The American Journal of Sports
The online version of this article can be found at:
2012 40: 2700 originally published online October 26, 2012 Am J Sports Med
Majumdar and Thomas M. Link
Pia M. Jungmann, Xiaojuan Li, Lorenzo Nardo, Karupppasamy Subburaj, Wilson Lin, C. Benjamin Ma, Sharmila
Morphological Evaluation With 3.0-T MRI
Do Cartilage Repair Procedures Prevent Degenerative Meniscus Changes?: Longitudinal T
On behalf of:
American Orthopaedic Society for Sports Medicine
can be found at:
The American Journal of Sports Medicine
Additional services and information for
What is This?
- Oct 26, 2012 OnlineFirst Version of Record
- Nov 28, 2012Version of Record >>
at UCSF LIBRARY & CKM on March 4, 2013ajs.sagepub.comDownloaded from
Do Cartilage Repair Procedures Prevent
Degenerative Meniscus Changes?
Longitudinal T1rand Morphological Evaluation
With 3.0-T MRI
Pia M. Jungmann,*yzMD, Xiaojuan Li,yPhD, Lorenzo Nardo,yMD,
Karupppasamy Subburaj,yPhD, Wilson Lin,yC. Benjamin Ma,§MD,
Sharmila Majumdar,yPhD, and Thomas M. Link,yMD, PhD
Investigation performed at Musculoskeletal and Quantitative Imaging Research Group (MQIR),
Department of Radiology and Biomedical Imaging, University of California–San Francisco,
San Francisco, California
Background: Cartilage repair (CR) procedures are widely accepted for treatment of isolated cartilage defects in the knee joint.
However, it is not well known whether these procedures prevent degenerative joint disease.
Hypothesis: Cartilage repair procedures prevent accelerated qualitative and quantitative progression of meniscus degeneration
in individuals with focal cartilage defects.
Study Design: Cohort study; Level of evidence, 2.
Methods: Ninety-four subjects were studied. Cartilage repair procedures were performed on 34 patients (osteochondral trans-
plantation, n = 16; microfracture, n = 18); 34 controls were matched. An additional 13 patients received CR and anterior cruciate
ligament (ACL) reconstruction (CR&ACL), and 13 patients received only ACL reconstruction. Magnetic resonance imaging at 3.0-
tesla with T1rmapping and sagittal fat-saturated intermediate-weighted fast spin echo (FSE) sequences was performed to quan-
titatively and qualitatively analyze menisci (Whole-Organ Magnetic Resonance Imaging Score [WORMS] assessment). Patients in
the CR and CR&ACL groups were examined 4 months (n = 34; n = 13), 1 year (n = 21; n = 8), and 2 years (n = 9; n = 5) after CR.
Control subjects were scanned at baseline and after 1 and 2 years, ACL patients after 1 and 2 years.
Results: At baseline, global meniscus T1rvalues (mean 6 SEM) were higher in individuals with CR (14.2 6 0.5 ms; P = .004) and in
individuals with CR&ACL (17.1 6 0.9 ms; P\.001) when compared with controls (12.8 6 0.6 ms). After 2 years, there was a statistical
difference between T1rat the overlying meniscus above cartilage defects (16.4 6 1.0 ms) and T1rof the subgroup of control knees with-
out cartilage defects (12.1 6 0.8 ms; P \ .001) and a statistical trend to the CR group (13.3 6 1.0 ms; P = .09). At baseline, 35% of
subjects with CR showed morphological meniscus tears at the overlying meniscus; 10% of CR subjects showed an increase in the
WORMS meniscus score within the first year, and none progressed in the second year. Control subjects with (without) cartilage defects
showed meniscus tears in 30% (5%) at baseline; 38% (19%) increased within the first year, and 15% (10%) within the second year.
Conclusion: This study demonstrated more severe meniscus degeneration after CR surgery compared with controls. However,
progression of T1rvalues was not observed from 1 to 2 years after surgery. These results suggest that CR may prevent degen-
erative meniscus changes.
Keywords: cartilage repair; meniscus; T1r; WORMS; 3.0-T MRI
Articular cartilage defects have limited potential to regen-
erate and are associated with the early onset of osteoar-
thritis (OA).8Over the past decade, cartilage repair (CR)
has been increasingly used to treat focal cartilage defects
of the knee.35The most common technique is microfracture
(Mfx), which is used for smaller lesions that do not affect
the subchondralbone.For largerregions, besides
autologous chondrocyte implantation, osteochondral trans-
plantation (OCT) is the procedure of choice.41
Noninvasive magnetic resonance imaging (MRI) is the
most important diagnostic tool for monitoring the postoper-
ative course of these patients.43,48The current standard in
clinical practice is 1.5-T MRI of the knee,20,29but evidence
suggests that 3.0-T MRI may be more advantageous, yield-
ing a stronger magnetic field strength, which allows for
thinner sections, higher plane spatial resolution, and
increased signal-to-noise ratio. In addition, 3.0-T MRI is
more sensitive in diagnosing meniscus lesions, a known
contributor to the early onset of OA.1,40
The American Journal of Sports Medicine, Vol. 40, No. 12
? 2012 The Author(s)
at UCSF LIBRARY & CKM on March 4, 2013ajs.sagepub.comDownloaded from
Quantitative T1rrelaxation time measurements reflect
early degenerative changes in the biochemical composition
of cartilage, such as proteoglycan loss and increase in
water content.1,24,25,40It has recently also been applied to
quantitatively and noninvasively detect meniscus degener-
ation.5,39Although the exact factors that contribute to
a higher meniscus T1rin subjects with OA and a lower
meniscus T1rin healthy subjects are not yet clear, 1 study
found a positive correlation between cartilage degenera-
tion and increased T1rvalues in the meniscus.38
Several studies have examined results after CR on
a descriptive level. The challenge remains to demonstrate
that CR can prevent joint degeneration beyond other meth-
ods by clinically validating imaging outcomes.48A few studies
have used 3.0-T MRI for follow-up, but little is known about
meniscus degeneration as an outcome parameter after CR.
However, evaluation of the effect of meniscus degeneration
is crucial to ensure quality control and the development of
future treatment guidelines in patients after CR.
As T1rrelaxation time measurements of CR tissue have
been described previously,14the purpose of this study was
to evaluate meniscus degeneration, as measured by mor-
phological assessment and quantitative T1r meniscus
measurements, at multiple longitudinal time points in
patients who received CR surgery compared with controls.
