Masquelet Technique: Effects of Topical Mechanical Stability on the Formation of Masquelet Membrane in a Rabbit Radial Defect Model


 Purpose The Masquelet technique is a commonly treatment strategy for segmental bone defects. Still, the exact generating mechanism is unknown. This study intends to explore the effects of topical mechanical stability on the formation of Masquelet membrane. Methods Thirty-six New Zealand white rabbits were evenly randomized into three groups. Segmental left radius shaft defect (length, 9 mm) was created in all rabbits. Bone defects were filled with polymethylmethacrylate (PMMA) in Non-fixation group, and with PMMA fixed with plates in Fixation group, and subjected to no disposal in control group. The stability of PMMA and plates were monitored via X-ray. And the membranes were excised for further Histological, IHC (Ki67, COL I, and CD31), and Western-Blotting analysis (RUNX2, ALP, VEGFA, and TGF-β1), 4 and 6 weeks post-operatively Results X-ray revealed no sign of plates loosening, or shift of PMMA. Pathological examinations suggested that vascularized and osteogenic membranes were formed around PMMA. IHC and Western-Blotting analysis revealed that both Fixation and Non-fixation group exerted significant effects on the expression of Ki67, COL I, and CD31 positive cells, as well as the protein expression of osteogenic (RUNX2, ALP) and angiogenic (VEGFA, TGF-β1) factors. And compared with membrane in Non-fixation group, Fixing PMMA spacer with plates caused a significant increase in osteogenic and angiogenic expression, 4 and 6 weeks post-operatively. Conclusion The present study indicates that rigid fixation provided by internal plate in Masquelet technique positively alters the quality of membrane formed surrounding PMMA, in terms of significantly osteogenic and angiogenic potential.


Introduction
Segmental bone defects, caused by trauma, osteomyelitis, and tumor resection, has become a major health problem obfuscating orthopedist (1)(2)(3). One strategy of growing concern in bone defect repairing is the Masquelet or induced membrane technique (4)(5)(6). Mixed results are shown in previous animal or human studies (1,3,7,8). The key to bone repair relies on a vascularized and osteoblastic membrane formed during the first stage of Masquelet technique (9,10). Still, the exact generating and regulatory mechanism of Masquelet membrane are unknown. In order to solve these problems, several studies show that the membrane has structural characteristics and osteogenic potentials. It is not just offering a physical barrier and a rich vascular network, but also numerous osteoprogenitor cells and a local source of key biochemical factors, such as BMP2, and TGF-β (11)(12)(13). And previous researches have demonstrated that selecting a new model, such as spacer alternatives, micro-topography modification, local antibiotic usage, would affect the structure and function of membrane (13)(14)(15)(16)(17)(18).
During the first phase of Masquelet, the bone defects filling with bone cement (polymethyl methacrylate, PMMA) spacer are fixed with internal plate, intramedullary nail, external fixator, casting, or nothing (12,17,(19)(20)(21). But it is not sure that which fixation method is more appropriate for membrane formation in the first stage of Masquelet (13,(22)(23)(24). It is known that the stiffness of fixation construct directly affects the jiggle of fracture fragments, subsequently impacts the healing of bone fracture. A relatively high level of rigid fixation results in intramembranous ossification, while less rigid fixation results in cartilaginous callus and endochonrdral ossification (25,26). And there is no literature reported regarding whether there are differences in the membrane formation among various fixation 5 pattern.
This study intends to explore the effects of topical mechanical stability on the formation of Masquelet membrane. Based on previous studies, we hypothesize that an increase in fixation strength will results in superior Masquelet membrane with significance in osteogenic and angiogenic factors expression.

Animal models
All procedures were done with the approval of our Medical Ethics Committee. Thirtysix New Zealand white rabbits, weighing approximately 2500-3000 g and aged 24-28 weeks, were evenly randomized into three groups (Fixation group, Non-fixation group, and control group). Segmental left radius shaft defect (length, 9 mm) was created in all rabbits. And the bone defects were filled with PMMA (length, 9 mm) only (Non-fixation group, n = 12), or PMMA fixed with plates (Synthes, USA) (Fixation group, n = 12), respectively. In the control group, the defects were subjected to no disposal (control group, n = 12).
Xylazine hydrochloride was given to rabbits by intramuscular injection, and Iidocaine was used for additional regional anesthesia before osteotomy. All rabbits were arranged on fixed frames, and the left forelimbs were shaved, prepared, and draped before surgery. After the skin, superficial fascia and muscles of the left forelimb were incised, exposing the shaft of radius. A 9 mm defect was created at approximately middle radius shaft with a high-speed power drill, and subsequently a cylindrical PMMA spacer (3 mmØ×9 mm) was placed in the defect in Fixation and Non-fixation group. A six-hole, 2.5 mm stainless steel plates (Synthes, USA) was applied to the radius shaft and secured in place with two proximal and two distal Immediately, 4 and 6 weeks post-operatively, X-ray was used to evaluate the location of implants. Half of animals were sacrificed at 4 weeks after operation, and the other half were sacrificed at 6 weeks through air embolism. Through the former incision, the skin and muscles were dissected. Membranes formed around PMMA were excised completely. The membranes were divided into two parts, which were respectively fixed in 10% paraformaldehyde for later pathological examination and shock frozen (− 80 °C) for later Western-Blotting analysis. (Fig. 1E-F)

