Abstract
Introduction:
New reconstructive and less invasive methods have been searched to optimize bone formation and osseointegration of dental implants in maxillary sinus augmentation.
Purpose:
The aim of the presented ovine split-mouth study was to compare bovine bone mineral (BBM) alone and in combination with mesenchymal stem cells (MSCs) regarding their potential in sinus augmentation.
Material and Methods:
Bilateral sinus floor augmentations were performed in six adult sheep. BBM and MSCs were placed into the test side and only BBM in the contra-lateral control side of each sheep. Animals were sacrificed after 8 and 16 weeks. Augmentation sites were analyzed by computed tomography, histology, and histomorphometry.
Results:
The initial volumes of both sides were similar and did not change significantly with time. A tight connection between the particles of BBM and the new bone was observed histologically. Bone formation was significantly (p = 0.027) faster by 49% in the test sides.
Conclusion:
The combination of BBM and MSCs accelerated new bone formation in this model of maxillary sinus augmentation. This could allow early placement of implants.
Introduction
The selection of the most appropriate scaffolds is an important step toward the regeneration of hard tissue. Bovine bone mineral (BBM) is one of the most widely used scaffolds employed in sinus augmentation procedures.9–11 It has similar physical properties to human cancellous bone, both in its morphological structure and its mineral composition. 12 Better bone-forming kinetics has been reported in BBM associated with 25% autogenous bone compared to BBM alone. 13 The rate of added autogenous bone (10–20%) to biomaterial did not significantly influence new bone formation (NBF). 14 However, if BBM is placed alone into the sinus to fill the cavity, the time required to obtain optimal bone formation can be from 6 to 8 months or longer.15,16
For this reason, there is interest in developing a surgical technique that does not require harvesting of autogenous bone but which results in adequate bone formation within short time, allowing early installation of implants in cases of atrophic maxilla and mandible.
New bone formation can be improved by adding autologous adult mesenchymal stem cells (MSCs) originating from bone marrow to biomaterial.17,18 MSCs or bone marrow stromal cells are defined as multipotent progenitor cells with the ability to generate several tissues, including bone. 19
The purpose of this study was to investigate the influence of MSCs in combination with BBM on bone regeneration in an animal model.
Materials and Methods
Animals
This study was approved by the animal trial council of the State Administration in Baden Württemberg (Germany). Six female adult sheep weighing 91.5 ± 11.4 kg (mean ± standard deviation) were included in the study. All animals were kept under the same conditions and observed by a specialist for veterinary surgery.
Study design
A split-mouth design was chosen for the controlled parallel-group study. Six animals underwent bilateral sinus augmentation. One side was augmented with BBM only (Bio-Oss®; Geistlich Biomaterials, Wolhusen, Switzerland). In the test side MSCs were added. Three animals each were assigned for an observation period of 8 and 16 weeks.
Surgical procedure
All animals received preanesthetic medication consisting of 0.01 mg kg−1 atropine (Atropuinsulfat Braun; B. Braun Melsungen, Melsungen, Germany), 20 mg kg−1 ketamine hydrochloride (10% Ketamin; Essex, München, Germany), and 0.1 mg kg−1 xylazine (Rompun®; Bayer Vital, Leverkusen, Germany) intramuscularly. Anesthesia was induced intravenously with 2–4 mg kg−1 propofol (1% Propofol; MCT Fresenius; Fresenius Kabi, Bad Homburg, Germany). After tracheal intubation, anesthesia was maintained using isoflurane (Forene®; Abbott, Wiesbaden, Germany) at an endtidal concentration of 1.8% in oxygen/air and a bolus of 0.002 mg (kg·h)−1 fentanyl citrate (Fentanyl-Janssen; Janssen-Cilag, Neuss, Germany) followed by continuous infusion of 0.001 mg (kg·h)−1 fentanyl citrate. Mechanical ventilation was provided by an anesthesia ventilator (Cato, Dräger, Lübeck, Germany). Respiratory rates and inspired O2 concentration were adjusted to maintain endexpiratory CO2, tension between 38 and 42 mm Hg, and percutaneous O2 saturation between 95% and 100%. To ensure the fluid balance, intravenous Ringer's solution (10 mL [kg·h]−1) was administered continuously. Postoperative analgesia was achieved with subcutaneous carprofen (Rimadyl®; Pfizer, Karlsruhe, Germany) (4 mg kg−1 in 24 h) for 4 days, first dose given after induction of anesthesia.
