Patterns of femorotibial cartilage loss in knees with neutral, varus, and valgus alignment.
ABSTRACT Malalignment is known to alter medial-to-lateral femorotibial load distribution and to affect osteoarthritis (OA) progression in the mechanically stressed compartment. We investigated the pattern of cartilage loss in neutral, varus, and valgus knees.
Alignment was measured from full-limb radiographs in 174 participants with symptomatic knee OA. Coronal magnetic resonance images were acquired at baseline and a mean +/- SD of 26.6 +/- 5.4 months later. The weight-bearing femorotibial cartilages were segmented from paired images. Cartilage volume, surface area, and thickness were determined in total cartilage plates and defined subregions using proprietary software.
The medial-to-lateral ratio of femorotibial cartilage loss was 1.4:1 in neutral knees (n = 74), 3.7:1 in varus knees (n = 57), and 1:6.0 in valgus knees (n = 43). The relative contribution of cartilage thickness change tended to be greater in knees with mild cartilage loss, whereas the increase of denuded area was greater in knees with accelerated cartilage loss. In both varus and neutral knees, the greatest changes were observed in the same subregions of the medial femorotibial compartment (central and external medial tibia, and central medial femur). In valgus and neutral knees, the subregions with the greatest changes in the lateral femorotibial compartment were also similar (internal and central lateral tibia, external lateral femur).
The medial-to-lateral rate of femorotibial cartilage loss strongly depended on alignment. Subregions of greater-than-average cartilage loss within the stressed compartment were, however, similar in neutral, varus, and valgus knees. This indicates that the medial-to-lateral loading pattern is different, but that the (sub)regional loading pattern may not differ substantially between neutral and malaligned knees.
- [Show abstract] [Hide abstract]
ABSTRACT: Objective Anti-catabolic disease modifying drugs (DMOADs) aim to reduce cartilage loss in knee osteoarthritis (KOA). Testing such drugs in clinical trials requires sufficient rates of loss in the study participants to occur, preferably at a mild disease stage where cartilage can be preserved. Here we analyze a “progression” model in mild radiographic KOA (RKOA), based on contra-lateral radiographic status. Methods We studied 837 participants (62.4±9yrs; 30±4.9kg/m²; 61.8% women) from the Osteoarthritis Initiative (OAI) with mild to moderate RKOA (Kellgren Lawrence grade [KLG] 2 to 3) and with/without OARSI atlas radiographic joint space narrowing (JSN). These had quantitative measurements of subregional femorotibial cartilage thickness from magnetic resonance imaging (MRI) at baseline and 1-year follow-up. They were stratified by contra-lateral knee status: no (KLG 0/1), definite (KLG2) and moderate RKOA (KLG 3/4). Results KLG2 knees with JSN and moderate contra-lateral RKOA had (p=0.008) greater maximum subregional cartilage loss -220μm [95% confidence interval -255, -184μm] than those without contra-lateral RKOA -164μm [-187, -140μm]. Their rate of subregional cartilage loss was similar and not significantly different (p=0.61) to that in KLG 3 knees without contra-lateral RKOA (-232μm; [-266;-198μm]). The effect of contra-lateral RKOA status was less in KLG2 knees without JSN, and in KLG3 knees. Conclusion KLG2 knees with JSN and moderate contra-lateral RKOA, display relatively high rates of subregional femorotibial cartilage loss, despite being at a relatively mild stage of RKOA. They may therefore provide a unique opportunity for recruitment in clinical trials that explore the efficacy of anti-catabolic DMOADs on structural progression.Osteoarthritis and Cartilage 09/2014; · 4.66 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Imaging in clinical trials is used to evaluate subject eligibility, and/or efficacy of intervention, supporting decision making in drug development by ascertaining treatment effects on joint structure. This review focusses on imaging of bone and cartilage in clinical trials of (knee) osteoarthritis. We narratively review the full-text literature on imaging of bone and cartilage, adding primary experience in the implementation of imaging methods in clinical trials. Aims and constraints of applying imaging in clinical trials are outlined. The specific uses of semi-quantitative and quantitative imaging biomarkers of bone and cartilage in osteoarthritis trials are summarized, focusing on radiography and magnetic resonance imaging (MRI). Studies having compared both imaging methodologies directly and those having established a relationship between imaging biomarkers and clinical outcomes are highlighted. To make this review of practical use, recommendations are provided as to which imaging protocols are ideal for capturing specific aspects of bone and cartilage tissue, and pitfalls in their usage are highlighted. Further, the longitudinal sensitivity to change, of different imaging methods is reported for various patient strata. From these power calculations can be accomplished, provided the strength of the treatment effect is known. In conclusion, current imaging methodologies provide powerful tools for scoring and measuring morphological and compositional aspects of most articular tissues, capturing longitudinal change with reasonable to excellent sensitivity. When employed properly, imaging has tremendous potential for ascertaining treatment effects on various joint structures, potentially over shorter time scales than required for demonstrating effects on clinical outcomes.Osteoarthritis and Cartilage 10/2014; 22(10):1516-1532. · 4.66 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Cartilage morphometry based on magnetic resonance images (MRIs) is an emerging outcome measure for clinical trials among patients with knee osteoarthritis (KOA). However, current methods for cartilage morphometry take many hours per knee and require extensive training on the use of the associated software. In this study we tested the feasibility, reliability, and construct validity of a novel osteoarthritis cartilage damage quantification method (Cartilage Damage Index [CDI]) that utilizes informative locations on knee MRIs.BMC Musculoskeletal Disorders 08/2014; 15(1):264. · 1.90 Impact Factor
Patterns of Femorotibial Cartilage Loss in Knees
With Neutral, Varus, and Valgus Alignment
FELIX ECKSTEIN,1WOLFGANG WIRTH,2MARTIN HUDELMAIER,1VERENA STEIN,3
VERENA LENGFELDER,3SEPTEMBER CAHUE,4MEREDITH MARSHALL,4POTTUMARTHI PRASAD,5
AND LEENA SHARMA4
Objective. Malalignment is known to alter medial-to-lateral femorotibial load distribution and to affect osteoarthritis
(OA) progression in the mechanically stressed compartment. We investigated the pattern of cartilage loss in neutral,
varus, and valgus knees.
