You are on page 1of 9

The American Journal of Sports


Clinical Outcomes After Cell-Seeded Autologous Chondrocyte Implantation of the Knee: When Can
Success or Failure Be Predicted?
Jan M. Pestka, Gerrit Bode, Gian Salzmann, Mathias Steinwachs, Hagen Schmal, Norbert P. Sdkamp and Philipp
Am J Sports Med 2014 42: 208 originally published online October 28, 2013
DOI: 10.1177/0363546513507768

The online version of this article can be found at:

Published by:

On behalf of:
American Orthopaedic Society for Sports Medicine

Additional services and information for The American Journal of Sports Medicine can be found at:

Email Alerts:




>> Version of Record - Jan 2, 2014

OnlineFirst Version of Record - Oct 28, 2013

What is This?

Downloaded from by guest on April 8, 2016

Clinical Outcomes After Cell-Seeded
Autologous Chondrocyte Implantation
of the Knee
When Can Success or Failure Be Predicted?
Jan M. Pestka,*y MD, Gerrit Bode,y MD, Gian Salzmann,y MD, Mathias Steinwachs,y MD,
Hagen Schmal,y MD, PhD, Norbert P. Sudkamp,y MD, PhD, and Philipp Niemeyer,y MD, PhD
Investigation performed at the Department of Orthopedic Surgery and Traumatology,
Freiburg University Hospital, Freiburg, Germany

Background: Autologous chondrocyte implantation (ACI) has been associated with satisfying results. Still, it remains unclear
when success or failure after ACI can be estimated.
Purpose: To evaluate the clinical outcomes of cell-seeded collagen matrixsupported ACI (ACI-Cs) for the treatment of cartilage
defects of the knee at 36 months and to determine a time point after ACI-Cs at which success or failure can be estimated.
Study Design: Cohort study; Level of evidence, 3.
Methods: A total of 80 patients with isolated full-thickness cartilage defects of the knee joint treated with ACI-Cs were prospec-
tively assessed before surgery as well as postoperatively by use of the International Knee Documentation Committee (IKDC) score
and Lysholm knee score.
Results: Preoperative IKDC and Lysholm scores increased from 49.6 and 59.5, respectively, to 79.1 and 83.5, respectively, at
36 months. Only half the patients (46.6%) with poor IKDC scores (ie, \70) at 6 months postoperatively showed continued
poor or fair scores at 36 months follow-up. The probability of poor scores at 36 months after surgery further increased to
0.61 and 0.81, respectively, when scores were persistent at 12 and 24 months. All 3 patients (100%) with good IKDC scores
(ie, 81-90) at 6 months after surgery showed constant or even improved scores at 36 months follow-up. Ninety-one percent
of patients with good and excellent scores at 12 months and 83% of patients with good and excellent scores at 24 months (a
total of 23 and 37 patients, respectively) were able to maintain these scores at 36 months follow-up. Similar results were obtained
for the Lysholm score.
Conclusion: With regard to the improvements in functional outcomes after ACI-Cs at 36 months after surgery, the technique
described here appears to lead to satisfying and stable clinical results. This study helps the treating physician to predict the like-
liness of further clinical improvements or constant unsatisfactory results after ACI. In patients with good/excellent scores shortly
after surgery, deterioration of the knees condition is rarely found. For patients with poor and fair postoperative scores, clinical
outcomes are more difficult to predict, especially during the first year after the procedure.
Keywords: autologous chondrocyte implantation; articular cartilage; knee joint; IKDC; Lysholm; success; failure

Autologous chondrocyte implantation (ACI) is a well-

*Address correspondence to Jan M. Pestka, MD, Department of established and accepted surgical procedure for the treat-
Orthopedic Surgery and Traumatology, Freiburg University Hospital, ment of cartilage defects of the knee joint.3,6,13 Short-
Hugstetter Str. 55, D - 79095 Freiburg, Germany (e-mail: jan.pestka@ term and midterm results regarding this procedure have
been studied in depth and reveal satisfying outcomes.3,13
Department of Orthopedic Surgery and Traumatology, Freiburg Uni-
versity Hospital, Freiburg, Germany.
Over the years, the initial first-generation ACI technique
One or more of the authors has declared the following potential con- in which a periosteal flap was used to cover and seal the
flict of interest or source of funding: Funding was provided by Deutsche defect has been modified because of harvest site morbidity
Forschungsgesellschaft (DFG) and the AO Foundation Research Fund. and graft hypertrophy.7,20 This led to the development of
second- and third-generation techniques in which in sec-
The American Journal of Sports Medicine, Vol. 42, No. 1
DOI: 10.1177/0363546513507768
ond-generation ACI, a collagen membrane is used to cover
2013 The Author(s) the cartilage defect after chondrocyte application to the

