The effect of defect localization on spontaneous repair of osteochondral defects in a Göttingen minipig model: a retrospective analysis of the medial patellar groove versus the medial femoral condyle

Abstract

Various animal models for experimental osteochondral defect healing have been used in orthopaedic research. Two main defect locations were chosen: the patellar groove or the central part of the medial femoral condyles (MFC). To date, it is not clear whether both locations display similar patterns in critical size osteochondral defect healing. We retrospectively analysed both locations in our minipig model hypothesizing that they show similar healing pattern. Thirty-five defects were analysed after three or 12 months. Osteochondral defects were 10 mm deep and 6.3 mm (MFC, n = 19) in diameter or 8 mm and 5.4 mm, respectively (trochlear groove [TG], n = 16). Semi-quantitative histological scoring and histomorphological evaluation were carried out. Both defect locations showed fillings of fibrous and fibrocartilage-like repair tissue. The osseous defect was closed by endochondral bone formation in the MFC. Semi-quantitative scoring did not show differences, whereas qualitative histomorphological analysis more frequently showed cartilaginous repair tissue in MFC defects. There was more frequent subchondral bone cyst formation in MFC location (P = 0.05), TG defects resulted in lower postoperative pain. Both defect localizations are suitable for studies on osteochondral healing. Since regenerating with less hyaline-like repair tissue and less subchondral cyst formation, TG is more favourable for experimental osteochondral defect healing in this model.

Keywords

Osteochondral defect cartilage repair minipig

The healing of cartilage defects is still a challenging problem in the clinical setting. Many operative procedures have been established, such as microfracturing,¹ osteochondral autograft transplantation² and autologous chondrocyte transplantation,³ but each of them has shown distinct limitation in clinical experience.⁴ Subsequently, there is a focus on tissue engineering-based strategies in current research to improve articular cartilage repair including stem cell, scaffold and growth factor-based technologies. For a detailed review the reader is referred to Hunziker.⁵ As a matter of fact any new treatment protocol has to be evaluated in an appropriate animal model before being brought into clinical practice. For this purpose various animal models were employed to investigate experimental defect healing. The smallest animals used were mice⁶ and rats,⁷ being followed by rabbits.⁸ For more advanced preclinical studies, large animal models such as dogs,⁹ goats,¹⁰ pigs¹¹ or horses¹² were used.

It has been shown that spontaneous healing of osteochondral defects is dependent on defect size,¹³ depth,¹⁴ geometry¹⁵ and age¹⁶ of the individual, but little is known about the influence of defect location on the natural healing response of osteochondral defects in the knee joint.

A larger animal used in our and other investigators’ work dealing with osteochondral articular cartilage defect repair is the Göttingen minipig (GMP) model.^11,17–22 In earlier studies investigating various scaffolds to improve repair tissue quality, we evaluated experimental osteochondral defect healing using two different localizations in the knee of miniature pigs: either in the central portion of the medial femoral condyle (MFC) or the medial facet of the trochlear groove (TG). This is the first report on different healing patterns of osteochondral defects created either in the MFC or the TG. To our knowledge there are only rare comparative studies available evaluating the impact of different defect localizations on osteochondral defect repair in one species. We retrospectively analysed the histological outcome of spontaneous osteochondral defect healing and hypothesized that both localizations will display a similar healing pattern.

Methods

Animals

All animal care, housing and treatment was performed according to the German Animal Welfare Act of 25 May 1998, after receiving written allowance for the experiments by the animal rights protection authorities in Baden Württemberg, Germany (Regierunspraesidium Karlsruhe, AZ: 35-9185.81/5/00, 35-9185.81/151/00).

The outcome parameters of 35 mature Göttingen minipigs (Elegaard Soro Landevej 302, DK-4261 Dalmose, Denmark) of either sex (12 male; 23 female) were retrospectively used for this study. The average age and weight of the minipigs were 32 months (22–49 months) and 34.1 kg (21–46 kg), respectively, at the beginning of the study. Animals were held in big indoor runs for at least one week for acclimatization with unrestricted movement of food and water ad libitum before starting the experiment (for details concerning minipig housing and strain see Gotterbarm et al. ¹¹).

