The American Journal of Sports

Medicine http://ajs.sagepub.com/

Cartilage Restoration of the Knee: A Systematic Review and Meta-Analysis of Level 1 Studies
Raman Mundi, Asheesh Bedi, Linda Chow, Sarah Crouch, Nicole Simunovic, Elizabeth Sibilsky Enselman and Olufemi
R. Ayeni
Am J Sports Med published online July 2, 2015
DOI: 10.1177/0363546515589167

The online version of this article can be found at:

http://ajs.sagepub.com/content/early/2015/07/01/0363546515589167

Published by:

http://www.sagepublications.com

On behalf of:
American Orthopaedic Society for Sports Medicine

Additional services and information for The American Journal of Sports Medicine can be found at:
Published online July 2, 2015 in advance of the print journal.

P<P
Email Alerts: http://ajs.sagepub.com/cgi/alerts

Subscriptions: http://ajs.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

>> OnlineFirst Version of Record - Jul 2, 2015

What is This?

Downloaded from ajs.sagepub.com at Purdue University on July 5, 2015

 AJSM PreView, published on July 2, 2015 as doi:10.1177/0363546515589167

Clinical Sports Medicine Update

Cartilage Restoration of the Knee

A Systematic Review and Meta-Analysis
of Level 1 Studies
Raman Mundi,* MD, Asheesh Bedi,y MD, Linda Chow,* BHSc, Sarah Crouch,z BSc,
Nicole Simunovic,z MSc, Elizabeth Sibilsky Enselman,y MEd, AT, ATC,
and Olufemi R. Ayeni,*# MD, MSc, FRCSC
Investigation performed at the Division of Orthopaedic Surgery, McMaster University,
Hamilton, Ontario, Canada

Background: Focal cartilage defects of the knee are a substantial cause of pain and disability in active patients. There has been
an emergence of randomized controlled trials evaluating surgical techniques to manage such injuries, including marrow stimula-
tion (MS), autologous chondrocyte implantation (ACI), and osteochondral autograft transfer (OAT).
Purpose: A meta-analysis was conducted to determine if any single technique provides superior clinical results at intermediate
follow-up.
Study Design: Systematic review and meta-analysis of randomized controlled trials.
Methods: The MEDLINE, EMBASE, and Cochrane Library databases were systematically searched and supplemented with manual
searches of PubMed and reference lists. Eligible studies consisted exclusively of randomized controlled trials comparing MS, ACI, or
OAT techniques in patients with focal cartilage defects of the knee. The primary outcome of interest was function (Lysholm score,
International Knee Documentation Committee score, Knee Osteoarthritis Outcome Score) and pain at 24 months postoperatively. A
meta-analysis using standardized mean differences was performed to provide a pooled estimate of effect comparing treatments.
Results: A total of 12 eligible randomized trials with a cumulative sample size of 765 patients (62% males) and a mean (6SD)
lesion size of 3.9 6 1.3 cm2 were included in this review. There were 5 trials comparing ACI with MS, 3 comparing ACI with
OAT, and 3 evaluating different generations of ACI. In a pooled analysis comparing ACI with MS, there was no difference in out-
comes at 24-month follow-up for function (standardized mean difference, 0.47 [95% CI, –0.19 to 1.13]; P = .16) or pain (standard-
ized mean difference, –0.13 [95% CI, –0.39 to 0.13]; P = .33). The comparisons of ACI to OAT or between different generations of
ACI were not amenable to pooled analysis. Overall, 5 of the 6 trials concluded that there was no significant difference in functional
outcomes between ACI and OAT or between generations of ACI.
Conclusion: There is no significant difference between MS, ACI, and OAT in improving function and pain at intermediate-term
follow-up. Further randomized trials with long-term outcomes are warranted.
Keywords: knee cartilage defect; autologous chondrocyte implantation; microfracture; osteochondral autograft transfer