We hypothesized that patients who underwent CR would
have higher meniscus T1rvalues, indicating more degener-
ative meniscus changes at baseline, but no further menis-
cus degeneration during follow-up.
MATERIALS AND METHODS
Ninety-four subjects were analyzed in this study. Thirty-four
subjects were treated with cartilage repair procedures (CR
group) for isolated posttraumatic or degenerative full-thick-
ness cartilage defects in the knee (graded III and IV according
to the International Cartilage Repair Society [ICRS] classifi-
cation). A control cohort (control group; n = 34) was recruited
that was matched for sex and for Kellgren-Lawrence (KL)
score.18Thirteen additional subjects received CR as well as
anterior cruciate ligament (ACL) reconstruction (CR&ACL
group). A final additional 13 subjects received only ACL recon-
struction (ACL group). The study was approved by the local
Institutional Review Board and conducted in accordance
with the Committee for Human Research at the University
of California–San Francisco (UCSF). All subjects gave written
informed consent before participation in the study. Data were
prospective and nonrandomized.
The indication for CR surgery was made in consultation with
the patient and confirmed during arthroscopy of the affected
knee joint. Exclusion criteria for CR procedures were as fol-
lows: uncontained large defects of several joint regions, sig-
nificant degenerative changes of the affected joint (KL
score .2), noncorrectable ligamentous instability, varus or
valgus malalignment of .5?, muscle loss, presence of inflam-
matory or metabolic disorders, obesity (body mass index
.30 kg/m2), and age .55 years. Additional exclusion criteria
for this study were MR contraindications and retropatellar
CR. None of the patients received CR at 2 sites, nor did
any receive revision surgery during the observation period.
The indication for ACL reconstruction was subacute, com-
plete ACL rupture by clinically diagnosed anterior-posterior
laxity (Lachman grades 2 to 3) with confirmation by MRI.
All procedures were performed by a single surgeon
(C.B.M.). Microfracture was used for smaller lesions (\3
cm2). For mostly larger defects, patients received OCT. The
cartilage defect area was debrided until its edges were com-
pletely surrounded by healthy cartilage. Microfracture sur-
gery was performed during arthroscopy as previously
described.2,47The cartilage defects treated by OCT (using
the Osteochondral Autograft Transfer System [OATS];
Arthrex Inc, Naples, Florida) were assessed arthroscopically
and subsequently treated by an arthrotomy of the knee.2
Osteochondral grafts were harvested from the nonweight-
bearing, nonarticulating intercondylar notch region of the
trochlea of the same knee during surgery. A mean (6SD) of
2.0 6 1.0 transplanted cylinders was used. The ACL recon-
struction was performed with single-bundle hamstring or
patellar tendon graft.22During the postoperative period,
weightbearing was limited to 15 kg for 6 weeks (3 wk if
only ACL reconstruction) and was gradually increased to
reach full weightbearing after 8-12 weeks. Subjects under-
went physical therapy to strengthen the joint during follow-
Standard standing anteroposterior plain radiographs of
the knee were obtained in all subjects at baseline. All
subjects were scanned with a 3.0-T Signa HDx MR scanner
(GE Healthcare, Milwaukee, Wisconsin) using an 8-
channel phased array transmit/receive knee coil (Invivo,
Orlando, Florida). For semiquantitative Whole-Organ
Magnetic Resonance Imaging Score (WORMS) assess-
ment,37an intermediate-weighted fat-saturated fast spin
echo (FSE) sequence (repetition time [TR]/echo time
[TE] = 4300/51 ms, field of view [FOV] = 14 cm, matrix =
*Address correspondence to Pia M. Jungmann, MD, Institute of Radiology, Technical University of Munich, Ismaninger Strasse 22, 81675 Munich, Ger-
many (e-mail: email@example.com).
yMusculoskeletal and Quantitative Imaging Research, Department of Radiology and Biomedical Imaging, University of California–San Francisco, San
zInstitute of Radiology, Technical University of Munich, Germany.
§Department of Orthopaedic Surgery, University of California–San Francisco, San Francisco, California.
One or more of the authors has declared the following potential conflict of interest or source of funding: This research was supported by the National
Institutes of Health (grants R01 AR46905, K25 AR053633, P50 AR060752, and UO1 AR059507) and the Osteoarthritis Initiative (OAI; N01-AR-2-2258, N01-
AR-2-2259, N01-AR-2-2260, N01-AR-2-2261, and N01-AR-2-2262).
Vol. 40, No. 12, 2012Cartilage Repair and Meniscus Degeneration 2701
at UCSF LIBRARY & CKM on March 4, 2013ajs.sagepub.comDownloaded from
512 3 256, slice thickness = 2.5 mm, gap = 0.5 mm) was
used. Sagittal 3D T1r sequences were used to quantify
the meniscus relaxation time.6,27,39A spin-lock technique
was followed by a spoiled gradient recalled (SPGR) acquisi-
tion using transient signals evolving toward steady state
with the following parameters: repetition time/echo time
(TR/TE) = 9.3/3.7 ms, time of recovery = 1500 ms, field of
view (FOV) = 14 cm, matrix = 256 3 192, slice thickness
= 3 mm, bandwidth = 31.25 kHz, views per segment =
48, time of spin-lock (TSL) = 0/10/40 ms, frequency of
spin-lock (FSL) = 500 Hz.26Parallel imaging with array
spatial sensitivity technique (ASSET) was performed
with an acceleration factor of 2.
