Autologous osteoperiosteal graft shows superior outcomes to bone graft in shelf acetabuloplasty for acetabular dysplasia

Abstract

Aims

Endoscopic shelf acetabuloplasty (ESA) has been reported as an effective approach to prevent the progression of borderline developmental dysplasia of the hip (DDH), typically utilizing autologous iliac bone graft (BG) to address acetabular bone defects. However, BG fails to restore articular cartilage. Autologous osteoperiosteal grafting has shown promising results in repairing osteochondral defects, yet its application in DDH remains unreported. This study aimed to verify the efficacy of autologous osteoperiosteal graft (OPG) in ESA for DDH treatment in a rabbit model.

Methods

Acetabular dysplasia was induced in 30 female rabbits through six weeks of hindlimb cast immobilization. Subsequently, these animals were randomly allocated into three groups: 1) DDH group (no surgical intervention), 2) BG group (shelf acetabuloplasty with autologous bone grafting), and 3) OPG group (shelf acetabuloplasty with osteoperiosteal grafting). Another ten female normal rabbits served as controls. At four and eight weeks postoperatively, graft transformation and acetabular cartilage status were assessed via gross morphology, histology, and quantitative reverse transcriptase polymerase chain reaction (qRT-PCR). Acetabular coverage and subchondral bone quality of the acetabular roof was evaluated by micro-CT, while nanoindentation tests were performed to assess the biomechanical properties of regenerated tissues.

Results

Hyaline-like cartilage formation was found macroscopically and histologically in the OPG group while no cartilaginous tissue was detected in the BG group. qRT-PCR showed that the OPG group expressed more cartilage marker genes than the BG group. OPG-regenerated tissue achieved better biomechanical properties while no difference of improved acetabular coverage and subchondral bone quality was found in both graft groups. Moreover, OPG could alleviate acetabular cartilage degeneration much more significantly than BG macroscopically, histologically, and genetically.

Conclusion

Our study revealed that autologous OPG outperformed conventional BG shelf acetabuloplasty in a rabbit DDH model, suggesting its potential as a superior clinical option for DDH treatment during ESA surgery.

Cite this article: Bone Joint Res 2026;15(6):617–631.

Article focus

• Our study investigated the efficacy of autologous osteoperiosteal graft (OPG) in shelf acetabuloplasty for developmental dysplasia of the hip (DDH) treatment in a rabbit model.

Key messages

• Our findings demonstrate the superior efficacy of autologous OPG over conventional bone graft (BG) in shelf acetabuloplasty using a rabbit model.

• The work provides preclinical evidence supporting the use of autologous OPG in shelf acetabuloplasty, which could influence surgical strategies for young athletes and active patients with borderline acetabular dysplasia.

Strengths and limitations

• This study is the first to employ a rabbit DDH model to evaluate the therapeutic effect of OPG in shelf acetabuloplasty.

• The anatomical differences in acetabular structure and weightbearing patterns between rabbits and humans limit direct clinical translation, underscoring the need for human studies.

Introduction

Developmental dysplasia of the hip (DDH) is characterized by various structural abnormalities in hip development, including malformations of the femoral head and acetabulum, hip subluxation, and dislocation.^1,2 With the adoption of ultrasound screening and early intervention for DDH, the prevalence of DDH in children has been gradually declining.² However, due to regional variations in ultrasound screening and differences in examiner experience, delayed diagnosis of DDH occurs in many countries.³ Missed diagnoses and delayed treatment also contribute to the considerable number of DDH patients among adolescents and young adults. A large population study in China reported an incidence of DDH of 1.52% in adults, with rates of 0.75% in males and 2.07% in females.⁴

Early diagnosis and treatment of DDH are crucial for maintaining normal hip function.^5,6 Currently, periacetabular osteotomy (PAO) serves as the predominant surgical approach for older children and adults with DDH, particularly in moderate to severe cases.^7-9 For mild or borderline hip dysplasia, several surgical options have been proposed, including PAO, shelf acetabuloplasty, and capsular plication.⁷ PAO is a technically complex procedure with high demands, and is associated with significant trauma and complications, including nerve injury, poor acetabular correction, and prolonged rehabilitation.^10-12 In contrast, shelf acetabuloplasty is a relatively simpler procedure that increases femoral head coverage by adding a bony shelf at the outer edge of the acetabulum,^13,14 thereby enhancing the weightbearing area of the hip joint.^15-17 Endoscopic shelf acetabuloplasty (ESA), a minimally invasive procedure, has recently been employed for treating adult DDH, particularly in patients with borderline hip dysplasia, as well as those with concomitant femoroacetabular impingement and labrum injuries.^18-20 While ESA has shown promising clinical outcomes, including pain relief and functional improvement, controversies remain regarding its long-term efficacy.^21,22