Defect localization and surgery

All animals were operated in aseptic conditions under general anaesthesia on both hind legs either at the MFC or the medial facet of the TG as described elsewhere, creating exactly one defect per knee joint.^18,20 In both locations the defect diameter was chosen to allow exact fitting in the area of the corresponding joint surface. One side was left empty, the other one was treated with various experimental procedures, reported elsewhere, see above.

To access the MFC, a medially-placed paramedian skin incision was set at each knee, followed by soft tissue preparation to the knee capsule which was opened with a 2 cm incision directly placed over the MFC. The location for the osteochondral lesion was chosen centrally and midline on the central weight-bearing portion of the medial condyle, 10 mm distal of the intercondylar notch. The cartilage was cut to the subchondral bone plate perpendicular to the joint surface with a circular cutter to avoid cartilage tearing. A diamond bone cutting system (Surgical Diamond Instruments [SDI]; MedArtis AG, Germany) was used to create a 6.3 mm wide and 10 mm deep cylindrical osteochondral defect. The diamond cutter was rinsed with 0.9% sterile saline to avoid heat damage to the adjacent bone.

For the TG defect localization, a median skin incision was placed between the tibial tuberositas and the distal patellar pole while extending the knee. After soft tissue preparation, the patellar tendon and the inferior patellar fat pad were split longitudinally, exposing the patellar groove. The articular cartilage on the medial trochlear facet was cut 10 mm above the intercondylar notch as described above with a 5.4 mm circular cutter, and the osteochondral defect of 5.4 mm × 8 mm was created with a bone wet grinding device. After extraction of the osteochondral plug in either localization, the defect was rinsed with saline solution and any bone or cartilage debris was removed carefully and the wound was closed in layers.

The animals of both groups were allowed immediate, full weight bearing postoperatively. Animals were followed-up closely during the first 10 days after operation, with special concern for postoperative pain levels and abnormal gait (limping). If necessary, intramuscular pain medication was administered with 0.01 mg/kg buprenorphine (Temgesic^©, Essex Pharma GmbH, Germany). Animals were sacrificed by an intravenously administered overdose (40 mL) of barbiturate (Eunarcon^®, Upjohn GmbH, Germany) either three or 12 months postoperatively and the hind legs were dissected to obtain the knees for further preparation.

Histological evaluation

Following defect preparation, a cube of approximately 2 × 2 × 2 cm was cut out of the MFC or the medial trochlea, respectively, including the repair zone in the centre. This cube was cut dividing the round repair zone into 2/3:1/3. Formalin-fixed specimens were decalcified in 0.5 mmol/L ethylenediaminetetraacetic acid (pH = 7.4) for four to six weeks followed by dehydration and paraffin embedding. Serial sections (5 μm) were stained with Safranin O/Fast green according to standard protocols. For the histological evaluation according to Mainil-Varlet et al. ²³ a modification of the semi-quantitative scoring system originally described by O'Driscoll et al. ²⁴ was used, reaching a maximum of 34 points (Table 1). The evaluation was carried out blindly by two of the authors (TG and MJ). Additional interest was kept on subchondral cyst formation categorizing into ‘yes’ or ‘no’. Cysts were defined as remaining subchondral bone cavities covered by a ‘bridge’ of newly formed bone, and not only as partially filled defects.

Table 1
Histological grading scale by Mainil-Varlet reaching a maximum of 34 points

Category Points

(I) Tissue morphology

Mostly hyaline cartilage 3

Mostly fibrocartilage 2

Mostly non-cartilage 1

Non-cartilage only 0

(MFC: 0.67; TG: 0.44)

(II) Matrix staining

Normal or nearly normal 3

Moderate 2

Slight 1

None 0

(MFC: 0.56; TG: 0.44)

(III) Structural integrity

Normal 4

Beginning of a columnar organization 3

No organization 2

Cysts or disruptions 1

Severe disintegration 0

(MFC: 1.56; TG: 1.56)

(IV) Chondrocyte clustering

No clusters 2

<25% of the cells 1

25–100% of the cells 0

(MFC: 0.11; TG: 0.0)