Images were obtained at the clinically important time
points of 4 months (4.0 6 1.1 mo; 34/34 CR subjects: n =
16 OCT, n = 18 Mfx; 13/13 CR&ACL subjects: n = 2 OCT,
n = 11 Mfx) and 1 year after CR (11.8 6 2.8 mo; 21/34
CR subjects: n = 10 OCT, n = 11 Mfx; 8/13 CR&ACL sub-
jects: n = 2 OCT, n = 6 Mfx). The MR studies were obtained
2 years (24.6 6 1.2 mo) after surgery in 9 of 34 CR subjects
(n = 5 OCT, n = 4 Mfx) and 5 of 13 CR&ACL subjects (n = 2
OCT, n = 3 Mfx). Control subjects without ACL reconstruc-
tion (n = 34) were scanned at baseline and after 1 and 2
years. Patients with only ACL reconstruction (n = 13)
were scanned 1 and 2 years after surgery (see the Appen-
dix, available in the online version of this article at
Images were evaluated separately by 2 musculoskeletal radi-
ologists (P.M.J., 4 y of experience; L.N., 6 y of experience); if
scores were not identical, consensus reading by both radiolog-
ists and another independent radiologist (T.M.L., 22 y of
experience) was performed. Images were reviewed on picture
archiving communication system (PACS) workstations (Agfa,
Ridgefield Park, New Jersey). Regarding plain knee radio-
graphs, subjects with a KL score of more than 2 were
excluded from CR surgery and from this study. A UCSF-mod-
ified WORMS system, as presented in Table 1, was used to
assess morphological abnormalities.38The medial and lateral
menisci were separated into the following compartments:
anterior horn, meniscus body, and posterior horn. For preva-
lence analysis, ‘‘no meniscus defect,’’ ‘‘simple tear,’’ and ‘‘com-
plex tear’’ were differentiated. For progression analysis, any
increase of entire meniscus WORMS score was considered
as ‘‘progression.’’ Controls were divided into subjects with
and those without morphological cartilage defect. Bone mar-
row lesions (BMLs) of the compartment with cartilage repair
were graded according to the WORMS score and separated
into BMLs ?2 cm (WORMS score ?2) and BMLs .2 cm
(WORMS score 3).
The T1rsequences were transferred to a remote work-
station (SPARC; Sun Microsystems, Mountain View, Cali-
fornia) and analyzed by using software developed at our
institution with an interactive display language (IDL;
Research Systems, Boulder, Colorado) environment. Seg-
mentation of the anterior and posterior horns of the medial
and lateral meniscus in every section was performed by 1
radiologist and supervised by a senior radiologist.39
Sagittal imaging precluded the meniscus body segmenta-
tion. The T1rmaps were reconstructed by fitting the T1r
images pixel by pixel using a Levenberg Marquardt
mono-exponential fitting algorithm developed in-house.53
Reproducibility was calculated in a randomly selected sam-
ple of 10 image datasets for each compartment. For
WORMS measurements, each subregion of the images
was graded twice, by 2 radiologists on 2 separate occasions.
Linear weighted Cohen k values were calculated. Interob-
server k was 0.89 for cartilage defects. Intraobserver k
was 0.91 and 0.95. For bone marrow abnormalities, inter-
observer k was 0.80, and intraobserver k was 0.81 and
0.87. Interobserver k was 0.80 for meniscus defects. Intra-
observer k was 0.89 and 0.95. The mean coefficient of
Morphological Meniscus and Cartilage Grading and
Definitions Based on WORMS Scoringa
A. Meniscus Grading
WORMS grade of meniscus partsb
0 = normal meniscus
1 = intrasubstance abnormalities
2 = nondisplaced meniscus tear
3 = displaced or complex tear
4 = complete meniscus destruction/maceration
Total WORMS gradec
0 = grade 0 in all meniscus parts
1 = no grade .1 in any part
2 = grade 2 in 1 part
3 = grade 2 in .1 part
4 = grade 3 in 1 or more parts
5 = grade 4 in 1 part
6 = grade 4 in .1 part
No meniscus tear = entire meniscus grade of \2 (no tear)
Simple meniscus tear = entire meniscus grade of 2
Complex meniscus tear = entire meniscus grade of .2
No progression = no increase of the entire meniscus
grade over time
Progression = increase of the entire meniscus grade over time
B. Cartilage Grading
WORMS cartilage score
0 = no cartilage abnormality
1 = intrasubstance cartilage abnormalities
.1 = morphological cartilae lesion with volume loss
WORMS 0 or 1 = no cartilage defect
WORMS .1 = cartilage defect
aWORMS, Whole-Organ Magnetic Resonance Imaging Score.
bAnterior horn, posterior horn, and body of each meniscus was
cMedial and lateral meniscus were assessed separately.
2702 Jungmann et alThe American Journal of Sports Medicine
at UCSF LIBRARY & CKM on March 4, 2013ajs.sagepub.comDownloaded from
variation (%), determined for T1r measurements of the
meniscus in our laboratory, was 4.1%.5,53
Mean T1rvalues were calculated for both menisci and glob-
ally (mean of the value for medial and lateral meniscus)
from the segmented regions of interest. Statistical process-
ing was performed with JMP software, version 9 (SAS
Institute, Cary, North Carolina). Statistics were obtained
applying multivariate regression models that adjusted in
1 model for KL score, sex, and age by adding these varia-
bles as covariates for each of the analyses. For T1rvalue
measurements, 1-way analysis of variance (ANOVA) and
2-way Student t test were applied. For morphological anal-
ysis, the Mann-Whitney U test was used. Results were con-
sidered as significantly different if P \ .05. Values in the
Results section are presented as mean 6 standard error
of the mean (SEM) unless stated otherwise.
Of the 94 subjects in this study, 34 were treated with only
CR (21 male, 13 female), 13 with CR and ACL reconstruc-
tion, and 13 with only ACL reconstruction. The CR group
and control group (n = 34) were matched for sex and KL
score (Table 2). Age was significantly different between
the CR and the control group (mean 6 SD, 35 6 11 y vs
47 6 11 y, respectively). Cartilage repair was performed
20 of 34 times at the medial femoral condyle, 10 of 34 times
at the lateral femoral condyle, and 4 of 34 times at the
trochlea (only Mfx). Screening controls for cartilage defects
revealed that 13 of 34 control subjects had cartilage defects
(medial, 7/34; lateral, 6/34) at baseline.
Baseline Meniscus T1rAnalysis
At baseline, patients without surgery (control group) had the
lowest T1rvalues (global T1r: 12.8 6 0.6 ms) (Table 3, Figure
1). Considering both menisci separately, the medial meniscus
showed slightly higher values than the lateral meniscus in
controls at baseline (13.1 6 0.6 ms vs 12.5 6 0.6 ms, P =
.07). The CR group showed a significantly higher global T1r
value of 14.2 6 0.5 ms 4 months after surgery (P = .004 vs
controls). The CR&ACL group showed the highest T1rvalues
4 months after surgery (17.1 6 0.9 ms; P\.001 vs controls).