Autologous iliac bone is typically used as a graft in ESA surgery,^18,20,22 providing support on the lateral side of the acetabulum. However, it is challenging to form hyaline cartilage locally, leading to uncertain long-term efficacy and a high risk of hip osteoarthritis.^21-23 Consequently, there is an urgent need for a method that can effectively repair bone defects and restore the overlying articular cartilage. Although several treatment methods for osteochondral defects exist,^24-26 an ideal clinical repair strategy is still lacking. Autologous or allogenic osteochondral transplantation is considered an effective option,²⁷ yet it faces challenges such as limited donor availability, secondary injury, and immune rejection. Autologous osteoperiosteal transplantation has proven effective in repairing articular osteochondral lesions.^28,29 Numerous studies have reported its efficacy in treating osteochondral lesions of the talus, with improvements in cartilage damage and joint function postoperatively.^29-32 However, the use of autologous osteoperiosteal graft (OPG) to repair acetabular defects in DDH has not been previously reported. Our study aimed to verify the efficacy of autologous OPG in treating DDH. We innovatively performed the experiment in a rabbit DDH model and hypothesized that autologous OPG would be more effective in repairing acetabular defects in DDH than bone graft (BG).

Methods

Experimental animals and grouping

The experimental animals were provided by Jiangsu Kanion Laboratory (China), and all experimental protocols were approved by the Ethics Committee of Jiangsu Kanion (approval number: 2022072605), and performed according to the ARRIVE guidelines. An ARRIVE checklist is included in the Supplementary Material to show that the ARRIVE guidelines were adhered to in this study. A total of 40 female New Zealand white rabbits aged six weeks (1 to 1.5 kg) were used in our study. The resource equation approach was used to determine the sample size for the study. To achieve the error degree of freedom (DF) between ten and 20, a sample size of five was allocated to each group at each timepoint. As the experimental design included two observation time points (four and eigh weeks), each group ultimately comprised ten rabbits, resulting in a total of 40 rabbits across all four experimental groups. A two-stage randomization was performed for grouping. Ten rabbits were randomly assigned to the non-immobilized control group. The remaining 30 underwent six-week cast immobilization and were subsequently block-randomized after cast removal into three experimental groups: the DDH group (no surgical intervention), the BG group (autologous bone graft for shelf acetabuloplasty), and the OPG group (autologous osteoperiosteal graft for shelf acetabuloplasty), with ten rabbits per group.

Establishment of animal model

Each of the 30 rabbits was fixed in a straight-leg swaddling position in the left leg using a cylinder cast, while the remaining ten rabbits were not fixed. Radiograph examination of the hip joint was performed at two, four, and six weeks after cast immobilization (Figure 1a). The acetabulums of cast-fixed rabbits were compared with those of the normal rabbits, and the centre edge (CE) angle and Sharp’s angle were measured (Figure 1b). Radiological analysis demonstrated that acetabular dysplasia was successfully developed in the rabbits subjected to four weeks’ cast fixation with progressive deterioration observed at six weeks, manifested by femoral head eccentricity and concomitant subluxation. Compared with the normal rabbits, the CE angle decreased significantly, and the Sharp’s angle increased significantly at four and six weeks post-fixation (Figure 1c), demonstrating the successful establishment of the acetabular dysplasia model.

Fig. 1.

Establishment of a rabbit model of acetabular dysplasia. a) Radiograph examinations were performed at two, four, and six weeks to observe hip development in the control and developmental dysplasia of the hip (DDH) groups (control groups: without cast immobilization; DDH groups: the left lower limbs of rabbits were immobilized by cast with knee extension). The femoral head appeared eccentric, and hip congruency was decreased after four weeks of immobilization (yellow arrowhead). Subluxation of the left femoral head occurred after six weeks of immobilization (red arrowhead). b) Measurement of CE and Sharp’s angle on the radiograph images of the control group and DDH group at six weeks. Centre edge (CE) angle is the angle between the vertical line through the centre of the femoral head and the line between the centre of the femoral head to the lateral edge of the acetabulum, and Sharp’s angle is the angle between the line connecting the tip of the bilateral obturator and the line between the tip of the obturator and the lateral edge of the acetabulum. c) Comparison of CE and Sharp’s angles between the DDH and control groups at two, four, and six weeks. Data are presented as histograms using the mean value with SD. N = 10. ***p < 0.001, independent-samples t-test. ns, non-significant; w, weeks.

