Long-Term Evaluation of Allogenic Chondrocyte-Loaded PVA–PCL IPN Scaffolds for Articular Cartilage Repair in Rabbits

Abstract

Objective

This study tested the long-term efficacy of two synthetic scaffolds for osteochondral defects and compare the outcomes with that of an established technique that uses monolayer cultured chondrocytes in a rabbit model.

Methods

Articular cartilage defect was created in both knees of 18 rabbits and divided into three groups of six in each. The defects in first group receiving cells loaded on Scaffold A (polyvinyl alcohol–polycaprolactone semi-interpenetrating polymer network (Monophasic, PVA–PCL semi-IPN), the second on Scaffold B (biphasic, PVA–PCL incorporated with bioglass as the lower layer), and the third group received chondrocytes alone. One animal from each group was sacrificed at 2 months and the rest at 1 year. O’Driscoll’s score measured the quality of cartilage repair.

Results

The histological outcome had good scores (22, 20, and 19) for all three groups at 2 months. At 1-year follow-up, the chondrocyte alone group had the best scores (mean 20.0 ± 1.4), while the group treated by PVA–PCL semi-IPN scaffolds fared better (mean 15 ± 4.2) than the group that received biphasic scaffolds (mean 11.8 ± 5.9). In all three groups, defects treated without cells scored less than the transplant.

Conclusion

These results indicate that while these scaffolds with chondrocytes perform well initially, their late outcome is disappointing. We propose that for all scaffold-based tissue repairs, a long-term evaluation should be mandatory. The slow degrading scaffolds need further modifications to improve the milieu for long-term growth of chondrocytes and their hyaline phenotype for the better incorporation of tissue-engineered constructs.

Introduction

Articular cartilage is a hyaline tissue of extraordinary mechanical strength, elasticity, and durability, even though it may be only a few millimeters thick. However, its intricate structure, low cellularity, and lack of blood supply impair regeneration following injury [1, 2]. Damaged articular cartilage thus usually does not wholly reconstitute itself—instead, it is replaced with fibrocartilage which has poor durability. Current clinical modalities of treatment, including microfractures, abrasion chondroplasty, mosaicplasty, and autologous chondrocyte transplantation, have not been shown conclusively to maintain lasting healing, although short-term results in some studies have been good [3]. The lack of good long-term results has stimulated research into articular cartilage regeneration to find a way to improve the quality of cartilage repair [3, 4]. A multitude of research studies is being carried out to address this problem.

One significant advance in cartilage tissue engineering has been the use of scaffolds for chondrocyte delivery. These provide a three-dimensional framework for cells to grow while maintaining the cartilage phenotype. They also promote cell adhesion and can provide mechanical strength. Scaffolds include those derived from synthetic polymers, natural polymers, copolymers, and composites [5]. Although a wide array is commercially available, none have proven long-term efficacy, and the quest for the ideal scaffold for cartilage engineering continues [6].

In the past, one of the authors has developed a synthetic 3D porous semi-interpenetrating polymer scaffold made of polyvinyl alcohol and polycaprolactone (PVA–PCL semi-IPN scaffold) [7, 8]. It is shown to be non-toxic and has supported chondrocyte growth and extracellular matrix synthesis [7]. A biphasic scaffold which, in addition to the PVA–PCL semi-IPN as the upper layer, has incorporated bioglass as the lower layer to contact sub-chondral bone when implanted in articular cartilage defects, has also been developed further by her to provide a completely synthetic osteochondral mimic. A similar biphasic scaffold has been shown to hold a promising future in cartilage engineering [9]. In the present study, we have evaluated these polymer scaffolds for the regeneration of osteochondral defects in a rabbit knee model using chondrocyte transplants as the benchmark.