In comparing T1r values of the overlying meniscus
above the cartilage repair regions with the overlying
meniscus above untreated cartilage defect regions in the
control subgroup with cartilage defects at baseline, no sig-
nificant difference was detected at baseline (14.7 6 0.7 ms
vs 14.8 6 0.9 ms) (Figure 2). However, differences with
a statistical trend were found between the control sub-
group with cartilage defects and the global meniscus T1r
of the control subgroup without cartilage defects (12.5 6
0.8 ms; P = .06), and a significant difference was found
between the CR group and the control subgroup without
cartilage defects (P = .001). In contrast, for each of the 4
separately segmented meniscus parts, the difference in
T1rbetween the CR group and the control group was not
significant at baseline because of different CR locations.
Meniscus T1rat Follow-up
Global meniscus T1rin the CR group did not increase in the
first (14.1 6 0.7 ms) or second year (13.2 6 0.9 ms) after sur-
gery (Figure 1). Global T1rvalues in the control group were
stable during follow-up, 13.0 6 0.6 ms after 1 year and
13.1 6 0.6 ms after 2 years. The CR&ACL group did not
show a further increase over time but still showed a signifi-
cant difference from the control group after 2 years (P = .04).
In the subgroup of controls without cartilage defects
(n = 21), global T1rvalues were stable over time, and there
was no significant difference in T1r of the overlying
Epidemiological Data of the Analyzed Groupsa
Parameter OverallControlsCR ACLACL&CR
No. of patients
Age 6 SD, y
KL score (0:1:2)
40 6 12
47 6 11
35 6 11
37 6 9
32 6 11
aACL, anterior cruciate ligament; CR, cartilage repair; KL, Kellgren-Lawrence; SD, standard deviation.
Global Meniscus T1rRelaxation Time Values
After CR and After ACL Reconstruction and CR
Versus Without Surgerya
Surgery TypenGlobal T1r, Mean 6 SEM, ms
12.8 6 0.6
14.2 6 0.5
17.1 6 0.9
aCR, cartilage repair; ACL, anterior cruciate ligament. P overall
(analysis of variance) = .002.
bStatistically significant compared to no surgery (P \ .05).
Vol. 40, No. 12, 2012 Cartilage Repair and Meniscus Degeneration 2703
at UCSF LIBRARY & CKM on March 4, 2013 ajs.sagepub.comDownloaded from
meniscus above the cartilage repair region (n = 9) after 2
years (12.1 6 0.8 ms vs 13.3 6 1.0 ms; P = .11) (Figure
2). However, there was a statistical trend toward a differ-
ence between T1rof the overlying meniscus above the car-
tilage repair region and T1r of the overlying meniscus
above the untreated cartilage defect (16.4 61.0 ms) after
2 years (P = .09) and a statistical difference between T1r
of the overlying meniscus above the untreated cartilage
defect and global T1rof the control subgroups without car-
tilage defects after 2 years (P \ .001).
Absolute T1rprogression was also calculated between
the different time points and showed the same trend, but
no statistically significant difference (P . .05).
Comparison of Mfx and OCT in the CR Group
Subjects with Mfx (n = 18) showed higher T1rat the over-
lying meniscus than OCT subjects (n = 16) (Figure 3). T1r
decreased at the 1- and 2-year time points in subjects
with OCT. In subjects with Mfx, T1r decreased only in
the second year of follow-up. However, these results did
not show any significant difference.
Correlation of Bone Marrow Lesions With Meniscus T1r
At baseline, CR subjects with BMLs ?2 cm (correlating
with a WORMS score ?2) showed slightly higher T1rval-
ues at the overlying meniscus (15.5 6 1.3 ms; 14/34) than
CR subjects with BMLs .2 cm (14.1 6 1.0 ms; 20/34; P =
.09). However, after 2 years, BMLs ?2 cm (n = 4) were
associated with lower meniscus T1r values than BMLs
.2 cm (n = 5; 12.1 6 1.8 ms vs 14.5 6 2.1 ms; P = .10). Pres-
ence of large BMLs 1 year after CR was not significantly
associated with meniscus T1rafter 2 years (P = .20).
Patients with only ACL reconstruction (ACL group, n = 13)
had a global meniscus T1rvalue of 14.3 6 0.8 ms 1 year after
surgery (P = .03 vs control group). The medial meniscus
showed lower T1rvalues than the lateral meniscus (13.8 6
0.8 ms vs 14.6 6 0.8 ms, P = .20). In patients with only
ACL reconstruction, both the medial and lateral meniscus
showed higher T1rvalues at the 2-year follow-up time point
compared with the 1-year time point (medial: 15.7 6 1.0 ms;
P = .03; lateral: 15.7 6 0.9 ms; P . .05). The absolute differ-
ence of the T1rvalues between the 2 time points was higher
in the medial meniscus. At the 2-year time point, global
meniscus T1rin the ACL group (n = 13) was significantly
higher than in the CR group (n = 34; P = .009).
Morphological Meniscus Lesions at Baseline
At baseline, 15% (5/34) and 20% (7/34) of the CR group had
simple and complex morphological meniscus tears, respec-
tively, at the overlying meniscus (Table 4 and Figure 4).
Figure 1. Global meniscus T1rvalues 0.3, 1, and 2 y after
cartilage repair (CR) compared with subjects 0.3, 1, and 2 y
after combined CR and anterior cruciate ligament recon-
struction (CR&ACL) and with controls at baseline and after
1 and 2 y. *P \ .05, compared with the control group at
the corresponding time point.
Figure 2. Global meniscus T1rvalues of the control subgroup
without cartilage defect at baseline and after 1 and 2 y, com-
pared with meniscus T1rat the overlying meniscus above the
cartilage defect in the control subgroup with cartilage defect
(Defect) and compared with meniscus T1r at the overlying
meniscus above the cartilage repair region in the cartilage
repair group (CR) 0.3, 1, and 2 y after surgery. *P\.05, com-
pared with the control subgroup without cartilage defect (No)
at the according time point. **P = .06.
Figure 3. The T1rvalues of the overlying meniscus above the
repair region in subjects with osteochondral transplantation
(OCT) and subjects with microfracture procedure (Mfx) 0.3,
1, and 2 y after surgery. Differences between the groups
were not significant (P . .05).