Surgical procedure

Overall, 20 rabbits in the BG and OPG groups underwent the surgery. Before the procedure, buprenorphine (0.05 mg/kg) was used as a preoperative analgesic. The rabbits were anaesthetized through intramuscular injection of xylazine (5 mg/kg) and ketamine (30 mg/kg), followed by isoflurane inhalation anaesthesia during the surgical procedure. The rabbits were positioned in a lateral decubitus position, and the left hip area was sterilized with iodophor after shaving. A skin incision was made on the iliac crest surface, and the inner and outer muscles of the iliac plate were stripped. A square iliac bone with periosteum and a side length of 5 to 8 mm was cut with an osteotome (Figure 2a) and trimmed into OPG (5 × 5 × 2 mm, retaining the periosteum; Figures 2f and 2g) and BG (5 × 5 × 2 mm, completely removing the periosteum; Figures 2d and 2e). An incision was made posterior to the greater trochanter of the left hip, followed by sequential division of the skin and subcutaneous fascia. A portion of the gluteus maximus muscle was incised, and the underlying gluteus medius muscle was bluntly separated to expose the hip capsule. Different surgical approaches were used depending on the type of the graft. 1) BG group: the dysplastic region of the acetabulum was identified. Following careful removal of the overlying soft-tissue while preserving the integrity of the hip capsule, the graft was precisely positioned outside the capsule. The graft was securely fixed to the superolateral aspect of the acetabulum using two 0.8 mm Kirschner (K-) wires (Yutong Medical, China; Figure 2b). 2) OPG group: the joint capsule overlying the dysplastic area was incised. The graft was positioned intra-articularly and securely anchored to the acetabular rim using two 0.8 mm K-wires (Yutong Medical). Following graft placement, the joint capsule was then tightly sutured to the periphery of the graft (Figure 2c). 3) Subsequent wound closure was achieved in a layered fashion: deep soft tissues were approximated with 4/0 Vicryl (Ethicon, USA) absorbable sutures, while skin closure was completed using 3/0 Mersilk (Ethicon) non-absorbable sutures. Upon full recovery from anaesthesia, the rabbits were administered a subcutaneous dose of meloxicam (0.2 mg/kg), which was repeated once daily for three days. Additionally, to prevent postoperative infection, they received intramuscular injections of amikacin sulfate (1.5 mg/kg) once daily for three days. After the surgery, all animals were kept in separate cages and allowed to move freely with access to a standard diet and water. Wound care was provided with daily follow-up. Rabbits with implant infection and significant displacement that failed to heal were excluded; the remaining animals were euthanized at four and eight weeks post-surgery, and their hip joints were harvested for subsequent analysis.

Fig. 2.

Shelf acetabuloplasty procedure in the rabbit acetabular dysplasia model. a) Autologous iliac bone was harvested (yellow circle). A square iliac bone with periosteum and a side length of 5 to 8 mm was cut with an osteotome. b) Transplantation of autologous bone graft (BG) (yellow circle). The autologous BG was positioned outside the capsule and fixed to the superolateral aspect of the acetabulum using 0.8 mm Kirschner wires. c) Transplantation of autologous osteoperiosteal graft (OPG) (yellow circle). The autologous OPG was positioned intra-articularly and securely anchored to the acetabular rim using 0.8 mm Kirschner wires. Subsequently, the joint capsule was sutured. d) and e) BG, 5 × 5 × 2 mm, completely removing the periosteum. f) and g) OPG, 5 × 5 × 2 mm, retaining the periosteum.

Gross morphological observation

Following euthanasia, acetabular specimens were harvested for macroscopic assessment of morphological features and articular cartilage integrity. Articular cartilage degeneration was evaluated and scored by three independent blinded professionals using the Osteoarthritis Research Society International (OARSI) macroscopic assessment scale (Supplementary Table i).³³

Micro-CT examination

Rabbit acetabula were dissected to remove soft tissues and fixed in 10% paraformaldehyde solution for 48 hours. High-resolution micro-CT (SkyScan 1176; Bruker, USA) was used to scan the bone structure with a resolution of 18 μm. A cubic area of 5 × 10 × 5 mm containing the acetabular roof and the adjacent graft at the superolateral margin of the acetabulum were selected as the region of interest. CTAn software (V1.23; Bruker) was used for bone structure analysis, and CTVox (V3.3; Bruker) was used for 3D reconstruction. Bone volume (BV) and bone volume fraction (BV/TV) were measured in the area of interest, and the characteristics of the subchondral bone trabeculae in the acetabular roof were also analyzed.

Histological evaluation

After micro-CT examination, acetabular sample blocks were decalcified in 10% ethylenediamine tetraacetic acid (EDTA) for approximately 30 days. The graft area and acetabular roof were sectioned and embedded in paraffin. The trimmed wax blocks were cut into 5 μm thick sections, and haematoxylin and eosin (HE) and Safranin O-fast green staining were performed. Tissue morphology was observed under a microscope, and images were captured and analyzed. The Mankin scale was used to evaluate the degenerative changes in the acetabular roof cartilage (Supplementary Table ii). The evaluations were performed by three independent professionals with good experience of cartilage histology and blinded to the experiment.

