Anatomical patella instability risk factors on MRI show sensitivity without specificity in patients with patellofemoral instability: a systematic review

Abstract

Objective The purpose of this systematic review is to assess whether the current literature distinguishes between 9 common anatomic patellofemoral instability (PFI) imaging measurements to determine if they are consistent among the group of patients with PFI, and if a diagnostic difference exists between the control and patients with PFI in these imaging measurements.

Methods Following PRISMA guidelines, a systematic review of PubMed and CINAHL/SPORTDiscus was performed to identify all studies measuring PFI anatomic imaging factors from January 1980 to February 2015. Study inclusion criteria were: English language, MRI studies, imaging of PFI anatomic factors. References of the included articles were reviewed and those that met inclusion criteria were added to the search results. PFI group was defined as the study stating that the cohort had at least 1 objective patient episode of lateral patellar dislocation. Studies that did not report sample size, mean and SD were excluded, as well as those that used osseous (vs cartilage) landmarks, or who did not define their measurement landmarks. Meta-analysis software was used to determine weighted mean and 95% CI for PFI and control groups for each measurement. Of the 135 articles identified, 22 met the inclusion criteria.

Results Caton-Deschamps patella height index and trochlear depth measurement showed non-overlapping CIs between the control and PFI groups. Insall-Salvati ratio, lateral patella tilt, tibial tubercle-trochlear groove distance and trochlear facet asymmetry, although having statistically different means, had overlaped between the overall mean 95% CI of the 2 groups. Although few studies were available for each measurement, the number of total participants per measurement included is considerable: PFI (range 40–685); controls (range 94–671).

Conclusions A systematic search of current literature with meta-analysis reveals wide variation of PFI imaging measurements used for clinical decision-making within the controls and PFI groups. This systematic review demonstrates that appropriate abnormality thresholds of anatomic patella instability imaging factors exist for the PFI group, indicating sensitivity. The wide variation in the majority of measurements, especially in the control group, suggests poor specificity in these measurements. Trochlear depth has the best discrimination in evaluation between the control and the PFI groups.

Level of evidence Level III, systematic review.

Full text

Anatomical patella instability risk factors on MRI
show sensitivity without specificity in patients with
patellofemoral instability: a systematic review
Taylor J Ridley,
1
Betina Bremer Hinckel,
2
Bradley M Kruckeberg,
3
Julie Agel,
1
Elizabeth A Arendt
1
▸Additional material is
published online only. To view
please visit the journal online
(http://dx.doi.org/10.1136/
jisakos-2015-000015).
1
Department of Orthopaedic
Surgery, University of
Minnesota, Minneapolis,
Minnesota, USA
2
Medical School, University of
Minnesota, Minneapolis,
Minnesota, USA
3
Institute of Orthopedics and
Traumatology of the Clinical
Hospital, Medical School,
University of São Paulo,
Minneapolis, Minnesota, USA
Correspondence to
Dr Elizabeth A Arendt,
Department of Orthopaedic
Surgery, University of
Minnesota, 2450 Riverside
Av. So., Ste. R200,
Minneapolis 55454, MN, USA;
arend001@umn.edu
Received 8 February 2016
Revised 18 April 2016
Accepted 22 April 2016
To cite: Ridley TJ, Bremer
Hinckel B, Kruckeberg BM,
et al.JISAKOS Published
Online First: [please include
Day Month Year]
doi:10.1136/jisakos-2015-
000015
ABSTRACT
Purpose The purpose of this systematic review is to
assess whether the current literature distinguishes
between 9 common anatomic patellofemoral instability
(PFI) imaging measurements to determine if they are
consistent among the group of patients with PFI, and if
a diagnostic difference exists between the control and
patients with PFI in these imaging measurements.
Methods Following PRISMA guidelines, a systematic
review of PubMed and CINAHL/SPORTDiscus was
performed to identify all studies measuring PFI anatomic
imaging factors from January 1980 to February 2015.
Study inclusion criteria were: English language, MRI
studies, imaging of PFI anatomic factors. References of
the included articles were reviewed and those that met
inclusion criteria were added to the search results. PFI
group was defined as the study stating that the cohort
had at least 1 objective patient episode of lateral patellar
dislocation. Studies that did not report sample size, mean
and SD were excluded, as well as those that used
osseous (vs cartilage) landmarks, or who did not define
their measurement landmarks. Meta-analysis software
was used to determine weighted mean and 95% CI for
PFI and control groups for each measurement. Of the
135 articles identified, 22 met the inclusion criteria.
Results Caton-Deschamps patella height index and
trochlear depth measurement showed non-overlapping CIs
between the control and PFI groups. Insall-Salvati ratio,
lateral patella tilt, tibial tubercle-trochlear groove distance
and trochlear facet asymmetry, although having statistically
different means, had overlaped between the overall mean
95% CI of the 2 groups. Although few studies were
available for each measurement, the number of total
participants per measurement included is considerable: PFI
(range 40–685); controls (range 94–671).
Conclusions A systematic search of current literature
with meta-analysis reveals wide variation of PFI imaging
measurements used for clinical decision-making within
the controls and PFI groups. This systematic review
demonstrates that appropriate abnormality thresholds of
anatomic patella instability imaging factors exist for the
PFI group, indicating sensitivity. The wide variation in
the majority of measurements, especially in the control
group, suggests poor specificity in these measurements.
Trochlear depth has the best discrimination in evaluation
between the control and the PFI groups.
Level of evidence Level III, systematic review.
INTRODUCTION
Patellofemoral instability (PFI) is a debilitating con-
dition that primarily affects young athletes, being a
major cause of traumatic haemarthrosis in children.
Patellofemoral (PF) anatomy is complex due to
numerous bony dysplastic variations. Characterising
anatomic variations through imaging has identified
anatomic risk factors associated with patella
What is already known?
▸Anatomic risk factors associated with
patellofemoral instability are trochlear
dysplasia, patella alta and increased lateral
quadriceps vector.
▸Anatomic factors have historically been
evaluated by radiographs and physical
examination.
▸Risk factor measurements by radiographs are
different between patellofemoral disease and
control populations.
▸MRI permits visualisation of associated injuries
(meniscal or ligamentous) as well as functional
‘tendinous-cartilaginous’measurement of the
risk factors.
▸With the advent of digital imaging techniques
measuring to 0.01 sensitivity in linear
measurements, and the lack of consensus in
defining the landmarks points on MRI and/or
slice used for measurement, allows for greater
variability than was present in plain
radiographs.
What are the new findings?
▸Many studies do not clearly report the
measurement landmarks within their material
and methods.
▸There is a wide variability in MRI risk factor
measurements within and between the control
and patellofemoral groups.
▸On meta-analysis, the trochlear depth and
Caton-Deschamps index had no overlap in the
95% CI between the control and patellofemoral
groups.
▸Abnormality thresholds of patella instability risk
factor imaging measurements for the PFI group
appear to be appropriate, making these
imaging measurements sensitive without
specificity.
Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015 Copyright © 2016 ISAKOS 1
Systematic review
Journal of ISAKOS: Joint Disorders & Orthopaedic Sports Medicine Publish Ahead of Print, published on May 18, 2016 as doi:10.1136/jisakos-2015-000015
Copyright 2016 by International Society of Arthroscopy, Knee Surgery and Orthopaedic Sports Medicine.
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