Materials and Methods

This study was conducted after obtaining the approval from the institutional review board, Christian Medical College, Vellore (IRB approval No: 7111), and the experimentation on rabbits was performed after the institutional animal ethics committee (IAEC approval No: 07/2010) clearance. The New Zealand male white rabbits aged 4 months and weighing 2.5–3.5 kg were recruited from a government certified animal husbandry for the study. These were housed in individual cages with standardized access to food and water, and maintained throughout as per the approved institutional protocol. The animals were quarantined for 1 month before experimentation. All animals for sacrifice were euthanized with a high dose of anesthesia. The carcasses were disposed of according to the institutional protocol for safe disposal of biological waste.

Scaffolds

Monophasic: These were 3D porous semi-interpenetrating polymer networks made of polyvinyl alcohol and polycaprolactone (PVA–PCL semi-IPN scaffold). Their cytocompatibility, biodegradability, and suitability for growing chondrocytes and support chondrogenesis have already been studied and published [7].

Biphasic: These were composites of PVA–PCL semi-IPN and bioglass as the lower layer in contact with the sub-chondral bone. An emulsion freeze-drying process was used, and evaluation of the physicochemical properties of this scaffold showed it supported cartilage tissue engineering [7]. A cell cytotoxicity study had also already been done. This bioactive composite scaffold has been shown to favor mineralization when tested in simulated body fluid after 14 days [10].

Study Groups

There were three study groups. Group A comprised rabbits whose cartilage defects were repaired with chondrocytes seeded on the monophasic scaffold. Group B rabbits had their defects repaired with chondrocytes seeded on the biphasic scaffold, and Group C animals were subjected to chondrocyte transplant without a scaffold. The opposite knees in Group A and B were treated with scaffold alone and untreated defect in Group C served as an internal control for the defects treated with cells. Group C acted as the benchmark for assessing the efficacy of the scaffolds used in the other two study groups. We used allogeneic cells as these have been shown in the previous studies to be minimally immunogenic [11]. Allogeneic chondrocyte implantation has been reasonably successful in regenerating articular cartilage according to previously published studies in rabbits [12, 13].

Harvest of the knees was done for one animal in each group after 2 months for assessing short-term healing, and in five animals after 12 months to assess long-term healing. All knees were scored histologically as per the protocol and the results compared at the end of the study.

Chondrocyte Isolation and Culture

Following euthanasia, cartilage was collected from 2–4 month-old rabbits by gentle scraping with a no.15 scalpel from the knee joints under aseptic and moist conditions. The tissues were transferred to the lab in 50 ml centrifuge tubes, each containing 15 ml of Phosphate Buffered Saline (PBS). Isolation of chondrocytes from these specimens was done as per the previously published protocol [14]. Isolated chondrocytes were cultured as a monolayer at a high seeding density of 1 × 10⁴ cells/cm² in a T75 cm² culture flask containing Dulbecco’s Modified Eagle Medium/Hams F12 (DMEM/ F12), 10% fetal bovine serum, and 62 μg/ml ascorbic acid.

Preparation of Chondrocytes for Transplantation

When confluent (6th day), the cultured cells were trypsinized and pelleted by centrifuging at 2400 rpm for 10 min at room temperature. One ml of DMEM/F12 was added to each pellet before transporting to the animal facility at room temperature in sterile sealed microcentrifuge tubes. In the animal facility, the medium was removed and the pellets re-suspended at a concentration of 1 × 10⁶ cells/30 µl of PBS.

Disease Model and Methodology

Every rabbit that was to receive a chondrocyte transplant was given two identically sited cartilage defects, one in each knee. One defect was used for the test (with cells), while the other served as a control (without cells). The test and control knee allotment was quasi-randomized and blinded. An independent basic research scientist who participated in the study, but was blinded for the surgery and the quality of defect, decided the side for test and control. This step was added to avoid the possibility of operator bias during defect creation and implantation.

