Joint homeostasis
THE DISCREPANCY BETWEEN OLD AND FRESH DEFECTS IN
CARTILAGE REPAIR
D. B. F. Saris, W. J. A. Dhert, A. J. Verbout
From the University Medical Center, Utrecht, The Netherlands
he discrepancy between successful experimental
studies of cartilage repair and the clinical results is
unexplained. We have evaluated the effect of metabolic
alterations in joint homeostasis owing to an articular
defect on the outcome of cartilage repair using tissue
engineering methods. We used 21 adolescent Dutch
goats divided into three groups. The control knees were
left untreated while the contralateral knee was
randomised to receive either no treatment (N), early
treatment (E) or late treatment (L). The metabolism of
proteoglycans in the surrounding joint surface was
determined and correlated with the O’Driscoll score
used to quantify the histological aspect of the repair of
the defect.
Synthesis of proteoglycan (PG) was increased in all
groups. The release of glucosaminoglycan (GAG) was
significantly higher in the untreated but not after early
transplantation (1.3 v 1.8 NS). The cartilage repair
scores in the early treatment group were not as good as
those of the normal control group, but were significantly
better when compared with both the untreated defects
and the late treated defects. Defects which had been
treated late showed a significantly decreased score when
compared with those which had had early treatment or
the normal control group and did not differ (p = 0.12)
from those with no treatment. The histological and
biochemical scores closely resembled the macroscopic
and functional parameters which showed a significant
deterioration for the late treated group and those
without treatment compared with animals treated early.
Thus, tissue-engineered cartilage repair is negatively
influenced by altered matrix metabolism. Early
treatment showed significantly better results for repair
T
D. B. F. Saris, MD, PhD, Orthopaedic Surgeon
W. J. A. Dhert, MD, PhD, Director of Orthopaedic Research
A. J. Verbout, MD, PhD, Professor and Orthopaedic Surgeon
Department of Orthopaedics, University Medical Center, POB 85500, 3508
GA, Utrecht, The Netherlands.
Correspondence should be sent to Dr D. B. F. Saris.
©2003 British Editorial Society of Bone and Joint Surgery
doi:10.1302/0301-620X.85B7.13745 $2.00
VOL. 85-B, No. 7, SEPTEMBER 2003
of cartilage than late or no treatment, with a concurrent
decrease in the detrimental disturbance of cartilage
metabolism which constituted a protective effect on the
articulation.
J Bone Joint Surg [Br] 2003;85-B:1067-76.
Received 7 August 2002; Accepted after revision 14 February 2003
The incidence of cartilage lesions, the lack of adequate healing and subsequent indications for surgical intervention
have been described by various authors.1-4 Durable restoration of damaged articular cartilage is a valuable but as yet
unachieved goal.5-14 The need for cartilage repair and the
use of tissue-engineering strategies for the restoration of a
defect of the articular cartilage have been well established.
With regard to the latter, various strategies for the repair of
tissue-engineered cartilage have been conceived and established by extensive basic scientific investigations in vitro,
experimental studies in vivo and an increasing number of
short- to mid-term clinical outcome studies.1,5,12,13,15-18
Review of the literature suggests, however, that there is an
obvious discrepancy between the favourable outcome demonstrated in preclinical research and that in clinical practice
in which the results have not yet been reproduced. Only a
few examples of reliable long-term clinical results are available and the initial results which sparked enthusiasm in the
field are significantly less favourable at long-term followup.15,19,20
Most of the good, basic scientific results originate from
an optimally-controlled laboratory environment or animal
experimental work with normal joints in which a fresh cartilage defect has been treated by tissue engineering. Studies
in vivo with only a periosteal/perichondrial flap or with cultured chondrocytes under a periosteal or collagen cover
demonstrate restoration of the articular surface with histological and biochemical analyses indicating the presence of
regeneration tissue in the defects resembling hyaline cartilage. In the clinical setting more disappointing results are
seen. There is a tendency for incomplete bonding between
the tissue-engineered construct and the wall of the original
defect. Surface restoration is only partial. The periosteal flap
may undergo hypertrophy and calcification of the outer surface.18 Eventually, the new matrix shows loss of normal
metabolic activity. It is therefore possible that factors in the
1067
1068
D. B. F. SARIS, W. J. A. DHERT, A. J. VERBOUT
a
b
d
c
e
Fig. 1
Fig. 1a – Photograph of an operative view of the standardised defect in the medial femoral condyle. Fig. 1b –
Photograph of the defect treated with a periosteal graft obtained from the proximal tibia, sutured into the bottom
of the defect, with the cambium layer facing into the joint. Fig. 1c – The macroscopic aspect showing the natural
‘healing’ process after ten weeks of unlimited weight-bearing and movement of an untreated defect. Fig. 1d –
Photomicrograph of best result seen after periosteal transplantation in the ‘early treatment’ group. Fig. 1e – Diagram of the histological section in Figure 1d showing the margin of the cartilage defect and near perfect restoration of the surface of the subchondral bone and the tidemark of the cartilage.
environment surrounding the defect cause the discrepancy
between the results.