Radiography
The stability of PMMA and plates (i.e. PMMA migration and plates loosening) were monitored via X-ray in all rabbits immediately, 4 and 6 weeks post-operatively.
Subsequently, the specimens were stained with haematoxylin/eosin (HE) and Alizarin red S to evaluate the histopathological characteristics.
Then, the processes of antigen retrieval, incubation of primary antibody (CD31, Ki67, and COL I), incubation of secondary antibody (anti-mouse, rabbit, mouse-IgG antibody-HRP polymer) were in progress. Finally, after DAB coloration, hematoxylin redyeing, and Alcohol dehydration, IHC sections were imaged under microscope (DMIL-LED, LEICA, Germany). An independent observer blinded to the group constituency analysed the images.

Western blot analysis
Osteogenic proteins (RUNX2 and ALP) and angiogenic proteins (VEGF and TGF-β1) were measured using Western blot analysis in 4 and 6 weeks postoperatively.

Statistical analysis
All data were reported as means with standard deviations, and analyzed by SPSS 13.0 software (SPSS Inc, Chicago, IL, USA). Differences were considered as significant at P ≤ 0.05. For normally distributed data, Student t test or one-way analysis of variance was used to compare differences between 2 different groups or among more than 2 groups. Results 8 X-ray measurement Immediately after surgery, as presented in Fig. 2A-C, segmental left radius shaft defects were fixed with PMMA and plates, PMMA only, or nothing, suggesting that these models were successful. Both plates and PMMA were radiopaque so that the bone defects and the implant materials were clearly visible. Compared with immediate and 4 weeks postoperative radiographs ( Fig. 2A-F), there was no sign of plates loosening, or shift of PMMA. As shown in Fig. 2F, no radiopaque growth was observed in the control group 4 weeks postoperatively. And the reduction of both ends of fracture was in accordance with that immediately after operation indicated that there was no significant movement between the fracture ends.

Gross histology
Gross observation showed the translucent and elastic membrane was formed around PMMA (Fig. 1F). In the control group, there was no membrane formation but callus tissue surrounding with connective tissue. The induced membrane (Fixation group and Non-fixation group) and connective tissue (control group) was removed completely for further testing. As shown in Fig. 3, H&E staining revealed an intensive fibrous, cell-rich, and vascularized tissue in PMMA groups (Fixation group and Non-fixation group), while just little vessel like structure was observed in control group. Compared with membrane in Non-fixation group, there were more micro-vessels in Fixation group. And with the extending of implantation time, the micro-vessels had a tendency to increase and to be more mature. The osteogenesis activity was assessed via Alizarin red staining for calcium deposition. The results showed that the reaction of calcifying nodules was positive in Fixation group and Non-fixation group (Fig. 4), indicating excellent osteogenic activity.

Osteogenesis
Content of osteogenesis in the induced membranes and in the connective tissue were measured using IHC analysis by detection COL I. As shown in Fig. 6, the positive cells displayed brown yellow or brown particles, and mainly scattered in the membrane. The qualitative IHC analysis shown that the expression of COL I increased evidently than that of the control group, and that there was no obvious difference in the percentage of COL I between the two PMMA groups in 4weeks and 6 weeks after operation.
As shown in Fig. 7, Western blot analysis revealed that the membrane extracted from Fixation group and Non-fixation group both exhibited higher protein expression of Runx2 and ALP compared to the control group, 4 and 6 weeks postoperatively.
Additionally, in comparison to membrane in Non-fixation group, an increasing trend in the protein expression of Runx2 and ALP was observed in Fixation group, although the difference did not reach significance in the IM group on day 7 (Fig. 2B-D). Western blot analysis also revealed that both the IM and BMP-2 exert a significant effect on the protein expression of Runx2, Col 1, and OCN, particularly in the BMP-2 group (Fig. 2E and 2F). Western blot was almost consistent with the IHC results.

Vascularization
Content of vascularization in the induced membranes and in the connective tissue were measured using IHC analysis by detection the percentage of CD31 positive cells. As shown in Fig. 8, the positive cells displayed brown yellow particles, mainly scattered, and formed tube-like vascular structure in the membrane. The results indicated that CD31 positive cells were relatively high in the induced membranes, but that in the connective tissue was little. The comparison between Fixation group and Non-fixation group did not reveal any significant differences postoperatively.
And with the extending of implantation time, the CD31 positive cells formed the more mature organized micro-vessels in 6 weeks after operation.

