Background

In orthopedic surgery, the reconstruction of a large bone defect due to trauma, infection, or tumors remains challenging. Distraction osteogenesis using the Ilizarov method [1, 2] and vascularized fibula grafts [3, 4] continue to be limited treatment options. The Masquelet technique, also known as the induced membrane (IM) technique, provides a third option for reconstruction of large bone defects [5,6,7], and is comprised of a two-staged surgery. In the first stage, bone defects due to debridement are filled with a polymethyl methacrylate (PMMA) cement spacer. A bioactive IM then forms around the cement spacer. After the IM has formed, the cement is removed, and autologous bone grafting is performed in the space of the bone defect surrounded by the IM. The latter serves as a conduit for cells and provides a favorable environment for bone graft osseointegration. IM formed around the cement spacer is a key component of Masquelet reconstruction surgery [5,6,7,8]. Previous studies have shown that IM possesses osteogenic and osteoinductive properties, and is richly vascular [9,10,11,12,13,14,15,16].

Bone morphogenetic proteins (BMPs) are representative osteoinductive factors. They stimulate the proliferation and differentiation of mesenchymal stem cells or chondro-/osteo-progenitor cells involved in endochondral or intramembranous ossification [17, 18]. Furthermore, the expression of BMPs during fracture repair has also been reported [19,20,21]. Endogenous BMPs are important for bone regeneration and repair [22]. Exogenous BMPs such as BMP-2 and 7 have been applied to promote bone regeneration and repair for open fractures and non-union [23, 24]. Recently, BMPs have been utilized as osteoinductive factors as per the ‘diamond concept’ when using the Masquelet technique [25,26,27,28].

The expression of BMP-2 protein has been detected using immunohistochemistry, enzyme-linked immunosorbent assay, and western blotting in human [15, 29] and animal [9,10,11, 30] IM samples. The gene expression of BMP-2, 3b, 6, 7, 10, and 14 has also been detected in human IM samples [29, 31]. However, the presence of BMP proteins other than BMP-2 has not yet been observed using immunohistochemistry in IM samples. Therefore, this study aimed to investigate the presence of BMPs (BMP-4, 6, 7, and 9) using histological samples of human IM.

Methods

Ethical approval

This study was performed in accordance with the ethical standards laid down by the 1964 Helsinki Declaration and its later amendments, and was approved by the Ethics Committee of our university. Due to the study’s retrospective design, the requirement for informed consent was waived.

Patient inclusion

Six patients from our department whose bone defects had been treated using the Masquelet technique were included. The affected sites were the femur (n = 3) and the tibia (n = 3).

Histological specimens

During the second surgery (removal of the cement and bone grafting), 1 cm2 pieces of the IM that were in contact with the cement spacer were harvested and immersed in 10% neutral buffered formalin. Samples were embedded in paraffin, and histological sections (4 µm in thickness) were made.

Histological analyses

Histological sections were stained with hematoxylin and eosin and analyzed by two clinical pathologists. The number of blood vessels per 1 mm2 within the IM was counted in locations with the highest capillary density. Histological findings of inflammation, foreign body reaction, and fibrosis were assessed using a semi-quantified grading scale ranging from 0 to 3, with grade 3 indicating the highest degree of inflammation, foreign body reaction, and fibrosis, and grade 0 indicating an absence of such findings.

Immunohistochemistry

After deparaffinization, the sections were incubated overnight at 4 °C with anti-BMP4 primary antibody (1:100 dilution, GTX100875, GeneTex Inc., Hsinchu City, Taiwan), anti-BMP6 primary antibody (1:100 dilution, ab155963, Abcam, Cambridge, MA, USA), anti-BMP7 primary antibody (1:100 dilution, ab84684, Abcam), or anti-BMP9 primary antibody (1:100 dilution, ab35088, Abcam), and subsequently treated with peroxidase-labeled anti-rabbit immunoglobulin (Histofine® Simple Stain MAX PO, Nichirei Bioscience, Tokyo, Japan) at room temperature for 60 min. The signal was observed as the development of a brown reaction product with the peroxidase substrate 3,3′-diaminobenzidine (Histofine® Simple Stain 3,3-Diaminobenzidine Solution, Nichirei Bioscience). The sections were counterstained with hematoxylin and examined using a BZ-X700 confocal microscope (Keyence Corporation, Osaka, Japan). Phosphate-buffered saline (PBS) was used instead of primary antibodies to stain the negative control samples. Formalin-fixed paraffin-embedded human tissue sections (catalog number CS812148, case ID CU0000012835, 63-year-old, male, bone, distal femur) were obtained from OriGene Technologies (Rockville, MD, USA), and were used as the positive control.