Methods. Alignment was measured from full-limb radiographs in 174 participants with symptomatic knee OA. Coronal
magnetic resonance images were acquired at baseline and a mean ? SD of 26.6 ? 5.4 months later. The weight-bearing
femorotibial cartilages were segmented from paired images. Cartilage volume, surface area, and thickness were deter-
mined in total cartilage plates and defined subregions using proprietary software.
Results. The medial-to-lateral ratio of femorotibial cartilage loss was 1.4:1 in neutral knees (n ? 74), 3.7:1 in varus knees
(n ? 57), and 1:6.0 in valgus knees (n ? 43). The relative contribution of cartilage thickness change tended to be greater
in knees with mild cartilage loss, whereas the increase of denuded area was greater in knees with accelerated cartilage
loss. In both varus and neutral knees, the greatest changes were observed in the same subregions of the medial
femorotibial compartment (central and external medial tibia, and central medial femur). In valgus and neutral knees, the
subregions with the greatest changes in the lateral femorotibial compartment were also similar (internal and central
lateral tibia, external lateral femur).
Conclusion. The medial-to-lateral rate of femorotibial cartilage loss strongly depended on alignment. Subregions of
greater-than-average cartilage loss within the stressed compartment were, however, similar in neutral, varus, and valgus
knees. This indicates that the medial-to-lateral loading pattern is different, but that the (sub)regional loading pattern may
not differ substantially between neutral and malaligned knees.
Malalignment is known to alter the medial-to-lateral load
distribution in the femorotibial joint, with greater loads
being transferred through the medial femorotibial com-
partment in neutral and varus knees, and with relatively
greater loads through the lateral compartment in valgus
knees (1–6). Malalignment has also been identified as an
important risk factor for structural progression of femo-
rotibial osteoarthritis (OA) as observed on weight-bearing
radiographs (7–10). However, radiography has important
limitations. Measurements of joint space narrowing (JSN)
in the less-loaded compartment may not be meaningful
due to pseudo-widening (11), and JSN measurement may
reflect a change not only in cartilage but also in the me-
In contrast, magnetic resonance imaging (MRI) permits
one to obtain insight into the pattern of femorotibial car-
tilage loss in malaligned knees: cartilage morphology can
be accurately assessed in the tibia and femur of the medial
and lateral compartments, respectively (15–18), and in
Supported by the NIH/National Institute of Arthritis and
Musculoskeletal and Skin Diseases (grants R01-AR-48216,
R01-AR-48748, and P60-AR-48098).
1Felix Eckstein, MD, Martin Hudelmaier, MD: Paracelsus
Medical University, Salzburg, Austria, and Chondromet-
rics, Ainring, Germany;
metrics, Ainring, Germany, and Ludwig Maximilians Uni-
versita ¨t Mu ¨nchen, Munich, Germany;3Verena Stein, MD,
Verena Lengfelder, MD: Paracelsus Medical University, Sal-
zburg, Austria;4September Cahue, MD, Meredith Marshall,
MD, Leena Sharma, MD: Feinberg School of Medicine, North-
western University, Chicago, Illinois;
MD: Evanston Northwestern Healthcare, Evanston, Illinois.
2Wolfgang Wirth, PhD: Chondro-
Dr. Eckstein has received consultant fees, speaking fees,
and/or honoraria (less than $10,000) from Novo Nordisk and
(more than $10,000 each) from Pfizer, Merck, and Wyeth. Dr.
Wirth has served as a freelancer for Chondrometrics, Ainring,
Germany, and received a fee (more than $10,000).
Address correspondence to Felix Eckstein, MD, Institute
of Anatomy and Musculoskeletal Research, Paracelsus
Medical University, Strubergasse 21, A5020 Salzburg, Aus-
tria. E-mail: felix.Eckstein@pmu.ac.at.
Submitted for publication March 13, 2008; accepted in
revised form July 14, 2008.
Arthritis & Rheumatism (Arthritis Care & Research)
Vol. 59, No. 11, November 15, 2008, pp 1563–1570
© 2008, American College of Rheumatology
specific subregions within the femorotibial cartilage plates
(19–21). Subregional cartilage loss in OA may be related to
the specific pattern of mechanical stress within the femo-
rotibial cartilage plates, and its analysis may support the
development of targeted strategies to improve the load/
stress distribution and to prevent OA progression in mal-
Using MRI, Cicuttini et al (22) demonstrated a signifi-
cant relationship between femoral cartilage loss and varus
angulation, with less evidence of this relationship in tibial
cartilage, whereas another (cross-sectional) study showed
a higher correlation between malalignment and femoral
cartilage loss (23). Recently, we reported that varus mal-
alignment increased the risk of cartilage loss in medial
cartilage surfaces after adjusting for age, sex, body mass
index, medial meniscal damage and extrusion, and lateral
laxity (24). The spatial pattern of femorotibial cartilage
loss, specifically changes in anatomically defined subre-
gions, however, has not been previously examined. In the
present study, we investigated whether 1) magnitudes of
tibial and femoral cartilage loss are affected differently by
alignment in the stressed femorotibial compartment, 2)
loss of cartilage in malaligned knees is primarily due to
reductions in cartilage thickness or cartilage area, and 3)
the subregional pattern of femorotibial cartilage loss dif-
fers between neutral and malaligned knees.