Defect Creation

The animals were sedated with ketamine 15 mg/kg and anaesthetized with xylazine 0.2 mg/kg. The operation site was shaved and the knee was exposed after preparing and covering the limb with sterile surgical drapes. A medial parapatellar incision was used. The extensor expansion was released, and the patella dislocated laterally. The joint capsule and synovium were opened and the knee flexed to expose the femoral condylar articular surface. A 3 mm-diameter defect of 3 mm depth was made in the medial femoral condyle with a drill. The defect was washed with saline before transplantation (Fig. 1a).

Fig. 1.

In the left image, the defect created in the weight-bearing portion of the medial femoral condyle is seen (a); on the right, cells loaded onto the monophasic scaffold (Scaffold a) have been implanted by press-fit technique into the defect (see arrow) (b)

Transplantation

Scaffold seeded with cells was pressed into the defect, and in the contralateral knee, the scaffold was implanted without loading the chondrocytes (Fig. 2b). Rabbits that received cultured chondrocytes without a scaffold (Group C) had their test defects subsequently sealed with fibrin glue to prevent the escape of transplanted cells. In this group, the opposite defect was left untreated.

Fig. 2.

Gross and microscopic images in the short-term (2 months) follow-up. Left (monophasic scaffold group), Center (biphasic scaffold group), and Right (allogenic chondrocytes group) (Black arrow indicate defect area). Histological images represent the area of regenerated cartilage stained with Safranin O dye (× 4, magnification)

After transplantation, the wounds were closed in layers and covered with sterile dressings. The animals were allowed to ambulate independently as tolerated in the cage environment. Analgesia with meloxicam 5 mg/kg was given for the first postoperative day. Injection cloxacillin 25 mg/kg was given for 3 days after surgery to prevent wound infection. Post-operatively, all wounds were regularly inspected until healing was completed.

Gross Morphologic Evaluation at Sacrifice

All knee joints were removed intact by the surgeon. Preoperative photographs of the defects were taken and served as a record of the gross pathological features. Features noted included the color, smoothness, and luster of reparative changes, the degree to which these filled the defects, and their pattern of transition with surrounding healthy cartilage. The joint margins were examined for the presence of synovitis and osteophytes. These observations helped to correlate the gross and microscopic appearance at the time of histopathological scoring.

Tissue Processing and Histopathological Analysis

The specimens were initially fixed for 48 h in 10% buffered formalin and then decalcified in a 3:1 mixture of 10% buffered formalin and formic acid. After softening, sagittal blocks of tissue incorporating the area of the cartilage defect were cut from each specimen and subjected to standard paraffin embedding procedure [15]. Representative 5 µm sections were prepared and stained with Safranin O-Fast Green. A senior histopathologist independently reviewed an average of four sections per site and the quality of cartilage repair scored according to the O’Driscoll system [16]. The pathologist was blinded to the source of the slide.

Statistics

We used the Statistical Package for Social Sciences (SPSS) software, version 16. Unpaired t-test for comparing the individual mean scores and one-way analysis of variance was used for comparing all the mean scores.

Results

The final analysis included all transplanted animals. No infection or arthrofibrosis was found in the animals’ knees. One animal died at 11 months due to medical problems that were unrelated to the experiment.

In vitro results: Chondrocytes were seeded at a density of 1 × 10⁴ cells/cm² in the culture flask and 8.17 × 10⁶ cells were isolated at harvest. Cells reached confluence on the 6th day of culture. Viability during chondrocytes harvest was 95%.

Gross Morphology

The untreated defect alone and scaffold alone treated defects had irregularly filled defects with undulating surfaces and loss of luster (Fig. 2). In the short-term follow-up in Group A, the limb that received monophasic scaffold + cells, the defect is filled with a yellow tissue; while in limb treated with scaffold alone the defect though well integrated with surrounded tissue, showed pannus invading from the hypertrophic synovium. In Group B, the limb that received biphasic scaffold + cells, the defect was invaded by pannus, whereas the scaffold alone treated limb showed eburnated bone at the margin of the unfilled defect. In Group C, the limb with defect treated with allogeneic chondrocyte was replaced by glistening bulging articular cartilage blending imperceptibly with surrounding native cartilage in contrast to the depressed area in the untreated knee.