Concept of joint homeostasis. An articulating joint has a
complex design with many essential components and a multitude of interactions between structures such as synovium,
cartilage, menisci, synovial fluid, ligaments and subchondral bone.5,7,21,22 These anatomical structures are influenced by factors such as movement, loading, alignment,
weight, age, hormonal influences, etc. It is evident that this
complex environment must be rigorously regulated. Furthermore, metabolic control must be flexible because the
external environment of the cells is not constant.23 Studies
of a wide range of organisms have shown that there are a
number of mechanisms which control the physiological
equilibrium, also referred to as homeostasis. Vogel24 suggested that ‘the crux of a feedback system is the ability to
adjust what it does, depending on conditions outside itself,
where those conditions include the result of its own actions.
In a strictly mechanical and formal sense, it has self-awareness’. By normal joint homeostasis we mean the stable equilibrium of the synovium and cartilage matrix without
inflammation in a well-functioning articulation. When joint
homeostasis is disturbed, this equilibrium is changed and a
myriad of intra-articular factors, such as inflammatory,
molecular or cellular components, come into play.
Most patients have a long history of complaints before
they seek treatment for their joint damage. For instance,
effusion and pain indicate that some degree of synovitis and
matrix degradation is present. The disturbed local environment provides a condition somewhat different from that in
the undamaged, healthy joint.
We have tested the hypothesis that this altered environment constitutes a change in joint homeostasis which provides an explantation for the discrepancy between the
promising in vivo and the variable clinical results. In the
light of this hypothesis, the initial in vitro findings of Rodrigo et al25 are of interest. They compared the effect of samples of synovial fluid from the knees of 25 patients with a
fresh or old traumatic chondral defect in a chick limb-bud
assay. Samples from 11 of 17 acutely injured knees stimulated chondrogenesis, four were inhibitory, and two showed
no effect. Samples from six of eight chronically injured
knees inhibited chondrogenesis; the other two stimulated
chondrogenesis. They suggested that synovial fluid may
contain factors which stimulate the healing of cartilage in
the acute period after traumatic injury, but that this effect
THE JOURNAL OF BONE AND JOINT SURGERY
JOINT HOMEOSTASIS
1069
Table I. Details of the macroscopic findings according to the criteria of O’Driscoll et al.30-32
Five parameters were scored from 0 to 2 points for each knee of all animals in the three groups.
The figures are the number of animals in a group which were given corresponding scores
Range of movement (degrees)
Full
<20 decrease
>20 decrease
Intra-articular fibrosis
None
Minor
Major
Restoration of contour
Complete
Partial
None
Cartilage erosion
None
Graft
Graft + cartilage
Appearance
Translucent
Opaque
Discoloured/irregular
No treatment*
Early treatment
Late treatment
C
C
E
C
(n = 7) (n = 7)
(n = 6)
(n = 6)
(n = 6) (n = 6)
2
1
0
7
0
0
4
3
0
6
0
0
6
0
0
6
0
0
4
2
0
2
1
0
7
0
0
2
3
2
5
1
0
1
5
0
6
0
0
1
4
1
2
1
0
7
0
0
0
3
4
6
0
0
2
3
1
6
0
0
1
3
2
2
1
0
6
0
1
0
0
7
5
1
0
1
2
3
4
2
0
0
3
3
2
1
0
6
1
0
0
3
4
6
0
0
1
4
1
5
0
1
0
3
3
E
E
*C, control; E, experimental
may become inhibitory in chronic lesions. The full extent of
altered joint homeostasis cannot be examined effectively in
a controlled environment since these conditions only mimic
part of the natural organism and not all of the intricate regulatory functions are reproduced.
Following the approach suggested by our concept of
joint homeostasis not only the damaged cartilage surface
needs to be treated, but the entire joint has to be targeted in
tissue-engineered regeneration. This implies addressing
synovitis, meniscal damage, ligament stability, limb alignment and normalising metabolic activity. The establishment
of a true biological joint reconstruction may be of great
importance for a large patient population for whom currently there is no reliable treatment.
We hypothesise that late treatment of an old defect, which
has initiated degeneration of cartilage and caused changes in
joint homeostasis, will have a negative effect on the process
of the repair of cartilage. To test this, we used a goat model
and compared the outcome of repair from periosteal chondrogenesis. We examined three groups as follows: an
untreated defect, a defect treated by early periosteal transplantation and a defect treated late when homeostasis was
disturbed. The histological outcome of repair was correlated
with the biochemical parameters of cartilage metabolism to
identify a possible effect of metabolic alterations or altered
joint homeostasis on the process of repair of the cartilage.
Materials and Methods
Experimental design. We used 21 six-month-old, female
Dutch milk goats from a commercial vendor and kept them
in group housing for a minimum of two weeks before surVOL. 85-B, No. 7, SEPTEMBER 2003
gery. They were aged between 5.9 and 6.5 months and
weighed 20.7 to 23.9 kg. All experiments were approved
and monitored by the institutional animal experimental
Ethics Committee. In all animals a defect was made in the
medial femoral condyle26 (Fig. 1a), and the contralateral
knee was left untreated to serve as a control. Subsequently,
the animals were randomised into one of three groups as
follows:
1) No treatment. The defect was left untouched and natural
healing was studied. Unrestricted weight-bearing and movement were allowed.