2704Jungmann et al The American Journal of Sports Medicine
at UCSF LIBRARY & CKM on March 4, 2013ajs.sagepub.comDownloaded from
The control subgroup without cartilage defects (21/34)
showed significantly fewer meniscus tears (0% [0/21] simple
tears; 5% [1/21] complex tears; P = .001). The control sub-
group with untreated cartilage defects (13/34; 15% (2/13)
simple tears; 15% (2/13) complex tears) showed no signifi-
cant difference of meniscus tears compared to the CR group
(P = .84), but significantly more meniscus tears than the
control subgroup without cartilage defects (P = .006). The
CR&ACL group showed more meniscus tears than all other
groups (medial meniscus 92% [12/13]; lateral meniscus 69%
[9/13]). Subjects in the ACL group (only ACL reconstruction)
had medial meniscus tears in 35% (5/13) and lateral menis-
cus tears in 31% (4/13) at 1 year after surgery.
Incidence (in percentages) of Morphological
Meniscus Defects in Different Groupsa
No surgery (all controls)
with cartilage defects
without cartilage defect
aIn the cartilage repair group, the overlying meniscus above the
cartilage repair region was analyzed. In the control subcohort with
cartilage defect, the overlying meniscus above the defect was ana-
lyzed. In the control subcohort without cartilage defect, the menis-
cus with the higher Whole-Organ Magnetic Resonance Imaging
Score (WORMS) was considered.
Figure 4. Prevalence of morphological meniscus lesions.
Green indicates no meniscus lesions; yellow, simple; and
red, complex meniscus lesions. (A) Lesions at the overlying
meniscus above cartilage repair regions in the cartilage
repair group (CR) and above cartilage defects in the control
subgroup with cartilage defect (Defect) were compared
with the control subgroup without cartilage defect (No). (B)
Lesions at the medial (MM) and lateral (LM) meniscus in the
group with reconstruction of the anterior cruciate ligament
(ACL; 1 y after surgery) and the group with combined CR
and ACL surgery (CR&ACL), respectively.
Morphological Meniscus Changes Measured
by WORMS Scoring Between 2 Time Pointsa
Group; Time Points, yn Progression, %b
Control subgroup with defect; 0-1
Control subgroup with defect; 1-2
Control subgroup without defect; 0-1
Control subgroup without defect; 1-2
aWORMS, Whole-Organ Magnetic Resonance Imaging Score;
CR, cartilage repair.
bPercentage of subjects who showed an increase in the entire
meniscus grading for the overlying meniscus above the cartilage
repair region or the cartilage defect region, respectively. In con-
trols without cartilage defect, any increase of either meniscus
was considered as progression.
Figure 5. T1rcolor maps of the anterior and posterior horn of
the medial meniscus of 1-y and 2-y follow-up time points,
overlaid with the first-echo images. (A) Control subject with
cartilage defect at the medial femoral condyle, who did not
receive a cartilage repair (CR) procedure. (B) CR subject
with osteochondral transplantation at the medial femoral
condyle. Blue color indicates low, and red color indicates
high meniscus T1rvalues. Subjects with untreated cartilage
lesions showed a greater increase in T1rvalues over time
compared with the subjects with CR.
Vol. 40, No. 12, 2012Cartilage Repair and Meniscus Degeneration 2705
at UCSF LIBRARY & CKM on March 4, 2013ajs.sagepub.com Downloaded from
Progression of Morphological Meniscus Lesions
Any increase in the entire WORMS score was considered as
progression of morphological meniscus lesions (Table 5).
During follow-up, 10% of the CR group (2/21) showed an
increase at the overlying meniscus in the first postoperative
year; none progressed in the second postoperative year (0/9).
Within the subgroup of control subjects without cartilage
defects, 19% (4/21) increased in the first and 10% (2/21)
increased in the second year. Within the subgroup with
untreated cartilage defects, 38% (5/13) increased in the first
and 15% (2/13) increased in the second year. However, the
differences between the groups were not significant (P .
.05). In the CR&ACL group, 14% and 29% of subjects,
respectively, showed an increase at the medial and lateral
meniscus within the first postoperative year. Within the
second year, 20% showed an increase at the medial and
20% at the lateral meniscus. An increase of morphological
meniscus lesions was found in 19% of subjects with ACL
surgery between the first and second postoperative year.
This longitudinal study analyzed meniscus degeneration as
an outcome parameter after cartilage reconstruction proce-
dures. Noninvasive MRI is used for monitoring the postop-
erative course of these patients to ensure quality control
and development of future treatment guidelines. In this
study, evaluation of the menisci was performed semiquanti-
tatively by morphological analysis (WORMS) and quantita-
tively by meniscus T1r relaxation time measurements at
several time points during a 2-year 3.0-T MRI follow-up.
Four months after surgery, patients with CR and ACL
reconstruction had a significantly higher meniscus T1r
than controls, whereas T1r was the highest in subjects
who received both surgeries. During follow-up, individuals
with only ACL reconstruction and controls with cartilage
defects showed a further increase in T1rvalues, whereas
T1rvalues in CR subjects did not increase (Figure 5). In con-
sideration of the methodological limitations, these findings
suggest that individuals with CR surgery may benefit
from this procedure, as it appears to prevent meniscus
degeneration and potentially, the early onset of OA.
Apart from autologous chondrocyte implantation,16,34Mfx
and OCT are 2 alternatively applied CR procedures.2,41How-
ever, the outcome after these procedures with respect to pre-
vention of further degenerative changes and early onset of
OA of the knee is unclear. Most studies have evaluated clin-
ical outcomes, and few have considered using MRI at follow-
up.4,48,50However, meniscus evaluation remains an impor-
tant parameter with respect to evaluation of progression of
early and advanced OA.45In fact, previous studies have
used meniscus evaluation to assess the risk of OA as an out-
come after surgery, particularly ACL reconstruction.19,51By
evaluating the menisci, we confirmed previously reported
findings that there are more simple and complex meniscus
defects in individuals who have undergone ACL and CR sur-
gery.10,30Cartilage defects usually coincide with degenera-
tion of the overlying meniscus, as confirmed in our study45;
patients with cartilage defects or patients after repair of car-
tilage defects had more meniscus lesions. In addition, in our
study we detected higher meniscus T1rvalues in individuals
with CR, ACL reconstruction, and untreated cartilage
defects. Meniscus degeneration can be quantified and contin-
uously monitored by T1rmapping. T1rhas recently been used
not only for noninvasive biochemical analysis of cartilage, but
also for detection and monitoring of meniscus degenera-
tion.9,39T1r measurements of cartilage repair areas have
been performed, and results have previously been reported.14
Studies have shown that cartilage degeneration correlates
with meniscus degeneration.53Even though T2relaxation
time remains the more widely used technique in biochemi-
cally assessing cartilage integrity, studies have shown that
lesions.36,42For meniscus evaluation, both T2and T1rmeas-
urements appear to be useful.53Rauscher et al39found that
high meniscus T1rvalues were associated with osteoarthritic
knees. In our current study, we found that subjects who
underwent CR procedures showed no further increase of
T1rvalues in the meniscus over time. Consequently, if menis-
cus T1rcorrelates with cartilage degeneration, it suggests
that CR surgery may halt the progression of intrameniscal
degeneration and further joint degeneration.