Discussion
There were many researches regarding that the stability of fracture broken end directly affects the healing of bone fracture (25)(26)(27)(28). While during the first phase of Masquelet, the bone defects filling with PMMA spacer are fixed with internal plate, intramedullary nail, external fixator, casting, or nothing (17,(19)(20)(21). Each fixed mode could result in a different mechanical environment, subsequently have significant influence on bone repair. No study has been done yet to evaluate the effect of PMMA spacer stability on subsequently bone formation in the membrane cavity. This was the first study to investigate the influencing of topical mechanical stability of fracture broken end on the subsequently formation of Masquelet membrane. In order to explore the specific mechanisms, two fixation variables, rigid fixation offered by plates and relative fixation by suturing, were selected for further research.
The rabbit models of bone defects were commonly used in laboratory experiments, and stainless-steel internal plate was usually applied for the fixation of bone defect ends (9,29,30). Through X-ray measurement postoperatively we confirmed that the Masquelet models with plate internal fixation was successfully made. The X-ray results suggested no significant movement between the fracture ends, even in the control group without PMMA and plate. Due to the integrity of ulna and interosseous membrane, there was no visible displacement of fracture segments (19). And due to the supporting of PMMA in defects, and suture tension of fascia, PMMA in Nonfixation group was in relatively stable pattern with little distortions (30). Still, the jiggle of fracture fragments was existed theoretically, which might have impacts on subsequently membrane formation.
Through gross histology we found that a translucent and membrane-like fibrotic capsule was formed around PMMA, and that connective tissue formed in the bone defects in the control group, which is consistent with the literature reported (15,19).
H&E and Alizarin red staining results suggested that the intensive fibrous, cell-rich, calcium-positive, and vascularized membrane around PMMA would be more conducive to osteogenic activity.
Content of proliferation was assessed by detection the percentage of Ki67 positive cells. Ki67 is a core protein seen in proliferating cells, which exhibits good morphological properties of cell proliferation (31). The proliferation was slightly higher in the induced membranes compared to the control groups, and a slightly increased proliferative activity was observed in membranes of Fixation group in 12 comparison to that of Non-fixation group, suggesting that the mechanical stability offered by plates was beneficial to cell proliferation in the induced membrane. The results was corresponding to previous research that the cell proliferation was relatively active in the membrane (15,19). Besides, our research found that a qualitative trend toward increased proliferation was observed in PMMA groups between 4 weeks and 6 weeks.
The Col I was selected for osteogenesis qualitatively. And the expression levels of ALP and RUNX2 were used for the quantitative analysis of osteogenic activity, of which ALP is an early sign of osteoblast differentiation and maturation, and RUNX2 plays a central role in coordinating multiple signals involved in osteoblast activity (32). By comparing results between PMMA groups and the control group, osteogenic activity of the induced membranes was obviously expressed during the whole observation period. Similar animal results were obtained from previous studies. It is well-known that the different mechanical stabilities around bone defects lead to different callus responses and osteoblast activities (33). And current clinical findings indicate that appropriate fixation of the bone defect is desirable in Masquelet technique, with the use of internal fixator, external fixator, or plaster cast (1,3,7). Still, the exact mechanism remains unknown. In our research, an increasing trend in the protein expression of Runx2 and ALP was observed in Fixation group, in comparison to Non-fixation group. These results revealed that the rigid fixation provided by internal plate could further promote osteogenesis, which implying that the use of fixation in Masquelet technique has potential clinical benefits.
The detection of CD31 positive cells represents the formation of new vessels. And the vascularity of the membrane was quantitatively evaluated by detecting the level 13 of VEGF and TGF-β1 (16,34). Previous researches have reported that the tube-like vascular structure is formed in the induced membrane around PMMA, and the highly vascular membrane has many similarities to the periosteum (13,35). The present study indicated that CD31 positive cells were relatively high in the induced membranes, and the higher level of CD31 expression signals greater capacity of vascularization. And the relative expression levels of VEGF and TGF-β1 in membrane around PMMA were higher than those in connective tissue of the control group.
Besides, significant differences were found between Fixation and Non-fixation group, suggesting that the rigid fixation provided by internal plate could further promote vascularization too.

Conclusion
The results of the current study clearly indicate that rigid fixation provided by

Ethics approval and consent to participate
The medical ethics committee of Xiangya Hospital Central South University gave ethical approval of this research.

Consent for publication
All patients gave written informed consent to involvement in this study, including consent to use data from medical records and radiographs.

Availability of data and material
Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.

Competing interests
No conflict of interest exits in the submission of this manuscript, and all authors are fully involved in the study and preparation of the manuscript.  Representative H&E staining of Masquelet membrane (magnification * 100). H&E staining sh  Western-Blotting analysis of osteogenic proteins (RUNX2 and ALP) and angiogenic proteins (V