Patients
Between February 1992 and March 2000, fourteen consecutive patients
(twenty-two knees) with diastrophic dysplasia were scheduled to undergo
primary total knee arthroplasty for the treatment of severe osteoarthritis.
Twenty-one arthroplasties (in fourteen patients) were successfully performed.
One knee underwent arthrodesis instead of arthroplasty. All operations were
performed by the same senior orthopaedic surgeon (T.P.). The indications for
arthroplasty were severe knee pain, considerable difficulty in walking and
performing normal daily activities, and radiographic signs of severe
osteoarthritis (Figs.
1-A,1-B,1-C,1-D,1-E,1-F).
The average age of the patients at the time of the arthroplasty was forty-four
years (range, thirty to sixty-six years). The series included two men and
twelve women. The mean height of the patients was 1.25 m (range, 1.14 to 1.42
m), and the mean weight was 43 kg (range, 32 to 55 kg). Previous surgical
procedures performed on these patients included eleven total hip
arthroplasties (ten bilateral), one patellectomy (not in the knee treated with
the arthroplasty), and one proximal tibial osteotomy.
All fourteen patients (twenty-one knees) were followed with clinical and
radiographic examinations. The mean duration of follow-up was 3.4 years
(range, 2.0 to 9.0 years). Five patients underwent bilateral arthroplasty as
separate operations, and two patients underwent a one-stage bilateral
procedure. Two patients had one-stage ipsilateral hip and knee
replacements.
This study was approved by the local ethics committee, and all patients
provided written informed consent.
Study Design
Patient data were recorded prospectively. All patients underwent detailed
clinical examinations of the hip and knee joints, and standard anteroposterior
and lateral radiographs of the hips and knees were made before and after the
operation. The mechanical axis was determined preoperatively on standing
hip-to-ankle radiographs. Follow-up examinations were performed at one, two,
and three years and at two-year intervals thereafter. The last follow-up
examination was performed by two independent observers (I.H. and V.R.).
Preoperatively and at all follow-up visits, the outcome was evaluated
according to the Hungerford knee scoring
system19 and the
Knee Society radiographic scoring
system20. Walking
distance, as reported by the patient, was recorded preoperatively and
postoperatively.
Preoperative Data and Operative Technique
All knees had severe degeneration of the tibiofemoral and patellofemoral
joints. Typical deformities of the knees were flexion contracture, excessive
valgus alignment with mediolateral and anteroposterior instability, and
subluxation or chronic dislocation of the inferiorly located patella. Flexion
contractures ranged from 0° to 45° (mean, 21°), and valgus
deformities ranged from 5° to 40° (mean, 18°). Ten patellae were
chronically dislocated, and eight of the remaining patellae were subluxated.
All knees had abnormal mediolateral instability, and eleven of them had
anteroposterior laxity. Eleven knees had medial or lateral tibiofemoral
subluxation (>10 mm of incongruity), which was associated with excessive
anterior slope of the tibial plateau. Loose bodies were noted in six
knees.
The most severe flexion contractures were combined with a dislocated
patella and severe valgus alignment. The mean metaphyseal angulation was
12° (range, 0° to 48°) in the distal part of the femur and 9°
(range, 0° to 18°) in the proximal part of the tibia. Two knees had
pronounced angular (apex anterior) and varus deformities of the distal part of
the femur requiring a distal femoral osteotomy to insert the prosthesis. The
patella was typically located inferiorly. The mean distance between the
tibiofemoral joint line and the distal apex of the patella was 9 mm (range, 2
to 20 mm).
All patients had the operation under spinal or epidural anesthesia. The
length and conical shape of the thigh and the commonly needed
soft-tissue-release procedures often prevented the use of a tourniquet, which
was utilized for only six patients (seven knees). All knees were approached
through a medial parapatellar incision. The anterior cruciate ligament was
either rudimentary (twelve knees) or absent (nine knees) in all patients.
Overgrowth of the medial femoral condyle or hypoplasia of the lateral condyle
was seen in the knees with combined severe flexion contracture, valgus
alignment, and a dislocated patella. The remnants of the cartilage of the
tibial plateau were often detached from the subchondral bone. The patella was
of the bipartite type in five knees; two were typical, and three were
horseshoe-shaped on the coronal radiograph and double-layered with a vertical
cleft on the sagittal radiograph.
Three knees were suitable for a total condylar prosthesis (AGC V2; Biomet,
Warsaw, Indiana). Eight knees required a more constrained, T-3 model (AGC
Dual; Biomet), and ten knees were treated with a rotating-hinge prosthesis
(Endomodel; Waldemar Link, Hamburg, Germany). The stems of the constrained
prosthesis were always shortened intraoperatively and were often made thinner
and even bent to fit the dysplastic bone. Because of curving of the metaphysis
and deformity of the knee joints, we were not able to use intramedullary
guides or special resection instruments. Bone resections were therefore
performed with a freehand technique. Anatomical landmarks were unreliable for
determining external rotation because of the distorted and severely deformed
anatomy. Therefore, rotational alignment was estimated clinically with a goal
of achieving 3° of external rotation of the femoral component.