instability. Measurement of these risk factors is central to clinical
treatment, that is, defining which anatomic abnormalities to
address with surgical correction. Three main anatomic risk
factors associated with PFI are trochlear dysplasia, patella alta
and increased lateral quadriceps vector.
1
These anatomic factors
have historically been evaluated by radiographs and physical
examination; however, advanced (slice) imaging techniques such
as CT, and recently MRI, have allowed measurements of these
anatomic factors to be made with more precision, as well as
more objective ways to characterise trochlea dysplasia. MRI
offers the additional advantage of identifying acute osteochon-
dral injuries, cartilage lesions, the location of medial PF liga-
ment injuries and other associated injuries (meniscal or ligament
injuries).
2–5
Moreover, MRI visualises cartilage structures enab-
ling a more anatomic representation of PF congruence as well as
a functional ‘tendinous-cartilaginous’measurement.
6–8
Though
studies of anatomic PFI factors by MRI measurements have
been analysed, the threshold measurements reported in the
initial studies of Dejour et al,
1
made on radiographs and CT
imaging, remain the gold standard for clinical decision-making.
It is not known if these historic threshold measurements are dir-
ectly transferable to MRI measurements.
To date, the most measured anatomic factor analysed by slice
imaging is the measurement of the lateral quadriceps angle or
tibial tubercle-trochlear groove (TT-TG) distance. Dejour et al
1
using CT, found an average TT-TG distance of 12.7±3.4 mm in
their control group and 19.8±1.6 mm in their PFI group; 3.5%
of the control group had values >20 mm while 56% of the PFI
group had values >20 mm. The authors concluded that 20 mm
was the threshold for pathological instability. This value has
stood as a clinical standard for decades. Recent studies involving
both CT and MRI measurements have shown mean TT-TG dis-
tance in patients with PFI to be considerably less than the his-
toric 20 mm pathological threshold.
9–15
For the controls,
including both CT and MRI measurements ‘normal’values have
been shown ranging from 9.4±0.6 to 14.4±2.9 mm.
116–19
Cartilage bone mismatch has been shown to be significant in
anatomical
20 21
and MRI studies.
21 22
Similarly, the tendinous-
cartilaginous measures of the TT-TG distance have differed
from the osseous ones, primarily due to the more lateral inser-
tion of the tendon relative to the more distal point of the bony
tibial tuberosity.
23 24
The purpose of this systematic review is to examine existing
literature of nine current anatomic PFI imaging measurements
using MRI in order to determine:
1. The number of articles reporting anatomic PFI imaging
measurements;
2. The consistency in the literature for each anatomic PFI
imaging measurements in patients with PFI and control
knees;
3. If a diagnostic difference exists between patients with PFI
and control knees for each measurement analysed.
MATERIALS AND METHODS
The literature review was carried out according to the PRISMA
guidelines (Registration #CRD42016029517).
25 26
Both a
PubMed and CINAHL/SPORTDiscus search was performed to
identify all available studies measuring imaging risk factors for
PFI using MRI from January 1980 until February 2015. The
PubMed search yielded 83 citations using keywords ‘patella(r)
dislocation’,‘patella(r) instability’,‘patella’,‘knee joint’,‘joint
instability’,‘magnetic resonance imaging’,‘MRI’,‘risk factors’,
and excluding any review articles, case studies, or surgical tech-
nique. A CINAHL with SPORTDiscus search using the same
keywords yielded 77 citations, 25 of them overlapping with the
PubMed search. Inclusion criteria from these searches were as
follows: English language, MRI studies and imaging of PFI ana-
tomic factors. References of the included articles were reviewed
and those that met inclusion criteria were added to the search.
Exclusion criteria were: heterogeneity of PFI group without a
subdivision of their results; heterogeneity of the control group
without a subdivision of their results; no mean/standard devi-
ation reported; measurements were not included in the review;
osseous landmarks or unclear landmark were used. Nine mea-
surements were included in our review: Insall-Salvati (IS) ratio,
modified IS (MIS) ratio, Caton-Deschamps (CD) index, lateral
patellar tilt, TT-TG distance, sulcus angle (SA), trochlear facet
asymmetry (TFA), trochlear depth (TD) and lateral trochlear
inclination angle (LTIA). Measurements of patella height,
TT-TG distance, SA and patellar tilt are well-established ana-
tomic risk factors associated with PFI.
27 28
MRI measurements
of trochlear dysplasia is best defined by TD, TFA and lateral
inclination angle
729
(table 1). Articles were included when the
MRI measurements used cartilaginous and tendinous reference
points for their measurement techniques rather than bone.
Osseous landmarks were accepted only for the IS ratio and
lateral patellar tilt measurements. Measurement technique was
assessed by the authors’description of the measurements, or by
viewing the measurements within the figures provided in
manuscript.
Reasons for exclusion were recorded and are specified in
figure 1. All articles were reviewed by two researchers and veri-
fied by a third.
The data were evaluated by creating a pool of data for each
anatomic PFI imaging measurement. Mean, SD and sample size
were recorded for each group: control group (patients without
PF symptoms/disease/or injury) and PFI group (patients with at
least one episode of lateral patellar dislocation; table 2). Owing
to inconsistency in the literature concerning patients with ‘sub-
luxation’, patients were included in the PFI group only if the
authors reported the definition of patella instability to be at
least one patient episode of objective lateral patellar dislocation.
MRIs were reportedly taken in full or near-full extension in a
majority of studies, although this was not specified by all papers
and was not an exclusion criterion.
For the TT-TG distance, 16 studies were excluded for either
using the tibial tubercle as an osseous landmark (5
studies),
18 30–33
or no clear description of measurement para-
meters reported (11 studies).
9–11 14 15 27 34–38
For the trochlear measurements (SA, TD, facet asymmetry
and LTIA), 14 studies were excluded from at least one of the
measurements for using osseous landmarks or due to the lack of
a clear description of measurement parameters
reported.
37911 12293032343538–41
All of the control data were from patients with knee symp-
toms other than PF; for example, meniscal, cartilage and/or liga-
mentous injuries.
69–12 15 17 30 37 38 42–44
Only two papers used
a strict definition of normals (asymptomatic volunteers).
41 45
Mean age and sex were usually reported, but the data were
rarely stratified by these variables. Balcarek et al
9
reported
means stratified by patient sex, Salzmann et al
22
reported means
stratified by trochlear dysplasia classification and Seeley et al
32
reported means stratified by the presence or absence of an
injury recurrence; the stratified results were analysed as separate
populations.
Weighted means were determined using Open Meta Analyst
software from Brown University’s Center for Evidence-Based
Medicine.
44
A continuous random-effect model was used that
2 Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