Downloaded from by guest on April 8, 2016
Vol. 42, No. 1, 2014 Predicting Success or Failure of ACI at 36 Months 209

cartilage defect.7 In brief, in third-generation ACI, chondro- and scores were considered excellent if .90.9 Subgroups
cytes are embedded into the 3-dimensional collagen struc- for the Lysholm score consisted of poor scores of \65 points
ture of the membrane, which is subsequently fitted into to fair scores ranging from 65 to 83 points; good scores
the cartilage defect.8 In comparison with all 3 generations, were defined as 84 to 94 points, whereas excellent scores
cell-seeded collagen matrixsupported ACI (ACI-Cs) repre- were defined as 95 points.2 Follow-up evaluations were per-
sents a novel technical development. Use of a collagen-based formed at 6, 12, 24, and 36 months after surgery. At the time
membrane avoids graft hypertrophy and reduces operative of data collection, 80 patients had reached a follow-up of 36
time. In contrast to other third-generation ACI techniques, months. Initial clinical results of 59 patients at 24 months
the cells are seeded within the membrane, not beneath it. have been reported earlier.20 Complete preoperative and
Chondrocytes are transferred manually onto the membrane postoperative data were available for 80 of 83 patients
just before implantation by the surgeon. As the seeded (follow-up rate, 96.4%). Failure was defined as an improve-
membrane is initially attached mechanically while stable ment in the IKDC and/or Lysholm score of less than 10%
onto the cartilage defect floor, it can progressively fill or worsening knee function at 36 months follow-up in com-
the defect, while the construct can adapt to the local parison with preoperative scores.
topology.27,28 In all patients, indications for ACI were determined
Only recently have preliminary 2-year results of 59 con- during routine arthroscopic surgery of the affected knee
secutive patients treated with ACI-Cs for cartilage defects joint. Generally, ACI was performed in defects that were
of the knee joint been reported.20 These are noteworthy, as greater than 3 cm2 in size, while in smaller defects, arthro-
the technique resulted in a significant improvement in scopic microfracturing was preferred. Exclusion criteria for
patients clinical results after 24 months in comparison ACI-Cs in this study were corresponding (kissing) carti-
with preoperative results. Absolute International Knee lage lesions, uncontained defects, and defects of the sub-
Documentation Committee (IKDC) and Lysholm scores chondral bone plate exceeding a depth of 3 to 4 mm.
were in the range of earlier reported data for the use of During arthroscopic surgery, chondrocytes were harvested
ACI for full-thickness cartilage defects of the knee by dif- using a standardized cartilage biopsy tool (Storz, Tuttlin-
ferent technical modifications.1,10,18 Furthermore, while gen, Germany) from the intercondylar notch.20 In all
a number of studies reported on the dynamics of postoper- patients, ACI-Cs was performed as described earlier.27,28
ative results both from a histological as well as from a clin- In brief, between 1 and 2 million chondrocytes/cm2 per
ical point of view up to 3 years after surgery, it was defect (CartiGro Autologous Chondrocytes, MetreonBio-
uncertain if superior or inferior clinical results could be products GmbH, Freiburg, Germany) were applied 5 to
expected over the course of the third year after ACI- 10 minutes before transplantation to the rough side of
Cs.4,26 Therefore, the present study was initiated to evalu- a porcine type I/III collagen membrane (Chondro-Gide,
ate midterm outcomes including a total of 80 patients con- Geistlich, Wolhusen, Switzerland). Details of the surgical
secutively followed for a period of 36 months after ACI-Cs. technique have been reported.28
The aim of the present study was to present prospective After surgery, a standardized protocol11,20 was applied
clinical results of a large cohort of 80 patients treated with that included continuous passive motion from day 1 after
ACI-Cs and to further complete our previously presented surgery to 6 weeks postoperatively for up to 4 hours per
2-year results.20 Additionally, this study aimed to evaluate day. Limited weightbearing was recommended for 6 weeks
different time points in the clinical course after ACI when after ACI-Cs. Weightbearing was increased to full weight-
definite success or failure after ACI can be estimated. The bearing 9 weeks after surgery. Individual limits of flexion
patients individual course was evaluated with regard to the were additionally established depending on the exact
probability of final failure or success based upon IKDC and defect location to avoid early exposure of the regenerative
Lysholm scores at different time points. This seems of high cartilage to axial compression and shear forces.
clinical relevance, as it might support important decisions
during the clinical course after ACI such as indications for Statistical Analysis
revision surgery or patients return to physical activity.
Significant differences between individual time points of
the present study as well as cohorts of patients with
MATERIALS AND METHODS respect to the defect location (medial and lateral femoral
condyles, trochlea, and patella) were evaluated using
Patient Cohort 1-way analyses of variance (ANOVA) for all parametric
data (IKDC and Lysholm scores). The significance level
Between October 2005 and September 2008, 83 consecutive was P \ .05 for all analyses. P values for results after 36
patients with circumscribed, full-thickness cartilage defects months in comparison with different time points preopera-
of the knee joint, graded III and IV according to the Interna- tively and 6, 12, and 24 months postoperatively were calcu-
tional Cartilage Repair Society (ICRS) classification,24 were lated using the Wilcoxon signed-rank test to compare
treated with ACI-Cs. All patients were enrolled in this study paired samples. For all statistical analyses, the software
before surgery and evaluated using the Lysholm16 and IKDC SPSS version 16.0 (IBM, Armonk, New York) was used
scores.9 Previously defined subgroups for the IKDC score together with GraphPad Prism software (version 5, Graph-
were applied in which poor scores were \70, fair scores Pad Software, La Jolla, California) and Microsoft Excel XP
ranged from 71 to 80, good scores ranged from 81 to 90, (Microsoft, Redmond, Washington).

