323 / 460
P. Beaufils, M. Thaunat, D. Passeron, P. Boisrenoult, N Pujol
322
Two different situations have to be considered:
1) Patella is pre operatively well centered
The goal of the procedure is to maintain a
well aligned patella. In this occurrence,
approach is mainly dictated by other factors
such as previous surgery or frontal deformity:
medial approaches in varus knees, lateral
approach in valgus knees. Respect of several
well defined criteria (adapted rotation of
femoral and tibial implants, thickness of the
surfaced patella, global AP whidth, patellar
hight…) should allow to obtain a proper post-
operative patellar alignment and thus to avoid
a complementary lateral retinacular release.
Lateral release in this occurrence is always
the witness of a technical error. Patellar
tracking is not influenced by the medial or
lateral approach. One should however
underline the interest of the subvastus
approach, first described by Gernez and then
by Hofmann [21] or midvastus [16] compared
with the standardmedial parapatellar approach
in terms of quadriceps recovery [13, 14, 17],
immediate functional result and patellar
tracking. These approaches would limit the
risk of associated retinacular release [8, 20,
24]. Matsueda and Gustilo [24] compared the
medial parapatellar approach with the medial
infra-vastus approach and obtained better
centering with the latter (83% versus 63%).
But these approaches are not indicated in
cases of expected difficulties of exposure:
flexum deformity, patella baja, obesity…
2) Patella is laterraly subluxated
The goal is to restore a normal patellar tracking
and thus to correct bony or ligamentous
abnormalities. As it is, when considering the
frontal plane: correction of a varus or valgus
deformity to obtain a well aligned knee
(table 1). Chan and Gill [12] consider that each
2° of lateral tilt observed preoperatively
increases the post-operative tilt 1°. Bindelglass
and Vince [8] come to the same conclusions.
Pushing this comparison between axial plane
(the patellar alignment) and frontal plane (the
frontal knee axis), one can find the same factors
of deformity (lateralization) than those of the
frontal deformity :
a.
Cartilaginous lateral patellar wear which
pushes the patella laterally It is the problem
of the cartilaginous correction which is
probably the easiest factor to correct.
b.
Ligamentous abnormalities (lateral retina
culum contraction, medial retinaculum
stretchening). It is the problem of the liga
ment balance
c.
Bony ablormalities (patellofemoral dys
plasia): it is the problem of the correction
of bony deformities in the ligamentous
envelope.
In this occurrence of subluxated patella, the
choice of the approach may have two interests:
- Direct influence:
for example on the
ligamentous balance. Why to propose a
medial approach in case of lateral subluxated
patella, that is to say an approach which goes
through the “convexity”. Do we approach a
varus knee with lateral ligamentous
stretchening using a lateral approach, or a
valgus knee with medial stretchening using a
medial approach?
- Indirect Influence
allowing a better positioning
of femoral and tibial implants: in the same
way as for a valgus knee: lateral approach
allows a better assessment and better
correction of lateral bony abnormalities.
For all these reasons we propose the use of the
lateral approach with parapatellar lateral
arthrotomy when the patella is laterally
subluxated, whatever the frontal axis (varus or
valgus knee). In this occurrence, the use of a
medial approach dramatically increases the
risk of lateral retinaculum release. This release
considerably increases the risk of patellar
fracture [31], even though the effects on patellar
blood supply remain a subject of debate [23,
31, 34, 36]. The real effect on patellar tilt
remains uncertain [25].
Lateral approach has a bad reputation, with an
increased risk of extensor apparatus rupture. In
reality, there is no more intra operative or post-
operative morbidity compared with standard
medial parapatellar approach [27, 35].