weights means by the DerSimonian-Laird method (1986). For
each measure, meta-analyses were performed on the PFI group
and the control group with differences reported. For each vari-
able, determination of a weighted mean required that there be
at least two studies available that contained measurements from
the PFI and control groups.
All analyses were reviewed for heterogeneity. All met the
assumptions of heterogeneity except TD for the control group
Table 1 MR measurements of knee anatomic factors for PF instability (University of MN)
Measurement Description Figure
IS ratio Measure on sagittal cut with greatest patellar length. Line (B) the longest sagittal
diameter of the patella; line (A) from lower point of line B to superior aspect of insertion
of patellar tendon on tibial tuberosity.
IS ratio=A÷B
CD index Measure on sagittal cut with greatest patellar length. Line (D) measurement of the length
of the cartilage articular surface of the patella; line (C) from lower point of line D to the
anterior corner of the superior tibial joint surface.
CD index=C÷D
Measure on sagittal cut with greatest patellar length. Line (D) measurement of the length
of the patella cartilage articular surface. Strike a reference line at 90° to inferior end of
line D. Line (E) parallel to line (D), most superior femoral articular cartilage to reference
line.
PTI=E÷D
TT-TG distance
(mm)
Measure on axial view (requires 2 slice imaging views). Baseline (F) across posterior
condyles at the level of the most proximal slice where there is cartilage across both
trochlear condyles; line (A) at 90° to baseline through the deepest point of TG cartilage.
Superimpose a second image of TT; line (B) parallel to line (A) at midline of first cut
where patellar tendon insertion onto tibia. Measure distance (C) between lines (A)
and (B).
TT-TG distance=C
Sulcus angle
(degrees)
Measure on axial view at the level of the most proximal slice where there is cartilage
across both trochlear condyles. Angle formed by the 2 lines (D and E) from the deepest
Continued
Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015 3
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

and MIS ratio in the control and PFI groups. MIS ratio (n=2)
and LTI (n=3) were unable to have heterogeneity assessed due
to the small number of papers.
RESULTS
Twenty-one studies contributing at least one measurement of the
anatomic factors were included in the meta-analyses. The mea-
surements from each individual study can be found in table 3.
IS ratio
IS ratio was reported in 8 studies (N=671) for the control group
and 10 studies (N=685) for the PFI group. Mean IS ratio ranged
from 1.02–1.24 in the control group to 1.18–1.40 in the
PFI group. The control group weighted IS ratio was 1.10
(95% CI 1.06 to 1.15). PFI weighted IS ratio was 1.25 (95%
CI 1.22 to 1.29). The weighted mean difference was 0.16 (95%
CI 0.14 to 0.19).
Table 1 Continued
Measurement Description Figure
point of the TG cartilage and extending to highest points of lateral and medial trochlear
facets cartilage.
TD (mm) Measure on axial view at the level of the most proximal slice where there is cartilage
across both trochlear condyles. Drop lines at 90° to baseline (F) along posterior condyles.
Average the lengths of the medial (A) and lateral (C) trochlear facets cartilaginous surface,
and subtract the length of the TG (B).
TD=((A+C)÷2)−B
TFA Measure on axial view at the level of the most proximal slice where there is cartilage
across both trochlear condyles. Measure the distance between the deepest point of the TG
to the highest point of the medial trochlear facet cartilage and to the highest point to the
lateral facet cartilage. Calculate the ratio of medial trochlear facet length (D) to lateral
trochlear facet length (E).
TFA=D÷E
LTI angle
(degrees)
Measure on axial view at the level of the most proximal slice where there is cartilage
across both trochlear condyles. Baseline (F) tangent to the posterior aspect of femoral
condyles. Line (G) tangent to the cartilage of the lateral facet. Angle (H) measured
between lines (F) and (G).
LTI angle=H°
PT (degrees) Measure on axial view (may require 2 slice imaging views). Baseline (F) across posterior
condyles; angle line (G) at the cut with the greatest patellar width through bony medial/
lateral patella midpoint (may be on another MR slice).
PT=H°
CD, Caton-Deschamps; IS, Insall-Salvati; LTIA, lateral trochlear inclination; MN, Minnesota; PF, patellofemoral; PT, patellar tilt; PTI, patellotrochlear index; TD, trochlear depth; TFA,
trochlear facet asymmetry; TT-TG, tibial tubercle-trochlear groove.
4 Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