Immunohistochemical staining

Immunohistochemical staining was performed to evaluate the status of the regenerated graft tissues and cartilage in the acetabular roof. Sections were deparaffinized in xylene and rehydrated through a graded ethanol series. Antigen retrieval was performed using citrate buffer to expose epitopes. Endogenous peroxidase activity was quenched by incubating sections with 3% hydrogen peroxide for 15 minutes at room temperature. To prevent nonspecific binding, sections were circumscribed with a hydrophobic barrier (Dako pen; Agilent technologies, USA) and blocked with 3% bovine serum albumin (BSA) for 30 minutes at room temperature. Primary antibodies targeting type 2 collagen A1 (COL-2A1; SCbio, USA), platelet-derived growth factor receptor α (PDGFR-α; Abcam, UK), matrix metalloproteinase 13 (MMP-13; SCbio), and type 10 collagen (COL-10A1; Abcam) were applied and incubated overnight at 4°C in a humidified chamber. Following three five-minute washes with phosphate-buffered saline (PBS), sections were incubated with species-matched horseradish peroxidase (HRP)-conjugated secondary antibodies for one hour at room temperature. Immunoreactivity was visualized using 3,3'-diaminobenzidine (DAB) substrate, with reaction times optimized for each antibody under microscopic monitoring. Following counterstaining with haematoxylin for ten seconds, the slides were dehydrated through a graded ethanol series (70%, 80%, 95%, and 100% ethanol), cleared in xylene, and mounted with a coverslip for microscopic examination. Findings were quantitatively analyzed using ImageJ software (v.1.53; National Institutes of Health, USA).

Real-time fluorescence qRT-PCR

Acetabulum tissue samples were collected four weeks post-surgery. Total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, USA), and messenger RNA (mRNA) was reverse transcribed into complementary DNA (cDNA). Polymerase chain reaction (PCR) was performed using a universal SYBR Green premix (Thermo Fisher Scientific) and CFX96 real-time PCR system (Bio-Rad, USA). The mRNA expression levels of aggrecan (ACAN), COL-2A1, PDGFR-α, MMP-13, a disintegrin and metalloproteinase metallopeptidase with thrombospondin type 1 motif 5 (ADAMTS-5), and COL-10A1 were detected. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the reference gene to calculate the relative mRNA expression levels of each target gene. Primer sequences are shown in supplementary materials (Supplementary Table iii).

Biomechanical tests

Nanoindentation tests were performed on the newly formed articular surface tissues in the BG and OPG groups using a nanoindentation instrument (NHT²; Anton Paar, Austria). Samples were fixed on the table, and the maximum load was set at 15 mN, with a loading and unloading rate of 30 mN/min and a load holding time of ten seconds. Five points were measured for each sample. The load-displacement curve was drawn, and the reduced modulus of the tissues were calculated.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 9.5 software (GraphPad Software, USA). For comparisons between two independent groups, the independent-samples t-test was employed. When analyzing data from three or more groups, we first verified the assumptions of normality and equal variance. For datasets satisfying these assumptions, one-way analysis of variance (ANOVA) was conducted followed by Tukey’s post-hoc test for multiple comparisons. In cases where homogeneity of variance was not met, we applied Brown-Forsythe and Welch ANOVA tests. All quantified data are presented as mean (SD). Differences with a p-value < 0.05 were considered statistically significant.

Results

Two rabbits were excluded from the experiment due to postoperative complications: one in the OPG group developed an implant-associated infection, and one in the BG group experienced complete graft displacement secondary to internal fixation failure.

Macroscopic evaluation of graft transformation and acetabular degeneration

At four and eight weeks postoperatively, the dysplastic area of the acetabulums in the OPG and BG groups showed significant improvement in macroscopic appearance compared to the DDH group (Figure 3a). In the DDH group, a pronounced acetabular defect was observed predominantly in the superolateral region of the acetabulum. In contrast, both the BG and OPG groups displayed complete acetabulum coverage restoration, with the grafts showing seamless integration and precise alignment with the adjacent host tissue.

Fig. 3.

Macroscopic evaluation of acetabulum at four and eight weeks after surgery. a) Gross representative images of the acetabulum of the control, developmental dysplasia of the hip (DDH), bone graft (BG), and osteoperiosteal graft (OPG) groups. White arrows indicate the acetabular defect, yellow arrows indicate the graft, and red triangles indicate the acetabular roof. b) Evaluation of acetabular roof cartilage degeneration according to the Osteoarthritis Research Society International macroscopic scoring system. Data are presented as mean (SD). n = 4 (four weeks); n = 5 (eight weeks). *p < 0.05; **p < 0.01; ***p < 0.001. Il, ilium; Is, ischium; ns, not significant; w, weeks.

Regarding the newly formed articular surface derived from the graft, the OPG group exhibited a smooth morphology with no discernible boundary between the regenerated and native cartilage. Furthermore, the OPG group demonstrated near normal appearance as compared to the adjacent host cartilage, while in the BG group, the capsule tissues beneath the BG still exhibited fibrous tissue appearance.

To further investigate whether shelf acetabuloplasty using different grafts can reduce acetabular degeneration, we performed gross observation of the acetabular roof (Figure 3a). At four and eight weeks postoperatively, the cartilage of the acetabular roof in the DDH group was eroded. However, the acetabular roof in the BG and OPG groups were similar to those in the control group macroscopically, with little observed cartilage degeneration. At each postoperative timepoint, the OARSI macroscopic scores related to cartilage damage were significantly lower in the BG and OPG groups than those in the DDH group (Figure 3b). At four weeks postoperatively, there were no significant difference between the OPG, BG, and control groups. At eight weeks postoperatively, the OARSI scores were slightly lower in the OPG group than those in the BG group, while no significant difference was observed between the OPG and control groups.