V Formation of tidemark

Complete 4

76–90% 3

50–75% 2

25–49% 1

<25% 0

(MFC: 0.33; TG: 0.0)

(VI) Subchondral bone formation

Good 2

Slight 1

No formation 0

(MFC: 0.22; TG: 0.44)

(VII) Architecture of the surface

Normal 3

Slight fibrillation or irregularity 2

Moderate fibrillation or irregularity 1

Severe fibrillation or disruption 0

(MFC: 2.0; TG: 2.67)

(VIII) Filling of the defect

91–110% 4

76–90% or 110–125% 3

51–75% 2

26–50% 1

<25% 0

(MFC: 1.0; TG: 1.11)

(IX) Lateral integration

Bonded at both ends of graft 2

Bonded at one end/partially bonded at both sides 1

Not bonded 0

(MFC: 1.89; TG: 1.89)

(X) Basal integration

91–100% 3

70–90% 2

50–70% 1

<50% 0

(MFC: 0.0; TG: 0.0)

(XI) Inflammation

No inflammation 4

Slight inflammation 2

Strong inflammation 0

(MFC: 0.0; TG: 0.0)

Means of each subheading are given in brackets for the medial femoral condyle (MFC) and trochlear groove (TG) defects

Statistical analysis

Since the variance of the evaluated groups was significantly different, non-parametric tests were utilized: the Mann-Whitney U-test for independent samples was used for evaluating differences in the histopathological score of the two defect localizations. The Fisher's exact test was performed to find out significance between the cyst formations of the two defect localizations. Significance was reported if P ≤ 0.05.

Data analysis was performed using SPSS for windows 10.0 (SPSS Inc, Chicago, IL, USA).

Results

After three and 12 months, respectively, nine and 10 MFC defects were available, whereas nine and seven TG defects could be examined.

Clinical observation revealed that animals with osteochondral defects in the MFC showed prolonged postoperative recovery in terms of higher pain levels with limping up to 7–10 days, whereas only 3–5 days of limping was seen in the TG group.

Semi-quantitative histological evaluation with the modified O'Driscoll score did not show any statistical differences in the three-month group (median/interquartile range [IQR] 13/8.5 in the MFC vs. 14/6.5 in the TG). After 12 months, the modified O'Driscoll score values improved for both defect localizations showing significance only for the MFC group (P = 0.028; Figure 1). A slight tendency with improved repair tissue quality was seen towards the femoral condyle defect, with a median/IQR of 23/6.25 compared with 21/12 in the TG group, not reaching significance however (P = 0.23).

Figure 1
Box plot showing median, 25 and 75 quartiles of the histopathological scoring revealing no significant differences between the two defect localizations after either three or 12 months. A significant change of the histopathological score values was found between the three and 12 month interval in the femoral condyle group (^# P = 0.028). Empty boxes: femoral condyle; filled boxes: trochlear groove

New trabecular bone formation was conducted by enchondral ossification beginning at the defect edges in the three-month group. This process started to ‘bridge’ the osteochondral defect in the upper third of the defect immediately after three months in the MFC group and allowed neo-cartilage tissue formation towards the joint cavity (Figure 2A). Anyhow up to three months no defect was ‘bridged’ completely. Complete restoration of the subchondral bone plate after 12 months resulted in subchondral bone cyst formation, which did not regenerate until 12 months postoperatively (Figure 3).

Figure 2
(A) Enchondral ossification three months postoperatively at the defect rim of femoral condyle group: an ossification front is closing the upper part of the defect leaving chondrocytes at the defect surface, restoring the chondral part of the defect. (B) Nearly no enchondral ossification was seen three months postoperatively in the trochlea groove group

Figure 3
Defect healing by enchondral ossification in the femoral condyle group was showing nearly almost complete restoration of the subchondral bone after one year, even though a bone cyst (see asterisks) resulted in the deeper parts of the defect

In TG defects enchondral bone formation at the upper defect edges was also found after three months, but on a much lesser extent compared with MFC defects (Figure 2B). In contrast, complete closure of the subchondral bone cavity by new bone formation creating a subchondral bone cyst was rarely seen at three and 12 months. Going along with this finding, less glycosaminoglycan (GAG)-positive regeneration tissue formation in the TG group was seen. Statistical analyses revealed bone cyst formation significantly more often after 12 months in MFC defects when compared with TG defects (70% in MFC vs. 14.3% in TG defects, P = 0.05).