PARTICIPANTS AND METHODS
Participants were recruited from the community and were
members of a cohort of participants in a natural history
study of knee OA (Mechanical Factors in Arthritis of the
Knee Study, second cycle) (24). Inclusion criteria were as
follows: definite tibiofemoral osteophyte presence (Kell-
gren/Lawrence [K/L] radiographic grade 2 or higher) in 1
or both knees and Likert category of at least “a little diffi-
culty” for ?2 items on the Western Ontario and McMaster
Universities Osteoarthritis Index physical function scale
(24). Approval was obtained from the Office for the Pro-
tection of Research Subjects-Institutional Review Boards
of Northwestern University and Evanston Northwestern
To determine the K/L grade, bilateral, anteroposterior,
semiflexed, weight-bearing knee radiographs with fluoro-
scopic control were obtained for all participants at base-
line (25). To assess the hip-knee-ankle angle, a single an-
teroposterior radiograph of both legs was obtained (8,26).
Alignment was analyzed as a continuous variable (hip-
knee-ankle angle) as described previously (24). Knees with
a hip-knee-ankle angle of ?2° to ?2° were classified as
neutral, those with an angle ?2° as varus, and those with
an angle less than ?2° as valgus.
All participants underwent MRI of both knees using a
commercial knee coil and 1 of 2 whole-body scanners,
either a 1.5 Tesla Symphony at Northwestern University
(Siemens, Erlangen, Germany) or a 3.0 Tesla Genesis Signa
Scanner at Evanston Northwestern Healthcare (GE Health-
care Technologies, Waukesha, WI). Baseline and followup
MRIs were always performed on the same scanner. The
mean ? SD observation interval was 26.6 ? 5.4 months
(range 14–50 months). Previously validated (16–18,27)
coronal spoiled gradient echo sequences with water exci-
tation were acquired with a slice thickness of 1.5 mm and
an in-plane resolution of 0.31 mm ? 0.31 mm. The repe-
tition time, echo time, and flip angle, respectively, were
18.6 msec, 9.3 msec, and 15° at 1.5T and 12.2 msec, 5.8
msec, and 9° at 3.0T. The MRI data were sent to the image
analysis center, quality controlled, and converted to a
proprietary format (Chondrometrics, Ainring, Germany).
One knee per participant was studied. When good-qual-
ity baseline and followup data sets without artifacts were
available for both knees, the dominant knee was included
(157 dominant, 17 nondominant, 145 right, and 29 left).
The mean ? SD age of the 174 participants (76% women,
24% men) was 66 ? 11.1 years and the mean body mass
index was 30.1 ? 5.9 kg/m2. Most knees had a K/L grade of
2 (41%) or 3 (33%) at baseline. A total of 74 knees showed
Figure 1. Coronal magnetic resonance image, acquired with a
spoiled gradient echo sequence with water excitation. The top
image shows the femorotibial cartilage plates: MT ? medial tibia;
LT ? lateral tibia; cMF ? medial weight-bearing femur; cLF ?
lateral weight-bearing femur. The bottom image shows an en-
larged part of the top image in the medial compartment: the top
bracket shows over which region the area of the cartilage surface
(AC) is segmented in the MT. The bottom brackets show where the
total area of subchondral bone (tAB) is segmented in the MT. The
part of the tAB that is covered by AC is the cartilaginous area of
bone (cAB). The part not coverd by AC is the denuded area of
bone (dAB). Cartilage thickness is measured both over the cAB
and over the entire tAB, with inclusion of dAB areas with 0-mm
1564Eckstein et al
neutral alignment (mean ?SD 0.2 ? 1.3°), 57 showed varus
malalignment (6.5 ? 4.1°, maximum 19°), and 43 showed
valgus malalignment (?5.3 ? 2.5°, minimum ?13°).
Segmentation of the femorotibial cartilages was per-
formed by 10 readers with formal training in cartilage
segmentation using proprietary software (Chondromet-
rics). Images were read in pairs with blinding to acquisi-
tion order. Segmentation involved manual tracing (Figure
1) of the total subchondral bone area and the cartilage joint
surface area of the medial tibia, the lateral tibia, the central
(weight-bearing) medial femoral condyle, and the central
lateral femoral condyle (28–32). Quality control of all seg-
mentations was performed by one reader (FE). The seg-
mentations were used to compute the total area of sub-
chondral bone, the cartilage surface area, the part of the
subchondral bone covered with cartilage, the denuded
subchondral bone area, the cartilage volume, the cartilage
thickness over the cartilage covered area not including
denuded areas, and the cartilage thickness over the entire
subchondral bone area, including denuded areas with
0-mm cartilage thickness (Figure 1) (28). Changes were
computed for the medial femorotibial compartment
(MFTC) and lateral femorotibial compartment (LFTC) by
summing values of the medial tibia and femur and the
lateral tibia and femur, respectively, at baseline and fol-
lowup (30,31). In a next step, 5 subregions (central, inter-
nal, external, anterior, posterior) were determined based
on the subchondral bone area in the tibiae (Figure 1), with
the central subregion occupying 20% of the total subchon-
dral bone area (Figures 2A and B) (21). The central tibial
region was defined by a perpendicular cylinder around the
center of gravity of the tibial subchondral bone area, the
diameter being adapted to its individual shape (21). Be-
cause the weight-bearing femoral condyles are limited in
their anteroposterior extension (femoral trochlea anteri-
orly and posterior femoral condyle posteriorly), they were
divided into a central, internal, and external strip-like
region of interest, respectively (Figures 2A and C), each
occupying 33.3% of the subchondral bone area (21). Car-
tilage thickness was determined in all subregions.