Follow-up time points at 4 months, 1 year, and 2 years
after CR surgery were chosen because of their clinical rel-
evance. Most failures occur during the early postoperative
period.31,49The healing process, involving cell prolifera-
tion, matrix production, and matrix remodeling, is espe-
cially crucial within the first postoperative year.17This
outcome is also reflected by meniscus T1rvalues, which
show a particular improvement not within the first, but
within the second year. On the contrary, clinically, many
patients report a benefit of the intervention for only 2
years.32,44Although OCT is a more invasive, open proce-
dure including arthrotomy of the knee joint, clinically,
a better outcome up to 3 years after surgery was reported
for OCT compared to Mfx.12Interestingly, for OCT, we
found a slight decrease of T1rvalues already within the
first year of follow-up, whereas Mfx led to a decrease of
T1rvalues only within the second year of follow-up. This
finding is concordant with clinical findings.23MRI has
been found to correlate with clinical outcome after Mfx.32
Although there is no definite correlation between CR his-
tology and clinical outcome, there is evidence that in
patients with good histological results, there is less therapy
failure.21In the case of OCT, preexisting cartilage is trans-
planted. This technique showed an improvement in MR
findings postoperatively between 4 months and 1 year.28
However, some histological changes to fibrous cartilage
and modification of the repair area have been observed. In
the case of CR procedures that involve bone marrow stimu-
lation, meniscus status improved over time. This finding is
consistent with the literature, since a technique of bone
marrow stimulation has been described that allows a better
healing of the meniscus after meniscus repair.11
The BMLs did not correlate with meniscus T1r at 4
months or 1 year after surgery. However, persisting huge
BMLs 2 years after surgery showed a trend toward a statis-
tical correlation with a higher meniscus T1r. This finding
2706 Jungmann et alThe American Journal of Sports Medicine
at UCSF LIBRARY & CKM on March 4, 2013 ajs.sagepub.comDownloaded from
supports the presumption that a persisting BML is associ-
ated with outcome after CR.52However, BMLs after 1 year
did not predict 2-year meniscus T1r.
Meniscus integrity is crucial for proper knee joint func-
tioning and shock-absorption. An influence of meniscus
lesions on further OA progression has been observed.33
Meniscus defects and OA have also been associated with
higher and less homogeneous cartilage T2relaxation time
values.15Our study showed that patients with CR demon-
strated improved meniscus T1r values over time, which
suggests that CR plays a role in halting OA progression
as a postoperative outcome. In this context, in particular
the findings regarding control subjects with and without
morphological cartilage defects seem important. Subjects
with untreated morphological cartilage defects had higher
T1r values at follow-up time points than at baseline,
whereas controls without defects and CR subjects did not.
Meniscus T1rin subjects with only ACL reconstruction
(ACL group) increased over time. It is known that although
the ACL is reconstructed, kinematics may not be completely
restored. Anterior-posterior tibial translation is usually well
restored, whereas rotational stability was observed to be still
pathological after ACL reconstruction.13,46Therefore, ACL-
reconstructed subjects still suffer from increased degenera-
tion of the knee, as detected by T1rin our study. One year
after surgery, particularly in the lateral meniscus, the ACL
group showed higher T1rvalues, which is consistent with
previous T1rfindings,53as well as clinical MR findings that
result from the kissing bone bruise. However, during the sec-
ond year after ACL reconstruction, T1r values increased
more at the medial meniscus than at the lateral meniscus,
indicating a faster degeneration of the medial meniscus after
ACL reconstruction, probably because of rotational instabil-
ity.3In control patients, the medial meniscus showed slightly
higher T1rvalues and more morphological meniscus lesions,
which is consistent with prior studies.45We found that
patients with combined ACL and CR procedures showed
the highest T1rvalues, with slightly lower values at follow-
up. This result supports previously reported findings of rea-
sonable outcomes for combined surgery.7
There are several limitations of the present study. First,
not all patients came back for 1- and 2-year follow-up. The
low rate of follow-up was the result of a young patient cli-
entele with high mobility, who are leaving the area and are
not able to attend for follow-up visits, as well as limited
scan times at our institution and missing or not reaching
the right follow-up time point. Second, age was signifi-
cantly different between the groups; since age is known
as 1 of the most important risk factors for OA, results
were adjusted for this parameter. Third, hamstring and
patellar tendon grafts (bone-tendon-bone, BTB) were not
differentiated, since the focus of this study was on CR
and the number of subjects in each group would have
been too small for further analysis. In the ACL group, 9
patients received BTB grafts and 4 patients received ham-
string grafts; in the CR&ACL group, 4 patients received
BTB grafts and 9 patients received hamstring grafts. It
may be interesting to evaluate the influence of different
techniques in future studies. Fourth, results were not
adjusted for lower limb alignment, which could potentially
also slightly influence the results. However, subjects with
an axis deviation .5? were excluded. Finally, only 1 clini-
cal sequence was used, because of scan-time limitations.
In this study we used 3.0-T MRI meniscus T1rrelaxation
time measurements along with morphological meniscus
assessment in a cross-sectional, 2-year longitudinal analysis
of individuals who underwent cartilage resurfacing proce-
dures and compared the findings with normal controls.
Meniscus T1rvalues were higher in individuals with CR
or cartilage defects at baseline compared with individuals
without defects. Although T1rdid not increase at the fol-
low-up time points in CR patients, increasing T1rvalues
were detected in patients with untreated cartilage defects
or ACL reconstruction. Compared with controls, morpholog-
ical meniscus defects showed a lower progression during the
second year of follow-up after CR. These results suggest
that patients with focal cartilage defects may benefit from
cartilage repair procedures with regard to prevention of fur-
ther meniscus matrix degeneration and consequently, pre-
vention of early OA in the knee.
