Radiographic Results
The lateral center-edge angle, the anterior center-edge angle,
and the acetabular index improved between the preoperative and postoperative
evaluations (Table III). The Shenton line was restored in
forty-one (62 percent) of the sixty-six hips with available radiographs.
Lateralization of the femoral head was reduced postoperatively to
an average of ten millimeters (range, -9 to +24 millimeters)
compared with eleven millimeters (range, -3 to +18 millimeters)
in the unaffected hips. All of the differences between the preoperative
and postoperative values were significant, with a p value between 0.001
and 0.0001.
The hip joint was preserved in fifty-eight (82 percent)
of the seventy-one cases that were followed for an average of 11.3 years.
Thirteen hip joints were revised subsequently, to either a total
hip arthroplasty (twelve) or a hip fusion. These procedures were
performed at an average of 6.1 years (range, one to 13.2 years)
after the periacetabular osteotomy. In four of these cases, total
hip replacement was done after an interval of ten years or more.
Clinical Results
The average Merle d’Aubigné and Postel7 score for the fifty-eight preserved
hip joints increased from 14.6 points (range, 7 to 17 points) preoperatively
to 16.3 points (range, 12 to 18 points) at the time of the last
follow-up (p < 0.0001) (Table IV). There was a significant improvement
in the scores for pain and walking. The range of motion, especially
flexion and internal rotation, decreased in the majority of the
hip joints (Table IV).
Six of the fifty-eight preserved hip joints had only a fair result (12,
13, or 14 points). Thus, fifty-two (73 percent) of the seventy-one
hip joints were preserved and had a good or excellent clinical rating.
Osteoarthritis
The immediate preoperative radiographs of fifty-five of the preserved
hip joints were available and could be graded for osteoarthritis.
Twenty-seven (49 percent) had no preoperative signs of osteoarthritis
(Table V). At
the time of the last follow-up, forty-four hip joints (80 percent)
had no or grade-1 osteoarthritis (Table V). Degenerative signs had progressed
in fourteen hips (25 percent), including ten with progression from
no to grade-1 osteoarthritis. In four hips, there was a radiographic
improvement of one grade. Complete preoperative radiographic documentation
was available for eleven of the thirteen nonpreserved hip joints.
Six of them had grade-2 osteoarthritis prior to the periacetabular
osteotomy (Table V).
Of a total of fifty-one hip joints with no or grade-1 osteoarthritis preoperatively,
five (10 percent) had undergone total hip arthroplasty, whereas
six (46 percent) of thirteen hip joints with grade-2 osteoarthritis
had been replaced by the time of the last follow-up. The
two hip joints with grade-3 preoperative osteoarthritis were preserved
at the time of the last follow-up.
Labral Lesions
Although arthrotomy was not done routinely in this early period,
a labral lesion was detected in fifteen hips (21 percent), which
had a significantly worse outcome (p < 0.001). Twenty-seven
(38 percent) of the seventy-one hip joints had a positive impingement
test on the initial physical examination. Over the years, the results
of this test have been found to be closely associated with pathological changes
of the acetabular rim8 and detectable
labral lesions on magnetic resonance arthrography9.
The labral lesion was left untreated in twelve hips in the present
series. Resuturing of the torn labrum in two hips failed, and joint
revision with resection of the labrum was performed later while
still preserving the hip.
Additional Intertrochanteric Osteotomy
A simultaneous intertrochanteric osteotomy was performed in sixteen
hip joints (21 percent) as a second step after the periacetabular
osteotomy. The decision to carry out this procedure was predominantly
based on the findings on the intraoperative radiographs made after
the acetabular correction was performed. In many of these sixteen
hips, one could suspect additional benefit from an intertrochanteric
osteotomy on the basis of the preoperative radiographs.
In addition to the routine intraoperative anteroposterior pelvic
radiographs, an abduction or adduction radiograph was made after
the acetabular correction to simulate the effects of an additional
intertrochanteric osteotomy in hips with insufficient joint congruency, joint-space
width, or femoral head coverage. Thirteen of the sixteen hip joints
underwent an abduction osteotomy; two, an adduction osteotomy; and
one, an extension osteotomy. However, seven of the thirteen abduction
osteotomies were performed to reverse the effects of a former adduction
osteotomy.
Prognostic Factors
The Wilcoxon signed rank-sum test was used to evaluate
the significance of differences between preoperative and postoperative clinical
and radiographic variables. Patient groups with different variables
(for example, preoperative osteoarthritis or a labral tear) were
compared with use of the Mann-Whitney U test. Logistic
regression analysis of the different variables was used to detect
independent factors that may have prognostic relevance. Significant
variables predicting a negative outcome were older age at the time
of surgery (p < 0.0001), presence and grade of osteoarthritis
(p < 0.0001), and presence of a labral lesion (p < 0.001).
A low anterior center-edge angle after correction and an acetabular
index outside of the range of 0 to 10 degrees after correction were
also associated with a less favorable outcome (p < 0.0058).
Complications
All major complications in this series occurred in the first
eighteen hips to be treated. The complications included an intra-articular
osteotomy in two hips. In one of them, the joint was replaced by
a total hip prosthesis nineteen months later. A fair result was
seen in the other patient, who died in a traffic accident six years
later. Three hip joints underwent repeat osteotomy: two, because
of a loss of correction due to too early weight-bearing, and one, because
of excessive lateralization of the fragment. Posterior subluxation
of the femoral head occurred in three hip joints with a lack of
posterior coverage postoperatively. In two of these hips, a subsequent
posterior shelfplasty was performed. One transient femoral nerve
palsy was observed in this series. There was no damage to major
blood vessels. Heterotopic ossification with restricted motion occurred in
four hips (6 percent). Subsequent successful resection of ectopic
bone was performed in two of them. Nonunion of a pubic osteotomy
site was seen in one patient. Extreme correction at the osteotomy
site had created a large gap. Consolidation did not occur despite
subsequent autogenous bone-grafting. An extensive study of the complications
after 508 Bernese periacetabular osteotomies was performed by Hussell
et al.10.