Cartilage defects of the knee often result in symptoms that

#
Address correspondence to Olufemi R. Ayeni, MD, MSc, FRCSC,
Division of Orthopaedic Surgery, Department of Surgery, McMaster Uni-
versity, 1200 Main Street West 4E15, Hamilton, ON L8N 3Z5, Canada
(email: femi.ayeni@gmail.com).
*Division of Orthopaedic Surgery, Department of Surgery, McMaster
University, Hamilton, Ontario, Canada.
y
MedSport, Department of Orthopaedic Surgery, University of Michi-
gan, Ann Arbor, Michigan, USA.
z increasing and nearing 300,000 in 2010 alone.30
Department of Clinical Epidemiology and Biostatistics, McMaster
University, Hamilton, Ontario, Canada.
One or more of the authors has declared the following potential con-
flict of interest or source of funding: A.B. has been a consultant for Smith
& Nephew and is a shareholder of A3 Surgical.
The American Journal of Sports Medicine, Vol. XX, No. X
DOI: 10.1177/0363546515589167
Ó 2015 The Author(s)
cause patients considerable pain and disability.1,25 Unfor-
tunately, these defects have a limited intrinsic capacity
for spontaneous healing due to the avascular and hypocel-
lular nature of articular cartilage.12 Surgical management
has accordingly been a mainstay in the management of
these injuries, with the estimated annual incidence of
knee cartilage procedures in the United States steadily
Current surgical restoration techniques can largely be
classified into 1 of 3 categories, including marrow stimula-
tion, cell-based implantation, and osteochondral graft-
ing.32,34 Marrow stimulation (MS) procedures such as
microfracture stimulate a healing response by exposing
the subchondral bone marrow and creating a blood clot
that fills the defect and facilitates the recruitment of

1
Downloaded from ajs.sagepub.com at Purdue University on July 5, 2015
2 Mundi et al
mesenchymal stem cells to repair it. The repair, however,
is composed of fibrocartilaginous tissue with inferior wear
characteristics compared with hyaline cartilage.12 Since its
inception, microfracture (MF) has become the most fre-
quently used procedure for knee cartilage repair of focal
defects.30 Autologous chondrocyte implantation (ACI) is
a cell-based method that first showed clinical promise in
1994 in a study conducted by Brittberg and colleagues.8 It
has earned a particular distinction from MS techniques
for its ability to create a hyaline-like cartilage repair that
better resembles articular cartilage.8 Second- and third-gen-
eration ACI techniques have since been introduced, in
which the previously used periosteal patch has been
replaced with a collagen cover (ACI-C) and scaffold, respec-
tively. This modification has been driven by an effort to min-
imize the complications (periosteal hypertrophy and
overgrowth), procedural invasiveness, and technical
demands associated with the harvest and use of a periosteal
flap cover (ACI-P).7,29 Whole-tissue transplantation techni-
ques, including osteochondral autograft transfer (OAT),
have also been developed as a means for providing a hyaline
repair. Although donor site morbidity remains a concern,
OAT has been successfully used for the management of
smaller lesions that extend into subchondral bone.32
Several randomized controlled trials (RCTs) comparing
the above-mentioned techniques for knee cartilage repair
have been published in recent years. However, there are
no recent systematic reviews and meta-analyses that have
focused exclusively on randomized trial evidence and com-
prehensively evaluated all 3 treatment approaches. Accord-
ingly, the objectives of the following systematic review and
meta-analysis were 2-fold: (1) to summarize the available
high-quality evidence comparing the outcomes of MS, ACI,
and OAT techniques for the restoration of focal knee carti-
lage defects and (2) to perform a pooled analysis of function
and pain outcomes comparing these techniques.
METHODS
Search Strategy
Two independent reviewers conducted a search for ran-
domized trials evaluating surgical techniques for cartilage
defects of the knee. Three databases were searched, includ-
ing MEDLINE, EMBASE, and the Cochrane Library (1950
to February 2013). The search terms included cartilage,
knee, knee joint, knee injuries, joint instability, cartilage
cell, and articular cartilage (see Appendix Table A1, avail-
able online at http://ajsm.sagepub.com/supplemental).
A 2-step screening process was used to select eligible trials.
Initially, all titles and abstracts produced by the search strat-
egy were screened in duplicate. Relevant studies and those
with questionable eligibility were retrieved for full-text review.
All discrepancies were resolved through discussion and con-
sensus with the senior authors (O.R.A. and A.B.). All referen-
ces of the included studies were further reviewed for possible
inclusion of studies missed by the initial search strategy.
During the preparation of this review, the search was
updated using MEDLINE (January 2013 to April 2014)
Downloaded from ajs.sagepub.com at Purdue University on July 5, 2015
The American Journal of Sports Medicine
and EMBASE (January 2013 to April 2014) with an identi-
cal search strategy. The search was further supplemented
with a manual search of PubMed.
Assessment of Study Eligibility
To be included in this review, potential studies were
required to meet the following eligibility criteria: (1) RCT
design; (2) included patients with isolated cartilage lesions
or chondral defects in the knee; (3) directly compared 2 sur-
gical treatment options between MS, ACI, and OAT or ACI
between generations; (4) study on human subjects; and (5)
published in the English language.
The exclusion criteria were (1) non-RCTs, (2) evalua-
tions of nonsurgical treatment of the knee, and (3) patients
with osteoarthritis.
Assessment of Methodological Quality
Methodological quality was assessed using the Grading of
Recommendations Assessment, Development and Evalua-
tion (GRADE) system.19 In addition to assessing the risk
of bias of individual trials, the GRADE system incorporates
several other factors in determining the overall quality of
evidence of the pooled effect estimate.2
Briefly, in the GRADE system, the quality of evidence can
be assigned to 1 of 4 possible levels: high, moderate, low, or
very low. Randomized trial evidence is initially designated
as ‘‘high quality,’’ whereas evidence relying on observational
studies is rated as ‘‘low quality.’’ The assigned quality of evi-
dence is then either rated up or down levels, depending on 5
important factors: (1) risk of bias, (2) inconsistency, (3) indi-
rectness, (4) imprecision, and (5) publication bias.2,24 Risk
of bias refers to an evaluation of potential study limitations,
such as allocation concealment, blinding, and loss to follow-
up, among other factors.24 Inconsistency assesses the degree
of similarity of results across trials and is judged by evaluat-
ing variation in point estimates, the degree of overlap in con-
fidence intervals, the magnitude of I 2, and statistical tests for
heterogeneity.21 Indirectness is a measure of how applicable
the studies are to both the patients and interventions of
interest and the relevance of study outcomes.22 Imprecision
evaluates the certainty surrounding our pooled effect esti-
mate and requires an assessment of the 95% confidence
interval of the pooled effect, as well as the cumulative sample
size of the analysis.20 Finally, publication bias is a judgment
based on the sample size of trials, funding sources, and
assessment of a funnel plot.23 The final GRADE assessment
therefore reflects our overall confidence in the pooled esti-
mate of effect and is not simply a statement of the internal
validity of the trials.
Data Abstraction
Data were abstracted from the full text of all eligible articles
using standardized data collection forms. Abstracted data
included title, author, year, location, surgical technique, sam-
ple size, percentage of male patients, mean age, lesion loca-
tion, lesion size, lesion grade (Outerbridge or International
Vol. XX, No. X, XXXX Knee Cartilage Restoration Meta-analysis 3