1. Akella SV, Regatte RR, Gougoutas AJ, et al. Proteoglycan-induced
changes in T1rho-relaxation of articular cartilage at 4T. Magn Reson
2. Bedi A, Feeley BT, Williams RJ 3rd. Management of articular cartilage
defects of the knee. J Bone Joint Surg Am. 2010;92:994-1009.
3. Bellabarba C, Bush-Joseph CA, Bach BR Jr. Patterns of meniscal
injury in the anterior cruciate-deficient knee: a review of the literature.
Am J Orthop (Belle Mead NJ). 1997;26:18-23.
4. Benthien JP, Schwaninger M, Behrens P. We do not have evidence
based methods for the treatment of cartilage defects in the knee.
Knee Surg Sports Traumatol Arthrosc. 2011;19:543-552.
5. Bolbos RI, Ma CB, Link TM, Majumdar S, Li X. In vivo T1rho quanti-
tative assessment of knee cartilage after anterior cruciate ligament
injury using 3 Tesla magnetic resonance imaging. Invest Radiol.
6. Brittain JH, Hu BS, Wright GA, Meyer CH, Macovski A, Nishimura
DG. Coronary angiography with magnetization-prepared T2 contrast.
Magn Reson Med. 1995;33:689-696.
7. Brophy RH, Zeltser D, Wright RW, Flanigan D. Anterior cruciate liga-
ment reconstruction and concomitant articular cartilage injury: inci-
dence and treatment. Arthroscopy. 2010;26:112-120.
8. Buckwalter JA. Articular cartilage injuries. Clin Orthop Relat Res.
9. Choi JA, Gold GE. MR imaging of articular cartilage physiology. Magn
Reson Imaging Clin N Am. 2011;19:249-282.
10. Farr J, Rawal A, Marberry KM. Concomitant meniscal allograft trans-
plantation and autologous chondrocyte implantation: minimum 2-
year follow-up. Am J Sports Med. 2007;35:1459-1466.
11. Freedman KB, Nho SJ, Cole BJ. Marrow stimulating technique to
augment meniscus repair. Arthroscopy. 2003;19:794-798.
12. Gudas R, Stankevicius E, Monastyreckiene E, Pranys D, Kalesinskas
RJ. Osteochondral autologous transplantation versus microfracture
for the treatment of articular cartilage defects in the knee joint in ath-
letes. Knee Surg Sports Traumatol Arthrosc. 2006;14:834-842.
13. Haughom B, Schairer W, Souza RB, Carpenter D, Ma CB, Li X.
Abnormal tibiofemoral kinematics following ACL reconstruction are
associated with early cartilage matrix degeneration measured by
MRI T1rho. Knee. 2012;19:482-487.
Vol. 40, No. 12, 2012 Cartilage Repair and Meniscus Degeneration 2707
at UCSF LIBRARY & CKM on March 4, 2013ajs.sagepub.comDownloaded from
14. Holtzman DJ, Theologis AA, Carballido-Gamio J, Majumdar S, Li X,
Benjamin C. T(1rho) and T(2) quantitative magnetic resonance imag-
ing analysis of cartilage regeneration following microfracture and
mosaicplasty cartilage resurfacing procedures. J Magn Reson Imag-
15. Joseph GB, Baum T, Carballido-Gamio J, et al. Texture analysis of
cartilage T2 maps: individuals with risk factors for OA have higher
and more heterogeneous knee cartilage MR T2 compared to normal
controls - data from the osteoarthritis initiative. Arthritis Res Ther.
16. Jungmann PM, Salzmann GM, Schmal H, Pestka JM, Sudkamp NP,
Niemeyer P. Autologous chondrocyte implantation for treatment of
cartilage defects of the knee: what predicts the need for reinterven-
tion? Am J Sports Med. 2012;40:58-67.
17. Kalson NS, Gikas PD, Briggs TW. Current strategies for knee carti-
lage repair. Int J Clin Pract. 2010;64:1444-1452.
18. Kellgren JH, Lawrence JS. Radiological assessment of osteo-
arthrosis. Ann Rheum Dis. 1957;16:494-502.
19. Kessler MA, Behrend H, Henz S, Stutz G, Rukavina A, Kuster MS.
Function, osteoarthritis and activity after ACL-rupture: 11 years
follow-up results of conservative versus reconstructive treatment.
Knee Surg Sports Traumatol Arthrosc. 2008;16:442-448.
20. Kijowski R, Blankenbaker DG, Davis KW, Shinki K, Kaplan LD, De
Smet AA. Comparison of 1.5- and 3.0-T MR imaging for evaluating
the articular cartilage of the knee joint. Radiology. 2009;250:839-848.
21. Knutsen G, Drogset JO, Engebretsen L, et al. A randomized trial
comparing autologous chondrocyte implantation with microfracture.
Findings at five years. J Bone Joint Surg Am. 2007;89:2105-2112.
22. Kothari A, Haughom B, Subburaj K, Feeley B, Li X, Ma CB. Evaluating
rotational kinematics of the knee in ACL reconstructed patients using
3.0Tesla magnetic resonance imaging. Knee. 2012;19:648-651.
23. Kreuz PC, Steinwachs MR, Erggelet C, et al. Results after microfrac-
ture of full-thickness chondral defects in different compartments in
the knee. Osteoarthritis Cartilage. 2006;14:1119-1125.
24. Li X, Benjamin Ma C, Link TM, et al. In vivo T(1rho) and T(2) mapping
of articular cartilage in osteoarthritis of the knee using 3 T MRI. Oste-
oarthritis Cartilage. 2007;15:789-797.
25. Li X, Cheng J, Lin K, et al. Quantitative MRI using T1rho and T2 in
human osteoarthritic cartilage specimens: correlation with biochem-
ical measurements and histology. Magn Reson Imaging. 2011;
26. Li X, Han ET, Busse RF, Majumdar S. In vivo T(1rho) mapping in car-
tilage using 3D magnetization-prepared angle-modulated partitioned
k-space spoiled gradient echo snapshots (3D MAPSS). Magn Reson
27. Li X, Han ET, Newitt D, Majumdar S. T1rho relaxation quantification
using spiral imaging: a preliminary study. Conf Proc IEEE Eng Med
Biol Soc. 2004;2:1032-1035.
28. Link TM, Mischung J, Wortler K, Burkart A, Rummeny EJ, Imhoff AB.
Normal and pathological MR findings in osteochondral autografts
with longitudinal follow-up. Eur Radiol. 2006;16:88-96.