Downloaded from by guest on April 8, 2016

210 Pestka et al The American Journal of Sports Medicine

TABLE 1 and significant increase in both the IKDC and Lysholm

Characteristics of Patients (N = 80) and Defects (N = 106)a scores was observed over time up to the studys end point
at 36 months. To objectify decreases or increases in scores
Mean 6 Standard during the course of observation, the Wilcoxon test was used.
n (%) Deviation Range

Sex, M/F 46/34 (57.5/42.5) IKDC and Lysholm Scores at Different Time Points
Age at ACI, y 37.9 6 8.7 17.7-57.6
BMI, kg/m2 24.2 6 3.3 18.3-35.2
in Relation to Defect Location
No. of defects
The IKDC and Lysholm scores were evaluated with respect
1 lesion 57 (71.3)
2 lesions 22 (27.5)
to various defect locations at different time points. Signifi-
.2 lesions 1 (1.3) cant differences with regard to the locations analyzed
Defect size, cm2 4.6 6 1.7 1.0-8.8 were found neither for the IKDC nor for the Lysholm score.
\2 5 (4.7) When differences in scores preoperatively and 36 months
2-4 44 (41.5) postoperatively with respect to defect location (D IKDC
.4 57 (53.8) and D Lysholm) were assessed, significant differences
Defect location were found (Figure 2). One-way ANOVA of D IKDC var-
MFC 43 (40.6) iances revealed significant differences between all locations
LFC 10 (9.4) analyzed (P \ .001). Similar results were obtained for D
Patella 31 (29.2)
Lysholm in which all locations analyzed showed significant
Trochlear groove 22 (20.8)
Defect origin
differences with the exception of the trochlea versus patella,
Degenerative 41 (51.3) where there was no statistical significance (P = .13).
Posttraumatic 26 (32.5)
Traumatic 13 (16.3)
Failure Rate
ACI, autologous chondrocyte implantation; BMI, body mass The failure rate was defined as an improvement in the
index; F, female; LFC, lateral femoral condyle; M, male; MFC,
IKDC and/or Lysholm score of less than 10% or worsening
medial femoral condyle.
knee function at 36 months follow-up in comparison with
preoperative scores. Immediate postoperative complica-
RESULTS tions (eg, infection) were not observed and therefore had
no effect on the failure rate. A minority of 10 patients
Patient Characteristics (12.5%) failed the procedure. The mean age of these
patients was 40.4 6 8.5 years, 50% were male, and all
A total of 80 patients were included in the present study. but 1 were nonsmokers. The mean defect size did not differ
The mean age at the time of surgery was 37.9 6 8.7 years, significantly from the entire cohorts, with a value of 4.3 6
with a majority of male patients (46 male vs 34 female). A 1.6 cm2 (P = .09). Six patients had singular defects; defects
significant number of patients were smokers (n = 16, were located at the patella in 3 cases, at the medial femoral
20.0%). In 80 patients, a total of 106 cartilage defects condyle in 2 cases, and in 1 case at the trochlear groove.
were treated with ACI-Cs, including 57 patients with a sin-
gle lesion, 22 patients with 2 lesions, and 1 patient with
more than 2 lesions. The mean defect size was 4.6 6 Probabilities for Constant or Improved Scores
1.7 cm2. Among those patients with isolated cartilage After 36 Months Follow-up
defects (n = 57), 23 patients suffered from defects located
To analyze if a time point could be estimated during our
on the medial femoral condyle (40.4%); in 22 cases, patellar
course of observation at which a steady state after ACI
defects were treated (38.6%); 8 defects were located on the
without further significant decreases or increases in sub-
lateral femoral condyle (14.0%); and in 4 patients, defects
jective scores was reached, the course of the patients indi-
of the trochlear groove were present (7.0%). In addition,
vidual IKDC and Lysholm scores was analyzed and
multiple defect locations were treated in 23 patients
assessed in relation to the scores obtained 36 months after
(28.8%). Detailed information on all patient characteristics
ACI. Not surprisingly, preoperative IKDC and Lysholm
is presented in Table 1.
scores were poor/fair. No patient had good or excellent
scores. A total of 35 (44.8%) of those cases failed to improve
Functional Clinical Outcomes as Assessed to good or excellent IKDC scores, whereas 41.3% of
by IKDC and Lysholm Scores patients maintained poor or fair Lysholm scores at the
end point of the study. Nearly half (46.6%) of the cases
A total of 83 patients were evaluated preoperatively. The with poor IKDC scores at 6 months after surgery did not
follow-up rate was 96.4%, with 80 patients evaluated at 6, improve to a higher score category when analyzed 36
12, 24, and 36 months after surgery by the IKDC and months after surgery. In 61.1% of all cases, patients with
Lysholm scores and included in this study according to the poor IKDC scores at 12 months after surgery still had
ICRS criteria at 36 months. Details of all scores can be found poor scores at 36 months after surgery. An improvement
in Tables 2 and 3 as well as Figure 1. In summary, a constant in IKDC scores between the time points of 24 and 36