In the long-term follow-up, Group A limb that received monophasic scaffold + cells had good surface shininess, no irregularity or depression, smooth transition with surrounding cartilage, and mild inflammation (Fig. 3). While the scaffold alone limbs showed an irregular surface and completely covered with pannus. In Group B, both biphasic scaffold + cells and scaffold alone were filled up with yellowish tissue appearing different from the surrounding native cartilage and moderate pannus. The surfaces of these defects were irregular and their margins were well defined. The Group C defects treated with allogenic chondrocytes were completely replaced by articular cartilage that integrated well with and was similar to surrounding native cartilage. A depressed defect was seen in untreated limbs.

Fig. 3.

Gross and microscopic images in the long-term (12 months) follow-up. Left (monophasic scaffold group), Center (biphasic scaffold group), and Right (allogenic chondrocytes group) (black arrow indicate defect area). Light microscopic images represent the area of regenerated cartilage stained with Safranin O dye (× 4, magnification)

Histopathology

One rabbit from each group sacrificed at 2 months had good O’Driscoll scores in the cell-treated defects, while the corresponding scaffold alone and untreated defects did poorly (Table 1). In Group A: In the transplanted area (monophasic scaffold + cells), thick sub-chondral bone is covered by fibrocartilage, with the deeper layers showing better GAG staining and a more rounded appearance of the chondrocytes (Fig. 2). In scaffold alone treated defects, it was filled with fibrocartilage and absence of GAG staining. In Group B: the transplant (biphasic scaffold + cells) showed loss of cartilage up to the tide mark with necrotic cartilage at the margin. The wall of the scaffold alone treated defect was replaced by fibrous tissue containing masses of persistent scaffold material. In Group C, the cells treated defect showed regenerating cartilage with loss of GAG, surface fibrillation, and sub-chondral sclerosis. The defect in the articular cartilage of the untreated defect was sealed by fibrous tissue.

Table 1.

Short-term results of allogeneic transplants of the three groups. O’Driscoll scores at 2 month follow-up

Group	Scaffold alone/untreated (n = 1)	Scaffold + cells (n = 1)
A. PVA–PCL scaffold	11	22
B. Biphasic scaffold	7	20
C. Cultured chondrocytes only	13	19

(Minimum score = 0 and maximum score = 24)

In the long-term follow-up (12 months), the defects treated with chondrocytes (Group C) showed significantly better overall O’Driscoll scores than those of Group A and B. Most rabbits that received biphasic scaffolds (Group B) had low scores at 12 months (Table 2). Histologically, defects in Group A treated with monophasic scaffold + cells showed normal cartilage at the edges with the defect surface replaced by hyaline cartilage albeit with loss of GAGs. The scaffold alone defects had a band of fibrous tissue on the surface and surrounding the implanted scaffold (Fig. 3). The defects in Group B treated with biphasic scaffold with cells exhibited markedly disorganized regenerating cartilage and bone with patches of reactive sclerosis and pannus. In Group B, defects treated with scaffold alone showed a complete loss of articular cartilage, with the embedded scaffold being covered by bone and a thin layer of fibrous tissue. Both monophasic and biphasic scaffolds were associated with local inflammation and synovitis (Fig. 4). In Group C, allogenic chondrocyte transplantation repaired the defect with a thinner layer of cartilage compared to adjacent normal cartilage; this neocartilage shows decreased safranin O staining and thickening of underlying trabecular bone. Whereas, the untreated defect filled with fibrous tissue had an absence of safranin O staining.

Table 2.