Seven out of 10 specimens with cyst formation in the MFC group showed nearly complete subchondral bone repair in the upper third of the defect. The regenerated bone was covered with cartilagineous repair tissue (in 8 out of 10 defects), with round or oval chondrocyte-like cells within lacunae surrounded by a GAG-rich extracellular matrix (ECM) (Figure 4A). ECM appeared to be hyaline-like, without prominent collagen bundles and in three specimens columnar radial orientation of the cells similar to native articular cartilage (Figure 4B) was found. The regeneration tissue of TG defects displayed a more fibrocartilaginous quality (in 5 out of 7) with no radial orientation of the chondrocyte-like cells and less GAG deposition in the ECM (Figures 4C and D).

Figure 4
(A) Cartilaginous repair tissue was seen in eight out of 10 specimens after one year in the femoral condyle group. Radial orientation of the chondrocyte-like cells was noted in some specimen (B = magnification of A). (C) In trochlear groove defects, fibrocartilaginous repair tissue was found with no radial orientation of the cells (D = magnification of C). A and C: bar = 1000 μm, B and D: bar = 400 μm

Discussion

Clinical repair of osteochondral defects is still a challenging problem. Various defect models in different species have been used recently to examine the natural healing response and to evaluate the benefit of new treatment modalities. The central part of the MFC and the TG have been extensively used in several larger experimental animal models including goats,^10,25 dogs,⁹ sheep²⁶ and pigs.^{19,21,27–29} Surprisingly, only a few studies are published comparing the effect of the defect location on the natural course of osteochondral healing in the knee joint in a larger animal model.^30,31

In the presented work we retrospectively analysed our data when both defect locations were used for experimental osteochondral defect repair in the GMP model. A remarkable finding was that osteochondral defects located in the MFC did show a significantly more frequent cystic subchondral bone cavity after complete restoration of the subchondral bone plate. In addition, articular cartilage repair tissue quality seemed to improve with time and at its best resulted in cartilage-like tissue in MFC defects compared with fibrocartilaginous tissue in TG defects. Jackson et al. ¹⁰ reported similar findings in the natural course of osteochondral defect healing in the MFC in goats with cystic subchondral bone lesions after one year. Based on their findings, the authors assumed that defect location has a major impact on the endogenous repair response. Niederauer used two defect locations in an osteochondral defect model in goats: the MFC and the patellar groove. He further compared four different implants, half of which were supported by autologous cartilaginous rib cells. The implants consisted of different mixtures of polylactid/polyglycolic acid (PGL) and either polyglycolic fibres, bioglass particles or medical grade calcium sulphate particles. Interestingly, on the one hand, no significant differences were found between the different implant groups in the histological or macroscopic scores. On the other hand, significant differences in score values were found between the two defect locations, demonstrating a better defect healing in the MFC – consistent with our finding.³¹ Unfortunately, no untreated control defects were included in this study.

Several factors have been identified to predominantly influence the natural healing response of artificial osteochondral defects. Among these are defect size,¹³ depth,¹⁴ geometry¹⁵ and the age of the individual animal.¹⁶ There are also published data indicating that the mechanical environment of the defect location does play an important role for spontaneous defect repair. Duda et al. ¹⁷ have linked the process of repair tissue formation in osteochondral defect healing to distinct local mechanical strain. Increased strain at the circumference of osteochondral defects resulted in bone apposition at the defect edges and bone resorption at the defect base. In accordance, our data showed similar histomorphological findings with bone apposition by enchondral bone formation at the upper defect circumference leading to a restoration of the subchondral bone plate. Duda et al. ¹⁷ used TG as the defect location and did not see any cyst formation.