The correlation between cartilage loss and the hip-knee-
ankle angle was determined across all knees by computing
the Pearson correlation coefficient. The mean ? SD of the
change from baseline to followup, the standardized re-
sponse mean (SRM; mean/SD of change), and the signifi-
cance of change (paired t-test, without correction for mul-
tiple testing) were calculated for each cartilage plate,
subregion, parameter, and alignment group after adjusting
individual changes to a 12-month observation period. The
percentage mean change was obtained by relating the
mean changes (i.e., mm) to the mean baseline values. Dif-
ferences in changes of cartilage thickness between varus
and neutral knees and between valgus and neutral knees
were tested using an unpaired 2-sided t-test. Differences
(in changes) between subregions of each cartilage plate
were tested using an analysis of variance (ANOVA) of
The correlation of the hip-knee-ankle angle (continuous
variable) with cartilage volume loss and reduction in car-
tilage thickness across all medial tibiae was r ? 0.36 and
r ? 0.40, respectively (both P ? 0.001), whereas the coef-
ficients were 0.10 and 0.11, respectively (not significant),
in the medial weight-bearing femur. The correlation was
0.07 and 0.10, respectively (not significant), in the lateral
tibia, and 0.18 (P ? 0.05) and 0.21 (P ? 0.01), respectively,
in the lateral femur.
Neutral knees. A significant reduction in cartilage vol-
ume was observed in all 4 femorotibial cartilage plates,
ranging from ?0.8% (lateral tibia) to ?1.5% (medial fe-
mur) (Table 1). Changes in cartilage thickness tended to be
greater than changes in cartilage volume (P ? 0.047 for the
medial tibia), with annual rates between ?0.9% (lateral
tibia) and ?1.7% (medial femur). The highest SRM was
observed for changes in cartilage thickness in the medial
tibia (?0.50; P ? 0.0001). The medial-to-lateral ratio of
cartilage loss (MFTC versus LFTC) was 1.4:1.
In all 4 plates, the reduction in cartilage thickness ex-
clusive of denuded areas tended to make a relatively stron-
Figure 2. Image showing femorotibial subregions. A, Posterior
view of femorotibial subchondral bone areas (tibia at the bottom,
weight bearing femur at the top), with subregions displayed by
different gray values. B, Superior view of the tibial subchondral
bone area, with subregions labeled. C, Inferior view of the femoral
subchondral bone area, with subregions labeled. cMF ? central
(weight-bearing) medial femoral condyle; ccMF ? central cMF;
ecMF ? external cMF; icMF ? internal cMF; MT ? medial tibia;
cMT ? central MT; eMT ? external MT; iMT ? internal MT;
aMT ? anterior MT; pMT ? posterior MT; LT ? lateral tibia;
pLT ? posterior LT; aLT ? anterior LT; eLT ? external LT; iLT ?
internal LT; cLT ? central LT; cLF ? central lateral femoral
condyle; ecLF ? external cLF; ccLF ? central cLF; icLF ? internal
Femorotibial Cartilage Loss Patterns in Malaligned Knees1565
ger contribution to the cartilage loss than the reduction in
cartilage area or the increase in denuded area (Table 1).
The difference was, however, not statistically significant.
In the medial tibia, the rate of change between (sub)re-
gions differed significantly (P ? 0.01 by ANOVA). The
central (?1.5%, SRM ?0.49), external (?1.7%, SRM
?0.41), and anterior subregions (?1.4%, SRM ?0.41)
showed significant changes between baseline and fol-
lowup, whereas the posterior and internal subregions did
not (Table 1 and Figure 3). In the medial femur, the central
subregion displayed the highest rate of change (?2.2%)
and the internal subregion the highest SRM (?0.42),
whereas the external subregion failed to show significant
changes (Figure 3). There were, however, no significant
differences in the rate of change between subregions (P ?
0.12 by ANOVA). In the lateral tibia, the internal (?1.8%,
SRM ?0.55), central (?1.1%, SRM ?0.25), and anterior
subregions (?1.1%, SRM ?0.36) showed significant
change, whereas the posterior and external subregions did
not. However, the differences between subregions did not
reach statistical significance (P ? 0.07). In the lateral fe-
mur, all 3 subregions showed significant rates of change
and similar SRMs (?0.26 to ?0.27), with no significant
differences in the rate of change between the subregions
(P ? 0.93) (Table 1 and Figure 3).
Varus knees. The reduction in cartilage thickness in the
medial tibia (?2.6%, SRM ?0.52; P ? 0.0001) was signif-
icantly greater (P ? 0.04) than that in neutral knees. The
medial femur also showed a greater reduction in cartilage
thickness (?2.6%, SRM ?0.49; P ? 0.002) than in the
neutral knees (Table 2). The lateral tibia displayed a sig-
nificant change between baseline and followup (?1.2%,
SRM ?0.54; P ? 0.001), whereas the lateral femur did not
(Table 2). In the medial femur, the lateral tibia, and the
lateral femur, the rate of change did not differ significantly
from that in neutral knees (P ? 0.61, 0.30, and 0.07, re-
spectively). The ratio of medial to lateral cartilage loss was
3.7:1. In the medial tibia, the reduction in cartilage area
made a stronger contribution than the loss in thickness
(P ? 0.012), but in the medial femur (P ? 0.42) and the
lateral tibia (P ? 0.11), a trend toward the opposite was
observed (Table 2).
In the medial tibia, there was a highly significant differ-
ence in the rate of change between the subregions (P ?