For the IS ratio, two of the control population had all their
measurements outside the 95% CI of the overall control mean.
Two of the PFI population had 95% CIs where the lower
bound was within the 95% CI for the control population
(figure 2).
MIS ratio studies were reported in two studies (N=124) for
the control group and one study
11
in the PFI group. Mean MIS
ratio ranged from 1.64 to 1.65 in the control group versus 1.72
in the PFI group. The weighted mean was 1.64 (95% CI 1.63 to
1.65) for the control group (figure 3).
CD index
CD index was reported in four studies (N=353) for the control
group and in three studies (N=361) for the PFI group. Mean
CD ratio ranged from 0.92–1.13 in the control group to 1.18–
1.29 in the PFI group. Meta-analysis revealed a weighted mean
ratio of 1.03 (95% CI 0.93 to 1.13) and 1.24 (95% CI 1.17 to
1.31) in the control and PFI groups, respectively. The weighted
mean difference was 0.19 (95% CI 0.13 to 0.24).
There was no overlap in CD ratio between the PFI group and
the control group (figure 4).
Patella tilt
Patella tilt was reported in five studies (N=239) for the control
group and two studies (N=175) for the PFI group. Mean
patella tilt ranged from 6.68° to 17.0° in the control group
while the mean range was 15.38°–24.03°. Control group
weighted mean angle was 9.19° (95% CI 6.58° to 11.8°). PFI
group weighted mean angle was 19.7° (95% CI 11.3° to 28.2°).
Pub Med
83
CINAHL
77
21
studies used
39 studies
whole
paper evaluated
160
135
abstracts
evaluated
25 excluded for overlap
Did not meet inclusion criteria
• 13 No MRI
• 20 Non English
• 63 Without PFI MRI measurement
variables
• 96 Total
Inclusions and exclusions
• 14 Included by refere nce review
• Exclusions
– 7 Heterogeneous PF disease* (patellar
subluxation or patellar pain instead of
instability; controls together with
patients with patellofemoral disorders
– 15 No mean / standard deviation
– 3 Measurement not included in review
– 7 MRI methodology not defined or not
consistent with out stated criteria
*Unable to isolate homogeneous cohort
Figure 1 Flow chart of articles during the selection process. PFI,
patellofemoral index.
Table 2 Articles selected that included populations for each MRI anatomic instability imaging measurement in the control and PFI groups: total
21 studies
Article (year)
Control group
(sample size)
PFI group
(sample size) IS ratio
Modified
IS ratio CD index Patellar tilt TT-TG distance SA TFA TD LTIA
Zaidi et al
40
26 Y
Toms et al
46
44 Y
Balcarek et al
38
73 73 Y
Balcarek et al
9
149 99 Y
Salzmann et al
22
21 Y
Pandit et al
17
100 Y
Nicolaas et al
14
51 Y Y
Petri et al
35
40 Y Y Y Y
Wagner et al
36
50 Y
Seeley et al
32
111 Y Y Y Y Y
Guilbert et al
15
45 135 Y Y
Charles et al
11
81 40 Y Y Y Y
Kohlitz et al
12
186 186 Y Y Y Y
Regalado et al
41
19 Y Y
Mehl et al
30
43 Y Y Y Y Y Y
Hingelbaum et al
42
200 54 Y
Izadpanah et al
45
8 Y
Dornacher et al
43
38 61 Y
Steensen et al
10
120 60 Y
Skelley et al
37
53 63 Y
Dickens et al
27
495 76 Y
CD, Caton-Deschamps; IS, Insall-Salvati; LTI, lateral trochlear inclination; PFI, patellofemoral index; SA, sulcus angle; trochlear depth; TFA, trochlear facet asymmetry;
TT-TG, tibial tubercle-trochlear groove.
Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015 5
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

Table 3 Summary of MRI measurements in the control versus patellofemoral groups
Measurement Group Participants/studies Mean (95% CI)
IS* Control 671 (8) 1.10 (1.06 to 1.15)
PFI 685 (10) 1.25 (1.22 to 1.29)
Difference 1067†(6) 0.16 (0.14 to 0.19)
Modified Insall-Salvati ratio* Control 124 (2) 1.64 (1.63 to 1.65)
PFI
Difference
40 (1) 1.72‡NA
NA§
CD* Control 353 (4) 1.03 (0.93 to 1.13)
PFI 361 (3) 1.24 (1.17 to 1.31)
Difference 671†(3) 0.19 (0.13 to 0.24)
Patella tilt (degrees) Control 239 (5) 9.19 (6.58 to 11.8)
PFI 175 (2) 19.7 (11.3 to 28.2)
Difference 301†(2) 12.4 (5.30 to 19.5)
TT-TG distance (mm) Control 841 (5) 9.23 (8.22 to 10.2)
PFI 191 (3) 13.9 (11.6 to 16.1)
Difference 924†(3) 5.07 (3.17 to 6.98)
Sulcus angle (degrees) Control 94 (2) 149 (136 to 162)
PFI 216 (7) 157 (152 to 162)
Difference NA§
Trochlear facet asymmetry (%)* Control 227 (2) 62.0 (55.2 to 68.8)
PFI 337 (4) 47.7 (42.3 to 53.0)
Difference NA§
Trochlear depth (mm)* Control 239 (2) 4.52 (4.38 to 4.65)
PFI 360 (4) 2.59 (2.21 to 2.96)
Difference 488†(2) 2.19 (1.72 to 2.67)
Lateral trochlear inclination angle (degrees) Control 43 (1) 20.6‡NA
PFI 111 (2) 13.7 (11.9 to 15.4)
Difference NA§
Calculated means for: IS ratio, modified IS ratio, CD index, patella tilt, TT-TG distance, sulcus angle, trochlear facet asymmetry, trochlear depth and lateral trochlear inclination angle.
*No overlap of CI between two groups.
†Sum of controls and PFI participants.
‡One study reported, meta-analysis not available.
§No meta-analysis possible; <2 studies with both control and patients with PFI.
CD, Caton-Deschamps; IS, Insall-Salvati; NA, not available; PFI, patellofemoral index; TT-TG, tibial tubercle-trochlear groove.
Figure 2 Graphical representation
using forest plots for Insall-Salvati ratio
including the control and PFI groups.
PFI, patellofemoral index.
6 Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

The weighted mean difference was 12.4° (95% CI 5.30° to
19.6°).
One of the control populations was completely within the
95% CI for the PFI population interval (figure 5).
TT-TG distance
TT-TG distance was reported in five studies (N=841) for con-
trols and three studies (N=191) for the PFI group. Mean
TT-TG distance ranged from 7.5 to 11.6 mm in the control
group and ranged from 12.2 to 16.0 mm in the PFI group. The
weighted mean distance for the control group was 9.23 mm
(95% CI 8.22 to 10.2), while the PFI group weighted mean dis-
tance was 13.9 mm (95% CI 11.6 to 16.1). The weighted mean
difference was 5.07 mm (95% CI 3.17 to 6.98).
One of the control population had the upper bound of the
95% CI within the lower bound of the 95% CI for the PFI
group (figure 6).
Sulcus angle
Two studies reported the mean SA for controls (N=94) and
seven studies reported PFI groups (N=216). Mean SA ranged
from 143° to 156° in the control group and 147° to 169° in the
PFI group. Weighted mean SA was 149° (95% CI 136° to 162°)
for the control group and 157° (95% CI 152° to 162°) for the
PFI group.
The 95% CI for the control population was extremely wide
such that five of the PFI populations were within most of the
mean control CI (figure 7).
Trochlear facet asymmetry
Two studies reported mean facet asymmetry in the control
group (N=227) and four studies in the PFI group (N=337).
The mean TFA in the control group ranged from 58.4% to
65.3% and from 41.1% to 52.7% for the PFI group. Weighted
mean was 62.0% (95% CI 55.2% to 68.8%) and 47.61%
(95% CI 42.3% to 53.0%) for the control and PFI groups,
respectively. No weighted mean difference was available, as
there was only one study with both control and patients with
PFI.
Two studies have the 95% CI upper limit for the PFI group
overlapping the lower 95% CI of the control group. The two
studies for the control group have 95% CI that do not cross the
mean (figure 8).
Trochlear depth
TD was reported in two studies for the control group (N=239)
and four studies for the PFI group (N=360). Mean depth in the
control group ranged from 4.5 to 4.6 mm and from 2.1 to
3.09 mm in the PFI group. Control group weighted mean dis-
tance was 4.52 mm (95% CI 4.38 to 4.65) while the PFI group
weighted mean distance was 2.59 mm (95% CI 2.21 to 2.96).
Figure 3 Graphical representation
using forest plots for modified
Insall-Salvati ratio including the control
and PFI groups. NA, not available;
PFI, patellofemoral index.
Figure 4 Graphical representation
using forest plots for Caton-Deschamps
index including the control and PFI
groups. PFI, patellofemoral index.
Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015 7
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