Based on earlier published data, the tibiofemoral and patellofemoral joint compartments have to be considered as fully weight-bearing in humans and larger animals with a similar amount of mechanical load. Kuster et al. ³² presented a literature review of tibiofemoral and patellofemoral joint loads in humans. According to this review, tibiofemoral forces are described to be as high as 2.5–8 times the body weight (BW) during level or downhill walking. Nearly the same forces are reached in the patellofemoral joint ranging between 1.3 and 7 times the BW, respectively. Elaborate studies in larger animal models confirmed these findings, even though with generally lower joint loading in quadrupeds. Stress loads of 1.5–3.5 BW patellofemoral were described in cats³³ and 2.1 BW tibiofemoral in the sheep model.³⁴ Although no data are available in the minipig model, we consider joint load in both locations as similar. As defect size and geometry did not differ substantially in our studies, we interpret our findings based on the distinct shape and geometry of the articulating surfaces in each joint compartment: a rather plane and congruent articulating surface in the TG in contrast to a more convex surface in the MFC. Brown et al. ³⁵ showed evidence that osteochondral defects in the MFC display a contact stress concentration at the defect rim independent of the defect radius, but with an elevation of the radial contact stress gradient in larger defects of up to 6 mm. There are no data available on the stress distribution in TG defects, but we speculate that based on the geometry of the articulating surfaces (more flat and congruent compared with the convexity of the tibiofemoral joint), actual peak rim stress and therefore the radial stress gradient is significantly lower. As contact stress itself is one of the mediators leading to an increase in bone apposition,¹⁷ we speculate that lower mechanical load at the defect rim does result in a decreased amount of bone and tissue formation in TG defects when compared with the MFC. As bone apposition, resorption and tissue differentiation do take place in a more balanced fashion, the formation of cystic bone cavities does not occur. Further experiments have to be conducted to quantify and compare contact forces in osteochondral defects in both knee compartments. Computerized models of cartilage development and maintenance like in the work of Carter and Wong may further help understand the input of biomechanical load on cartilage healing.³⁶

In both localizations incomplete filling of the articular cartilage defect void with fibrous and fibrocartilage-like repair tissue was present. Semi-quantitative scoring did not show any differences in mean total score values, whereas qualitative histomorphological analysis revealed cartilage-like repair tissue in MFC defects more frequently after one year. Tissue formation is a subtle and complex process during defect restoration and the process of bone and cartilage formation is linked to each other. A possible explanation for the better chondral regeneration in the MFC group may be the fact that the MFC defects display an earlier closure of the bony defect with a full restoration of the subchondral bone plate starting from the defect rim. As reported by Huang et al. ³⁷ in 2006, it also seems possible that the chondroprogenitor cells at the periphery (i.e. near the original joint cavity) have not proceeded down the enchondral cascade and positively influenced the chondral regeneration.

In conclusion both defect localizations are principally suitable for experimental osteochondral defect repair in the GMP model. In experimental cartilage healing, low or even no self-repair capacity is desirable, in order to increase the differences between potentially beneficial cartilage-healing protocols and the empty defects. The somewhat better histomorphological results in cartilage repair and the longer limping in the MFC group have to be considered therefore as a major drawback. Since osteochondral defects in the patellar groove displayed a more limited capacity for regenerating a cartilage-like repair tissue with reduced pain levels, this defect localization seems to be more favourable for experimental osteochondral defect repair. Exact differences in defect rim stress and shear forces at these two defect localizations are not evaluated so far. Further experiments in that direction could enlighten the role of the biomechanical load in the spontaneous repair of osteochondral defects however.