2) Early treatment. Cartilage repair in a knee with normal
homeostasis was studied in which the fresh defect was
immediately transplanted (Fig. 1b).
3) Late treatment. The defect was left untouched for a
period of ten weeks with free weight-bearing allowed. After
this period the cartilage scar was resected (Fig. 1c) and
transplantation was performed identically to that in the previous group. All animals were killed ten weeks after transplantation.
Operative technique. All animals received premedication
with detomidinehydrocholoride (10 mg/kg). Subsequent
induction of anaesthesia was achieved with thiopental (50
mg/kg) and anaesthesia was maintained throughout the surgical procedure on a Magill system with halothane 1% to
1.5% and a mixture of O2 and NO2 in a ratio of 1:2.
Buprenorfinehydrochloride (0.15 mg daily) for relief of pain
and antibiotic prophylaxis with amoxicillin (15 mg/kg)
were given for a period of five days. A medial parapatellar
exposure was used. After retracting the patella laterally, the
medial femoral condyle was exposed. A standardised fullthickness cartilage defect of 0.8 × 0.5 cm was made in the
1070
D. B. F. SARIS, W. J. A. DHERT, A. J. VERBOUT
Posterior
b b
Tibia
b b
b b
H
H
b b
b b
b b
b b
b b
b b
b b
b b
b b
b b
b b
b b
b b
H
H
b b
b b
b b
b b
Fig. 2
Diagrams of the harvest locations of cartilage samples for biochemical and histological analysis. Centrally (H), is the medial or
lateral femoral region with the opposing tibial cartilage which was
harvested for histological analysis and scoring to analyse the effect
of cartilage damage and the outcome of repair. The surrounding
tissue was harvested in a standardised sequence and samples distributed evenly into groups for biochemical (b) and histological
analysis to determine the onset of early degeneration of the cartilage.
Anterior
L
M
M
L
Anterior
b b
b b
Femur
b b
H
b b
b b
H
b b
b b
b b
b b
b b
b b
b b
b b
b b
b b
b b
H
H
b b
b b
b b
b b
Posterior
Table II. Details of the analysis data for the medial femoral condyle for the three study groups. Standardised sampling locations were
used from both femur and tibia, lateral and medial (see Fig. 2). Paired data per animal and location within the joint were statistically
compared between groups. The results from the various locations demonstrated considerable statistical similarity
PG synthesis
GAG release
Total GAG content
Group
C
E
p value
C
E
p value C
E
p value
No treatment
Early treatment
Late treatment
19.3 ± 1.9
10.8 ± 2.5
14.2 ± 0.8
24.1 ± 2.9
13.3 ± 2.5
18.8 ± 1.0
<0.018
<0.028
<0.015
2.0 ± 0.1
1.3 ± 0.4
2.3 ± 0.2
3.0 ± 0.2
1.8 ± 0.5
3.8 ± 0.2
<0.018
NS
<0.028
33.1 ± 2.3
30.9 ± 2.2
27.8 ± 1.8
<0.032
NS
<0.046
medial femoral condyle using a sharp tissue elevator (Fig.
1a). The periosteal tissue was taken from the proximal,
medial tibia using a standardised technique.27-29 From the
transplant thus acquired, a sample (2 × 10 mm) was sent for
histological analysis to confirm the adequacy of the periosteal cambium layer in the transplant. The tissue transplant
was sutured into the bottom of the defect with the cambium
layer facing outwards into the joint using non-traumatic 3.0
resorbable vicryl sutures through predrilled bone tunnels
(Fig. 1b).
Imaging studies. AP and lateral radiographs of the knee
(15 msec/25 kV) were obtained before and immediately
after operation and at intervals of five weeks during the
study as well as at completion of follow-up to visualise the
growth plate and generally detect postoperative complications as well as developing subchondral sclerosis.
Macroscopic outcome evaluation. In order to quantify intraarticular adhesion, stiffening and the general appearance of
35.9 ± 1.1
32.2 ± 2.3
32.2 ± 2.3
the joint, all knees were scored using the macroscopic outcome parameters as described by O’Driscoll et al30-32
(Table I). The in vivo parameters applied after completion
of follow-up contain five categories within which individual
values of 0 to 2 points were awarded.
Biochemical assays. For biochemical analysis, cartilage
explants were obtained from the weight-bearing articular
surface surrounding the defect (Fig. 2) in a standardised
fashion.33 These explants were cultured individually in
DMEM (D-MEM, Gibco 074-01600; 0.81 mM SO42-; 24
mM NaHCO3) supplemented with ascorbic acid (0.85
mmol/l), glutamine (2 mmol/l), penicillin (100 IU/ml),
streptomycin sulphate (100 IU/ml), and 10% heat-inactivated, pooled female goat serum, on standard 96-well plates
(200 µl culture medium/well, 37˚C, 5% CO2 in air) according to previously published procedures.33 The DNA content
was determined as a measure of cartilage cellularity. Samples were digested in papain (2 hours; 65˚C). DNA was
THE JOURNAL OF BONE AND JOINT SURGERY
JOINT HOMEOSTASIS
Table III.