Long-term results of allogeneic transplants of three groups. O’Driscoll scores at 12 month follow-up

Group	Scaffold alone/untreated (n = 5)	Scaffold + cells (n = 5)	Difference between two sides
A. PVA–PCL scaffold	8.8 (± 2.95)	15 (± 4.12)	P = 0.025
B. Biphasic scaffold	7.4 (± 4.8)	11.8 (± 5.9)	P = 0.234
C. Cultured chondrocytes only	13.2 (± 4.3)	20 (± 1.4)	P = 0.011

(Minimum score = 0 and maximum score = 24)

Fig. 4.

The synovium of animal shows inflammatory infiltrates around the remnants of the degraded scaffold. This probably was the cause for deterioration of results in the long-term 12-month follow-up

The one-way analysis of variance (ANOVA) test showed a significant difference between three mean scores of reparative cartilage tissue at 12 months of follow-up (p = 0.03). Upon doing the multiple range test following the ANOVA, there was a statistical difference when the mean score of Group C was compared separately with the scores for the monophasic (Group A) and biphasic scaffolds (Group B) (Table 2). However, the multiple range test revealed no statistical difference between the scores of these two scaffolds. The defects treated with cells showed a significantly higher score in Group A (p = 0.025) and Group C (p = 0.011) compared to the untreated defects.

Discussion

An ideal scaffold for articular cartilage engineering should be mechanically resilient, biocompatible, biodegradable, and supportive of chondrocyte growth. Various materials have been used for cartilage tissue engineering, including hydrogels, polymers, and decellularized tissue materials [5]. Polymers are made mechanically robust during the fabrication stage by modifying the ratio of their constituents. However, biodegradability gets compromised with added strength, and thus, they take a longer time to undergo hydrolysis. Also, techniques to enhance mechanical strength lead to reduced cell adhesion as rigid polymer scaffolds are inferior to hydrogels in supporting cell growth [17]. Scaffold fabrication and its biomimetic properties have been enhanced by adding a fiber network to improve the mechanical characteristics [18]. In the biphasic scaffold used in our study, high porosity, and a combination of hydrophilic and hydrophobic domains were used to enhance cell adhesion and cell–cell interaction [10].

Our study has shown that defects treated with cultured allogeneic chondrocytes without a scaffold heal better than those that receive monophasic or biphasic PVA–PCL semi-IPN scaffolds seeded with cultured cartilage cells. Defects treated with cultured allogeneic chondrocytes served as the positive control, and the opposite knee of untreated defect served as a negative control. Both types of scaffolds had good results at 2 months post-transplant but not in long-term follow-up. Formation of fibrocartilage at 1 year and weak bonding of the structure with native cartilage were noted in both scaffold groups. Both types of scaffold could not form cartilage on par with the positive controls in our study.

The good results in the short term indicate that, at least in the early post-transplant period, the three-dimensional milieu of these scaffolds is conducive to maintenance of chondrocyte phenotype and synthesis of glycosaminoglycans. However, the same could not be replicated at a follow-up of 1 year. By this time, monophasic scaffolds were mostly resorbed from the defect site with residual scaffold material being found only in synovial phagocytes. The synthetic biphasic polymer scaffolds did not show satisfactory incorporation either, with scaffold remnants persisting at the implantation site where they were often associated with a foreign body reaction. This reaction could be due to the presence of two different materials of bioglass and biopolymer PVA–PCL semi-IPN in the same scaffold with differing rates of resorption and tissue reaction. The scaffold remnants likely provoke chronic inflammation and cartilage degradation in an already compromised environment.

Shafiee et al. used a similar PVA/PCL scaffold and showed the regeneration of full-thickness articular cartilage defect in rabbits after 3 months follow-up [19]. The results of our study are at variance with the above study. The difference could be due to the duration of follow-up. Most animal studies have assessed the results of chondrocyte transplantation at 6–24 weeks [20–22]. Though the ICRS recommendations state that 3 months of follow-up in a rabbit cartilage injury model is adequate [23], Martinez Diaz et al. have shown that long-term studies are needed to determine the fate of an implanted scaffold including the time taken for its complete resorption and the quality of the reconstituted cartilage after resorption [24]. One year of a rabbit follow-up is close to a decade of human life [25]. Given the above observations, we assessed our results at 1 year after injury found that the results had degraded. The poor outcome which we see in the scaffold group may be contributed to by the inflammatory response that is seen because of the presence of scaffold degradation products in the long term.