0.001), with the central subregion (?3.1%, SRM ?0.50)
and external subregion (?5.1%, SRM ?0.43) showing the
highest rates of change (Figure 3). In the medial femur, the
changes in the 3 subregions did not differ significantly
from one another (P ? 0.47) (Table 2 and Figure 3). In the
lateral tibia, the internal (?2.3%, SRM ?0.72), central
(?1.5%, SRM ?0.44), and posterior subregions (?0.9%,
SRM ?0.23) displayed significant change, but the anterior
and external subregions did not. None of the subregions of
the lateral femur showed significant differences between
baseline and followup (Table 2).
Valgus knees. The reduction in cartilage thickness of
the medial tibia (?0.1%, SRM ?0.06; P ? 0.63) was not
significant and was significantly less than for neutral knees
(P ? 0.04). The medial femur also showed only little
cartilage thinning (?0.7%, SRM ?0.22; P ? 0.08), with
Table 1. Changes in cartilage morphology parameters from baseline to followup in neutral knees*
Medial femorotibial compartmentLateral femorotibial compartment
P†Parameter MC% SRM
Total cartilage plateMT.VC
? 0.05Cartilage subregions
* MC% ? mean change (percentage); SRM ? standardized response mean (mean change/SD of change); MT ? medial tibia; VC ? volume of cartilage;
LT ? lateral tibia; cMF ? weight-bearing medial femoral condyle; cLF ? weight-bearing lateral femoral condyle; MFTC ? medial femorotibial
compartment (MT ? cMF); LFTC ? lateral femorotibial compartment (LT ? cLF); ThCtAB ? thickness of the cartilage over the entire subchondral bone
area (including denuded areas with 0-mm cartilage thickness); ThCcAB ? thickness of the cartilage over the entire cAB; cAB ? cartilage covered bone
area; cMT ? central MT; cLT ? central LT; eMT ? external MT; eLT ? external LT; iMT ? internal MT; iLT ? internal LT; aMT ? anterior MT; aLT
? anterior LT; pMT ? posterior MT; pLT ? posterior LT; ccMF ? central cMF; ccLF ? central cLF; ecMF ? external cMF; ecLF ? external cLF; icMF
? internal cMF; icLF ? internal cLF. Abbreviations have been chosen in accordance with published nomenclature (28).
† Level of significance of change between baseline and followup.
1566Eckstein et al
values not significantly different (P ? 0.26) from neutral
knees (Table 3). In the lateral tibia, the change in cartilage
thickness (?3.0%, SRM ?0.71; P ? 0.0001) was signifi-
cantly higher (P ? 0.018) than in neutral knees. In the
lateral femur, changes in cartilage thickness were signifi-
cant (?1.9%, SRM ?0.42; P ? 0.0079), but did not differ
significantly (P ? 0.30) from neutral knees (Table 3). The
medial-to-lateral ratio of cartilage loss was 1:6.0. In both
the lateral tibia and the lateral femur, the reduction in
cartilage area tended to make a stronger contribution than
the loss in thickness (Table 3), but the difference did not
reach statistical significance (P ? 0.22 and 0.51, respec-
In the medial tibia and medial femur, none of the sub-
regions showed significant change with time. In the lateral
tibia, the changes in the subregions differed significantly
among each other (P ? 0.05). The central (?4.7%, SRM
?0.81), internal (?4.2%, SRM ?0.76), and posterior sub-
regions (?2.0%, SRM ?0.42) displayed significant differ-
ences between baseline and followup, but the anterior and
external regions did not (Figure 3). In the lateral femur, the
difference in the rate of change between the subregions
was also significant (P ? 0.01), with the greatest change in
the external subregion (?2.9%, SRM ?0.50; P ? 0.002)
Our study is the first to investigate patterns of cartilage loss
in knees with neutral, varus, and valgus alignment. We
found the medial-to-lateral ratio of femorotibial cartilage
loss to be strongly affected by alignment. Correlations be-
tween thickness changes and the hip-knee-ankle angle
were stronger for the tibia than for the femur in the medial,
but stronger for the femur than for the tibia in the lateral
femorotibial compartment. The relative contribution of
cartilage thickness change tended to be greater in mild
cartilage loss, whereas the increase of denuded area was
greater in accelerated cartilage loss in the mechanically
stressed compartments. In varus and neutral knees, the
greatest changes were observed in the same subregions of
the medial femorotibial compartment (central and external
medial tibia, central medial femur). Greater-than-average
changes were also observed in similar subregions of the
lateral femorotibial compartment in valgus and neutral
knees (internal and central lateral tibia, external lateral
femur), albeit the similarity in the subregional pattern of
cartilage loss was less obvious than in the medial compart-
Limitations of this study include the modest sample size
compared with some of the radiographic studies (10), the
fact that, given the larger number of regions, cartilage
plates, and parameters that were compared, it was not
feasible to fully correct for multiple statistical testing, and
the fact that the observation period varied between 14 and
50 months. To account for differences in time between
baseline and followup measurements, the changes were
normalized to 1 year, although it is more likely that carti-
lage thinning is not linear, but that periods of flares alter-
nate with periods of relative stability. Another limitation
is the lack of measurement of the dynamic loads (33),
because static measurements of malalignment can only
predict approximately 50% of the variability in knee ad-
duction moments (34). However, Thorp et al (35) reported
that static (knee alignment angle) and dynamic markers of
knee loads (knee adduction angular momentum) each ex-
plained the same proportion (18%) of the variability of
proximal tibial bone mineral density in subjects with knee
OA. A strength of the present study is that it is the first to
report correlations and rates of MRI-based cartilage loss
with alignment measurements taken from full-limb radio-
graphs. Although knee radiographs can capture approxi-
mately 50% of the variability, they only provide an esti-
mate of the hip-knee-ankle angle (36). It is well known that
the medial-to-lateral femorotibial load and stress distribu-
tion depend on alignment: in neutral knees, the medial
compartment is more highly loaded than the lateral com-
partment because of the stance phase knee adduction mo-
ment (1,6,37). This fits in well with the 1.4:1 ratio of
medial-to-lateral cartilage loss observed in neutral knees
Figure 3. The subregional changes of thickness of the cartilage
over the entire subchondral bone area (including denuded areas
with 0-mm cartilage thickness) (ThCtAB) of the medial femo-
rotibial compartment in neutral and varus knees, and of the lateral
compartment in neutral and valgus knees. Whereas the rate of
change was higher in the medial compartment in varus knees and
higher in the lateral compartment in valgus knees, the regional
pattern of cartilage loss was very similar to that in neutral knees.