Cartilage Repair Society [ICRS]), length of follow-up, and 913 Studies Identified
weightbearing protocol. MEDLINE: 610 studies
Based on a preliminary survey of the most commonly used EMBASE: 300 studies
outcome scales, outcome scores were recorded for the follow- Cohrane Database: 3 studies
ing: Lysholm knee score, International Knee Documentation
Committee (IKDC) knee form, and Knee injury Osteoarthri- Title Review Excluded Studies
• Duplicates: 90 studies
tis Outcome Score (KOOS). If none of these scales were used,
we documented results for the primary functional scale used
Abstract Review Excluded Studies
by the trial. Pain scores were also recorded when reported • Non-RCTs: 626 studies
independently.

Data Analysis
Descriptive statistics were calculated for the included stud-
ies using SPSS software (v20.0; SPSS Inc). Continuous
variables are reported as weighted means with standard
deviations, and categorical variables are reported as fre-
quencies or percentages.
A meta-analysis evaluating function and pain at 24
months postoperatively was carried out to directly compare
surgical techniques. It was determined a priori that a mini-
mum of 3 trials with reported effect sizes would be required
for pooling. A standardized mean difference was used as the
pooled summary statistic to account for differences in the
functional and pain outcome scales used by the individual
trials. We were prepared to do a sensitivity analysis for
any trials that were outliers with respect to sample size,
variation in surgical technique, or methodological quality.
Analysis was performed using final outcome scores, as
well as change scores (change from baseline to follow-up)
for studies in which sufficient data were reported. Scoring
scales were reversed as necessary to ensure consistency in
direction across all trials. An attempt was made to contact
authors via email for outcome data as required. In the event
that this was unsuccessful, data were retrieved from graphs
published in the manuscript when available using Graph-
Click (v3.0; Arizona Software).
All pooled analyses were performed using the random-
effects model with the use of Review Manager software
(RevMan) 5.3 (The Nordic Cochrane Centre, The Cochrane
Collaboration, 2012).
Assessment of Agreement
The interrater reliability analysis was performed using the
k statistic to determine the degree of agreement in the
study identification process. Agreement was deemed as
fair (k = 0.21-0.40), moderate (k = 0.41-0.60), substantial
(k = 0.61-0.80), or almost perfect (k = 0.81-1.00).31
RESULTS
Study Identification
The initial literature search yielded 913 studies, 15 of
which met the inclusion criteria.§ Among these 15 studies,
10 were original trials included for final review, as 4
§
References 3-6, 9, 17, 18, 26-28, 35-37, 39, 40.
197 Studies
Excluded Studies
Full-Text Review
• 188 studies
Hand Search of
Full-Text References,
PubMed Search, Addional Studies Idenfied: 7
MEDLINE/EMBASE Update • Hand search (5)
• PubMed (2)
16 Studies Included
(4 duplicate cohorts)
12 Trials
Figure 1. Screening and selection of trials. RCT, randomized
controlled trial.
studies were extended follow-ups of previously published
trials, and 1 study used a cohort of patients from another
trial.6,17,27,36,37 Two additional trials10,13 were identified
via the supplemental search, resulting in 12 trials included
for final review (Figure 1). The overall agreement between
the reviewers for the full-text screening was substantial
(k = 0.63; 95% CI, 0.46-0.80).
Study Characteristics
All 12 trials were published between 2003 and 2012, with 11
(92%) of these studies conducted in Europe. The number of
patients enrolled in each trial ranged from 21 to 118 with
a cumulative sample size of 765 patients and a mean
(6SD) sample size of 64 6 29. Across the trials, 62% of par-
ticipants were males, and the mean age of patients was 32 6
3 years. Most trials provided follow-up data at 24 months
(n = 8; 67%). The mean lesion size was 3.9 6 1.3 cm2,
with 6 trials studying patients with a mean lesion size
between 4 and 10 cm2 and 6 trials studying those with
lesions smaller than 4 cm2 (Appendix Table A2, available
online).
The most commonly evaluated procedure was ACI,
which was assessed in 11 trials (92%). Marrow stimulation
techniques were the second most commonly studied tech-
nique (n = 6; 50%). Of these trials, 5 assessed microfracture
and 1 evaluated mechanical abrasion (chondroplasty). A
total of 4 trials (33%) evaluated OAT.
With respect to head-to-head comparisons, 5 trials com-
pared ACI with MS (4 MF and 1 mechanical abrasion), 3 tri-
als compared ACI with OAT, and 3 evaluated different
generations of ACI techniques. Only 1 trial compared

