402 / 460
Magnetic Resonance Imaging in Patellofemoral Instability
401
Discussion
Previous authors have utilized MRI imaging to
perform measurements of patellar tilt [9, 23].
Patellar tilt proved to be an excellent group of
measurements for delineating between Controls
and those with instability. The confidence
intervals between the two groups were
separated by at least 5° for each measurement.
For the Angle of Fulkerson the means of
18.18°±0.56 and -3.5°±2.62 for Controls and
PFJDs respectively support the established
cutoff of <8° for pathology established by
Schutzer
et al.
[29]. Patellar Inclination Angle
is another well-studied measurement. Our
means of 8.10°±0.55 for Controls and
24.03°±2.42 for PFJDs differed from Dejour
et
al.
(10°±4.3 vs. 16°±3.3) CT based study [5].
Our results were very similar, however, to those
by Escala
et al.
whose MRI based study found
means of 9.2 and 21.7 for controls and PFJDs,
respectively [9]. This is an interesting result, as
patellar inclination anglewould not be predicted
to change between CT and MRI imaging as
articular cartilage is not involved in its
measurements. For the Angle of Laurin our
results
(Controls
10.10°±0.48;
PFJDs
-5.23°±2.96) again differed from Biedert’s CT
based study with means of -0.1° and -3.1° [7].
The Angle of Laurin relies on the articular
cartilage both on the trochlea and the patella;
thus this difference is to be expected. As
mentioned earlier, all our measurements
reflected an increase in the lateral tilt of the
patella. The increase in lateral tilt can be
attributed to laxity or even rupture of medial
soft tissue structures, specifically the medial
patellofemoral ligament (MPFL) [8]. Biedert
et
al.
indirectly showed that even a weak vastus
medialis could contribute to an increase in
lateral tilt by demonstrating a difference in
patellar angles with and without active
quadriceps contraction [7].
The significance of
patella alta
is that an
elevated patella will not engage the bony
architecture of the proximal trochlea that is
necessary to prevent lateralization of the
patella. Patellar station, demonstrated limited
success in terms of finding significant
differences. One could assume that MRI
inclusion of articular cartilage would make
some of these measurements unreliable,
however Miller
et al.
demonstrated that patellar
height measurements could be reliably recorded
on MRI [20]. The Insall-Salvati Ratio has a
well-documented pathological value of >1.2
for
patella alta
based on XR measures [30].
Our means of 1.08±0.02 for Controls and
1.26±0.03 for PFJDs follows this initial cutoff.
Our findings are also similar to what Escala
et
al.
found in their MRI based study (1.11 and
1.35) [9]. Caton-Deschamps Ratio was the only
other patellar height ratio found to be significant.
Our means of 1.13±0.02 and 1.29±0.03 for
Controls and PFJDs again reflect the standard
cutoff of >1.2 significant for
patella alta
[5,
10].
Trochlear morphology was the main focus of
our research and yielded some very interesting
results. Our Sulcus Angle means 148.48°±0.94
(Controls) and 165.67°±2.65 (PFJD) reflect a
difference from the classic cutoff of 145° [1,
5-7]. While they do reflect an increase in the
angle, likely due to inclusion of articular
cartilage, our patellofemoral instability patients
did not have quite as large sulcus angle as
found by Van Huyssteen
et al.
andAli
et al.
Van
Huyssten’s group showed cartilage based mean
sulcus angles of 186.5° [17], while Ali
et al.
published angles of 173° [18]. These differences
could be attributed to a difference in the
location where the measurement was made:
Van Huuyssteen
et al.
made their measurements
within 3mm of the start of articular cartilage
and Ali
et al.
based their location on the portion
with the greatest ventral prominence. It should
be noted that the mean sulcus angles at our
distal trochlea were closer to the classic cutoff.
But since the initial cutoff was established on a
Merchant view of the knee, one can see how
the values may vary. Trochlear Groove depth is
another classic measurement with a cutoff of
<4mm being pathological for patellofemoral
instability [5]. Other authors have reported that
MRI based trochlear depth measures have not
varied much from the initial CT or X-ray values
[9, 18, 21]. Our means of 6.47mm±0.24 for
Controls and 4.00mm±0.43 for PFJDs at the
proximal trochlea differ from the classic cutoff
but prove similar to the results of Escala
et al.