Clinical data

Data concerning patient sex, age, morbidity accounting for the bone defect, free flap application to the affected limb, affected site (bone), impregnation of antibiotics to the cement spacer, duration of cement placement, history of smoking, and a history of diabetes mellitus (DM) or peripheral artery disease (PAD) were obtained from medical records. The duration of cement placement was defined as the number of days from the first-stage surgery in which the cement spacer was placed to the second-stage surgery in which the cement spacer was removed and bone grafting was performed. The radiographic apparent bone gap (RABG) [32] was measured to determine the size of each patients’ bone defect using plain radiographs. Bony union in the enrolled patients was assessed radiologically and clinically. The time point of bony union assessment was set at 6 months after the second-stage surgery (autologous bone grafting). Radiological bony union was defined as corticalization of the grafted bone and absence of a gap between the grafted and original bone, observed in three or four cortices using orthogonal (anteroposterior and mediolateral) plain radiographs. Clinically, bony union was defined as the absence of pain on full weight-bearing. Bony union was defined as the achievement of both radiological and clinical bony union. Three experienced orthopedic trauma surgeons individually assessed bony union, which was considered to have been achieved when at least two surgeons agreed that it had occurred.

Results

Patient characteristics

Five men and one woman were included in the study (Table 1). The mean age was 53.0 ± 11.1 years (range, 42–72). The morbidities accounting for the bone defects were osteomyelitis (3 patients), infected nonunion (1 patient), non-infected nonunion (1 patient), and severely comminuted open fracture (1 patient). The femur (n = 3 patients) and the tibia (n = 3 patients) were the affected sites. The RABG was 8.5 ± 3.1 cm (range, 2.7–10.4). The mean cement placement duration was 77.5 ± 31.3 days (range, 46–126). Three patients underwent free flap surgery. Five patients received antibiotic impregnation into the cement. Three patients had a history of smoking, and no patients had a history of DM or PAD. All patients achieved a bony union.

Table 1 Patient characteristics
Full size table

Histological findings

IM formation was confirmed histologically in all patients; the histological findings are summarized in Table 2. Blood vessel formation was noted in all patients. The mean number of blood vessels per 1 mm2 was 35.0 ± 20.4 (range, 15–70). Inflammation, foreign body reaction, and fibrosis were observed in all patients. The histological grading of inflammation, fibrosis, and foreign body reaction is shown in Table 2. Osteogenesis inside the IM was observed in one patient. A two-layered structure was noted in all patients. A synovial-like structure at the surface that was in contact with the cement was identified in three patients, while fibrin deposition was observed in four patients.

Table 2 Summary of the histological findings
Full size table

Immunohistochemical findings

The immunohistochemical findings are summarized in Table 3. The presence of BMP-4, 6, 7, and 9 was observed in all IM samples. Representative immunohistochemical images are shown in Fig. 1 (Case 3, femur) and Fig. 2 (Case 4, tibia). Immunolocalization of BMP-4, 6, 7, and 9 was observed in blood vessels (Figs. 1a, b, 2a, b) and in fibroblasts (Figs. 1c, d, 2c, d). These findings were observed in all IM samples. Immunolocalization of BMP-4, 6, 7, and 9 was also observed in the bone within the IM in one sample in which osteogenesis inside the IM was observed (Fig. 2e, f). Finally, immunostaining of human bone tissue as a positive control demonstrated positive immunoreactivity for BMP-4, 6, 7, and 9 (Fig. 3).

Table 3 Summary of the immunohistochemical findings
Full size table
Fig. 1
figure 1figure 1

Representative immunohistochemical images of the induced membrane (Case 3, femur) a blood vessel, low-power field, b blood vessel, high-power field, c fibroblast, low-power field, d fibroblast, high-power field. The arrows indicate immuno-positive staining. BMP bone morphogenetic protein, H/E hematoxylin and eosin

Full size image
Fig. 2
figure 2figure 2

Representative immunohistochemical images of the induced membrane (Case 4, tibia) a blood vessel, low-power field, b blood vessel, high-power field, c fibroblast, low-power field, d fibroblast, high-power field, e bone inside the induced membrane, low-power field, f bone inside the induced membrane, high-power field. The arrows indicate immuno-positive staining. BMP bone morphogenetic protein, H/E hematoxylin and eosin

Full size image
Fig. 3
figure 3

Immunostaining of human bone tissue as a positive control. Left side, low-power field; right side, high-power field. The arrows indicate immuno-positive staining. BMP bone morphogenetic protein

Full size image

Discussion

Recent advances in the Masquelet technique have included innovations using different materials instead of PMMA cement to induce the IM, different techniques of autograft harvesting and placement, and using bone substitutes to supplement the autograft [33]. Histological analyses have also been performed and factors affecting to the vascularity of the IM have been reported [34]. Vitiello et al. [35] recently reported foreign body reaction and IM formation following silver-coated knee megaprosthesis reconstruction.