Supracondylar correction osteotomy was performed if the distal part of the
femoral diaphysis was severely curved anteriorly and in varus angulation. To
avoid the problems associated with patella infera, the resection level of the
tibia was low, just near the tibial tuberosity. Extensive release of the
posterior aspect of the capsule and division of the posterior cruciate
ligament, popliteus tendon, iliotibial tract, and lateral collateral ligament
were often needed to overcome the flexion contractures and valgus deformities.
Distal femoral correction osteotomies were performed in two knees (one
patient) because of severe metaphyseal angulation. The popliteus tendon was
tenotomized in four knees. The lateral collateral ligament was tenotomized in
two valgus knees, and both were treated with a hinged prosthesis.
In most knees, realignment of the extensor apparatus and the patellofemoral
joint required wide lateral release, beginning just proximal to the tibial
tuberosity and extending nearly to the greater trochanter, as well as medial
capsulorrhaphy combined with advancement of the vastus medialis muscle over
the patella. The indication for advancement and duplication of the vastus
medialis muscle was a dislocated patella and a severely stretched vastus
medialis. The patella was resurfaced if severe patellofemoral osteoarthritis
with anterior knee pain was present preoperatively, and if, during the
surgery, the remnant of the patella was sufficiently thick (>10 mm) to
permit resurfacing after excision of the inner part of the double-layered
patella. The patella was resurfaced in thirteen knees. An extensive lateral
release was performed in nineteen knees. The vastus medialis was advanced over
the patella in five knees. The inner part of the bipartite patella was
resected in three knees. The remaining patellar fragment was inadequate for
resurfacing of two of these patellae as the fragments were only 7 and 8 mm
thick. In the knees treated with a constrained femoral or tibial component,
the stem was modified to fit the bowed femur or tibia. In four knees, the
dimensions of both the tibial plateau and the tibial stem were reduced
perioperatively by the surgeon, using a grinding machine.
The average duration of the operation was 210 minutes (range, ninety to 420
minutes), and the average perioperative blood loss was 850 mL (range, 50 to
4200 mL).
Prophylactic antibiotic therapy (with cefuroxime) was given two hours
before the operation and was continued for forty-eight hours thereafter.
Thromboembolic prophylaxis with subcutaneous heparin or enoxaparin was
administered on the day of the operation and was continued daily over the
period of the hospital stay. All patients except for one treated with
bilateral femoral osteotomy were allowed to walk with full weight-bearing
after the surgery. A continuous-passive-motion machine was used daily during
the hospital stay (seven to ten days).
Radiographic Evaluation
The radiographs were analyzed for component position and sizing,
radiolucent lines, and osteolysis. The distance between the tibiofemoral joint
line and the distal apex of the patella was measured as well. Radiolucent
lines were characterized by width and were classified by location according to
the zones described by the Knee
Society20.
Heterotopic ossification was noted. Definitely loose components were defined
as those with a complete radiolucent line or radiographic evidence of femoral
or tibial subsidence of =2
mm20.
Statistical Analysis
The Wilcoxon signed rank-sum test was used to compare preoperative and
postoperative Hungerford knee scores. P values of =0.05 were considered
significant.
Clinical Results
The average preoperative Hungerford knee pain and total scores were 5.8
points (range, 0 to 25 points) and 46 points (range, 25 to 65 points),
respectively. These scores improved significantly, to an average of 49 points
(range, 40 to 50 points) and 81 points (range, 60 to 95 points) at the
one-year follow-up visit (p < 0.001 for both comparisons). At the time of
the final follow-up, all of the pain scores were 50 points, and the total
scores averaged 83 points (range, 60 to 95 points). At the final examination,
all knees were pain-free. The mean walking distance increased from 380 m
(range, 20 to 1000 m) preoperatively to 1300 m (range, 250 to 3000 m) at the
final follow-up examination (p = 0.013). The average arc of motion was 83°
(range, 20° to 120°) preoperatively and 75° (range, 30° to
100°) at the time of the final follow-up. Flexion contractures decreased
from a mean of 21° (range, 0° to 45°) preoperatively to 7°
(range, 0° to 15°) at the final follow-up examination (p = 0.012).
Radiographic Results
There were no complete radiolucent lines around any component in any knee
during the follow-up period, and all radiolucent lines were <2 mm wide.