Histological evaluation of graft transformation and acetabular roof degeneration

To further evaluate the transformation capacities of different grafts, we observed the regenerated tissues in the graft region of the BG and OPG groups using HE and Safranin O-fast green staining at four and eight weeks after surgery (Figure 4a). In the BG group, the articular capsule beneath the graft remained without detachment. We found that it was hypocellular and mainly contained fibrous connective tissue without signs of cartilage transformation, as Safranin O staining revealed little red staining, suggesting an absence of cartilage proteoglycan matrix formation. In the OPG group, the periosteum on the graft surface was gradually differentiated into vital red staining tissue, suggesting abundant proteoglycan formation which closely resembled the normal hyaline cartilage.

Fig. 4.

Histological evaluation of the graft and acetabular roof at four and eight weeks postoperatively. a) Histological evaluation of the graft area by haematoxylin and eosin (HE) and Safranin O-Fast Green (SO) staining. The black arrows indicate the graft sites, and the yellow arrows indicate the regenerated articular surface tissues. Notably, the regenerated tissues originate from the joint capsule in the bone graft (BG) group, whereas they are derived from the periosteum in the osteoperiosteal graft (OPG) group. b) and c) HE and SO staining of the acetabular roof in the developmental dysplasia of the hip (DDH), BG, OPG, and control groups at four and eight weeks postoperatively. The image below is a magnified view of the red box in the image above. d) Evaluation of cartilage degeneration of the acetabular roof in the DDH, BG, OPG, and control groups according to the Mankin scale. Red scale bar: 1 cm; black scale bar: 400 µm. Data are presented as mean (SD). N = 4. *p < 0.05; **p < 0.01; ***p < 0.001. ns, not significant; w, weeks.

We further explored whether using BG or OPG in the shelf acetabuloplasty procedure could alleviate cartilage degeneration of the acetabular roof. The sections of the acetabular roof from the DDH, BG, OPG and control groups were examined using HE, and Safranin-O staining (Figures 4b and 4c). In the DDH group, the chondrocytes were arranged in a disorganized manner in the acetabular roof, and the cellular density and the cartilage proteoglycan matrix staining intensity were both significantly reduced compared to those in the other three groups. In contrast, both the BG and OPG groups demonstrated higher chondrocyte density in the acetabular roof, with more regular cellular organization and preserved proteoglycan staining, closely resembling those of the normal control group. The Mankin score evaluation of acetabular roof cartilage in the different treatment groups demonstrated significant improvements were obtained in both the OPG and BG groups as compared to the DDH group. However, no statistically significant difference was observed between the BG and OPG groups (Figure 4d). These findings suggested that both BG and OPG treatments could achieve comparable efficacy in mitigating cartilage degeneration of the acetabular roof in the DDH animals.

Immunohistochemical evaluation of graft transformation and acetabular roof degeneration

To further explore the histological difference of the graft tissues between the BG and OPG groups, we conducted immunohistochemical staining for PDGFR-α and COL-2A1 on sections from the grafts in the BG and OPG groups (Figure 5a). Similar to the control group, the OPG graft exhibited more COL-2A1 staining-positive tissue formation. In contrast, the joint capsule beneath the BG graft showed negative COL-2A1 staining but intensive staining for PDGFR-α, a characteristic marker of fibrous tissue. Further quantitative analysis demonstrated that COL-2A1 staining intensity was similar between the OPG and control groups, but which were markedly stronger than that in the BG group. Conversely, PDGFR-α staining was significantly elevated in the BG group when compared to both the OPG and control groups (Figure 5c). These results indicated that the OPG-grafted tissue could transform into a tissue with similar histological features as hyaline cartilage, whereas the capsular tissue beneath the BG graft showed no sign of cartilage transformation.

Fig. 5.

Immunohistochemical evaluation of the graft area and acetabular roof at eight weeks after surgery. a) Representative immunohistochemical staining images of platelet‑derived growth factor receptor alpha (PDGFR-α) and type 2 collagen A1 (COL-2A1) expression in the graft area in the control, bone graft (BG), and osteoperiosteal graft (OPG) groups. The red triangles indicate the grafts, the yellow arrows indicate the regenerated tissues originated from the joint capsule in the BG group, and the black arrows indicate the regenerated tissues derived from the periosteum in the OPG group. Scale bar: 1 mm. b) Representative immunohistochemical staining images of type 10 collagen A1 (COL-10A1), matrix metalloproteinase 13 (MMP-13), and COL-2A1 expression in the acetabular roof cartilage in the control, developmental dysplasia of the hip (DDH), BG, and OPG groups. Scale bar: 400 µm. c) Quantitative analyses of PDGFR-α and COL-2A1 staining positive cells in the graft area. d) Quantitative analyses of COL-2A1, COL-10A1, and MMP-13 staining positive cells in the acetabular roof cartilage. Data are presented as mean (SD). n = 4. *p < 0.05, post-hoc test; **p < 0.01, post-hoc test; ***p < 0.001, post-hoc test. ns, not significant.