MT ? medial tibia; cMF ? medial weight-bearing femur; LT ?
lateral tibia; cLF ? lateral weight-bearing femur; c ? central; e ?
external; i ? internal; a ? anterior; p ? posterior. The dotted lines
display the average change in cartilage thickness (ThCtAB)
throughout the entire cartilage plate, in order to visualize which
subregions show greater and which ones smaller changes than the
Femorotibial Cartilage Loss Patterns in Malaligned Knees 1567
in this study. Varus alignment is known to increase and
valgus alignment to decrease medial load (2–4,38). The
correlations of the hip-knee-ankle angle with medial ver-
sus lateral cartilage loss did therefore show the expected
relationships, and they were somewhat stronger for carti-
lage thickness than for cartilage volume. However, in con-
trast with a previous study (22), in the medial femorotibial
compartment the correlation was significant for the tibial
but not for the femoral cartilage, whereas in the lateral
femorotibial compartment it was significant for the femo-
ral but not for the tibial cartilage. The 3.7:1 medial-to-
lateral ratio of cartilage loss in varus knees reinforces the
notion that increased mechanical stress, caused by mal-
alignment, is an important risk factor of femorotibial OA
progression. The ratio of 1:6.0 in valgus knees is somewhat
surprising, because a previous study reported that greater
load was observed in the lateral femorotibial compartment
only in knees with severe valgus malalignment (3).
Previous longitudinal MRI studies have mainly focused
on the relationship between cartilage volume loss and
Table 2. Changes in cartilage morphology parameters from baseline to followup in varus knees*
Medial femorotibial compartment Lateral femorotibial compartment
Total cartilage plateMT.VC
* See Table 1 for definitions.
† Level of significance of change between baseline and followup.
Table 3. Changes in cartilage morphology parameters from baseline to followup in valgus knees*
Medial femorotibial compartmentLateral femorotibial compartment
Total cartilage plateMT.VC
* See Table 1 for definitions.
† Level of significance of change between baseline and followup.
1568 Eckstein et al
malalignment (22,39). Measurement of cartilage volume,
however, does not exploit the full capacity of MRI, as it
does not reveal the spatial pattern of cartilage loss (18–21).
In our study, we found somewhat higher correlations and
greater changes for cartilage thickness than for cartilage
(32,40,41), there was some increase in the subchondral
bone area between baseline and followup. The finding of a
greater reduction of cartilage thickness without denuded
areas than cartilage area in cases of mild cartilage loss, and
that of a greater reduction in cartilage area (? greater
increase in denuded area) than reduction in cartilage
thickness in cases of accelerated cartilage loss is intrigu-
ing. This should be examined in other cohorts (i.e., OA
Initiative: www.oai.ucsf.edu) (42) to identify whether this
relationship is specific to varus and valgus malalignment
or a more general phenomenon.
Of particular interest is that the magnitude, but not the
(relative) regional pattern of cartilage loss within the
stressed compartment differed between neutral, varus, and
valgus knees. Regions of maximal cartilage loss in the
medial femorotibial compartment were similar in varus
and in neutral knees, and those in the lateral femorotibial
compartment were similar (albeit to a lesser extent) be-
tween valgus and neutral knees. The only inconsistencies
were that in the lateral tibia, the internal subregion dis-
played the greatest changes in neutral knees and the cen-
tral subregions, the greatest changes in valgus knees, but
both regions showed greater changes than the other subre-
gions in both neutral and valgus knees. In addition, in the
lateral femur, the pattern was relatively diffuse in neutral
knees, with all subregions showing similar changes. Fi-
nally, the posterior region of the lateral tibia showed sig-
nificant changes in valgus and varus (but not in neutral)
knees, whereas the anterior subregions displayed signifi-
cant changes in neutral (but not in varus and valgus)
knees. Although it is speculative to draw conclusions from
patterns of cartilage loss about the regional patterns of
mechanical stress, these data may indicate that the stress
magnitude but not the relative stress distribution within
cartilage plates differs substantially between malaligned
knees. Therefore, it may be sufficient for treatment strate-
gies to improve the medial-to-lateral load–stress balance in
the femorotibial joint of malaligned knees, to prevent or
slow down OA progression.