1071
Details of the histological findings in each group
Nature of predominant tissue
Cellular morphology
Hyaline articular cartilage
Incompletely differentiated
Fibrous tissue or bone
Safranin O staining
Normal or near normal
Moderate
Slight
None
Structural characteristics
Surface regularity
Smooth and intact
Superficial, horizontal lamination
Fissures, 25% to 100% of the thickness
Severe disruption or fibrillation
Structural integrity
Normal
Slight disruption, including cysts
Severe disintegration
Thickness (related to normal cartilage, %)
100
50 to 100
0 to 50
Bonding to the adjacent tissue
Bonded at both sides and subchondral bone
Bonded partially
Not bonded
Hypocellularity
None
Slight
Moderate
Severe
Chondrocyte clustering
None
<25% of cells
25 to 100% of cells
Normal cellularity, no clusters, normal staining
Normal cellularity, mild clusters, moderate staining
Mild or moderate hypocellularity, slight staining
Severe hypocellularity, poor or no staining
stained with fluorescent dye (Hoechst 33528) and measured
using calf thymus DNA as a reference.34 The proteoglycan
content was determined by measuring the total amount of
glycosaminoglycans (GAG). Those in the papain digest
were precipitated and stained with Alcian Blue for subsequent photometric analysis. The GAG content was
expressed as GAG normalised to the wet weight of the cartilage explants (mg/g) (Table II). The rate of synthesis of proteoglycans (PG) was measured by evaluation of the
incorporation of sulphate after one hour of preincubation.
We added 148 kBq NA235SO42- (Dupont, NEX-041-H, carrier free) in 10 µl DMEM to 200 µl of incubation medium
containing the explants. After four hours of labelling, the
samples were digested in papain. GAGs were precipitated
by adding cetylpyridinium chloride (CPC). The 35SO42radioactivity was measured by liquid scintillation counting.
The rate of incorporation was calculated from the specific
VOL. 85-B, No. 7, SEPTEMBER 2003
No treatment
Early
treatment
Late treatment
C
C
C
E
E
E
(n = 7) (n = 7)
(n = 6) (n = 6)
(n = 6) (n = 6)
4
2
0
7
0
0
0
2
5
6
0
0
3
1
2
6
0
0
0
2
4
3
2
1
0
6
1
0
0
1
2
2
1
5
1
0
0
5
1
0
0
6
0
0
0
2
2
2
0
3
2
1
0
6
1
0
0
0
1
5
0
5
2
0
0
3
3
0
0
4
2
0
0
0
3
2
1
2
1
0
6
1
0
1
3
3
6
0
0
3
3
0
6
0
0
1
1
4
2
1
0
7
0
0
2
3
0
6
0
0
6
0
0
6
0
0
5
1
0
2
1
0
7
0
0
1
3
3
6
0
0
4
1
1
6
0
0
4
2
0
3
2
1
0
6
1
0
0
4
1
1
1
6
0
0
0
5
1
0
0
5
1
0
0
3
1
2
0
2
1
0
3
2
1
0
7
0
0
6
1
0
0
4
1
2
2
5
0
0
6
0
0
3
0
0
0
5
1
0
1
5
0
0
6
0
0
6
0
0
0
4
2
0
0
6
0
0
activity of the medium and normalised to the wet-weight of
the cartilage. Values were expressed in nmol of sulphate
incorporated per hour per gram of cartilage wet-weight
(nmol/h.g). Precipitation of Alcian Blue and scintillation
counting determined the release of both the total amount
and the newly synthesised PGs. Samples of cartilage were
labelled with 370 kBq/200 µl of 35S-sulphate as described
above. After four hours of labelling, the samples were rinsed
three times in fresh culture medium (37˚C) and incubated
for a subsequent three-day period in the absence of label.
The GAGs in the culture medium were precipitated with
Alcian Blue. The total amount of GAG released was determined by spectrophotometric quantification of blue staining. The percentage release was calculated from the total
amount of GAG released and the initial GAG content of the
tissues. The amount of newly formed GAG released was
1072
D. B. F. SARIS, W. J. A. DHERT, A. J. VERBOUT
Macroscopic repair score
12
No treatment
Early treatment
Late treatment
10
8
6
4
2
0
Control
Experimental
Control
Experimental
Control
Experimental
Fig. 3
Graphs showing the considerable differences with a significantly (p = 0.01) larger decrease (worsening) score in the
‘no treatment’ (-6.1) and ‘late treatment’ (-5.4) groups as compared with the ‘early treatment’ group (-3.6). Furthermore, the standard deviation in the ‘early treatment’ group was lower than in the two other groups, which may suggest a more reliable process of articular surface repair.
determined by scintillation counting of SO42- and is
expressed in nmol/g of wet weight during three days.33
Histological examination. On completion of the 10- or 20week follow-up period, all the animals were killed by an
overdose of thiopental. Synovial fluid, synovial lining tissue
and cartilage samples were taken according to a standard
explantation layout. From the centre of each sample of cartilage, multiple 3 µm sections were cut and stained with
either haematoxylin and eosin, Alcian Blue or Safranin OFast Green (Figs 1d and 1e). Generalised degeneration of
cartilage, filling of the defects and appearance of the synovial tissue were scored using the criteria suggested by
O’Driscoll (Table III) which have been previously validated
for use.35
Statistical analysis. In calculating the minimally required
sample size we selected a 90% power at a significance level
of 5% to detect a 30% difference in the means for the cartilage repair score. From previous experiments and literature
the SD in this analysis was estimated to be 20%. Thus six
animals were required and seven were included in each of
the three study groups. The difference in macroscopic outcome variables (Table I, Fig. 3) between the control and
experimental joints was compared between groups by
repeated-measures ANOVA. A hierarchical repeated-measure analysis of variance was applied with Student-NewmanKeuls post-hoc testing to analyse the difference between
means for the histological outcome scores (Table III). A
paired Wilcoxon test was used to determine potential significant differences between the control and experimental joint
within treatment groups for the biochemical parameters of
metabolism of PG.