Category	Points
(I) Tissue morphology
Mostly hyaline cartilage	3
Mostly fibrocartilage	2
Mostly non-cartilage	1
Non-cartilage only	0
(MFC: 0.67; TG: 0.44)
(II) Matrix staining
Normal or nearly normal	3
Moderate	2
Slight	1
None	0
(MFC: 0.56; TG: 0.44)
(III) Structural integrity
Normal	4
Beginning of a columnar organization	3
No organization	2
Cysts or disruptions	1
Severe disintegration	0
(MFC: 1.56; TG: 1.56)
(IV) Chondrocyte clustering
No clusters	2
<25% of the cells	1
25–100% of the cells	0
(MFC: 0.11; TG: 0.0)
V Formation of tidemark
Complete	4
76–90%	3
50–75%	2
25–49%	1
<25%	0
(MFC: 0.33; TG: 0.0)
(VI) Subchondral bone formation
Good	2
Slight	1
No formation	0
(MFC: 0.22; TG: 0.44)
(VII) Architecture of the surface
Normal	3
Slight fibrillation or irregularity	2
Moderate fibrillation or irregularity	1
Severe fibrillation or disruption	0
(MFC: 2.0; TG: 2.67)
(VIII) Filling of the defect
91–110%	4
76–90% or 110–125%	3
51–75%	2
26–50%	1
<25%	0
(MFC: 1.0; TG: 1.11)
(IX) Lateral integration
Bonded at both ends of graft	2
Bonded at one end/partially bonded at both sides	1
Not bonded	0
(MFC: 1.89; TG: 1.89)
(X) Basal integration
91–100%	3
70–90%	2
50–70%	1
<50%	0
(MFC: 0.0; TG: 0.0)
(XI) Inflammation
No inflammation	4
Slight inflammation	2
Strong inflammation	0
(MFC: 0.0; TG: 0.0)

Footnotes

ACKNOWLEDGEMENTS

We would like to thank Dr vet. Helga Lorenz for her expertise in animal anaesthesia and animal care. This study was funded partly by the Research Fund of the Department of Orthopaedic Surgery, University of Heidelberg, Germany.

References

Steadman

, Rodkey

, Briggs

, Rodrigo

. The microfracture technic in the management of complete cartilage defects in the knee joint. Orthopade 1999;28:26–32

Bobic

. Autologous osteo-chondral grafts in the management of articular cartilage lesions. Orthopade 1999;28:19–25

Brittberg

, Lindahl

, Nilsson

, Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 1994;331:889–95

Smith

, Knutsen

, Richardson

. A clinical review of cartilage repair techniques. J Bone Joint Surg Br 2005;87:445–9

Hunziker

. Articular cartilage repair: basic science and clinical progress. A review of the current status and prospects. Osteoarth Cartilage 2002;10:432–63

Angele

, Kujat

, Nerlich

, Engineering of osteochondral tissue with bone marrow mesenchymal progenitor cells in a derivatized hyaluronan-gelatin composite sponge. Tissue Eng 1999;5:545–54

, Prudhommeaux

, Meunier

, Effects of chitosan on rat knee cartilages. Biomaterials 1999;20:1937–44

Simank

, Sergi

, Jung

, Effects of local application of growth and differentiation factor-5 (GDF-5) in a full-thickness cartilage defect model. Growth Factors 2004;22:35–43

Cook

, Patron

, Salkeld

, Rueger

. Repair of articular cartilage defects with osteogenic protein-1 (BMP-7) in dogs. J Bone Joint Surg Am 2003;85-A(suppl.):3116–23

10.

Jackson

, Lalor

, Aberman

, Simon

. 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

11.

Gotterbarm

, Breusch

, Schneider

, Jung

. The minipig model for experimental chondral and osteochondral defect repair in tissue engineering: retrospective analysis of 180 defects. Lab Anim 2008;42:71–82

12.

Fortier

, Mohammed

, Lust

, Nixon

. Insulin-like growth factor-I enhances cell-based repair of articular cartilage. J Bone Joint Surg Br 2002;84:276–88

13.

Convery

, Akeson

, Keown

. The repair of large osteochondral defects. An experimental study in horses. Clin Orthop Relat Res 1972;82:253–62

14.

Hanie

, Sullins

, Powers

, Nelson

. Healing of full-thickness cartilage compared with full-thickness cartilage and subchondral bone defects in the equine third carpal bone. Equine Vet J 1992;24:382–6

15.

Lietman

, Miyamoto

, Brown

, The temporal sequence of spontaneous repair of osteochondral defects in the knees of rabbits is dependent on the geometry of the defect. J Bone Joint Surg Br 2002;84:600–6

16.

