Complications after MPFL reconstruction
163
these patients still had an average Kujala score
of 92.7 (72-100) indicating that this score is not
sensitive to a minor extensor lag. Smith
et al.
[26] did a comprehensive literature research on
the clinical and radiological results of MPFL
reconstructions. They could only find eight
papers, involving 186 MPFL reconstructions,
which met the criteria of their scoring system.
In only two of these eight papers did the authors
report on the post operative range of motion
and in both there were loss of flexion compared
to the non operated leg. None reported on loss
of active knee extension or extensor lag. Only
one paper, in this review, reported on quadriceps
atrophy with an incidence of 60% not
withstanding a mean Kujala score of 88.6 [7].
Loss of motion after an MPFL reconstruction is
directly related to the tension in the
reconstructed ligament. If it is too tight in
extension (femoral insertion too distal) there
will be an extensor lag although passive full
extension will not be affected. If it is too tight
in flexion (femoral insertion too proximal)
there will be loss of flexion both passive and
active; in this situation the patella might still be
unstable in extension.
Both the gracilis and semitendinosis tendons,
generally used, for reconstructing the MPFL is
stronger and stiffer than the original MPFL
[20]. The strength is a positive factor consi
dering the underlying predisposing factors
leading to the first dislocation. The stiffness on
the other hand can theoretically lead to overload
in the P-F joint especially in cases where the
reconstructed MPFL is not in the optimal
position.
In our technique of MPFL reconstruction we
try to recreate the normal non isometry of the
ligament so called “favourable anisometry”
[29].
This creates a ligament that is tight in extension
and lax in flexion. There is however the danger
that the ligament can be too tight in extension
resulting in an extensor lag. Post operative
quadriceps inhibition is very common and
should be distinguished from a permanent
extensor lag as a result of an over tight MPFL
reconstruction. At 3 months post operative
follow-up there was on average a 4° (5°-15°)
extensor lag, probably as a result of quads
inhibition, in 45% of our patients. This extensor
lag was however temporary and over the long
term only 4 out of the more than 200 cases had
permanent loss of active full extension which
was caused by an over tight reconstruction in
extension [27].
Elias [8] has shown experimentally that a too
proximal placed femoral position for the MPLF
graft will lead to increased patello femoral load
with potential overload of the articular surface
in the P-F joint. Loss of both active and passive
flexion will also be associated with this. In
techniques where the aim is to have an isometric
reconstructed MPFL, the danger of a having a
ligament that is too tight in flexion is increased.
Femoral insertions near or at the adductor
tubercle, although advocated by some authors,
should be avoided as this will lead to
reconstruction that is too tight in flexion and
too loose in extension [25, 27, 23].
In prevention of motion loss complications
special attention should be given to the
technique of determining the tension in the
reconstructed ligament. Beck showed [2] that
overtensioning can be avoided by applying low
loads to medial patellofemoral ligament recons
tructions, which reestablished normal transla
tion and patellofemoral contact pressures. The
aim of the MPFL reconstruction should be to
restore the tension in the MPFL to the same
tension that it had before being torn with a graft
that is stronger than the original ligament. If the
patella of the opposite knee is stable the amount
of transverse patella movement in the
reconstructed patella should be similar to that
of the uninjured knee. This can be achieved
intra operatively by draping both knees and
comparing the amount of transverse movement.
Fithian [13] advises adjusting the graft tension
in such a way that a 5 lb displacing force result
in 7-9mm of lateral displacement of the patella
We recommend that the isometry of the
ligament should be tested till the “favourable
anisometric point” is found by, using a guide
pin in the proposed femoral implantation site.
The “favourable anisometric point” would be a
point where the reconstructed ligament will be