Results
General findings. Two animals were lost to follow-up on
the day of the initial surgery because of anaesthesia-related,
early postoperative complications (one in the ‘early treat-
ment’ and one in the ‘late treatment’ group). Thus six animals remained for both periosteal transplantation groups
and seven in the control group. All other animals recovered
quickly and uneventfully from the anaesthesia and surgical
procedure. Within four hours of surgery, the animals were
able to walk with intermittent, partial weight-bearing on the
operated hind leg. From the fifth postoperative day, normal
weight-bearing was seen on all four legs. In some animals, a
brief period of limping and effusion was noted between
days 8 and 14, but there was no correlation with the experimental group. No fractures or other complications were
noted on radiological examination. All defects caused radiologically visible alteration in the subchondral bone, most
notably in the ‘no treatment’ group.
Macroscopic evaluation of explanted knees. The contour
of the articular surface was disturbed in both the ‘no treatment’ and the ‘late treatment’ groups with discoloration of
the surface of the cartilage. An increased frequency of
intra-articular fibrosis with decreased range of movement
resulted (Table I, Fig. 3). When comparing the difference
within each animal for the score of the operated v the control joint there was a clear decrease (worsening) in the outcome of macroscopic repair for all groups (no treatment 6.1, early treatment -3.6, late treatment -5.4). The decrease
of the score in the group in which the defect had been
treated early was significantly less (p = 0.01) than that in
the group which had not been treated, which suggests that
there was better joint function after early treatment than
when not treated. This was not the case in the joints treated
late (p = 0.41).
Biochemical analysis. The cartilage defects disturbed joint
homeostasis by causing reproducible early osteoarthritic
alterations in all groups. There was a differential pattern of
significantly increased synthesis of PG (p = 0.018, Fig. 4).
This was most pronounced in the cartilage of the area surrounding the defect in the medial femoral condyle as well as
THE JOURNAL OF BONE AND JOINT SURGERY
JOINT HOMEOSTASIS
1073
40
30
*
*
20
ns
30
E
*
10
C
PG synthesis [nmol/h.g]
PG synthesis [nmol/h.g]
C
0
E
*
20
*
*
10
0
FL
TL
FM
No
treatment
TM
Early
treatment
Late
treatment
Joint location
Fig. 4
Fig. 5
Histogram showing the PG synthesis by wet weight in the untreated defects,
as determined by 35SO4 uptake studies. There is a significantly increased
rate of synthesis in the surrounding cartilage after a standardised defect was
made in the medial femoral condyle (sampling location, both femur (F) and
tibia (T), either lateral (L) or medial (M); C, control; E, experimental; *,
p ≤ 0.05; ns, p > 0.05).
Histogram showing the rate of synthesis of PG in the cartilage of the medial
femoral condyle surrounding the defect for each of the three treatment
groups. There is a significant increase for all three groups. A similar effect
was found in all locations in the medial tibia and in the lateral femoral compartment of the ‘no treatment’ and ‘late treatment’ groups (C, control; E, experimental, *, p ≤ 0.05).
40
C
E
*
Total GAG content [mg/g]
GAG release [%]
4
*
ns
2
0
C
*
ns
*
Early
treatment
Late
treatment
E
30
20
10
0
No
treatment
Early
treatment
Late
treatment
No
treatment
Fig. 6
Fig. 7
Histogram showing that in both the ‘no treatment’ and ‘late treatment’
groups there is a significant increase in the release of GAG for both newly
formed and nascent GAGs from the medial compartment. There was no significant difference in the release of GAG in the group which was treated immediately (C, control; E, experimental; *, p ≤ 0.05; ns, p > 0.05).
Histogram showing a significant decrease in the total content of GAG in the
‘no treatment’ and the ‘late treatment’ groups but not in the ‘early treatment’
group. This indicates a protective effect on matrix metabolic alterations of
early intervention (C, control; E, experimental; *, p ≤ 0.05; ns, p > 0.05).
the opposing tibial surface, but was also present in the lateral joint compartment.
The total amount of synthesis of PG increased significantly and to a similar rate in all experimental groups (Table
II, Fig. 5). The release of GAGs from the matrix was significantly increased in the cartilage surrounding the ‘no treatment’ (p = 0.018) and ‘late treatment’ groups (p = 0.028),
while no significant loss was found in the ‘early treatment’
group (Table II, Fig. 6).
There was a significant decrease in the total GAG content
indicative of a loss of matrix integrity, in both the ‘no treat-
ment’ group (p = 0.032) as well as the ‘late treatment’ group
(p = 0.046, Table II, Fig. 7). We found no significant
increase in the release of GAGs nor was there a decrease in
the total GAG content in the ‘early treatment’ group.