To further assess the treatment effect of BG and OPG shelf acetabuloplasty on the acetabular roof cartilage degeneration after DDH, immunohistochemical analysis was performed on the DDH, BG, OPG, and control groups at eight weeks postoperatively (Figure 5b). The results revealed significant upregulation of MMP-13 and COL-10A1 expressions, which were the markers of cartilage degeneration and hypertrophy in the acetabular roof of the DDH group as compared to the control group, accompanied by marked downregulation of COL-2A1 expression, suggesting pronounced cartilage degradation and hypertrophy following DDH. In contrast, the OPG group exhibited significantly lower expressions of MMP-13 and COL-10A1 but higher COL-2 expression in the acetabular roof compared to the DDH group (Figure 5d). Similarly, the BG group exhibited reduced MMP-13 and elevated COL-2A1 levels relative to the DDH group, though no significant difference was observed in COL-10A1 expression between these two groups. These findings demonstrated that DDH is predisposed to acetabular roof cartilage degeneration, whereas the application of BG or OPG graft could effectively mitigate this degenerative process.

Micro-CT analysis on the bony coverage and subchondral bone quality of the acetabular roof

Significant bone loss and traumatic osteoarthritis often occur at the late stage of DDH, manifested as lower BV and inferior quality of the subchondral bone at the acetabular roof weight-bearing area. In order to investigate whether these adverse events could be decreased by BG and OPG shelf acetabuloplasty, micro-CT examination and 3D reconstruction were applied to analyze the acetabulum BV and subchondral trabecular bone microstructure (Figures 6a and 6c). At four and eight weeks postoperatively, the rabbits were killed and hip tissues were harvested. The acetabulum superolateral area (10 mm × 5 mm × 5 mm) was chosen as the region of interest for micro-CT analysis. Results indicated that the indices of BV and total BV/TV in the region of interest were significantly decreased in the DDH group than those of the other three groups (Figure 6b). Notably, while the indices were similar between BG and OPG groups, both showed significantly higher values compared to the DDH group at four and eight weeks postoperatively, demonstrating the efficacy of both shelf acetabuloplasty techniques in addressing acetabular bone deficiency in DDH. Similar results were obtained by 3D CT reconstruction, which revealed a large acetabular defect in the DDH group, compared with well-filled defects in the BG and OPG groups after grafting.s (Figure 6a). These results indicated that both BG and OPG shelf acetabuloplasty could restore acetabular coverage in the treatment of DDH.

Fig. 6.

Micro-CT scanning and quantitative analysis to evaluate changes in the bone volume (BV) and subchondral trabecular bone microstructure at four and eight weeks after surgery. a) 3D CT images of the acetabulums from the developmental dysplasia of the hip (DDH), bone graft (BG), and osteoperiosteal graft (OPG) groups at four and eight weeks postoperatively. Yellow cuboid (length 10 mm, width 5 mm, depth 5 mm) containing the dysplastic area and the graft in the BG and OPG groups was defined as the region of interest. b) Quantitative analyses of bone volume (BV) and total bone volume fraction (BV/TV) in the DDH, BG, and OPG groups. c) Coronal plane micro-CT images of the acetabulum from the control, DDH, BG, and OPG groups. Yellow circle indicates the subchondral bone of the acetabular roof. d) Quantitative analyses of the subchondral trabecular bone microstructure in the control, DDH, BG, and OPG groups. Data are presented as mean (SD). N = 4. *p < 0.05, post-hoc test; **p < 0.01, post-hoc test; ***p < 0.001, post-hoc test. ns, not significant; Tb.N, bone trabecular number; Tb.Sp, bone trabecular space; Tb.Th, bone trabecular thickness.

To further elucidate the impact of shelf acetabuloplasty on the bone quality in the weightbearing region of the acetabular roof, we analyzed the variations of the subchondral bone trabecular microstructure between the different treatment groups (Figure 6c). Our findings revealed a reduction in both the number and thickness of the subchondral trabecular bone in the DDH group. At four and eight weeks after shelf acetabuloplasty, significant enhancements in the number and thickness of the subchondral trabecular bone were achieved in the BG and OPG groups, but no notable differences of these indices were observed among the BG, OPG, and control groups. The trabecular spacing index in both the BG and OPG groups showed a significant decrease compared to the DDH group at eight weeks postoperatively. Interestingly, the trabecular spacing in the BG and OPG groups was also notably smaller than that in the control group at the same timepoint, likely due to new bone formation within the grafts (Figure 6d). These results underscore the efficacy of BG and OPG shelf acetabuloplasty on restoring acetabular bone defect and inferior subchondral bone quality of the acetabular roof following DDH.