Downloaded from ajs.sagepub.com at Purdue University on July 5, 2015

4 Mundi et al The American Journal of Sports Medicine

marrow stimulation (MF) with OAT. This latter study was

not included in the outcome analysis, as the paucity of com-
parator studies precluded any synthesis of findings across
trials.

Methodological Quality
The risk of bias assessment for each trial is presented in
Figure 2. Five of the 11 trials adequately reported their
randomization method, and 7 trials described a method
for concealing allocation. Most trials (n = 9) had a follow-
up of 80%. None of the trials were judged to be low risk
of bias, as all had at least 2 individual criteria that
3
were high or uncertain risk.
An overall GRADE assessment was performed for com-
4
parisons that were amenable to a pooled analysis, includ-
ing function and pain for ACI compared with MS. The
5
quality of evidence for our pooled comparison of function
in patients treated with ACI and MS was deemed ‘‘low.’’
9
This was due to the inconsistency of results across trials
and the imprecision of the pooled effect due to a small
10
cumulative sample size. The quality of evidence for our
pooled assessment of pain in patients treated with ACI
13
and MS was judged as ‘‘moderate,’’ due to imprecision of
the pooled effect estimate (Table 1).
26

Autologous Chondrocyte Implantation 28

Versus Marrow Stimulation
A total of 5 trials consisting of 338 patients compared ACI
with MS techniques (4 trials of MF; 1 trial of abra-
sion).4,9,28,35,39 Three trials used a scaffold for their ACI
technique,4,9,39 whereas 2 trials used an autologous perios-
teal flap.28,35 The mean age of patients was 34 6 2 years
across the 5 trials, with a mean lesion size of 3.4 6
1.0 cm2 for the 4 trials that reported this variable. Three
of the trials provided follow-up data on functional out-
comes at a mean of 24 months postoperatively.4,9,28 One
trial reported 12-month outcomes,39 and 1 trial provided
change scores at 24 months but final scores at 18
months.35,36 Three of the 5 studies also reported pain out-
comes9,28,35 (Appendix Table A3, available online). With
the exception of 1 trial, 4 studies concluded that ACI
resulted in an overall statistically superior improvement
in functional outcomes compared with MS.
Pooled analysis of final functional outcomes (Lysholm,
IKDC, KOOS) across all 5 studies, however, demonstrated
a statistically insignificant trend favoring ACI treatment
(standardized mean difference [SMD], 0.47 [95% CI,
–0.19 to 1.13]; P = .16). Heterogeneity across studies was
statistically significant (P \ .00001, I2 = 87%). Sensitivity
analysis with removal of the trial by Visna et al39—due
to variation in follow-up (12 months) and surgical tech-
nique in the MS group (abrasion)—did not alter the signif-
icance of the pooled estimate of effect (P = .40) (Figure 3).
Sufficient data were available from 3 trials to pool change
scores across the trials.9,28,35 The pooled effect comparing the
difference in improvement of function from baseline to follow-
36
39
40
Figure 2. Risk of bias assessment.
up between ACI and MF was statistically insignificant (SMD,
0.23 [95% CI, –0.31 to 0.76]; P = .40). Although this analysis
demonstrated less heterogeneity, it nevertheless remained
significant (P = .03, I 2 = 70%) (Figure 4).
Three trials reported pain score outcomes, with 2 trials
assessing pain at 24 months and 1 trial at 18 months.9,28,35
There was no difference in final pain score between those
patients treated with ACI or MF (SMD, –0.13 [95%
CI, –0.39 to 0.13]; P = .33). Testing for heterogeneity was
statistically insignificant (P = .61, I 2 = 0%) (Figure 5).
Autologous Chondrocyte Implantation
Versus Osteochondral Autograft Transfer
Three studies with a cumulative sample size of 187 patients,
a mean patient age of 31 6 2 years, and a mean lesion size of
3.80 6 1.2 cm2 compared ACI with OAT.5,10,26 All trials used
an autologous periosteal flap for the ACI procedure, and 2
studies specifically reported evaluating mosaicplasty as
the osteochondral technique.