Downloaded from by guest on April 8, 2016

Vol. 42, No. 1, 2014 Predicting Success or Failure of ACI at 36 Months 211

Number of Patients in Each Category for IKDC and Lysholm Scores Preoperatively
and at 6, 12, 24, and 36 Months Postoperativelya

Preoperatively 6 mo 12 mo 24 mo 36 mo

IKDC score
Poor (\70) 79 (95.2) 58 (72.5) 36 (45.0) 26 (32.5) 28 (35.0)
Fair (71-80) 4 (4.8) 19 (23.8) 21 (26.3) 17 (21.3) 11 (13.8)
Good (81-90) 0 (0) 3 (3.8) 19 (23.8) 23 (28.8) 23 (28.8)
Excellent (.90) 0 (0) 0 (0) 4 (5.9) 14 (17.5) 18 (22.5)
Lysholm score
Poor (\65) 52 (62.7) 19 (23.8) 20 (25.0) 11 (13.8) 8 (10.0)
Fair (65-83) 31 (37.3) 50 (62.5) 25 (31.3) 28 (35.0) 25 (31.3)
Good (84-94) 0 (0) 11 (13.8) 30 (37.5) 27 (33.8) 24 (30.0)
Excellent (.95) 0 (0) 0 (0) 5 (6.3) 14 (17.5) 23 (28.8)

Data are given as n (%). IKDC, International Knee Documentation Committee.

IKDC and Lysholm Scores Preoperatively and at 6, 12, 24, and 36 Months Postoperativelya

Preoperatively 6 mo 12 mo 24 mo 36 mo

No. of patients 83 80 80 80 80
Follow-up rate, % 100.0 96.4 96.4 96.4 96.4
IKDC scoreb 49.6 6 13.8 64.1 6 13.3 70.0 6 16.3 75.5 6 14.6 79.1 6 14.6
(16.1-75.3) (30.5-84.3) (19.3-97.7) (40.2-95.4) (50.4-97.3)
P value vs 36 mo \.001 \.001 \.001 \.001
Lysholm scoreb 59.5 6 11.6 70.6 6 12.9 76.0 6 16.6 81.3 6 14.2 83.5 6 13.7
(14.0-83.0) (31.0-94.0) (25.0-100.0) (36.0-100.0) (40.0-100.0)
P value vs 36 mo \.001 \.001 \.001 .020

Significant differences between various time points were calculated using the Wilcoxon signed-rank test to compare paired samples. Sig-
nificance level was P \ .05 for all analyses. IKDC, International Knee Documentation Committee.
Data are expressed as mean 6 standard deviation (range).