Gene level evaluation of graft transformation and acetabular roof degeneration

In order to investigate the chondrogenic differentiation capacity of the different grafts at the gene level, tissues from the grafts of both the BG and OPG groups were harvested at four weeks postoperatively. The expression levels of ACAN, COL-2A1, and PDGFR-α were analyzed by quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) with the values of these genes in the normal acetabular cartilage serving as the control groups (Figure 7a). The results revealed that the expression levels of the chondrogenesis-related genes, COL-2A1 and ACAN, were significantly higher in the OPG group than those in the BG group, although these values were still lower than those in the control groups. Conversely, the expression level of fibroblast-related gene PDGFR-α was higher in the BG group than that in both the OPG and control groups. These findings indicated that the OPG graft could achieve a chondrogenic differentiation profile more closely resembling the native cartilage than the BG graft genetically.

Fig. 7.

The expressions of genes involved in graft regeneration and acetabular roof degeneration were detected by quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) at four weeks after surgery. a) Relative expressions of type 2 collagen (COL-2A1), aggrecan (ACAN), and platelet‑derived growth factor receptor α (PDGFR-α) in the graft area of the control, bone graft (BG), and osteoperiosteal graft (OPG) groups. b) Relative expressions of the cartilage degeneration-related genes type 10 collagen (COL-10A1), matrix metalloproteinase 13 (MMP-13), and a disintegrin and metalloproteinase with thrombospondin type 1 motif 5 (ADAMTS-5) in the cartilage of the acetabular roof. N = 4. *p < 0.05, post-hoc test; **p < 0.01, post-hoc test. mRNA, messenger RNA; ns, not significant.

To investigate whether the BG or OPG graft could mitigate hip joint degeneration secondary to DDH at the gene level, tissues from the acetabular roof were collected and detected by qRT-PCR for the expression of cartilage degeneration-related genes COL-10A1, MMP-13, and ADAMTS-5 at four weeks postoperatively (Figure 7b). Of the four groups, we found that the highest expression of COL-10A1, MMP-13, and ADAMTS-5 in the DDH group, indicating that DDH could lead to significant cartilage degeneration if left untreated. As anticipated, the expressions of these genes were significantly decreased in the BG and OPG groups than those in the DDH group, indicating that either BG or OPG could mitigate hip cartilage degeneration following DDH.

Biomechanical evaluation of the grafted tissues

Biomechanical property serves as a critical indicator for assessing the quality of regenerated tissues. Thus, the tissues from the grafted area were harvested from both the BG and OPG groups and nano-indentation testing was performed at eight weeks postoperatively with the value of normal acetabular cartilage serving as the control result. Analysis of the load-displacement curves revealed that the curve of the OPG group closely approximated that of the control group, whereas the BG group displayed a lower and flatten curve with greater indentation depth even under equivalent loading conditions (Figure 8a). The quantitative assessment of the reduced elastic modulus demonstrated that articular regenerated tissues in the OPG group exhibited significantly higher mean values (10.65 MPa (SD 0.99)) than those in the BG group (7.99 MPa (SD 0.86)). Additionally, no statistically significant difference was observed between the OPG and control groups (Figure 8b). These findings collectively suggested that the regenerated tissues derived from the OPG graft could possess a similar mechanical strength as the healthy articular cartilage, and that the biomechanical strengthen of the regenerated graft tissue in the OPG group was superior to that in the BG group.

Fig. 8.

Biomechanical properties of the regenerated articular surface tissues at eight weeks after surgery. a) Comparative analysis of the load-displacement curves obtained from the osteoperiosteal graft (OPG), bone graft (BG), and control groups. b) Quantitative analysis of the reduced elastic modulus of articular regenerated tissues derived from BG and OPG, with the value of the normal hip joint cartilage as the control by the nanoindentation testing system. The data are expressed as mean (SD). N = 5 random testing positions for each group. ***p < 0.001, post-hoc test. ns, not significant.

Discussion

This study demonstrated that autologous OPG enhanced shelf acetabuloplasty is a promising approach for treating DDH in a rabbit model. The tissue regenerated from OPG was hyaline-like cartilage, and the OPG graft could restore the coverage of femoral head, improve the quality of the subchondral bone, and reduce cartilage degeneration in the acetabular roof.

Our study is the first to utilize a rabbit DDH model to verify the effect of OPG in the shelf acetabuloplasty procedure. We employed a method of straight-leg swaddling to induce the DDH model in rabbits, consistent with previously reported methods.^34,35 Unlike other studies that selected one-month-old rabbits, we chose six-week-old rabbits initially for six weeks of cast immobilization and thereafter for shelf acetabuloplasty. The primary concern was that one-month-old rabbits were too young to tolerate six weeks of immobilization, while three-month-old rabbits could have better surgical tolerance and a lower risk of death associated to surgical trauma. In humans, the female sex is a well-established high-risk factor for DDH.³⁶ While there is no evidence suggesting that male animals are unsuitable for studies on DDH, the selection of females more closely reflects the sex-specific distribution pattern observed in human DDH, thereby facilitating a more accurate modelling of its pathogenesis and pathological progression.