Downloaded from ajs.sagepub.com at Purdue University on July 5, 2015

Vol. XX, No. X, XXXX Knee Cartilage Restoration Meta-analysis 5

TABLE 1
Quality of Evidence Assessment: ACI Compared With MS for Knee Cartilage Defecta

Anticipated Absolute Effect

No. of Participants Quality of the Relative Risk Risk Difference
Outcome (Studies) Evidence (GRADEb) Effect (95% CI) With MS With ACI

Function at median 338 (5 RCTs) Lowc,d — — SMD = 0.47 higher (0.19 lower
24-mo follow-up to 1.13 higher)
Pain at median 228 (3 RCTs) Moderated — — SMD = 0.13 lower (0.39 lower
24-mo follow-up to 0.13 higher)

a
The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the
intervention (and its 95% CI). ACI, autologous chondrocyte implantation; MS, marrow stimulation; OR, odds ratio; RCT, randomized con-
trolled trial; RR, risk ratio; SMD, standardized mean difference.
b
GRADE Working Group grades of evidence: High quality = We are very confident that the true effect lies close to that of the estimate of
the effect. Moderate quality = We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the
effect, but there is a possibility that it is substantially different. Low quality = Confidence in the effect estimate is limited: the true effect may
be substantially different from the estimate of the effect. Very low quality = We have very little confidence in the effect estimate: the true
effect is likely to be substantially different from the estimate of effect.
c
Variation in point estimates, marginal overlap of confidence intervals, I2 = 86% to 87%, heterogeneity: P \ .00001. Rated down.
d
Wide confidence interval inclusive of no effect, pooled sample size \400 patients. Rated down.

P I
P

Figure 3. Forest plot of final scores: (A) autologous chondrocyte implantation (ACI) versus marrow stimulation (MS) and (B) ACI
versus microfracture (MF; sensitivity).

Follow-up varied from 12 to 24 months, and overall con- Chondrocyte Implantation—

clusions were discrepant among the trials. Two studies
found that treatment with ACI or OAT yielded comparable
functional outcomes as measured using the modified
Cincinnati rating system.5,10 One study with 24-month
follow-up, however, demonstrated superior Lysholm scores
with OAT treatment.26 Of note, the study by Dozin et al10
was weakened by a significant number of patients who
either did not receive the intended treatment or were lost
to follow-up before outcome assessment.
These trials were not amenable to a meaningful pooled
analysis as the studies varied in analyzing outcomes as
continuous or categorical data. None of the trials reported
pain scores independently.
Comparison of Generations
Three trials evaluated different generations of ACI, with
a single head-to-head comparison between each generation
(ACI-P, ACI-C, scaffold ACI).3,13,40 There were a total of
180 patients in the studies, with a mean patient age of
32 6 2 years and a mean cartilage defect size of 5.3 6
0.8 cm2.
Follow-up varied between 12 and 24 months, with all
studies concluding that overall, functional outcomes as
assessed with the modified Cincinnati score3,13 and IKDC
score40 are comparable across ACI techniques. One study,
however, found ACI-P was superior to matrix-induced