29. Link TM, Stahl R, Woertler K. Cartilage imaging: motivation, techni-
ques, current and future significance. Eur Radiol. 2007;17:1135-1146.
30. Magnussen RA, Mansour AA, Carey JL, Spindler KP. Meniscus sta-
tus at anterior cruciate ligament reconstruction associated with
radiographic signs of osteoarthritis at 5- to 10-year follow-up: a sys-
tematic review. J Knee Surg. 2009;22:347-357.
31. Miller BS, Steadman JR, Briggs KK, Rodrigo JJ, Rodkey WG. Patient
satisfaction and outcome after microfracture of the degenerative
knee. J Knee Surg. 2004;17:13-17.
32. Mithoefer K, McAdams T, Williams RJ, Kreuz PC, Mandelbaum BR.
Clinical efficacy of the microfracture technique for articular cartilage
repair in the knee: an evidence-based systematic analysis. Am J
Sports Med. 2009;37:2053-2063.
33. Muthuri SG, McWilliams DF, Doherty M, Zhang W. History of knee
injuries and knee osteoarthritis: a meta-analysis of observational
studies. Osteoarthritis Cartilage. 2011;19:1286-1293.
34. Niemeyer P, Kostler W, Salzmann GM, Lenz P, Kreuz PC, Sudkamp
NP. Autologous chondrocyte implantation for treatment of focal
cartilage defects in patients age 40 years and older: a matched-pair
analysis with 2-year follow-up. Am J Sports Med. 2010;38:2410-2416.
35. Niemeyer P, Pestka JM, Kreuz PC, et al. Characteristic complications
after autologous chondrocyte implantation for cartilage defects of the
knee joint. Am J Sports Med. 2008;36:2091-2099.
36. Oneto JM, Ellermann J, LaPrade RF. Longitudinal evaluation of car-
tilage repair tissue after microfracture using T2-mapping: a case
report with arthroscopic and MRI correlation. Knee Surg Sports Trau-
matol Arthrosc. 2010;18:1545-1550.
37. Peterfy CG, Guermazi A, Zaim S, et al. Whole-Organ Magnetic Res-
onance Imaging Score (WORMS) of the knee in osteoarthritis. Oste-
oarthritis Cartilage. 2004;12:177-190.
38. Peterfy CG, Schneider E, Nevitt M. The osteoarthritis initiative: report
on the design rationale for the magnetic resonance imaging protocol
for the knee. Osteoarthritis Cartilage. 2008;16:1433-1441.
39. Rauscher I, Stahl R, Cheng J, et al. Meniscal measurements of T1rho
and T2 at MR imaging in healthy subjects and patients with osteoar-
thritis. Radiology. 2008;249:591-600.
40. Regatte RR, Akella SV, Lonner JH, Kneeland JB, Reddy R. T1rho
relaxation mapping in human osteoarthritis (OA) cartilage: compari-
son of T1rho with T2. J Magn Reson Imaging. 2006;23:547-553.
41. Salzmann GM, Niemeyer P, Steinwachs M, Kreuz PC, Sudkamp NP,
Mayr HO. Cartilage repair approach and treatment characteristics
across the knee joint: a European survey. Arch Orthop Trauma
42. Salzmann GM, Paul J, Bauer JS, et al. T2 assessment and clinical
outcome following autologous matrix-assisted chondrocyte and
osteochondral autograft transplantation. Osteoarthritis Cartilage.
43. Sanders TG, Mentzer KD, Miller MD, Morrison WB, Campbell SE,
Penrod BJ. Autogenous osteochondral ‘‘plug’’ transfer for the treat-
ment of focal chondral defects: postoperative MR appearance with
clinical correlation. Skeletal Radiol. 2001;30:570-578.
44. Solheim E, Hegna J, Oyen J, Austgulen OK, Harlem T, Strand T.
Osteochondral autografting (mosaicplasty) in articular cartilage
defects in the knee: results at 5 to 9 years. Knee. 2010;17:84-87.
45. Stehling C, Lane NE, Nevitt MC, Lynch J, McCulloch CE, Link TM.
Subjects with higher physical activity levels have more severe focal
knee lesions diagnosed with 3T MRI: analysis of a non-symptomatic
cohort of the osteoarthritis initiative. Osteoarthritis Cartilage. 2010;
46. Tashman S, Collon D, Anderson K, Kolowich P, Anderst W. Abnormal
rotational knee motion during running after anterior cruciate ligament
reconstruction. Am J Sports Med. 2004;32:975-983.
47. Theologis AA, Schairer WW, Carballido-Gamio J, Majumdar S, Li X,
Ma CB. Longitudinal analysis of T(1rho) and T(2) quantitative MRI
of knee cartilage laminar organization following microfracture sur-
gery. Knee. 2012;19:652-657.
48. Trattnig S, Domayer S, Welsch GW, Mosher T, Eckstein F. MR imag-
ing of cartilage and its repair in the knee–a review. Eur Radiol.
49. Trattnig S, Millington SA, Szomolanyi P, Marlovits S. MR imaging of
osteochondral grafts and autologous chondrocyte implantation. Eur
50. Von Keudell A, Atzwanger J, Forstner R, Resch H, Hoffelner T, Mayer
M. Radiological evaluation of cartilage after microfracture treatment:
a long-term follow-up study. Eur J Radiol. 2012;81:1618-1624.
51. Wegrzyn J, Chouteau J, Philippot R, Fessy MH, Moyen B. Repeat
revision of anterior cruciate ligament reconstruction: a retrospective
review of management and outcome of 10 patients with an average
3-year follow-up. Am J Sports Med. 2009;37:776-785.
52. Welsch GH, Zak L, Mamisch TC, Resinger C, Marlovits S, Trattnig S.
Three-dimensional magnetic resonance observation of cartilage
repair tissue (MOCART) score assessed with an isotropic three-
dimensional true fast imaging with steady-state precession sequence
at 3.0 Tesla. Invest Radiol. 2009;44:603-612.
53. Zarins ZA, Bolbos RI, Pialat JB, et al. Cartilage and meniscus assessment
using T1rho and T2 measurements in healthy subjects and patients with
osteoarthritis. Osteoarthritis Cartilage. 2010;18:1408-1416.
For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav
2708Jungmann et alThe American Journal of Sports Medicine
at UCSF LIBRARY & CKM on March 4, 2013 ajs.sagepub.comDownloaded from