months was only observed in a minority of 19.2% of ACI for cartilage defects of the knee. Patients were evalu-
patients. Detailed data on decreases or increases in ated by means of the Lysholm and IKDC scores, which
IKDC scores are presented in Table 4. have been validated for patients who have undergone car-
If patients already had good (81-90) and excellent (.90) tilage repair and are recommended as appropriate outcome
IKDC scores as early as 6 months after surgery, no parameters in these patients by the ICRS.5
decrease in scores until the end point of this study at 36 Previous to ACI-Cs, matrix-associated ACI (MACI) had
months was found. A majority of patients (90.6%) with been performed. In this procedure, chondrocytes are
good and excellent scores at 12 months showed constant seeded on a collagen membrane several days before sur-
scores, whereas between 24 and 36 months, 82.6% of gery. During the operation, the membrane is cut to size
patients had similar or improved scores at the time of and implanted into the defect area.1 However, at first
follow-up. Results obtained for the Lysholm scores over glance its easy application can be associated with a number
the time course of this study showed similar trends as of disadvantages. Cell viability and the differentiation
observed for the IKDC scores (Table 4). state of the chondrocytes cannot be monitored after appli-
In summary, if scores were poor or fair at 6 months post- cation of the cells to the membrane. Additionally, cutting
operatively, only half the number of cases managed to further an appropriate size and manipulation of the cell-seeded
improve their scores toward the end point of this study. If membrane may result in a loss of chondrocytes or
poor or fair scores were still present at 12 and 24 months after decreased cell viability. Therefore, ACT-Cs was introduced
surgery, the majority of cases failed to further improve their to combine the technical attractiveness of MACI with opti-
scores. Detailed results can be found in Table 4. mal cell quality, easy handling, and easy application.
In the current study, ACI-Cs resulted in satisfying clin-
ical outcomes at 36 months follow-up, with a significant
DISCUSSION increase in both IKDC and Lysholm scores from before sur-
gery. The failure rate was 12.5%, which is slightly higher
This study represents a prospective case series of 80 con- than in our previous analysis, where 5.9% of all patients
secutive patients who were followed up to 36 months after failed the procedure.20 Still, these results are in line with

Downloaded from by guest on April 8, 2016

212 Pestka et al The American Journal of Sports Medicine

* 50
* * *
* 40

IKDC score (%)

20 MFC LFC Patella Trochlea

0 B
0 6 12 24 36
50 * * *
Follow up time (in months)
40 n. s.
B *
* - Lysholm
100 20
Lysholm score (points)

80 10

60 0
MFC LFC Patella Trochlea
Figure 2. D International Knee Documentation Committee
(IKDC) and D Lysholm: mean differences between scores
20 before surgery and at 36 months after surgery with respect
to various defect locations. (A) The greatest improvement
0 from preoperative to postoperative IKDC scores was
0 6 12 24 36 observed for cases with defects located at the lateral femoral
condyle (LFC). These were significantly different (P \ .05)
Follow up time (in months)
when compared with other defect locations such as the
trochlea, patella, or medial femoral condyle (MFC). Compar-
Figure 1. (A) A significant improvement in the International ison between the MFC and patella or trochlea showed signif-
Knee Documentation Committee (IKDC) score was observed icant differences (P \ .05). Significant differences between
preoperatively as well as after 6, 12, and 24 months when the defect locations at the trochlea and patella were addition-
compared with 36 months follow-up. (B) Lysholm scores ally found (P \ .05). (B) Analysis of D Lysholm similarly
constantly increased from preoperatively to 6, 12, and 24 revealed significant differences (all P \ .05) between all
months postoperatively and were significantly different from locations analyzed with the exception of the trochlea
scores at 36 months follow-up. *Significant differences versus patella, where there was no statistical significance
(P  .01) between individual time points. (P = .13).

those of other studies in which failure rates were reported additional increases in knee function might be supported
to be as low as 7.7%, and 80% to 90% of patients reported by the certainty that ACI-Cs is a biological procedure in
satisfying results postoperatively.8,17 which differentiation and maturation of regenerating car-
Concerning the overall results of the present study, sig- tilage last for a significant period of time. A possible expla-
nificant improvements in knee function as measured by nation for these clinical findings can be found in an
the IKDC and Lysholm scores were found between 12 experimental cartilage study that was able to show that
and 24 months after surgery as well as between 24 and after chondrocyte implantation, regenerated tissue under-
36 months after ACI. These findings are noteworthy, as goes a process of maturation that in the majority of cases