Downloaded from ajs.sagepub.com at Purdue University on July 5, 2015

6 Mundi et al
Figure 4. Forest plot of change scores: autologous chondrocyte implantation (ACI) versus microfracture (MF).
Figure 5. Forest plot of pain scores: autologous chondrocyte implantation (ACI) versus microfracture (MF).
ACI at 24 months in terms of Lysholm score.40 Only 1
study, comparing ACI-C with M-ACI, reported pain scores
and found no statistically significant difference between
the groups.3
DISCUSSION
This systematic review and meta-analysis represents a
comprehensive assessment of randomized trial evidence
evaluating the 3 major approaches of surgical treatment
(marrow stimulation, cell-based technology, and whole-
tissue transplantation) for symptomatic focal knee cartilage
defects. The findings of this review suggest that, based on
current evidence, no single technique is unequivocally supe-
rior in improving intermediate-term function and pain out-
comes. This conclusion stems from a pooled analysis that
demonstrated a statistically insignificant trend favoring
ACI over MF, as well as a review of trials in which there
was no significant superiority of any technique when com-
paring ACI with OAT and ACI across generations.
Several reviews have previously compared surgical
techniques for the management of knee cartilage defects,
yet notable differences in methodology exist between
such studies and the current review.14-16,25,33 Harris and
colleagues25 conducted a systematic review of 11 original
studies on ACI treatment, including both randomized tri-
als and observational studies. Our current review focused
exclusively on randomized trials and identified 2 trials
not included in the review by Harris et al. Furthermore,
our conclusion regarding the efficacy of ACI versus MS is
based on a meta-analysis. Despite 4 of 5 trials reporting
that ACI resulted in superior functional outcomes at 2
years compared with MF, the pooled effect size with
Downloaded from ajs.sagepub.com at Purdue University on July 5, 2015
The American Journal of Sports Medicine
studies appropriately weighted demonstrated that this
trend was insignificant. In addition to pooling final scores,
we also explored the pooled effect of change scores for the 3
trials that made such data available. This marginally
decreased heterogeneity, but the pooled standardized
mean difference remained insignificant.
There unfortunately remains a paucity of randomized
trials with extended follow-up periods, which precluded
us from performing a meaningful assessment of long-
term outcomes regarding function and pain. Only 2 studies
comparing ACI with MF published 5-year outcomes, both
of which found the 2 techniques to be comparable.27,38 To
that end, it has been demonstrated that any observed
short-term difference in treatment effect between ACI
and MF will ultimately diminish over time.33 Bentley
et al6 also performed a long-term analysis at 10 years of
their study comparing ACI with OAT (mosaicplasty) and
concluded that ACI results in superior functional outcomes
as measured by the Cincinnati scale. It must be noted, how-
ever, that only 15 of 48 patients randomized to mosaicplasty
were included in the functional assessment at 10 years. Fur-
thermore, patients waited a mean of 7.2 years from symp-
tom onset to treatment. Several studies have documented
prolonged delay of surgery as a strong prognostic indicator
for poor outcomes after cartilage restoration.3,38
There are several methodological strengths to the cur-
rent review. First and foremost, our review focused exclu-
sively on randomized trial evidence that was identified
through a comprehensive search strategy. Where feasible,
we also performed a meta-analysis using a standardized
mean difference to provide a pooled estimate of treatment
effect. Finally, in addition to assessing the risk of bias for
individual studies, the GRADE system was applied to eval-
uate the body of evidence as a whole based on several
Vol. XX, No. X, XXXX
important criteria. Despite such methodological strengths,
any review is limited by the methodological quality and
reporting of its component trials. Several trials insuffi-
ciently reported outcomes such that we relied on graphical
depictions for estimating treatment effects. Furthermore,
assessments of function and pain typically rely on subjec-
tive evaluations by patients themselves. Although it is
often impractical in surgical trials to blind patients, and
thus the outcome assessors of function and pain, it was cer-
tainly possible to blind data analysts. None of the included
trials reported blinding of their data analysts for these sub-
jective outcomes. It has also been recognized that orthopae-
dic trials are often underpowered to detect even large