A bilateral sinus augmentation procedure was carried out with an extra-oral approach after local anesthesia by infiltrating the surgical incision sites each with 10 mL articaine hydrochloride 1% added with epinephrine hydrochloride 0.0006% (1% Ultracain®–Suprarenin®; Sanofi Aventis, Frankfurt, Germany). Inferior of the lower orbital rim the anterior wall of the maxillary sinus was exposed by mobilization of the masseter muscle and the adhering periosteum before preparing of the lateral window.
Bone marrow was aspirated from the iliac crest. MSCs were extracted via Ficoll-separation. A volume of 3.5 cm3 of MSCs and BBM were transplanted into the right sinus (test). The left sinus (control) was augmented only with BBM. The wounds were closed with resorbable suture material. All the surgeries were carried out by the same surgeon. The surgical procedures for sinus augmentation and the follow-up were done according to Haas protocol. 20
Collection and processing of marrow cells
Sixty milliliters of bone marrow were aspirated with a heparinised biopsy needle, pipetted onto Ficoll (Sigma, St Louis, MO) at a 1:1 ratio, and centrifuged at 2400 rpms for 25 min. The interface layer was removed, resuspended in phosphate-buffered saline, and centrifugated (2000 rpm, 10 min, 300 g). For washing, the supernatant was pipetted off. This procedure was repeated. The resulting cell pellet was resuspended in phosphate-buffered saline. Viable nucleated cells were counted by trypan blue dye exclusion.
Colony forming unit assay
Cells were placed at four different densities (5000, 25,000, 50,000, 100,000 mononuclear cells [MNCs]/mL) in 96-well plates (Becton Dickinson, Los Angeles, CA) and cultured in MSCBM medium (Cambrex, Walkersville, MD) in a humidified atmosphere of 5% CO2 at 37°C for 7 days. The medium was initially changed after 24 h and then every second day. MSCs were selected on the basis of plastic adhesion and counted under a microscope.
Proof of pluripotency
The cells from the bone marrow concentrate were amplified and differentiated into three cell lineages according to the methods of Pittenger and coworkers to proof that the transplanted cells were actually pluripotent stem cells. 19 The lines consisted of the chondrogenic and adipogenic phenotype. Adipocytes were stained with oil red O, a lipophilic red dye. Chondrogenic potential was confirmed by immunostaining with mouse anti-human aggrecan antibodies. Osteogenic cells were characterized histologically by their expression of collagen type I and calcification marked with the von Kossa staining.
Computed tomography analysis
On the basis of the data of computed tomography, the volume of the augmentation was calculated by the program VoXim 4.3 (© IVS Solutions AG, Chemnitz, Germany).
Fluorescence marking
Ninety milliliters of Xylenolorange per kilogram of body weight was administered in the 6th and 7th week in the animals of the 8-week survival group and in the 14th and 15th week in the animals of the 16-week survival group. With this technique values for 5 and 13 weeks could be obtained by subtracting the fluorescence marked bone that was build in the 6th to the 8th and 14th to the 16th week from the total bone area of the 8- and 16-week groups.
Histological and histomorphometrical evaluation
After fixation in formalin and dehydration, the samples were infiltrated with resin (Heareus Kulzer, Hanau, Germany) and polymerized under UV light. Sections from the resin-embedded sinus were cut with a diamante micro saw (Microslice; IBS, Cambridge, GB, United Kingdom), placed on an acrylglas carrier (Maertin, Freiburg, Germany), and reduced to a thickness of 100 μm on a rotating grinding plate (Struers, Ballerup, Denmark). Six slides of each augmentation side per animal were selected and blinded evaluated. Analysis of the fluorescent markings was performed before staining with Azur II and Pararosanilin (Axiovert 135; Carl Zeiss AG, Oberkochen, Germany; AnalySISΛD Soft Imaging system; Olympus Europa GmbH, Hamburg, Germany).