Downloaded from by guest on April 8, 2016

Vol. 42, No. 1, 2014 Predicting Success or Failure of ACI at 36 Months 213

TABLE 4 scores and only 1 patient showed a good score at 36 months

Probabilities of Constant IKDC and Lysholm Scores after surgery. Numerous individual factors such as age at
at 36 Months in Relation to Clinical Results the time of implantation, sex, and defect origin, together
at 6, 12, and 24 Months Postoperativelya with general risk factors such as smoking and defect size,
have been identified that might possess an influence on
Probability at 36 mo outcomes after cartilage repair.14,21,24 Additionally, it has
IKDC Score Lysholm Score recently been reported that overweight and obesity are
negatively associated with satisfying results after ACI.12
Poor/fair scores However, no single factor can be associated with poor
Preoperatively 0.44 0.41 clinical outcomes after cartilage regenerative surgery. It
6 mo 0.51 0.46 therefore appears most likely that a combination of multi-
12 mo 0.61 0.60 ple factors influences regeneration after ACI in general
24 mo 0.81 0.77
and clinical outcomes in particular. However, as clinical
Good/excellent scores
results of both the IKDC and Lysholm scores significantly
Preoperatively 0.00 0.00
6 mo 1.00 0.91 improved from 6 to 36 months after surgery, the ACI
12 mo 0.91 0.83 procedure appears to be a success for the majority of
24 mo 0.89 0.93 patients.
This is especially important as most patients wish to
IKDC, International Knee Documentation Committee. receive information early after the operation regarding
general expectations toward success or failure. With the
help of this study, there may now be a possibility to discuss
takes longer than 18 months to complete.4,22 During this such probabilities with the patient. It is therefore our gen-
time, fibrous cartilage needs to be replaced by hyaline- eral recommendation to carefully assess and collect indi-
like tissue. Differentiation of cartilage replacement tissue vidual scores at set intervals of 6, 12, 24, and 36 months.
usually takes up to 24 months.22 After 5 years, complete In this context, it has been shown that clinical results after
filling and integration of the graft as measured by mag- this time appear to reach a plateau, with no significant fur-
netic resonance imaging (MRI) could be found in the major- ther changes either in knee function or patient satisfac-
ity of patients.17 Here, good clinical function appeared to be tion.29 In addition, satisfactory results at 2 to 3 years
associated with sufficient knee function. With regard to the after ACI appear to be associated with good long-term
assessment of clinical and subjective knee function after results.24,25 Therefore, from our point of view, the midterm
complete cartilage differentiation, further improvement results up to 3 years after surgery presented in this study
in knee function has been observed for a short interval can be used to appraise a first careful prognosis after ACI.
between 24 and 36 months postoperatively.26,29 In addi- Of course, more extensive follow-up data such as 5-year
tion, transplantation of articular chondrocytes has been results will be necessary to provide further insight into
associated with good clinical short-term and midterm the clinical course after ACI-Cs and to allow solid
results. predictions.
In contrast to patients with initial good and excellent On the basis of our data, patients with poor or fair
scores after ACI are those who reported poor and fair scores at 6 months after ACI might be informed that at
scores as early as 6 months after surgery. Over the course this early stage after surgery, chances of further improve-
of the following 30 months, nearly 50% failed to improve ment in knee function and overall satisfaction can be esti-
their knee function. This might distinguish the clinical mated at approximately 50%. For patients with good and
course of patients after ACI with the clinical course after excellent scores at 6 months after surgery, on the other
alternative techniques such as arthroscopic microfractur- hand, a significant deterioration in clinical results is
ing, which is generally recommended for smaller cartilage unlikely to occur over the course of the following 30
defects measuring less than 2 to 4 cm2.19 Data obtained for months. As the number of patients in this study with
microfracturing show that clinical results seem to deterio- good and excellent scores at 6 months after surgery was
rate between 24 and 36 months after surgery.19 In the cur- low (n = 3) and as a 50% chance of improvement after ini-
rent study, the mean defect size measured 4.6 cm2, which tial poor ACI results resembles that of tossing a coin, it is
was well above the size recommended for microfracturing. doubtful that such a discussion with the patient is profit-
Inferior results have been reported for microfracturing in able. It is therefore our recommendation to discuss the
defects of such a size.15 However, a total of 5 patients chances of success or failure at 1 year after surgery at
(4.7%) had cartilage lesions smaller than 2 cm2. It is an the earliest. With respect to our data, giving a prognosis
ongoing debate as to whether ACI is the best treatment at 1 year after surgery appears more reliable and less vari-
for these patients or if, in cases of failure, they are likely able. Most importantly, decision making regarding revi-
to fail other cartilage regenerative treatments as well. sion surgery should be postponed to 1 year after surgery
Interestingly, only 1 of those 5 patients with a single carti- at the earliest.
lage lesion of the medial femoral condyle reported a good Taken together, our overall recommendation is to dis-
Lysholm score and excellent IKDC score at 36 months after cuss the individual chances of success or failure at 1 year
surgery. The remaining 4 patients all had cartilage lesions after surgery at the earliest. As early as 6 months after
in 2 locations, of which 3 patients showed poor or fair surgery, patients with poor and fair scores show