Histological evaluation. Histological examination showed
that all periosteal flaps contained an adequate cambium
layer on transplantation into the bottom of the defect.
Safranin O-Fast Green staining of control and experimental
defects were used to quantify repair of cartilage using the
O’Driscoll score. These sections showed normal values in
control cartilage from the intact contralateral joints with a
VOL. 85-B, No. 7, SEPTEMBER 2003
1074
D. B. F. SARIS, W. J. A. DHERT, A. J. VERBOUT
Fig. 8a
Top row. Photographs showing histological sections of defects in the ‘no treatment’ group after ten weeks of followup (3 µm, Safranin O-Fast Green; ×200). Bottom row. Diagrams showing the location of the original defect. These
two typical samples show that the defects in the cartilage
which were left untreated either remained unchanged as on
the left, or had some degree of filling with fibrocartilage or
extruding cartilage from the normal cartilage rim as occurred on the right.
Fig. 8b
Top row. Photomicrographs of defects in the ‘early treatment’ group (3 µm, Safranin O-Fast Green, × 200). Bottom row. Diagrams of the sections depicting the original
defect. These two typical samples show how the fresh cartilage defect, which was treated early, showed a marked repair of the cartilage surface with a tissue resembling
hyaline cartilage. There are some chondrocyte clusters and
near normal PG staining throughout the matrix. Some remodelling of the subchondral bone is seen.
Fig. 8c
Top row. Photomicrographs of defects from the ‘late treatment’ group (3 mm, Safranin O-Fast Green × 200). Bottom row. Diagrams of the sections depicting the location
of the original defect. These two typical samples from the
‘late treatment’ group show an identical full-thickness defect as in the previous groups, but treated after ten weeks
of movement and loading on the previous superficial cartilage defect. There is irregular filling and hypertrophy of fibrocartilage with uneven staining of the matrix. Also,
signs of delamination and sidewall fissures were seen
more often than in the ‘early treatment’ group.
THE JOURNAL OF BONE AND JOINT SURGERY
JOINT HOMEOSTASIS
30
C
Cartilage repair score
ns
*
E
*
20
10
0
No
treatment
Early
treatment
Late
treatment
Fig. 9
Histogram showing the difference in the O’Driscoll semi-quantitative cartilage repair scores between the control and experimental joint for each of the
‘treatment’ groups. This clearly demonstrates that ‘no treatment’ results in
insufficient repair and that the tissue engineering outcome is not successful
in the ‘late treatment’ group, resembling the clinical condition. The ‘early
treatment’ group shows much better results, similar to those presented by
other authors.
median value of 24 (21 to 24) out of 24 points (Table III).
The defects in the ‘no treatment’ group did not heal and the
median cartilage repair score was low at 12 (6 to 15) (Fig.
8a). In the ‘early treatment’ group, the cartilage repair score
was significantly higher (p = 0.001) with a median of 21.5
(16 to 24) (Fig. 8b) as compared with that in the ‘no treatment’ group. So-called fresh defects treated early showed
reproducible formation of cartilage, with filling of the defect
by cartilaginous tissue, some restoration of the tidemark and
smoothing of the articular surface. There was a significant
decrease in the cartilage repair score with a median of 14
(10 to 16) in the ‘late treatment’ group (p = 0.005), in which
signs of surface irregularity, fibrillation, and loss of interface bonding, with considerable fibrous hypertrophy and
more synovitis were present (Fig. 8c). We found no significant difference (p = 0.12) between the defects treated late
and those left untreated (Fig. 9).
Discussion
The goat model has joint mechanics and a metabolic activity
of cartilage which are comparable to those of man. Adolescent animals were chosen since at this age physical growth
has ceased while maturation continues, and the anatomy of
the knee allows for reproducible surgical technique as well
as an adequate amount of material for subsequent analyses.26 As described earlier, chondrogenesis declines with
age.14 We decided to use these relatively young adults with
adequate periosteal chondrogenesis so that a reliable positive control group could be included in our study.
A defect of the cartilage in the medial femoral condyle
resulted in reproducible early degenerative changes with
significant histological and biochemical alterations. Using
VOL. 85-B, No. 7, SEPTEMBER 2003
1075
the macroscopic evaluation criteria described by O’Driscoll,
we found a significant advantage of early treatment as compared with no treatment. This suggests that normal joint
homeostasis is better retained with early treatment.
As an initial sign of matrix degeneration and disturbance
of joint homeostasis we found a significant increase in the
synthesis of PG in all groups. There was a decrease in the
total GAG content, even as early as a few weeks after the
initial injury. This was caused by the increased release of
matrix components in the ‘no treatment’ and ‘late treatment’ groups, but not in the ‘early treatment’ group. Interestingly, the signs of early osteoarthritis were in part
reversed in the ‘early treatment’ group in which the cartilage defect was treated immediately. This creates the suggestion of a protective effect from early treatment.