This study confirmed the presence of BMP-4, 6, 7, and 9 in IMs using human samples, and showed that these BMPs can serve as osteoinductive factors in the treatment of patients with bone defects using the Masquelet technique. This finding helps to clarify the mechanism through which IM promotes bone regeneration and repair in the treatment of bone defects using the Masquelet technique. Therefore, the presence of these BMPs as osteoinductive factors in IM may lead surgeons to omit the addition of exogenous BMPs to bone grafts. However, one animal study showed that the IM technique with BMP and a synthetic scaffold could heal a rat femoral critical size defect [36]. This finding suggested that the addition of exogenous BMP to the IM technique was also an effective option.

BMPs constitute the largest subdivision of the transforming growth factor-β (TGF-β) family of ligands, with nearly 30 distinct human proteins bearing the name [17, 18]. Among the various BMPs, we selected four for this study, based on the following reasons.

BMP-4 belongs to the same subgroup of bone-inducing BMPs as BMP-2, based on the homology of their amino acid sequences. BMP-2 is known to be the most representative osteoinductive factor and has been reported to be widely used clinically to treat bone fracture or nonunion [23, 37, 38]. BMP-4 has been detected along with BMP-2 in the area of endochondral ossification, particularly in the matrix between the newly formed osteoid in human fracture callus [19]. The expression of BMP-4 and Noggin, a major BMP antagonist in tissues, is highlighted in the newly formed callus tissue, thereby confirming the central role of BMP signaling in bone fracture repair [20].

BMP-7 has been clinically applied to treat nonunion of fractures. In a randomized controlled trial, the efficacy of recombinant human BMP-7 (rhBMP-7) in tibial nonunion involving 124 patients who received either autologous bone grafting or a device containing rhBMP-7 was tested [24]. The bone healing rate was found to be inferior in the rhBMP-7 treated group, albeit not statistically significant, and the bone healing capacity of rhBMP-7 was assessed as comparable to that of autologous bone grafting. The United States Food and Drug Administration and the European Medicines Agency have approved rhBMP-7 as a “humanitarian use device” for tibial nonunion. In addition, rhBMP-7 has been used off-label for various indications, including nonunion of the scaphoid, humerus, and clavicle [39,40,41,42].

BMP-6 is a paralog of BMP-7 with 87% similarity in their amino acid sequences. BMP-6 is more potent in promoting osteoblast differentiation in vitro and in inducing bone regeneration in vivo when compared with its closely related BMP-7 paralog. This is because of the reversible binding of BMP-6 to Noggin where, unlike BMP-7, BMP-6 may dissociate from Noggin and escape Noggin inhibition [43]. A novel rhBMP-6 containing osteogenic device aimed to accelerate bone regeneration has been developed and is being tested in clinical trials [18, 44, 45].

BMP-9 is a recent discovery in the BMP family. BMP-9 is resistant to Noggin, thus facilitating a more robust cellular differentiation of osteoprogenitor cells into preosteoblasts and osteoblasts [46]. It has been reported that BMP-9 stimulates callus formation in osteoporotic rats during fracture healing [47]. Aside from its osteogenic activity, BMP-9 is involved in a broad range of biological functions, including stem cell differentiation, angiogenesis, neurogenesis, tumorigenesis, and metabolic functions [48]. BMP-9 is likely to be a promising alternative to clinically available BMPs.

We found BMP-4, 6, 7, and 9 were present in all the human IM samples. In addition to immunolocalization of BMP-4, 6, 7, and 9 in blood vessels and fibroblasts, they were also observed within the bone inside the IM in one sample, in which osteogenesis inside the IM was observed. However, this histological finding of osteogenesis inside the IM was found in only one of the six samples. Other studies have also reported this finding in limited samples [14, 29]. We consider that BMP-4, 6, 7, and 9 present inside the IM and contributing to osteoinduction is an important finding.

The Masquelet technique is frequently applied to the lower extremities rather than to the upper extremities. The tibia and femur are long bones of the lower extremities and frequently become subject to bone defects. The volume of the surrounding soft tissue is greater in the femur than in the tibia, and can be considered to reflect vascularity, thereby potentially affecting the formation and properties of the IM. Therefore, we included patients with bone defects in the tibia and femur in this study. Our study results showed that IMs harvested from patients with tibia and femur bone defects expressed BMP-4, 6, 7, and 9.

The strength of this study was the use of human IM samples harvested from patients with bone defects treated using the Masquelet technique. In addition, we reported the presence of BMPs not previously found at the protein or the mRNA level. A limitation of this study was that specimens were harvested from a limited number of patients. Moreover, this was a retrospective study conducted at a single institution, and could therefore be susceptible to selection bias and limited generalizability.

Future studies are needed to further confirm findings. The BMP family, which plays a central role in osteoinduction, comprises many growth factors, while other factors also affect osteoinduction. Therefore, we intend to investigate these other BMPs and osteoinductive factors to further clarify the osteoinductive effect of the IM using human IM samples.

Conclusion

Osteoinductive factors, BMP-4, 6, 7, and 9 were found to be present in the IM of patients treated using the Masquelet technique.