Bos

, Verhaar

, van Osch

. Age-related differences in articular cartilage wound healing: a potential role for transforming growth factor beta1 in adult cartilage repair. Adv Exp Med Biol 2006;585:297–309

17.

Duda

, Maldonado

, Klein

, On the influence of mechanical conditions in osteochondral defect healing. J Biomech 2005;38:843–51

18.

Gotterbarm

, Richter

, Jung

, An in vivo study of a growth-factor enhanced, cell free, two-layered collagen-tricalcium phosphate in deep osteochondral defects. Biomaterials 2006;27:3387–95

19.

Gotterbarm

, Reitzel

, Schneider

, Integration of periosteum covered autogenous bone grafts with and without autologous chondrocytes. An animal experiment using the Gottinger minipig. Orthopade 2003;32:65–73

20.

Jung

, Tuischer

, Sergi

, Local application of a collagen type I/hyaluronate matrix and growth and differentiation factor 5 influences the closure of osteochondral defects in a minipig model by enchondral ossification. Growth Factors 2006;24:225–32

21.

Jung

, Gotterbarm

, Gruettgen

, Molecular characterization of spontaneous and growth-factor-augmented chondrogenesis in periosteum-bone tissue transferred into a joint. Histochem Cell Biol 2005;123:447–56

22.

Maruyama

, Hasegawa

, Sakano

, Experimental evaluation of the usefulness of osteochondral allograft for articular cartilage defect. J Orthop Sci 2003;8:560–6

23.

Mainil-Varlet

, Rieser

, Grogan

, Articular cartilage repair using a tissue-engineered cartilage-like implant: an animal study. Osteoarth Cartilage 2001;9(suppl. A):S6–15

24.

O'Driscoll

, Keeley

, Salter

. 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;70:595–606

25.

Louwerse

, Heyligers

, Klein-Nulend

, Use of recombinant human osteogenic protein-1 for the repair of subchondral defects in articular cartilage in goats. J Biomed Mater Res 2000;49:506–16

26.

Burks

, Greis

, Arnoczky

, Scher

. The use of a single osteochondral autograft plug in the treatment of a large osteochondral lesion in the femoral condyle: an experimental study in sheep. Am J Sports Med 2006;34:247–55

27.

Chang

, Pradhan

, Mitra

, The results of autogenous tibial periosteal transplants for full thickness cartilage defects in the knee joints of pigs. Ann Acad Med Singapore 1999;28:8–14

28.

Mahadev

, Mahara

, Chang

, Autogenous osteochondral morselised grafts for full thickness osteochondral defects in the knee joints of pigs. Singapore Med J 2001;42:410–16

29.

Zhou

, Wang

, Miao

, Repairing porcine knee joint osteochondral defects at non-weight bearing area by autologous BMSC. Zhonghua Yi Xue Za Zhi 2004;84:925–31

30.

DePalma

, McKeever

, Subin

. Process of repair of articular cartilage demonstrated by histology and autoradiography with tritiated thymidine. Clin Orthop Relat Res 1966;48:229–42

31.

Niederauer

, Slivka

, Leatherbury

, Evaluation of multiphase implants for repair of focal osteochondral defects in goats. Biomaterials 2000;21:2561–74

32.

Kuster

, Wood

, Stachowiak

, Gachter

. Joint load considerations in total knee replacement. J Bone Joint Surg Br 1997;79:109–13

33.

Hasler

, Herzog

, Ronsky

. Experimental evaluation of theoretical contact forces in the cat patellofemoral joint. J Biomech 1996;29:1201–5

34.

Taylor

, Ehrig

, Heller

, Tibio-femoral joint contact forces in sheep. J Biomech 2006;39:791–8

35.

Brown

, Pope

, Hale

, Effects of osteochondral defect size on cartilage contact stress. J Orthop Res 1991;9:559–67

36.

Carter

, Wong

. Modelling cartilage mechanobiology. Philos Trans R Soc Lond B Biol Sci 2003;358:1461–71

37.

Huang

, Durbhakula

, Angele

, Lunate arthroplasty with autologous mesenchymal stem cells in a rabbit model. J Bone Joint Surg Am 2006;88:744–52