Statistical analysis
For the parameters NBF, BBM, and marrow space, values were expressed in percent of the evaluated area. The evaluation of the histological slides generated six data points per sinus from which slopes of growth curves were estimated from a linear mixed model (function lme in package nlme under R). Data were log-transformed to compensate for unequal variances of the growth curves. Confidence intervals for the differences were computed from the contrast tables of the model.21,22 The confidence interval for the ratio of new bone was computed by a bootstrap with 4000 repeats.
Results
Colony forming unit assay
The number of MNC was 446 ± 348 × 106 mL−1 concentrated bone marrow aspirate. There were 37 ± 20 colony-forming units per million MNCs. No significant influence of the number of MNCs and colony-forming units on NBF could be found.
Proof of pluripotency
The cultured MSCs could be differentiated successfully into adipocytes as shown by oil red O staining, chondrocytes as shown by aggrecan immuno staining, and osteoblasts as shown by calcification and collagene type I activity (Fig. 1).

Proof of pluripotency. Plastic-adherent cells from both groups could be differentiated as follows: morphology of cultured MSCs stained with azur II (
Computed tomography volume analysis
The mean value of the filling volume was 2.38 cm3 for the control side and 2.42 cm3 for the test side. The 95% confidence interval of the difference in a paired t-test was −0.4 to 0.33, giving no evidence for an imbalance between the sides.
Histological results
The histological sections showed no signs of inflammation. The newly formed osseous lamellae appeared as vital bone tissue containing osteocytes inside the bone lacunae. In comparison to lamellar bone, the biomaterial could be easily identified by its size, shape, and color. In the Azur II–Pararosanilin staining, the newly formed bone appeared in a darker red than the BBM particles. The newly formed bone lamellae connected the biomaterial particles and stabilized the grafted complex. Bone formation appeared in the macro pores of the biomaterial as well. Blood vessels could be detected in the biopsy. The biomaterial with the newly formed bone was integrated well in the surrounding host bone (Fig. 2).

Histological specimens of the same animal after a healing time of 16 weeks. In the Azur II–Pararosanilin staining newly formed bone is stained in a magenta color (*sinusfloor, #lateral sinus wall, +cavum nasi). Connective tissue is marked blue. On the control side (bovine bone mineral only) new bone formation originated only from the sinus floor (
Histomorphometrical and fluorescence marking evaluation
In the test side, the NBF as estimated from log-transformed growth curves was 49% (95% confidence interval of 6% to 84%) faster than in the control side (p = 0.027) (Fig. 3).

New bone formation at implant by time; the vertical axis is on a log scale. The six data points from each sinus were time-jittered to better separate overlapping values. Solid line shows estimate from the mixed model. In the test side, the new bone formation as estimated from linear growth curves is 49% faster (95% confidence-interval of 6% to 84%) than that in the control side (p = 0.027).
Discussion
The standard surgical technique for sinus-floor elevation involves raising the sinus mucous membrane and filling the cavity with either bone or biomaterials, or a combination of these. The materials utilized comprise of autogenous bone,1,23,24 allogenous bone,25,26 xenogenous substitutes,2,15,27 synthetic materials, 3 or a combination of these materials. 4 Regardless of increased morbidity,28,29 autogenous bone is still the gold standard for reconstructive procedures.1,23 Although complications such as bleeding, infection, paresthesia, fractures, and pain have been described, iliac crest cortico-cancellous bone is the most suitable tissue in the reconstruction of large defects due to sufficient quantities available during harvesting.30,31
Lundgren et al. (2008) presented a new technique for maxillary floor augmentation performed only by raising the sinus membrane and not including any graft or biomaterial. A few months after simultaneous insertion of titanium implants, the amount of bone formation does not seem to differ when performing sinus membrane elevation with or without bone grafts. 32 This could be explained by the presence of osteoprogenitor cells in the blood coagulum, the neighboring bone structures, or the Schneiderian membrane. 33 However, up to now the indication of this technique has some limitations, due to the reason that it is necessary to have at least 4–5 mm of residual alveolar crest bone in the area to allow primary stability of the dental implants, which work at the same time as “tent pole” sustaining the sinus membrane.