The weighted mean difference was 2.19 mm (95% CI 1.72 to
2.67).
There is no overlap between the control and PFI groups
(figure 9).
Lateral trochlear inclination angle
One study was found to report on LTIA in controls with a mean
of 20.6° (N=43), and one study reported on LTIA in the PFI
group (N=111). Mean LTIA was reported in two subgroups
(presence or absence of recurring instability) in the PFI group
and ranged from 12.7° to 14.5°. PFI subgroups weighted mean
inclination angle was 13.7° (95% CI 11.9° to 15.4°; figure 10).
DISCUSSION
MRI measurements of PF anatomy are central to current clinical
treatment algorithms. The measurements chosen for inclusion in
this systematic review are historic PF imaging measurements as
well as newer measurements of trochlear shape defined by slice
imaging. Surgical technique has evolved in part due to better
understanding of the anatomy and biomechanics of the PF joint.
Commensurate with this improved anatomic knowledge aiding
surgical techniques, improved anatomic imaging also helps to
better define and stratify the injured population, helping to
refine outcome data leading to better patient care. MRI better
defines the articulating (cartilaginous) joint surfaces and tendin-
ous ligament insertion sites, rendering a more true anatomic
representation.
A review of studies that had at least one of the nine anatomic
imaging measurements revealed wide variability within and
between the control and PFI groups. TD and CD index were
the only two measurements that demonstrated no overlap
between the control and PFI groups, although both of these
measurements had one article/population that demonstrated a
mean and SD outside the 95% CI for the collective group. Two
other measurements demonstrated no overlap between the PFI
and control groups, but the sample sizes for these two measure-
ments were too small to render conclusions on their data (MIS
ratio and LTIA).
Central to the clinical use of imaging measurements is their
intervariability and intravariability values. Central to the clinical
use of imaging measurements is their intervariability and intra-
variability values. This has been reported in previous literature
for all measurements used.
811232447–51
Collectively these studies show that most clinicians can reli-
ably make these measurements. However, reliability of MRI
(slice) measurement depends on whether you are evaluating the
measuring technique of an agreed-on slice, or whether different
evaluators are choosing different slices. This can be most illus-
trative when measuring the SA; a 3 mm (typical slice thickness)
difference in a dysplastic proximal trochlea can result in a sig-
nificant difference in measured values. Of those papers report-
ing variability, none reported the method used to assess this
variability.
All measurements met the assumptions of heterogeneity
except trochlea depth for the control group and MIS ratio in
Figure 5 Graphical representation
using forest plots for patellar tilt
including the control and PFI group.
PFI, patellofemoral index.
Figure 6 Graphical representation
using forest plots for TT-TG including
the control and PFI groups. PFI,
patellofemoral index; TT-TG, tibial
tubercle-trochlear groove.
8 Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

the control and PFI groups, suggesting that for these two meas-
urement variables demonstrate more difference from each other
than would be expected by chance.
Two of the patellar height measures included in this study
were historically measured on true lateral radiographs (IS ratio
and CD index). There are reference ranges for patellar height
ratios measured on radiograph, with patella alta defined as CD
index >1.2,
52
IS ratio >1.2.
28 53 54
There is not yet consensus
on the reference range for normal patellar height on CT and
MRI. Published MRI studies consider IS ratio’s upper threshold
greater than measurements made on plain radiographs.
51 53
IS ratio was the most studied measurement variable that fit
our inclusion criteria. Within the IS ratio analysis, nine data sets
(five in the control group and four in the PFI group) had popu-
lations that did not cross the mean of their groups. The clinical
significance of this is debatable, but there are data sets where
the study’s entire population is an outlier to the combined
data set.
The studies measuring CD index showed good discrimination
within our population data set, suggesting CD index represents
true mean differences between the control and PFI groups.
Patella tilt has only two populations within the PFI group,
with large disparity between their mean values. Within the
control group, four of five data sets do not cross the mean of
the combined control group. The study by Guilbert et al
15
has
the largest cohort (n=180) in the control and PFI groups. Their
PFI group has no values crossing the traditional threshold of
pathological lateral patella tilt, that is, 20°.
TT-TG distance measured on CT or MRI and using the patel-
lar tendon or the tibial tubercle are not equivalent.
Measurements made using the tibial tubercle (bony landmark)
versus the patella tendon (tendinous landmark) are smaller by
1.0–2.9 mm;
23 24
MRI measurements (vs the CT) are smaller by
3.1–0.36 mm.
24 48
We chose to include only studies that used
the patella tendon as the distal landmark as the MRI has the
ability to image the soft tissue showing the true tendon inser-
tion. Current studies in the literature note that TT-TG dis-
tance measurements are strongly dependent on patient
positioning and that the placement of the knee in slight
flexion
24 55
and relative varus limb alignment
47
results in a
lower TT-TG distance. A controlled patient positioning proto-
col is encouraged to lessen variability among patients and
from centre to centre.
The results of this systematic review reveal that the average
TT-TG distance measurements between the PFI and control
groups were distinct; however, there was not a clear threshold
for increased risk of PFI. For the control population, one study
set with only eight participants has a very wide CI skewing the
data set. Removing this study would allow better discrimination
between controls and PFI groups for this measurement variable;
eliminating this study, the remaining studies have non-
overlapping CI.
MRI measurement of the TT-TG distance using the patellar
tendon as the distal reference results in smaller values, which
may result in a decrease of the abnormality threshold for this
anatomic factor.
49
Indeed, all of the values of the PFI group
Figure 7 Graphical representation
using forest plots for sulcus angle
including the control and PFI groups.
PFI, patellofemoral index.
Figure 8 Graphical representation
using forest plots for trochlear facet
asymmetry including the control and
PFI groups. PFI, patellofemoral index.
Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015 9
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