Downloaded from by guest on April 8, 2016

214 Pestka et al The American Journal of Sports Medicine

improvements, whereas the numbers of patients with good 5. Collins NJ, Misra D, Felson DT, Crossley KM, Roos EM. Measures of
and excellent scores are too low to give a reliable prognosis. knee function: International Knee Documentation Committee (IKDC)
Subjective Knee Evaluation Form, Knee Injury and Osteoarthritis Out-
As early as 12 months or as late as 24 months after sur-
come Score (KOOS), Knee Injury and Osteoarthritis Outcome Score
gery, this discussion can be much more goal oriented. Physical Function Short Form (KOOS-PS), Knee Outcome Survey
Concerning the limitations of this study, the lack of Activities of Daily Living Scale (KOS-ADL), Lysholm Knee Scoring
a control group treated with a different technique and Scale, Oxford Knee Score (OKS), Western Ontario and McMaster
the relatively short follow-up of only 36 months must be Universities Osteoarthritis Index (WOMAC), Activity Rating Scale
considered. Therefore, only limited conclusions can be (ARS), and Tegner Activity Score (TAS). Arthritis Care Res (Hoboken).
drawn concerning the long-term outcomes of patients trea- 2011;63 Suppl 11:S208-S228.
6. Erggelet C, Sittinger M, Lahm A. The arthroscopic implantation of
ted with ACI-Cs. This, however, we consider to be a general autologous chondrocytes for the treatment of full-thickness cartilage
limitation of most ACI studies in which few long-term defects of the knee joint. Arthroscopy. 2003;19:108-110.
follow-ups are available. Unlike studies on implantation 7. Gooding CR, Bartlett W, Bentley G, Skinner JA, Carrington R, Flana-
in hip arthroplasty, for example, where follow-up data gan A. A prospective, randomised study comparing two techniques
average 7 years, ACI in general and ACI-Cs in particular of autologous chondrocyte implantation for osteochondral defects
are relatively new techniques with consequently rather in the knee: periosteum covered versus type I/III collagen covered.
Knee. 2006;13:203-210.
limited follow-up times.23
8. Harris JD, Siston RA, Brophy RH, Lattermann C, Carey JL, Flanigan
An important general limitation of studies on cartilage DC. Failures, re-operations, and complications after autologous
treatment, which can also be found in this study, is the chondrocyte implantation: a systematic review. Osteoarthritis Carti-
lack of objective assessments, such as MRI and histology. lage. 2011;19:779-791.
Therefore, only subjective clinical results are presented. 9. Hefti F, Muller W, Jakob RP, Staubli HU. Evaluation of knee ligament
It is certainly important to encourage the use of objective injuries with the IKDC form. Knee Surg Sports Traumatol Arthrosc.
outcomes, but as shown in this study, the use of the
10. Henderson IJ, Tuy B, Connell D, Oakes B, Hettwer WH. Prospective
IKDC and Lysholm scoring systems provides sufficient clinical study of autologous chondrocyte implantation and correlation
data to predict the clinical course after ACI. with MRI at three and 12 months. J Bone Joint Surg Br. 2003;
11. Hirschmuller A, Baur H, Braun S, Kreuz PC, Sudkamp NP, Niemeyer
P. Rehabilitation after autologous chondrocyte implantation for iso-
CONCLUSION lated cartilage defects of the knee. Am J Sports Med. 2011;
In conclusion, the 3-year prospective clinical follow-up data 12. Jaiswal PK, Bentley G, Carrington RW, Skinner JA, Briggs TW. The
of 80 consecutive patients after ACI-Cs show promising adverse effect of elevated body mass index on outcome after autol-
results and a minimal failure rate. Timely sequential ogous chondrocyte implantation. J Bone Joint Surg Br. 2012;
data enabled us to compare clinical results at different
13. Jones DG, Peterson L. Autologous chondrocyte implantation. J Bone
postoperative time points. Poor clinical findings (IKDC Joint Surg Am. 2006;88:2502-2520.
\70) at 6 months postoperatively are associated with an 14. Jungmann PM, Salzmann GM, Schmal H, Pestka JM, Sudkamp NP,
almost 50% chance of constant poor results. Patients Niemeyer P. Autologous chondrocyte implantation for treatment of
with good or excellent results (IKDC .81-100) at 6 months cartilage defects of the knee: what predicts the need for reinterven-
after surgery generally maintain them, with a very high tion? Am J Sports Med. 2012;40:58-67.
probability (.80%) at 36 months after intervention. This 15. Knutsen G, Drogset JO, Engebretsen L, et al. A randomized trial
comparing autologous chondrocyte implantation with microfracture:
information can be beneficial for the treating physician
findings at five years. J Bone Joint Surg Am. 2007;89:2105-2112.
when taking care of postoperative ACI patients. As satisfy- 16. Lysholm J, Gillquist J. Evaluation of knee ligament surgery results
ing results after ACI can be predicted with a higher accu- with special emphasis on use of a scoring scale. Am J Sports Med.
racy, they might be discussed earlier with the patient than 1982;10:150-154.
cases of suspected failure. 17. Marlovits S, Aldrian S, Wondrasch B, et al. Clinical and radiological
outcomes 5 years after matrix-induced autologous chondrocyte
implantation in patients with symptomatic, traumatic chondral
defects. Am J Sports Med. 2012;40:2273-2280.
REFERENCES 18. McNickle AG, LHeureux DR, Yanke AB, Cole BJ. Outcomes of autol-
ogous chondrocyte implantation in a diverse patient population. Am
1. Behrens P, Bitter T, Kurz B, Russlies M. Matrix-associated autolo- J Sports Med. 2009;37:1344-1350.
gous chondrocyte transplantation/implantation (MACT/MACI): 5- 19. Mithoefer K, Williams RJ 3rd, Warren RF, et al. The microfracture
year follow-up. Knee. 2006;13:194-202. technique for the treatment of articular cartilage lesions in the
2. Briggs KK, Steadman JR, Hay CJ, Hines SL. Lysholm score and knee: a prospective cohort study. J Bone Joint Surg Am. 2005;
Tegner activity level in individuals with normal knees. Am J Sports 87:1911-1920.
Med. 2009;37:898-901. 20. Niemeyer P, Lenz P, Kreuz PC, et al. Chondrocyte-seeded type I/III
3. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L. collagen membrane for autologous chondrocyte transplantation: pro-
Treatment of deep cartilage defects in the knee with autologous spective 2-year results in patients with cartilage defects of the knee
chondrocyte transplantation. N Engl J Med. 1994;331:889-895. joint. Arthroscopy. 2010;26:1074-1082.
4. Brun P, Dickinson SC, Zavan B, Cortivo R, Hollander AP, Abatangelo 21. Niemeyer P, Salzmann G, Steinwachs M, et al. Presence of subchon-
G. Characteristics of repair tissue in second-look and third-look biop- dral bone marrow edema at the time of treatment represents a nega-
sies from patients treated with engineered cartilage: relationship to tive prognostic factor for early outcome after autologous
symptomatology and time after implantation. Arthritis Res Ther. chondrocyte implantation. Arch Orthop Trauma Surg. 2010;
2008;10:R132. 130:977-983.