differences in treatment effect.11 In the current review of
12 trials, only 4 trials discussed a minimal clinically impor-
tant difference, and 4 reported a sample size calculation.
The importance of adequately powering trials with the nec-
essary sample size cannot be overstressed. Despite includ-
ing 5 trials in our pooled analysis, the relatively small
sample size required a downward assessment of the qual-
ity of evidence for imprecision.
Although MS, ACI, and OAT are all generally efficacious
in improving symptoms in patients with focal knee cartilage
defects, current best evidence does not support any one
surgical technique as a superior method for improving
intermediate-term function and pain. Further high-quality
randomized trials that are adequately powered and provide
long-term follow-up are certainly warranted.
REFERENCES
1. Alford JW, Cole BJ. Cartilage restoration, part 1: basic science, his-
torical perspective, patient evaluation, and treatment options. Am J
Sports Med. 2005;33(2):295-306.
2. Balshem H, Helfand M, Schunemann HJ, et al. GRADE guidelines: 3.
Rating the quality of evidence. J Clin Epidemiol. 2011;64(4):401-406.
3. Bartlett W, Skinner JA, Gooding CR, et al. Autologous chondrocyte
implantation versus matrix-induced autologous chondrocyte implan-
tation for osteochondral defects of the knee: a prospective, rando-
mised study. J Bone Joint Surg Br. 2005;87(5):640-645.
4. Basad E, Ishaque B, Bachmann G, Sturz H, Steinmeyer J. Matrix-
induced autologous chondrocyte implantation versus microfracture
in the treatment of cartilage defects of the knee: a 2-year randomised
study. Knee Surg Sports Traumatol Arthrosc. 2010;18(4):519-527.
5. Bentley G, Biant LC, Carrington RW, et al. A prospective, randomised
comparison of autologous chondrocyte implantation versus mosaic-
plasty for osteochondral defects in the knee. J Bone Joint Surg Br.
2003;85(2):223-230.
6. Bentley G, Biant LC, Vijayan S, Macmull S, Skinner JA, Carrington
RW. Minimum ten-year results of a prospective randomised study
of autologous chondrocyte implantation versus mosaicplasty for
symptomatic articular cartilage lesions of the knee. J Bone Joint
Surg Br. 2012;94(4):504-509.
7. Brittberg M. Cell carriers as the next generation of cell therapy for car-
tilage repair: a review of the matrix-induced autologous chondrocyte
implantation procedure. Am J Sports Med. 2010;38(6):1259-1271.
8. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L.
Treatment of deep cartilage defects in the knee with autologous
chondrocyte transplantation. N Engl J Med. 1994;331(14):889-895.
9. Crawford DC, DeBerardino TM, Williams RJ III. NeoCart, an autolo-
gous cartilage tissue implant, compared with microfracture for treat-
ment of distal femoral cartilage lesions: an FDA phase-II prospective,
randomized clinical trial after two years. J Bone Joint Surg Am.
2012;94(11):979-989.
Downloaded from ajs.sagepub.com at Purdue University on July 5, 2015
Knee Cartilage Restoration Meta-analysis 7
10. Dozin B, Malpeli M, Cancedda R, et al. Comparative evaluation of
autologous chondrocyte implantation and mosaicplasty: a multicen-
tered randomized clinical trial. Clin J Sport Med. 2005;15(4):220-226.
11. Freedman KB, Back S, Bernstein J. Sample size and statistical power
of randomised, controlled trials in orthopaedics. J Bone Joint Surg
Br. 2001;83(3):397-402.
12. Gomoll AH, Minas T. The quality of healing: articular cartilage. Wound
Repair Regen. 2014;22(suppl 1):30-38.
13. Gooding CR, Bartlett W, Bentley G, Skinner JA, Carrington R,
Flanagan A. A prospective, randomised study comparing two techni-
ques of autologous chondrocyte implantation for osteochondral
defects in the knee: periosteum covered versus type I/III collagen
covered. Knee. 2006;13(3):203-210.
14. Goyal D, Goyal A, Keyhani S, Lee EH, Hui JH. Evidence-based status
of second- and third-generation autologous chondrocyte implanta-
tion over first generation: a systematic review of level I and II studies.
Arthroscopy. 2013;29(11):1872-1878.
15. Goyal D, Keyhani S, Goyal A, Lee EH, Hui JH, Vaziri AS. Evidence-
based status of osteochondral cylinder transfer techniques: a sys-
tematic review of level I and II studies. Arthroscopy. 2014;30(4):
497-505.
16. Goyal D, Keyhani S, Lee EH, Hui JH. Evidence-based status of
microfracture technique: a systematic review of level I and II studies.
Arthroscopy. 2013;29(9):1579-1588.
17. Gudas R, Gudaite A, Pocius A, et al. Ten-year follow-up of a prospec-
tive, randomized clinical study of mosaic osteochondral autologous
transplantation versus microfracture for the treatment of osteochon-
dral defects in the knee joint of athletes. Am J Sports Med.
2012;40(11):2499-2508.