In conclusion, we found that alignment direction pre-
dicted the medial-to-lateral ratio of femorotibial cartilage
loss. Medially, there was a significant correlation of tibial
(but not femoral) cartilage loss with the hip-knee-ankle
angle, and laterally, a significant correlation of femoral
(but not tibial) cartilage loss. When cartilage loss was mild
(i.e., neutral knees), it was dominated by a loss in cartilage
thickness (without including denuded areas), whereas
when it was more accelerated (i.e., stressed compartment
in varus and valgus knees), it was dominated by a decrease
in cartilage area (increase in denuded area). The subre-
gional pattern of cartilage loss in the mechanically stressed
compartment was similar between neutral, varus, and val-
gus knees. This indicates that, whereas medial-to-lateral
load distribution is strongly affected by alignment, the
(sub)regional loading pattern of femorotibial cartilage may
not differ substantially between varus, valgus, and neutral
We would like to thank the following readers: Gudrun
Goldmann, Linda Jakobi, Manuela Kunz, Dr. Susanne Mas-
chek, Sabine Mu ¨hlsimer, Franz Romeder, Annette Thebis,
and Dr. Barbara Wehr for dedicated data segmentation.
Dr. Eckstein had full access to all of the data in the study and
takes responsibility for the integrity of the data and the accuracy
of the data analysis.
Study design. Eckstein, Cahue, Marshall, Prasad, Sharma.
Acquisition of data. Cahue, Marshall, Prasad, Sharma.
Analysis and interpretation of data. Eckstein, Wirth, Hudelmaier,
Stein, Lengfelder, Sharma.
Manuscript preparation. Eckstein, Wirth, Hudelmaier, Stein,
Lengfelder, Cahue, Marshall, Prasad, Sharma.
Statistical analysis. Eckstein.
1. Morrison JB. The mechanics of the knee joint in relation to
normal walking. J Biomech 1970;3:51–61.
2. Maquet P. Mechanics and osteoarthritis of the patellofemoral
joint. Clin Orthop Relat Res 1979;144:70–3.
3. Johnson F, Leitl S, Waugh W. The distribution of load across
the knee: a comparison of static and dynamic measurements.
J Bone Joint Surg Br 1980;62:346–9.
4. Cooke TD, Sled EA, Scudamore RA. Frontal plane knee
alignment: a call for standardized measurement. J Rheumatol
5. Tetsworth K, Paley D. Malalignment and degenerative ar-
thropathy. Orthop Clin North Am 1994;25:367–77.
6. Andriacchi TP. Dynamics of knee malalignment. Orthop Clin
North Am 1994;25:395–403.
7. Sharma L, Lou C, Cahue S, Dunlop DD. The mechanism of the
effect of obesity in knee osteoarthritis: the mediating role of
malalignment. Arthritis Rheum 2000;43:568–75.
8. Sharma L, Song J, Felson DT, Cahue S, Shamiyeh E, Dunlop
DD. The role of knee alignment in disease progression and
functional decline in knee osteoarthritis. JAMA 2001;286:
9. Felson DT, McLaughlin S, Goggins J, LaValley MP, Gale ME,
Totterman S, et al. Bone marrow edema and its relation to
progression of knee osteoarthritis. Ann Intern Med 2003;139:
10. Brouwer GM, van Tol AW, Bergink AP, Belo JN, Bernsen RM,
Reijman M, et al. Association between valgus and varus align-
ment and the development and progression of radiographic
osteoarthritis of the knee. Arthritis Rheum 2007;56:1204–11.
11. Buckland-Wright JC, Macfarlane DG, Lynch JA, Jasani MK,
Bradshaw CR. Joint space width measures cartilage thickness
in osteoarthritis of the knee: high resolution plain film and
double contrast macroradiographic investigation. Ann Rheum
12. Gale DR, Chaisson CE, Totterman SM, Schwartz RK, Gale ME,
Felson D. Meniscal subluxation: association with osteoarthri-
tis and joint space narrowing. Osteoarthritis Cartilage 1999;7:
13. Hunter DJ, Zhang YQ, Tu X, LaValley M, Niu JB, Amin S, et al.
Change in joint space width: hyaline articular cartilage loss or
alteration in meniscus? Arthritis Rheum 2006;54:2488–95.
14. Adams JG, McAlindon T, Dimasi M, Carey J, Eustace S. Con-
tribution of meniscal extrusion and cartilage loss to joint
space narrowing in osteoarthritis. Clin Radiol 1999;54:502–6.
15. Burgkart R, Glaser C, Hyhlik-Durr A, Englmeier KH, Reiser M,
Eckstein F. Magnetic resonance imaging–based assessment of
Femorotibial Cartilage Loss Patterns in Malaligned Knees 1569
cartilage loss in severe osteoarthritis: accuracy, precision, and
diagnostic value. Arthritis Rheum 2001;44:2072–7.
16. Eckstein F, Cicuttini F, Raynauld JP, Waterton JC, Peterfy C.
Magnetic resonance imaging (MRI) of articular cartilage in
knee osteoarthritis (OA): morphological assessment. Osteoar-
thritis Cartilage 2006;14 Suppl A:A46–75.
17. Eckstein F, Burstein D, Link TM. Quantitative MRI of cartilage
and bone: degenerative changes in osteoarthritis. NMR
18. Graichen H, Eisenhart-Rothe R, Vogl T, Englmeier KH, Eck-
stein F. Quantitative assessment of cartilage status in osteo-
arthritis by quantitative magnetic resonance imaging: techni-
cal validation for use in analysis of cartilage volume and
further morphologic parameters. Arthritis Rheum 2004;50:
19. Koo S, Gold GE, Andriacchi TP. Considerations in measuring
cartilage thickness using MRI: factors influencing reproduc-
ibility and accuracy. Osteoarthritis Cartilage 2005;13:782–9.
20. Pelletier JP, Raynauld JP, Berthiaume MJ, Abram F, Choquette
D, Haraoui B, et al. Risk factors associated with the loss of
cartilage volume on weight-bearing areas in knee osteoarthri-
tis patients assessed by quantitative magnetic resonance
imaging: a longitudinal study. Arthritis Res Ther 2007;9:R74.