were under 18 mm, with the traditional upper value based on
CT measurements as 20 mm.
1
In concordance with previous studies utilising both CT and
MRI, our study has demonstrated a difference in the TT-TG
distance between control and patients with PFI with calculated
means of 9.23 mm and 13.9 mm, respectively. TT-TG distance
measurement value is not without its limitations. The reli-
ability of TT-TG distance and its usefulness depends on a
variety of factors. TT-TG distance has been observed to vary
among imaging modality, patient positioning and anatomic
landmarks.
23 24 47 48 55 56
Trochlear dysplasia has long been recognised as a risk factor
for PF instability.
157
It can be classified based on qualitative fea-
tures on the radiographs (crossing sign, supratrochlear bump
and double contour) by Dejour et al
1
classification or by quanti-
tative measures such as SA, TFA, TD and LTIA.
71129
In this review, trochlear dysplasia was defined by the SA, TD,
TFA and LTIA. No meta-analysis could be performed for the
LTIA due to the paucity of studies measuring this variable using
cartilaginous landmarks (control group n=2, PFI group n=1).
SA is the simplest of measurements to make; however, there is
potential for wide variation depending on which slice of the
proximal trochlea one uses. Within our data set, the control
group (n=2) showed a wide discrepancy in their mean values.
SAs within the PFI group are larger (representing a shallower
sulcus), with a mean SA of 157° for the PFI group. Larger
samples in the control group could increase the power of the
meta-analysis and show a difference in future analyses.
TD and TFA had larger samples; both measurements showed
statistically different weighted means between the control and
PFI groups. These results support the use of the TD and TFA to
differentiate dysplastic trochlea in patients with PFI from con-
trols patients. However, for TFA, three of four PFI studies have
95% CI that do not cross the overall mean ( figure 8); thus, there
is little consistency within and between each study population.
A review of studies that had at least one of the nine measure-
ments evaluated demonstrated wide variability with the control
and PFI groups. Patellar height measured by CD index and TD
were the only two measurements that demonstrated no overlap
between the control and PFI groups. With the exception of the
SA, all of the overlap was the control population crossing into
the PFI suggesting wide variation in the control ‘normal popula-
tion’. This means there is a subset of controls without symptom-
atic PFI whose anatomic instability factors suggest PFI. The
reverse seems not to be true, that is, there are normal people
who have abnormal (PFI) measurements. This indicates that the
anatomic imaging measurements for PFI demonstrate good sen-
sitivity without specificity.
The clinical algorithm for surgical management is dependent
on identifying dysplastic anatomy as defined by imaging mea-
surements. MRI is becoming the preferred slice-imaging tool
rather than CT scans in part due to its usefulness in identifying
cartilage status, the location of medial PF ligament injuries and
other associated injuries (meniscal or ligament injuries). With
increasing use of MRI as a tool to aid clinical assessment, docu-
mentation of the range of clinical measurements and their
thresholds in patients with PF instability versus controls is a
necessary component to optimise surgical decision-making in
patients with PF instability.
Limitations
The sample size could be greatly increased if studies clearly
reported the measurement landmarks within their material and
methods. Differences between males and females have been
shown,
912
as well as differences between mature and immature
or patient’s age.
27
These analyses could not distinguish
Figure 9 Graphical representation
using forest plots for trochlear depth
including the control and PFI groups.
PFI, patellofemoral index.
Figure 10 Graphical representation
using forest plots for LTIA including
the control and PFI groups. LTIA,
lateral trochlear inclination angle; NA,
not available; PFI, patellofemoral
index.
10 Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