In most other similar studies in rabbits, the articular injury was made in the femoral trochlear groove [13, 20, 26]. However, this region is not exposed to similar forces as the femoral condyles; hence, it does not accurately reflect the actual situation in human injuries. In our study, we made defects in the weight-bearing medial femoral condyle, thereby more closely simulating the real-life situation [27].

Preclinical [13, 28] and human [29, 30] studies have established cultured chondrocyte implantation as a good option in cartilage tissue engineering, and one that provides optimal results. This was apparent in the present study also. The other studies with PVA and PCL [19, 24] have shown comparable results in the short term.

Preoperative seeding 24 h prior to surgery resulted in friability of this scaffold and hence was not practiced. The possible benefits of preoperative seeding, therefore, could not be explored. Besides, our short-term group had only one animal to test the outcome as the main objective was to test the long-term efficacy of the scaffolds.

This study highlights the pitfalls of slow disintegrating scaffolds and bioglass use for regenerating bone in osteochondral defects in the long term. Scaffold research in osteochondral defect treatment is an active field, because the available current treatments provide a palliative care rather than a permanent solution. The complexity lies in the simultaneous regeneration of articular cartilage superficially and sub-chondral bone in depth. Similar to our study, many other studies have tried to investigate biphasic scaffolds that support cartilage and bone regeneration. While the reported outcome was excellent sub-chondral bone and superficial cartilage formation in the short term [31, 32], in the long-term outcomes were dismal because of vascularization and hypertrophic changes in the articular cartilage. In clinic, Shivji et al. confirmed that the biphasic scaffold failed to regenerate the defect and performed poorly in the long term [33]. This underlines the necessity for more research on understanding the scaffold, cartilage, and bone interaction and use of optimal scaffold to leverage the osteochondral repair in the long term.

Conclusions

In our study of indigenously made monophasic and biphasic novel polymer scaffolds in the regeneration of osteochondral defects in a rabbit knee model, PVA–PCL IPN scaffolds with and without a bioglass layer showed poor biodegradability and failed to maintain good quality hyaline cartilage in the long term when compared to allogenic chondrocyte transplants. Measures to improve the biodegradability of these scaffolds have to be explored. We consider that improving the degradation profile of such scaffolds might help in long-term maintenance of the hyaline phenotype of the cartilage. Notably, this study highlights the necessity of testing the scaffold in long-term follow-up before embark on the clinical studies based on short-term findings.

Author contributions

VM: Study design, analysis of data, gross examination of tissue, supervised the study, preparation of manuscript, and reviewed the manuscript. KR: Preparation of chondrocyte for transplantation, animal anesthesia, animal post-operation care, data analysis, and final manuscript preparation for journal communication. VD: Study design, carried out the animal surgery, data analysis, and preparation of manuscript. BB: Carried out the animal surgery and preparation of the manuscript. SKC: Assisted during the animal sacrifice. NW: Histopathology scoring, gross examination, and critically reviewed the manuscript. PDN: Design and developed the scaffold and reviewed the manuscript. All seven authors participated in this study and their contribution was essential to successfully complete this study.

Compliance with ethical standards

Conflict of interest

All the authors declared that there is no conflict of interest.

Ethical statement

This study was conducted after obtaining approval from the institutional review board, Christian Medical College, Vellore (IRB No: 7111), and the experimentation on rabbits were performed after institutional animal ethics committee (IAEC No: 07/2010) clearance.

Informed consent

For this type of study informed consent is not required.