Downloaded from by guest on April 8, 2016

Vol. 42, No. 1, 2014 Predicting Success or Failure of ACI at 36 Months 215

22. Ossendorf C, Kaps C, Kreuz PC, Burmester GR, Sittinger M, Erggelet 26. Saris DB, Vanlauwe J, Victor J, et al. Characterized chondrocyte
C. Treatment of posttraumatic and focal osteoarthritic cartilage implantation results in better structural repair when treating symp-
defects of the knee with autologous polymer-based three- tomatic cartilage defects of the knee in a randomized controlled trial
dimensional chondrocyte grafts: 2-year clinical results. Arthritis Res versus microfracture. Am J Sports Med. 2008;36:235-246.
Ther. 2007;9:R41. 27. Steinwachs M. New technique for cell-seeded collagen-matrix-supported
23. Perka C, Ludwig R, Stern S. [Total endoprosthetic management of autologous chondrocyte transplantation. Arthroscopy. 2009;25:208-211.
the hip joint after failed osteosynthesis of para-articular hip fracture]. 28. Steinwachs M, Peterson L, Bobic V, Verdonk P, Niemeyer P. Cell-
Z Orthop Ihre Grenzgeb. 2000;138:39-45. seeded collagen matrixsupported autologous chondrocyte trans-
24. Peterson L, Minas T, Brittberg M, Nilsson A, Sjogren-Jansson E, Lin- plantation (ACT-CS): a consensus statement on surgical technique.
dahl A. Two- to 9-year outcome after autologous chondrocyte trans- Cartilage. 2012;3:5-12.
plantation of the knee. Clin Orthop Relat Res. 2000;374:212-234. 29. Vanlauwe J, Saris DB, Victor J, Almqvist KF, Bellemans J, Luyten FP.
25. Peterson L, Vasiliadis HS, Brittberg M, Lindahl A. Autologous chon- Five-year outcome of characterized chondrocyte implantation versus
drocyte implantation: a long-term follow-up. Am J Sports Med. microfracture for symptomatic cartilage defects of the knee: early
2010;38:1117-1124. treatment matters. Am J Sports Med. 2011;39:2566-2574.

For reprints and permission queries, please visit SAGEs Web site at

Downloaded from by guest on April 8, 2016