Metabolic alterations in the ‘early treatment’ group were
considerably less than in the other groups. This may indicate that early restoration of the defect may allow the prevention of the degradation of cartilage and subsequently
provide an intra-articular environment more beneficial for
the formation of cartilage. We selected short-term follow-up
periods (10 and 20 weeks) since we wished to monitor early
derangement of the metabolism of the matrix and to determine if there was a negative effect on the formation of cartilage. This is warranted since there is no scientific evidence
for good repair at the long-term follow-up when the shortterm follow-up is poor.
In the presence of disturbed joint homeostasis, the formation of cartilage was significantly decreased and the outcome of tissue engineering became insufficient. The
histological findings in our ‘no treatment’ group are in line
with the data from Jackson et al26 on the natural healing of
goat defects and reconfirm that such cartilage defects do not
repair spontaneously. Furthermore, there was no significant
difference between the untreated defects and the defects
which were treated late. The ‘early treatment’ group showed
considerable restoration of the joint surface as demonstrated
by the O’Driscoll cartilage repair scores. Our normal cartilage, untreated defects and early treated defects have scores
which are comparable with data from cartilage repair studies by van Susante et al36,37 and Driesang and Hunziker38
and Niederauer et al.39
We therefore conclude that a disturbed intra-articular
environment negatively influences the formation of cartilage. These findings support our hypothesis that the metabolic aspects of joint homeostasis influence cartilage
formation.
This is of importance for at least three reasons. First,
most, if not all, patients currently treated by these methods
are known to have had a cartilage defect for a longer period
of time and thus cartilage degeneration is present. Our
results suggest an explanation for the discrepancy between
the positive results of experimental repair in healthy joints
and the as yet less favourable clinical results. However,
many questions remain. For instance, what can we learn
from a comparison between the long-term results of early
1076
D. B. F. SARIS, W. J. A. DHERT, A. J. VERBOUT
treatment and late treatment and how do joints respond to
pretreatment? The need for the ongoing investigation of
these effects is apparent.
Secondly, these findings indicate that there should be a
radical change in our treatment of these patients. We propose that clinical studies should be investigated to compare
prompt intervention by tissue engineering with current treatment. Immediate diagnosis and treatment after the occurrence of a cartilage defect may offer an opportunity to
improve results. Alternatively, we should determine methods of modulating joint homeostasis to create an environment more permissive for chondrogenesis before the
application of a tissue-engineering strategy.
Finally, but not less importantly, these data reiterate the
relevance of using appropriate animal models. The results of
our study suggest that the evaluation of tissue-engineering
techniques should be done using models of cartilage repair
comparable to those of the clinical situation. The sole investigation of fresh defects in healthy joints should not be considered to be of any predictive value for the success of the
clinical implementation of the technique studied.
The authors are grateful to H. Vosmeer for his assistance in all surgery and
animal care, and to F. Lafeber and M. Vianen for their support in the proteoglycan analysis.
No benefits in any form have been received or will be received from a
commercial party related directly or indirectly to the subject of this article.
References
1. Brittberg M, Lindahl A, Nilsson A, et al. Treatment of deep cartilage
defects in the knee with autologous chondrocyte transplantation. N Engl
J Med 1994;331:889-95.
2. Hardaker JWT, Garrett JWE, Bassett F. Evaluation of acute traumatic hemarthrosis of the knee joint. South Med J 1990;83:640-4.
3. Mankin HJ. The response of articular cartilage to mechanical injury. J
Bone Joint Surg [Am] 1982;64-A:460-6.
4. Radin EL, Ehrlich MG, Chernack R, Abernethy P, Paul IL, Rose
RM. Effect of repetitive impulsive loading on the knee joints of rabbits.
Clin Orthop 1978;131:288-93.
5. Buckwalter JA, Mankin HJ. Articular cartilage. Part I. Tissue design
and chondrocyte-matrix interactions. J Bone Joint Surg [Am] 1997;79A:600-11.
6. Brittberg M, Nilsson A, Lindahl A, Ohlsson C, Peterson I. Rabbit
articular cartilage defects with autologous cultured chondrocytes. Clin
Orthop 1996;326:270-83.
7. Buckwalter JA, Mankin H. Articular cartilage. Part II. Degeneration
and osteoarthrosis, repair, regeneration and transplantation. J Bone Joint
Surg [Am] 1997;79-A:612-32.
8. Mankin HJ, Buckwalter JA. Restoration of the osteoarthritic joint. J
Bone Joint Surg [Am] 1996;78-A:1-2.
9. O’Driscoll SW, Recklies AD, Poole AR. Chondrogenesis in periosteal
explants: an organ culture model for in vitro study. J Bone Joint Surg
[Am] 1994;76-A:1042-51.
10. Ritsila VA, Sontavirta S, Ahlopuros, et al. Periosteal and perichondral
grafting in reconstructive surgery. Clin Orthop 1994;302:259-65.
11. Minas T. The role of cartilage repair techniques, including chondrocyte
transplantation, in focal chondral knee damage. Instr Course Lect
1999;48:629-43.
12. Minas T, Peterson L. Advanced techniques in autologous chondrocyte
transplantation. Clin Sports Med 1999;18:13-44, v-vi.
13. O’Driscoll SW. The healing and regeneration of articular cartilage. J
Bone Joint Surg [Am] 1998;80-B:1795-812.