between these subsets due to the lack of stratification in most
studies.
Position of the knee (knee flexion and alignment) is not
always reported and can provide non-comparable
measurements.
24 47 55 56
Some of the ‘wideness’of the results is more apparent due
the precision of the individual measurements; typically linear
measurements were reported to 0.01 mm while angular mea-
surements were reported as whole numbers. Current MRI tech-
nology allows for more precision which precipitates more
potential variation.
Populations were not similar across all measurements.
The overall mean difference was calculated when there were
paired articles; the graphs presented the data on measurements
in the available literature that met our inclusion criteria.
Some studies were stratified, for example, sex, trochlear dys-
plasia classification; we combined them with non-stratified
populations for these analyses.
CONCLUSION
This systematic review is the first step towards refining MRI
measurement thresholds used in clinical management for surgi-
cal patella stabilisation, helping to define which dysplastic ele-
ments should be surgically addressed for optimal outcomes. To
accomplish this goal, the literature was reviewed using strict
inclusion criteria within two populations: PFI group and a
control group (non-injured). Non-overlapping CIs between
these groups would help establish the value of a particular risk
factor measurement.
TD, measured by slice imaging on MRI, has the best discrim-
ination between the control and PFI groups.
Taken as a whole, there appears to be appropriate abnormal-
ity thresholds of patella instability risk factor measurements for
the PFI group, making these measurements sensitive without
specificity. Many in the control population have an abnormal
measurement without disease; the challenge remains in working
towards good sensitivity in these measurements with improved
specificity. Agreement between PF clinicians on the most useful
imaging measurements and the most accurate measurement
methodology would help improve the literature in this field.
Journal editors mandating reporting of agreed-on measurements
in scientific papers are needed. A centralised repository with
standardised guidelines for measurements of each of these ana-
tomic factors would help to create a database of sufficient size
to better understand the value of imaging measurements in clin-
ical decisions for surgical stabilisation.
Acknowledgements The authors wish to acknowledge the assistance of Paul
Lender in the statistical analysis and creation of the figures.
Contributors BBH and EAA contributed to the concept and design of the work.
TJR, BBH, BMK, JA and EAA were involved in acquisition/analysis/interpretation of
the data. BBH, JA and EAA were involved in drafting/revising the work. EAA was
involved in final approval for accuracy and integrity of the work.
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed.
REFERENCES
1 Dejour H, Walch G, Nove-Josserand L, et al. Factors of patellar instability: an
anatomic radiographic study. Knee Surg Sports Traumatol Arthrosc 1994;2:19–26.
2 Guerrero P, Li X, Patel K, et al. Medial patellofemoral ligament injury patterns and
associated pathology in lateral patella dislocation: an MRI study. Sports Med Arthrosc
Rehabil Ther Technol 2009;1:1–17.
3 Balcarek P, Walde TA, Frosch S, et al. Patellar dislocations in children, adolescents
and adults: a comparative MRI study of medial patellofemoral ligament injury
patterns and trochlear groove anatomy. Eur J Radiol 2011;79:415–20.
4 Seeley MA, Knesek M, Vanderhave KL. Osteochondral injury after acute
patellar dislocation in children and adolescents. J Pediatr Orthop 2013;33:
511–18.
5 Farr J, Covell DJ, Lattermann C. Cartilage lesions in patellofemoral dislocations:
incidents/locations/when to treat. Sports Med Arthrosc 2012;20:181–6.
6 Biedert RM, Bachmann M. Anterior-posterior trochlear measurements of normal and
dysplastic trochlea by axial magnetic resonance imaging. Knee Surg Sports
Traumatol Arthrosc 2009;17:1225–30.
7Pfirrmann CW, Zanetti M, Romero J, et al. Femoral trochlear dysplasia: MR findings.
Radiology 2000;216:858–64.
8 Schoettle PB, Zanetti M, Seifert B, et al. The tibial tuberosity-trochlear groove
distance; a comparative study between CT and MRI scanning. Knee
2006;13:26–31.
9 Balcarek P, Jung K, Ammon J, et al. Anatomy of lateral patellar instability: trochlear
dysplasia and tibial tubercle-trochlear groove distance is more pronounced in
women who dislocate the patella. Am J Sports Med 2010;38:2320–7.
10 Steensen RN, Bentley JC, Trinh TQ, et al. The prevalence and combined prevalences
of anatomic factors associated with recurrent patellar dislocation: a magnetic
resonance imaging study. Am J Sports Med 2015;43:921–7.
11 Charles MD, Haloman S, Chen L, et al. Magnetic resonance imaging-based
topographical differences between control and recurrent patellofemoral instability
patients. Am J Sports Med 2013;41:374–84.
12 Köhlitz T, Scheffler S, Jung T, et al. Prevalence and patterns of anatomical risk
factors in patients after patellar dislocation: a case control study using MRI.
Eur Radiol 2013;23:1067–74.
13 Anley CM, Morris GV, Saithna A, et al.Defining the role of the tibial
tubercle-trochlear groove and tibial tubercle-posterior cruciate ligament distances in
the work-up of patients with patellofemoral disorders. Am J Sports Med
2015;43:1348–53.
14 Pennock AT, Alam M, Bastrom T. Variation in tibial tubercle-trochlear groove
measurement as a function of age, sex, size, and patellar instability. Am J Sports
Med 2014;42:389–93.
15 Guilbert S, Chassaing V, Radier C, et al. Axial MRI index of patellar engagement:
a new method to assess patellar instability. Orthop Traumatol Surg Res 2013;99:
S399–405.
16 Alemparte J, Ekdahl M, Burnier L, et al. Patellofemoral evaluation with radiographs
and computed tomography scans in 60 knees of asymptomatic subjects.
Arthroscopy 2007;23:170–7.
17 Pandit S, Frampton C, Stoddart J, et al. Magnetic resonance imaging assessment of
tibial tuberosity-trochlear groove distance: normal values for males and females.
Int Orthop 2011;35:1799–803.
18 Balcarek P, Jung K, Frosch KH, et al. Value of the tibial tuberosity-trochlear groove
distance in patellar instability in the young athlete. Am J Sports Med
2011;39:1756–61.
19 Tensho K, Akaoka Y, Shimodaira H, et al. What components comprise the
measurement of the tibial tuberosity-trochlear groove distance in a patellar
dislocation population? J Bone Joint Surg Am 2015;97:1441–8.
20 Shih YF, Bull AM, Amis AA. The cartilaginous and osseous geometry of the femoral
trochlear groove. Knee Surg Sports Traumatol Arthrosc 2004;12:300–6.
21 Staeubli HU, Bosshard C, Porcellini P, et al. Magnetic resonance imaging for
articular cartilage: cartilage-bone mismatch. Clin Sports Med 2002;21:
417–33, viii-ix.
22 Salzmann GM, Weber TS, Spang JT, et al. Comparison of native axial radiographs
with axial MR imaging for determination of the trochlear morphology in patients
with trochlear dysplasia. Arch Orthop Trauma Surg 2010;130:335–40.
23 Wilcox JJ, Snow BJ, Aoki SK, et al. Does landmark selection affect the reliability of
tibial tubercle-trochlear groove measurements using MRI? Clin Orthop Relat Res
2012;470:2253–60.
24 Hinckel BB, Gobbi RG, Filho EN, et al. Are the osseous and tendinous-cartilaginous
tibial tuberosity-trochlear groove distances the same on CT and MRI? Skeletal
Radiol 2015;44:1085–93.
25 Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews
and meta-analyses: the PRISMA statement. PLoS Med 2009;6:e1000097.
26 Shamseer L, Moher D, Clarke M, et al. Preferred reporting items for systematic
review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation.
BMJ 2015;349:g7647.
27 Dickens AJ, Morrell NT, Doering A, et al. Tibial tubercle-trochlear groove distance:
defining normal in a pediatric population. J Bone Joint Surg Am 2014;96:
318–24.
28 Insall J, Salvati E. Patella position in the normal knee joint. Radiology
1971;101:101–4.
29 Carrillon Y, Abidi H, Dejour D, et al. Patellar instability: assessment on MR images
by measuring the lateral trochlear inclination-initial experience. Radiology
2000;216:582–5.
30 Mehl J, Feucht MJ, Bode G, et al. Association between patellar cartilage defects
and patellofemoral geometry: a matched-pair MRI comparison of patients with and
without isolated patellar cartilage defects. Knee Surg Sports Traumatol Arthrosc
2016;24:838–46.
Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015 11
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

