Asixty-two-year-old woman was referred for the treatment of a failed total
hip prosthesis. When she was fourteen years old, she had an arthrodesis of the
right hip following several operations for the treatment of developmental
dysplasia of the hip. At the age of fifty-four, the fused hip was converted to
a noncemented total hip arthroplasty with use of a Wagner revision stem for
the femur. Three years later, she underwent ipsilateral total knee
replacement. The knee motion gradually decreased. At the time of the
evaluation, she was dependent on a wheelchair. During stance, the right lower
extremity would lengthen and shorten between 5 and 8 cm. Hip motion was
globally limited and painful. Knee motion was restricted to 20° of flexion
from full extension.
Radiographs made immediately following conversion of the fused hip to the
total hip arthroplasty showed a long Wagner revision stem with an imperfect
diaphyseal fit (Fig. 1). At the
time of the ipsilateral knee replacement, there was already widening of the
medullary cavity around the loose femoral component
(Fig. 2). Seven years following
the total hip arthroplasty, there was an impressive expansion of the medullary
canal with marked thinning of the diaphyseal cortex. The maximal femoral width
was double the original width on the anteroposterior radiograph and nearly
triple the width on the lateral projection
(Figs. 3-A and 3-B). At the
distal end of the expansion, 4 to 6 mm of sclerotic bone obliterated the
medullary canal. A limb-length discrepancy of 5 cm was determined from the
anteroposterior pelvic radiograph. Preoperative blood tests and cultures of
aspirates of the lesion and joint space revealed no evidence of infection.
Technique of Revision
The surgical technique involved the use of an acetabular reinforcement ring
and the implantation of a larger and longer Wagner revision stem. An extended
trochanteric osteotomy approach was used to remove the prosthesis. Revision of
the acetabular component proceeded without difficulty. The inner surfaces of
the femur were cleaned of membranous tissue. Intraoperative examination of the
tissue revealed connective tissue with few cellular elements and no evidence
of acute inflammation. The anterior and posterior femoral cortices were each
osteotomized from inside out into four separate longitudinal strips, taking
care to preserve their soft-tissue attachments
(Fig. 4-A).
To allow improved access to the sclerotic cap that obliterated the distal
femoral medullary canal and to facilitate reduction of the strips onto the new
prosthesis, the distal junction with the normal femoral shaft was cut. The
sclerotic bone was opened centrally and was gradually enlarged with femoral
reamers. Solid interference fixation of the stem with excellent rotational
stability was achieved in the distal part of the femur. By inserting the
largest and longest available stem to the proximal extent of the total knee
prosthesis and with use of the planned neck length, 3 cm of the limb-length
discrepancy observed on preoperative radiographs was corrected. The strips of
bone were secured to the inserted stem at the middle and at both ends with
double cerclage wires, again avoiding unnecessary soft-tissue stripping
(Fig. 4-B). Following reduction
of the hip, the bone strips moved closer to the prosthesis because of the
increased soft-tissue tension. The cerclage wires were progressively tightened
a final time, allowing the strips of bone to overlap
(Figs. 5-A and 5-B).
Final culture results of fluid and soft-tissue specimens obtained at the
time of surgery were negative for infection.
Immediately after surgery, the patient was mobilized with protected
weight-bearing. After proximal bone formation was evident, weight-bearing was
progressively increased as tolerated by the patient. Two years after surgery,
the proximal femoral bone had consolidated around the prosthesis and regained
substantial thickness. The prosthesis remained well fixed
(Figs. 6-A and 6-B).
Clinically, the patient was able to walk using one cane and she reported no
groin or thigh pain. The remaining limb-length discrepancy was 2 cm. Hip
motion remained moderately reduced in all directions, and she had a
Trendelenburg sign. Knee motion had increased to 40° of flexion.