14. O’Driscoll SW, Saris DB, Ito Y, Fitzsimmons JS. The chondrogenic
potential of periosteum decreases with age. J Orthop Res 2001;19:95103.
15. Bouwmeester P, Kuijer R, Terwindt-Rouwenhorst E, van der
Linden T, Bulstra S. Histological and biochemical evaluation of perichondrial transplants in human articular cartilage defects. J Orthop Res
1999;17:843-9.
16. Nehrer S, Spector M, Minas T. Histologic analysis of tissue after failed
cartilage repair procedures. Clin Orthop 1999;365:149-62.
17. Nehrer S, Minas T. Treatment of articular cartilage defects. Invest
Radiol 2000;35:639-46.
18. Peterson L, Minas T, Brittberg L, et al. Two- to 9-year outcome after
autologous chondrocyte transplantation of the knee. Clin Orthop
2000;374:212-34.
19. Hunziker EB. Articular cartilage repair: basic science and clinical
progress: a review of the current status and prospects. Osteoarthritis
Cartilage 2002;10:432-63.
20. Bouwmeester PS, Kuijer R, Homminga GN, Bulstra SK, Geesink
RG. A retrospective analysis of two independent prospective cartilage
repair studies: autogenous perichondrial grafting versus subchondral
drilling 10 years post-surgery. J Orthop Res 2002;20:267-73.
21. Hunter W. Of the structure and disease of articulating cartilages. Royal
Soc London, Phil Trans, 9:267-71, 1743.
22. Mankin HJ. The articular cartilages: a review. In: Instructional course
lectures of the American Academy of Orthopaedic Surgeons, pp. 204224. Edited, 204-224, St Louis, CV Mosby Co, 1970.
23. Stryer L. Biochemistry. Edited, New York: W. H. Freeman and Company, 1975.
24. Vogel J. Cats’ paws and catapults: mechanical worlds of nature and
people. Edited, New York: Norton & Company, 1998;229-48.
25. Rodrigo JJ, Steadman JR, Syftestad G, Benton H, Silliman J.
Effects of human knee synovial fluid on chondrogenesis in vitro. Am J
Knee Surg 1995;8:124-9.
26. Jackson DW, Lalor PA, Aberman HM, Simon TM. Spontaneous
repair of full-thickness defects of articular cartilage in a goat model: a
preliminary study. J Bone Joint Surg [Am] 2001;83-A:53-64.
27. O’Driscoll SW, Fitzsimmons JS. The importance of procedure specific
training in harvesting periosteum for chondrogenesis. Clin Orthop
2000;380:269-78.
28. Brownlow HC, Reed A, Joyner C, Simpson AH. Anatomical effects of
periosteal elevation. J Orthop Res 2000;18:500-2.
29. O’Driscoll SW, Meisami B, Miura Y, Fitzsimmons JS. Viability of
periosteal tissue obtained postmortem. Cell Transplant 1999;8:611-6.
30. O’Driscoll SW, Keeley FW, Salter RB. The chondrogenic potential of
free autogenous periosteal grafts for biological resurfacing of major fullthickness defects in joint surfaces under the influence of continuous passive motion: an experimental investigation in the rabbit. J Bone Joint
Surg [Am] 1986;68-A:1017-35.
31. O’Driscoll SW, Keeley FW, Salter RB. Durability of regenerated articular cartilage produced by free autogenous periosteal grafts in major
full-thickness defects in joint surfaces under the influence of continuous
passive motion: a follow-up report at one year. J Bone Joint Surg [Am]
1988;70-A:595-606.
32. O’Driscoll SW, Salter RB. The induction of neochondrogenesis in free
intra-articular periosteal autografts under the influence of continuous
passive motion: an experimental investigation in the rabbit. J Bone Joint
Surg [Am] 1984;66-A:1248-57.
33. Lafeber FP, Vander Kraan PM, Van Roy JL, Huber Bruning O,
Bijlsma JW. Articular cartilage explant culture: an appropriate invitro
system to compare osteoarthritic and normal human cartilage. Connect
Tissue Res 1993;29:287-99.
34. Labarca C, Paigen K. A simple, rapid, and sensitive DNA assay procedure. Anal Biochem 1980;102:344-52.
35. Moojen DJF, Saris DBF, et al. The correlation and reproducibility of histological scoring systems in cartilage repair. Tissue Eng 2002;8:630-8.
36. van Susante JL, Buma P, Homminga GN, van den Berg WB, Veth
RP. Chondrocyte-seeded hydroxyapatite for repair of large articular cartilage defects. A pilot study in the goat. Biomaterials 1998;19:2367-74.
37. van Susante JL, Buma P, Schuman L, et al. Resurfacing potential of
heterologous chondrocytes suspended in fibrin glue in large full-thickness defects of femoral articular cartilage: an experimental study in the
goat. Biomaterials 1999;20:1167-75.
38. Driesang IM, Hunziker EB. Delamination rates of tissue flaps used in
articular cartilage repair. J Orthop Res 2000;18:909-11.
39. Niederauer GG, Slivka MA, Leatherbury NC, et al. Evaluation of
multiphase implants for repair of focal osteochondral defects in goats.
Biomaterials 2000;21:2561-74.
THE JOURNAL OF BONE AND JOINT SURGERY