21. Wirth W, Eckstein F. A technique for regional analysis of
femorotibial cartilage thickness based on quantitative mag-
netic resonance imaging. IEEE Trans Med Imaging 2008;27:
22. Cicuttini F, Wluka A, Hankin J, Wang Y. Longitudinal study
of the relationship between knee angle and tibiofemoral car-
tilage volume in subjects with knee osteoarthritis. Rheuma-
tology (Oxford) 2004;43:321–4.
23. Von Eisenhart-Rothe R, Graichen H, Hudelmaier M, Vogl T,
Sharma L, Eckstein F. Femorotibial and patellar cartilage loss
in patients prior to total knee arthroplasty, heterogeneity, and
correlation with alignment of the knee. Ann Rheum Dis 2006;
24. Sharma L, Eckstein F, Song J, Guermazi A, Prasad P, Kapoor
D, et al. Relationship of meniscal damage, meniscal extrusion,
malalignment, and joint laxity to subsequent cartilage loss in
osteoarthritic knees. Arthritis Rheum 2008;58:1716–26.
25. Buckland-Wright C. Protocols for precise radio-anatomical
positioning of the tibiofemoral and patellofemoral compart-
ments of the knee. Osteoarthritis Cartilage 1995;3 Suppl
26. Moreland JR, Bassett LW, Hanker GJ. Radiographic analysis of
the axial alignment of the lower extremity. J Bone Joint Surg
27. Eckstein F, Charles HC, Buck RJ, Kraus VB, Remmers AE,
Hudelmaier M, et al. Accuracy and precision of quantitative
assessment of cartilage morphology by magnetic resonance
imaging at 3.0T. Arthritis Rheum 2005;52:3132–6.
28. Eckstein F, Ateshian G, Burgkart R, Burstein D, Cicuttini F,
Dardzinski B, et al. Proposal for a nomenclature for magnetic
resonance imaging based measures of articular cartilage in
osteoarthritis. Osteoarthritis Cartilage 2006;14:974–83.
29. Eckstein F, Hudelmaier M, Wirth W, Kiefer B, Jackson R, Yu
J, et al. Double echo steady state magnetic resonance imaging
of knee articular cartilage at 3 Tesla: a pilot study for the
Osteoarthritis Initiative. Ann Rheum Dis 2006;65:433–41.
30. Eckstein F, Kunz M, Hudelmaier M, Jackson R, Yu J, Eaton
CB, et al. Impact of coil design on the contrast-to-noise ratio,
precision, and consistency of quantitative cartilage morphom-
etry at 3 Tesla: a pilot study for the osteoarthritis initiative.
Magn Reson Med 2007;57:448–54.
31. Eckstein F, Kunz M, Schutzer M, Hudelmaier M, Jackson RD,
Yu J, et al. Two year longitudinal change and test-retest-
precision of knee cartilage morphology in a pilot study for the
osteoarthritis initiative. Osteoarthritis Cartilage 2007;15:
32. Eckstein F, Buck RJ, Burstein D, Charles HC, Crim J, Hudel-
maier M, et al. Precision of 3.0 Tesla quantitative magnetic
resonance imaging of cartilage morphology in a multi center
clinical trial. Ann Rheum Dis 2008. E-pub ahead of print.
33. Miyazaki T, Wada M, Kawahara H, Sato M, Baba H, Shimada
S. Dynamic load at baseline can predict radiographic disease
progression in medial compartment knee osteoarthritis. Ann
Rheum Dis 2002;61:617–22.
34. Hurwitz DE, Ryals AB, Case JP, Block JA, Andriacchi TP. The
knee adduction moment during gait in subjects with knee
osteoarthritis is more closely correlated with static alignment
than radiographic disease severity, toe out angle and pain.
J Orthop Res 2002;20:101–7.
35. Thorp LE, Wimmer MA, Block JA, Moisio KC, Shott S, Goker
B, et al. Bone mineral density in the proximal tibia varies as a
function of static alignment and knee adduction angular mo-
mentum in individuals with medial knee osteoarthritis. Bone
36. Kraus VB, Vail TP, Worrell T, McDaniel G. A comparative
assessment of alignment angle of the knee by radiographic
and physical examination methods. Arthritis Rheum 2005;52:
37. Harrington IJ. Static and dynamic loading patterns in knee
joints with deformities. J Bone Joint Surg Am 1983;65:247–59.
38. Hilding MB, Lanshammar H, Ryd L. A relationship between
dynamic and static assessments of knee joint load: gait ana-
lysis and radiography before and after knee replacement in 45
patients. Acta Orthop Scand 1995;66:317–20.
39. Zhai G, Ding C, Cicuttini F, Jones G. A longitudinal study of
the association between knee alignment and change in carti-
lage volume and chondral defects in a largely non-osteo-
arthritic population. J Rheumatol 2007;34:181–6.
40. Jones G, Ding C, Scott F, Glisson M, Cicuttini F. Early radio-
graphic osteoarthritis is associated with substantial changes
in cartilage volume and tibial bone surface area in both males
and females. Osteoarthritis Cartilage 2004;12:169–74.
41. Wang Y, Wluka AE, Cicuttini FM. The determinants of change
in tibial plateau bone area in osteoarthritic knees: a cohort
study. Arthritis Res Ther 2005;7:R687–93.
42. Eckstein F, Maschek S, Wirth W, Hudelmaier M, Hitzl W,
Wyman B, et al. One year change of knee cartilage morphol-
ogy in the first release of participants from the Osteoarthritis
Initiative progression subcohort: association with sex, body
mass index, symptoms, and radiographic OA status. Ann
Rheum Dis 2008. E-pub ahead of print.
1570 Eckstein et al