31 Seitlinger G, Scheurecker G, Högler R, et al. The position of the tibia tubercle in 0
degrees -90 degrees flexion: comparing patients with patella dislocation to healthy
volunteers. Knee Surg Sports Traumatol Arthrosc 2014;22:2396–400.
32 Seeley M, Bowman KF, Walsh C, et al. Magnetic resonance imaging of acute
patellar dislocation in children: patterns of injury and risk factors for recurrence.
J Pediatr Orthop 2012;32:145–55.
33 Park HJ, Ahn JH, Kim SS, et al. A new assessment of patellar instability using
coronal magnetic resonance images of the patella superimposed on the femur and
its clinical utility. J Comput Assist Tomogr 2013;37:470–4.
34 Seitlinger G, Scheurecker G, Högler R, et al. Tibial tubercle-posterior cruciate
ligament distance: a new measurement to define the position of the tibial tubercle
in patients with patellar dislocation. Am J Sports Med 2012;40:1119–25.
35 Petri M, von Falck C, Broese M, et al.Influence of rupture patterns of the medial
patellofemoral ligament (MPFL) on the outcome after operative treatment of
traumatic patellar dislocation. Knee Surg Sports Traumatol Arthrosc 2013;21:683–9.
36 Wagner D, Pfalzer F, Hingelbaum S, et al. The influence of risk factors on clinical
outcomes following anatomical medial patellofemoral ligament (MPFL) reconstruction
using the gracilis tendon. Knee Surg Sports Traumatol Arthrosc 2012;21:318–24.
37 Skelley N, Friedman M, McGinnis M, et al. Inter- and intraobserver reliability in the
MRI measurement of the tibial tubercle-trochlear groove distance and trochlea
dysplasia. Am J Sports Med 2015;43:873–8.
38 Balcarek P, Ammon J, Frosch S, et al. Magnetic resonance imaging characteristics of
the medial patellofemoral ligament lesion in acute lateral patellar dislocations
considering trochlear dysplasia, patella alta, and tibial tuberosity-trochlear groove
distance. Arthroscopy 2010;26:926–35.
39 Keser S, Savranlar A, Bayar A, et al. Is there a relationship between anterior knee
pain and femoral trochlear dysplasia? Assessment of lateral trochlear inclination by
magnetic resonance imaging. Knee Surg Sports Traumatol Arthrosc
2008;16:911–15.
40 Zaidi A, Babyn P, Astori I, et al. MRI of traumatic patellar dislocation in children.
Pediatr Radiol 2006;36:1163–70.
41 Regalado G, Lintula H, Eskelinen M, et al. Dynamic KINE-MRI in patellofemoral
instability in adolescents. Knee Surg Sports Traumatol Arthrosc 2014;22:2795–802.
42 Hingelbaum S, Best R, Huth J, et al. The TT-TG Index: a new knee size adjusted
measure method to determine the TT-TG distance. Knee Surg Sports Traumatol
Arthrosc 2014;22:2388–95.
43 Dornacher D, Reichel H, Lippacher S. Measurement of tibial tuberosity-trochlear
groove distance: evaluation of inter- and intraobserver correlation dependent on the
severity of trochlear dysplasia. Knee Surg Sports Traumatol Arthrosc 2014;22:2382–7.
44 Nicolaas L, Tigchelaar S, Koëter S. Patellofemoral evaluation with magnetic
resonance imaging in 51 knees of asymptomatic subjects. Knee Surg Sports
Traumatol Arthrosc 2011;19:1735–9.
45 Izadpanah K, Weitzel E, Vicari M, et al.Influence of knee flexion angle and
weight bearing on the tibial tuberosity-trochlear groove (TTTG) distance for
evaluation of patellofemoral alignment. Knee Surg Sports Traumatol Arthrosc
2014;22:2655–61.
46 Toms AP, Cahir J, Swift L, et al. Imaging the femoral sulcus with ultrasound, CT,
and MRI: reliability and generalizability in patients with patellar instability. Skeletal
Radiol 2009;38:329–38.
47 Ho CP, James EW, Surowiec RK, et al. Systematic technique-dependent differences
in CT versus MRI measurement of the tibial tubercle-trochlear groove distance.
Am J Sports Med 2015;43:675–82.
48 Camp CL, Stuart MJ, Krych AJ, et al. CT and MRI measurements of tibial
tubercle-trochlear groove distances are not equivalent in patients with patellar
instability. Am J Sports Med 2013;41:1835–40.
49 Hinckel BB, Gobbi RG, Kihara Filho EN, et al. Patellar tendon–trochlear groove
angle measurement: a new method for patellofemoral rotational analyses. Orthop J
Sports Med 2015;3.
50 Lee PP, Chalian M, Carrino JA, et al. Multimodality correlations of patellar height
measurement on X-ray, CT, and MRI. Skeletal Radiol 2012;41:1309–14.
51 Shabshin N, Schweitzer ME, Morrison WB, et al. MRI criteria for patella alta and
baja. Skeletal Radiol 2004;33:445–50.
52 Caton JH, Dejour D. Tibial tubercle osteotomy in patello-femoral instability and in
patellar height abnormality. Int Orthop 2010;34:305–9.
53 Miller TT, Staron RB, Feldman F. Patellar height on sagittal MR imaging of the
knee. AJR Am J Roentgenol 1996;167:339–41.
54 Dejour D, Saggin PR. Disorders of the patellofemoral joint. In: Scott W, ed. Insall &
Scott surgery of the knee-5th edition. Philadelphia: Elsevier-Churchill Livingstone,
2012:592–623.
55 Dietrich TJ, Betz M, Pfirrmann CW, et al. End-stage extension of the knee and its
influence on tibial tuberosity-trochlear groove distance (TTTG) in asymptomatic
volunteers. Knee Surg Sports Traumatol Arthrosc 2014;22:214–18.
56 Yao L, Gai N, Boutin RD. Axial scan orientation and the tibial tubercle-trochlear
groove distance: error analysis and correction. AJR Am J Roentgenol
2014;202:1291–6.
57 Malghem J, Maldague B. Depth insufficiency of the proximal trochlear groove on
lateral radiographs of the knee: relation to patellar dislocation. Radiology
1989;170:507–10.
12 Ridley TJ, et al.JISAKOS 2016;0:1–12. doi:10.1136/jisakos-2015-000015
Systematic review
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from

systematic review
patients with patellofemoral instability: a
MRI show sensitivity without specificity in
Anatomical patella instability risk factors on
and Elizabeth A Arendt
Taylor J Ridley, Betina Bremer Hinckel, Bradley M Kruckeberg, Julie Agel
published online May 18, 2016J ISAKOS
http://jisakos.bmj.com/content/early/2016/05/18/jisakos-2015-000015
Updated information and services can be found at:
These include:
References
#BIBL
http://jisakos.bmj.com/content/early/2016/05/18/jisakos-2015-000015
This article cites 55 articles, 13 of which you can access for free at:
service
Email alerting box at the top right corner of the online article.
Receive free email alerts when new articles cite this article. Sign up in the
Notes
http://group.bmj.com/group/rights-licensing/permissions
To request permissions go to:
http://journals.bmj.com/cgi/reprintform
To order reprints go to:
http://group.bmj.com/subscribe/
To subscribe to BMJ go to:
group.bmj.com on May 20, 2016 - Published by http://jisakos.bmj.com/Downloaded from