This further information could elucidate the limitations of growth factors used in the absence of adequate cellular components required for the production of bone. With regard to regenerative potential, MSCs showed superior outcomes compared to bone morphogenetic proteins loaded on the same delivery vehicle. 34
In humans the addition of platelet-rich plasma (PRP) seems not to improve neither maxillary bone volume nor clinical outcome when compared to conventional bone application. A systematical human-controlled clinical review shows no evidence for beneficial effects of PRP in sinus augmentation. 35 In the other hand, Choukroun et al. found comparable NBF in sinus augmentations between platelet-rich fibrin in combination with freeze-dried bone allograft after 4 months and freeze-dried bone allograft alone after 8 months. 36 Clotting is essential for the cells to release their factors. In addition, the use of MSCs in conjunction with BBM can enhance bone regeneration more than PRP. 37
Bone induction with the aid of osteogenic precursor cells could decrease the risks and morbidity associated with a donor site. Accordingly, alternatives for autogenous transplantation can be found in tissue engineering by means of autogenous osteoblast-like cells in combination with biomaterials to generate functional tissues. In cases of severe bone maxillary resorption, the use of MSCs associated to biomaterials could be a noninvasive and good alternative to stimulate faster bone formation.
Some recent studies reported a technique called “injectable bone.”7,38 This involves the morphogenesis of new bone tissue from isolated cells associated to biocompatible scaffolds. A few weeks before the operation, MSCs are isolated from the patient's iliac crest marrow aspirate. After an in vitro culture process, the cells are trypsinized and used for the implantation in human maxillary sinus. 39 Despite some positive results, the number of patients operated was low and the follow-up period of observation still short. Moreover, the procedures of separating, growing, and differentiating the limited number of stem cells in vitro makes the clinical application difficult and expensive. More studies using control groups and a higher number of animals or clinical trials are necessary for the evaluation of the effectiveness and reliability of this method when compared to the simple method of aspiration and separation of cells used in the present study.
In our study we compared the influence of MSCs in combination with BBM on bone regeneration for sinus lift augmentation. In this trial, MSCs were aspirated from the iliac crest and separated using the Ficoll method.40,41 MSCs pluripotency was observed in vitro.
In the present study, the addition of MSCs did not show a statistically significant difference in the final volume of the augmentation. However, it brought about a markedly faster bone formation of 49%. As a reviewer noted, there is evidence from Figure 3 that the median value at week 5 is lower in the test group compared to the control group. This suggests a biphasic structure of the growth curve, with an initial delay and an acceleration later. The hypothesis cannot be ruled out, but we could not find statistical evidence for nonlinearity of the growth curve on a logarithmic scale. This result supports the data from Hernandez-Alfaro et al., 42 who found accelerated bone formation rates induced by autologous bone marrow cells in a clinical feasibility study, using an ex vivo–cultured and autologous bone-marrow-derived cell product. Moreover, a tight connection between the particles of BBM and the new bone was observed histologically, and the newly formed bone was well integrated in the surrounding host bone, including vital bone tissue containing osteocytes inside the bone lacunae and normal vascularization.
The minimal invasiveness of bone marrow aspiration and local application of autologous MSCs associated with BBM can not only offer a safe and faster alternative method to the regenerate bone, but also eliminate general anesthesia and hospitalization, and decrease patient morbidity. 41
Conclusion
The combination of MSCs and BBM resulted in significantly faster NBF over time than BBM alone. This method could allow early insertion of dental implants. In other fields of regenerative medicine, this technique could help to regenerate bone with reduced donor-site morbidity.
Footnotes
Acknowledgments
The authors are indebted to Ute Hübner (Cell Laboratory, Department for Oral- and Maxillofacial Surgery, University of Freiburg, Freiburg, Germany), Heike Jahnke, Annette Lindner, and Dr. Heiner Nagursky (Hard Tissue Research Laboratory, Department for Oral- and Maxillofacial Surgery, University of Freiburg, Freiburg, Germany), and Dr. Rainer Schmelzeisen (Head, University of Freiburg, Freiburg, Germany) for excellent technical assistance. Dr. Dieter Menne (Menne Biomed Consulting, Tübingen, Germany) performed the statistical analysis.
Disclosure Statement
Laboratory work was financially supported by the Camlog Foundation, Basel, Switzerland. Financial and technical support was given by Geistlich Biomaterials.