There was no migration or subsidence of any component in any knee. Heterotopic
ossification was not observed.
Complications
Six complications were recorded. Two patellae had redislocated during the
follow-up period. In addition, both femoral osteotomies were followed by
nonunion, and a reoperation was required five and six months after the primary
surgery. There was one fracture of the lateral tibial condyle and one of the
medial femoral condyle. Except for the patellar dislocations, all
complications resolved. None of the twenty-one knee arthroplasties needed to
be revised during the three and one-half-year follow-up period.
Diastrophic dysplasia is caused by mutations in a sulphate transporter
gene, which encodes for a sulphate transporter membrane protein. It is located
in the distal long arm of chromosome 5'. Mutations are thought to result
in impaired sulphate uptake of the cells and reduced proteoglycan sulphation,
causing structural and functional defects in the
cartilage1,21.
This leads to early and rapid deformity and degeneration of all major joints.
Symptoms are, however, mainly seen in the weight-bearing hip and knee joints.
Osteoarthritis develops in all three compartments of the knee. Therefore,
total knee arthroplasty is often indicated in patients with diastrophic
dysplasia.
The diastrophic femur is often wide in the metaphysis and extremely narrow
in the diaphysis17.
The bones are short, and metaphyseal angulation is common. Furthermore, the
osseous surfaces frequently provide only a small contact area for prosthetic
support. These factors create numerous technical challenges. Use of
intramedullary alignment guides is rarely possible. Extramedullary guides
could have been used in some of our patients, but we chose to cut the bones
with a freehand technique. Radiographic evaluation of the mechanical axis is
unreliable in these patients, since they commonly have severe flexion
contractures of both hip and knee joints.
All intramedullary stem extensions required manual modification, either
bending or shortening, in order to allow seating of the component within the
deformed bones. In addition, four prosthetic tibial plateaus required
modification to avoid overhang on the small osseous tibial plateau. Because
the flexion deformity makes it difficult to estimate the level of resection of
the tibia accurately or to determine the size of the component preoperatively,
it is often necessary to customize the tibial components. Although the
structural properties of the prostheses may have been altered by these
modifications in our series, none of the prostheses have failed to date.
Custommade components might be a solution to this problem, but accurate
preoperative imaging would be required.
The range of motion of the diastrophic knee starts to diminish before the
age of five years and was very restricted in our patients at the time of the
total knee arthroplasties. The range of motion actually diminished further
postoperatively, perhaps as a result of the scarring from the extensive
releases that were necessary during the operation or as a result of
progression of the disease itself. This decrease in motion substantially
reduced the Hungerford knee scores, even though all patients reported no pain
at the final follow-up examination.
In order to improve patellar tracking and quadriceps function, we performed
extensive lateral releases, beginning at the level of the tibial tubercle and
extending nearly to the greater trochanter, as well as medial capsulorrhaphy
and advancement of the vastus medialis over the patella. In addition, we
resected the inner portion of the bipartite patellae in order to avoid painful
motion of the fibrous junction between these two parts. Also, since our
patients had short patellar ligaments with patella infera, the tibial cut was
always performed as distally as possible, near the level of the tibial
tubercle.
Complications were common in our series. Insertion of the components was
difficult because of the severe deformity of the bones. There were two
intraoperative fractures, one of the lateral tibial condyle and one of the
medial femoral condyle. One patient with severe bilateral metaphyseal
angulation required two distal femoral osteotomies, both of which were
followed by a nonunion requiring a reoperation. Use of cement fixation may
have contributed to these nonunions. Distal femoral osteotomies performed as a
first stage before the total knee arthroplasty might have prevented this
complication. However, in these patients, severe angular deformity of the
distal part of the femur was compensated for by deformity of the proximal part
of the tibia to enable walking preoperatively. Performing femoral corrective
osteotomies without arthroplasty would have made the axial deformity of the
knees so severe that walking would have been impossible during the interval
between the osteotomy and the arthroplasty. Custom devices may have obviated
the need for osteotomy.
Two patellae redislocated during the follow-up period. Both knees had had
substantial preoperative quadriceps atrophy and chronic patellar dislocation.
The final outcome was poor in one patient and fair in the other.
Total hip and knee arthroplasty with additional soft-tissue procedures can
restore the walking ability of patients with diastrophic
dysplasia12,17,18.
These procedures often are required at a young age, potentiating the need for
later revision. However, the low weight and activity level characteristic of
these patients should promote longevity of the prosthesis and durable
long-term fixation. Nevertheless, the abnormal gait patterns that are often
necessary to compensate for the skeletal deformities in these patients are a
reason for concern about stress across these prosthetic joints. Overall, we
believe, on the basis of this study and our experience, that total knee
arthroplasty is an effective procedure in patients with diastrophic
dysplasia.