18. Gudas R, Kalesinskas RJ, Kimtys V, et al. A prospective randomized
clinical study of mosaic osteochondral autologous transplantation
versus microfracture for the treatment of osteochondral defects
in the knee joint in young athletes. Arthroscopy. 2005;21(9):1066-
1075.
19. Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines: 1. Introduc-
tion-GRADE evidence profiles and summary of findings tables.
J Clin Epidemiol. 2011;64(4):383-394.
20. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines 6. Rating the
quality of evidence—imprecision. J Clin Epidemiol. 2011;64(12):
1283-1293.
21. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 7. Rating
the quality of evidence—inconsistency. J Clin Epidemiol. 2011;
64(12):1294-1302.
22. Guyatt GH, Oxman AD, Kunz R, et al. GRADE guidelines: 8. Rating
the quality of evidence—indirectness. J Clin Epidemiol. 2011;64(12):
1303-1310.
23. Guyatt GH, Oxman AD, Montori V, et al. GRADE guidelines: 5. Rating
the quality of evidence—publication bias. J Clin Epidemiol. 2011;
64(12):1277-1282.
24. Guyatt GH, Oxman AD, Vist G, et al. GRADE guidelines: 4. Rating the
quality of evidence—study limitations (risk of bias). J Clin Epidemiol.
2011;64(4):407-415.
25. Harris JD, Siston RA, Pan X, Flanigan DC. Autologous chondrocyte
implantation: a systematic review. J Bone Joint Surg Am.
2010;92(12):2220-2233.
26. Horas U, Pelinkovic D, Herr G, Aigner T, Schnettler R. Autologous
chondrocyte implantation and osteochondral cylinder transplantation
in cartilage repair of the knee joint: a prospective, comparative trial.
J Bone Joint Surg Am. 2003;85(2):185-192.
27. Knutsen G, Drogset JO, Engebretsen L, et al. A randomized
trial comparing autologous chondrocyte implantation with microfrac-
ture: findings at five years. J Bone Joint Surg Am. 2007;89(10):2105-
2112.
28. Knutsen G, Engebretsen L, Ludvigsen TC, et al. Autologous chondro-
cyte implantation compared with microfracture in the knee: a ran-
domized trial. J Bone Joint Surg Am. 2004;86(3):455-464.
29. Marlovits S, Zeller P, Singer P, Resinger C, Vecsei V. Cartilage repair:
generations of autologous chondrocyte transplantation. Eur J Radiol.
2006;57(1):24-31.
8 Mundi et al
30. McCormick F, Harris JD, Abrams GD, et al. Trends in the surgical
treatment of articular cartilage lesions in the United States: an anal-
ysis of a large private-payer database over a period of 8 years.
Arthroscopy. 2014;30(2):222-226.
31. McHugh ML. Interrater reliability: the kappa statistic. Biochem Med
(Zagreb). 2012;22(3):276-282.
32. Moran CJ, Pascual-Garrido C, Chubinskaya S, et al. Restoration of
articular cartilage. J Bone Joint Surg Am. 2014;96(4):336-344.
33. Negrin LL, Vecsei V. Do meta-analyses reveal time-dependent differ-
ences between the clinical outcomes achieved by microfracture and
autologous chondrocyte implantation in the treatment of cartilage
defects of the knee? J Orthop Sci. 2013;18(6):940-948.
34. Safran MR, Seiber K. The evidence for surgical repair of articular
cartilage in the knee. J Am Acad Orthop Surg. 2010;18(5):
259-266.
35. Saris DB, Vanlauwe J, Victor J, et al. Characterized chondrocyte
implantation results in better structural repair when treating
symptomatic cartilage defects of the knee in a randomized con-
trolled trial versus microfracture. Am J Sports Med. 2008;36(2):
235-246.
For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav
Downloaded from ajs.sagepub.com at Purdue University on July 5, 2015
The American Journal of Sports Medicine
36. Saris DB, Vanlauwe J, Victor J, et al. Treatment of symptomatic cartilage
defects of the knee: characterized chondrocyte implantation results in
better clinical outcome at 36 months in a randomized trial compared to
microfracture. Am J Sports Med. 2009;37(suppl 1):10S-19S.
37. Van Assche D, Staes F, Van Caspel D, et al. Autologous chondrocyte
implantation versus microfracture for knee cartilage injury: a prospec-
tive randomized trial, with 2-year follow-up. Knee Surg Sports Trau-
matol Arthrosc. 2010;18(4):486-495.
38. Vanlauwe J, Saris DB, Victor J, et al. Five-year outcome of character-
ized chondrocyte implantation versus microfracture for symptomatic
cartilage defects of the knee: early treatment matters. Am J Sports
Med. 2011;39(12):2566-2574.
39. Visna P, Pasa L, Cizmar I, Hart R, Hoch J. Treatment of deep carti-
lage defects of the knee using autologous chondrograft transplanta-
tion and by abrasive techniques—a randomized controlled study.
Acta Chir Belg. 2004;104(6):709-714.
40. Zeifang F, Oberle D, Nierhoff C, Richter W, Moradi B, Schmitt H. Autolo-
gous chondrocyte implantation using the original periosteum-cover tech-
nique versus matrix-associated autologous chondrocyte implantation:
a randomized clinical trial. Am J Sports Med. 2010;38(5):924-933.