Currently reported shelf acetabuloplasty procedures typically involve harvesting an autologous BG, implanting it at the osteochondral defect site on the acetabular rim (usually anterosuperior), and placing it outside the joint capsule.^20-23 In our technique of autologous OPG for shelf acetabuloplasty, the graft is similarly placed at the osteochondral defect of the acetabular rim. Unlike the BG, however, the OPG is positioned inside the hip joint capsule, in closer proximity to the femoral head, with the periosteal surface oriented toward the joint cavity and femoral head. This positioning facilitates chondrogenic differentiation of the periosteum, ultimately contributing to the formation of a new acetabular articular surface. A primary intraoperative challenge relates to the management of the acetabular labrum. Our current approach involves preserving the labrum without transection, with the graft placed directly over it. This preserves labral integrity and maintains joint seal. However, this method may leave a slight gap between the graft and the femoral head, which could have a mild impact on joint stability and periosteal chondrogenesis. An alternative method involves resecting the labrum at the graft site, placing the graft directly against the acetabular rim, and subsequently reconstructing the labral tissue over the outer edge of the graft. This technique provides better continuity with the acetabular articular surface and closer contact with the femoral head. However, it is technically more complex, and the functional outcome of the reconstructed labrum remains uncertain. Given the relatively small acetabulum and thin labrum in rabbits, performing labral reconstruction is particularly challenging. Consequently, we did not adopt this approach. We propose that labral reconstruction would be more appropriately studied in a large animal model.³⁷

Osteoarthritis frequently develops in the late stage of DDH, serving as a primary cause of pain and joint dysfunction,^38,39 and ultimately constitutes the main indication for total hip arthroplasty.⁴⁰ Cartilage degeneration and subchondral bone structural abnormalities are the main pathological changes in hip osteoarthritis.^41,42 In our study, we demonstrated cartilage erosion in the acetabular roof of the DDH group at macroscopic, histological, and genetic levels, and structural abnormalities in the underlying subchondral trabecular bone through micro-CT analysis. We further found that cartilage degeneration in the acetabular roof could be alleviated, and the structure of the subchondral trabecular bone could be strengthened after BG and OPG transplantation. These results demonstrated again that either OPG or BG shelf acetabuloplasty could effectively alleviate cartilage degeneration and restore the normal subchondral trabecular bone structure in the acetabular roof.

Bone grafting is frequently used to treat bone defects, but bone lacks the ability to differentiate into cartilage. The periosteum is a highly vascularized thin layer of tissue covering the bone surface, containing a large number of bone stem cells with excellent osteogenic differenciation capacity.^43,44 Numerous studies have also found that the periosteum can differentiate into cartilage.^45-47 As expected, our histological studies confirmed that the periosteum on the OPG surface could differentiate into large quantities of Safranin O staining-positive tissues (proteoglycan) and hyaline cartilage matrixes type II collagen while the articular capsule underneath BG remained Safranin O negative staining at four and eight weeks postoperatively together with high expression of PDGFR-α. Furthermore, our biomechanical and genetic studies confirmed that the tissues regenerated from OPG presented highly similar characteristics to hyaline cartilage than those from BG. In summary, OPG was superior to BG when applied in shelf acetabuloplasty histologically, biomechanically, and genetically.

Apart from the presence of periosteum on the graft surface, another key distinction between OPG and BG lies in their implantation sites. The OPG is implanted intra-articularly, where the hypoxic environment of the joint cavity and the contact stress from the articular surface are more conducive to promoting articular cartilage formation.⁴⁸ In contrast, the BG is placed extra-articularly and relies on fibrocartilaginous metaplasia of the joint capsule to augment the weightbearing area. However, the abundance of surrounding soft tissue predisposes the graft to greater absorption, while the predominantly fibrous nature of the joint capsule limits its potential for fibrocartilaginous transformation – a finding that has been substantiated in our study.

Our study represents a preclinical investigation using a rabbit model with a relatively short-term follow-up of eight weeks. It primarily demonstrates the superiority of OPG over BG in shelf acetabuloplasty at the level of graft tissue differentiation. These findings indicate the potential translational value of autologous OPG for clinical application in ESA surgery.

Our in vivo study in the rabbit model has several limitations. First, the observation period was short, and the sample size was small. Second, the rabbit model of acetabular dysplasia induced by cast immobilization was established in six-week-old rabbits, thus, the model may not fully reflect the pathophysiological process of acetabular dysplasia after birth. Third, the acetabular labrum, an important structure for hip stability, was not included in our study due to the differences in acetabular anatomical structure and weightbearing patterns between rabbits and humans.

In conclusion, our study demonstrated the advantages of autologous OPG over traditional BG shelf acetabuloplasty for treating DDH in a rabbit model of acetabular dysplasia. Although both OPG and BG shelf acetabuloplasty could to some degree ameliorate cartilage degeneration and restore the normal subchondral trabecular bone structure in the acetabular roof, OPG could more effectively alleviate hip degeneration by restoration of hyaline-like cartilage through periosteum oriented chondrogenic regeneration in the weight-bearing area of the acetabulum. Our study provides a reliable basis for further clinical application of autologous OPG in ESA surgery.

Data Availability

The data that support the findings for this study are available to other researchers from the corresponding author upon reasonable request.