Now Reading
Endocultivation: continuous application of rhBMP-2 via mini-osmotic pumps to induce bone formation at extraskeletal sites

Endocultivation: continuous application of rhBMP-2 via mini-osmotic pumps to induce bone formation at extraskeletal sites



International Journal of Oral & Maxillofacial Surgery, 2017-05-01, Volume 46, Issue 5, Pages 655-661, Copyright © 2017 International Association of Oral and Maxillofacial Surgeons


Abstract

The continuous presence of recombinant human bone morphogenetic protein 2 (rhBMP-2) inside a scaffold may be crucial to the outcome in bone tissue engineering. This study investigated whether the release of the growth factor rhBMP-2 via different continuous application schemes influences histomorphological aspects of the hard and soft tissues induced. Three-dimensionally printed hydroxyapatite scaffolds were implanted into one latissimus dorsi muscle of 42 female Lewis rats. Simultaneously implanted mini-osmotic pumps were used to provide a continuous application of rhBMP-2 over 1, 2, or 4 weeks (total dose 200 μg). A reference group received rhBMP-2 at the time of implantation only, and a control group received only block implantation. Bone density and histological examinations were performed after 8 weeks. No significant difference in bone density was found between the groups; however, the blood vessel count differed significantly between the groups receiving continuous treatments and both the control group and simultaneous rhBMP-2 treatment group ( P < 0.0001). Soft tissue types were distributed differently among the study groups. RhBMP-2 application via mini-osmotic pumps is as suitable for inducing bone formation as a single application at the time of implantation. The time interval over which rhBMP-2 was administered had no impact on the amount of new bone formation, probably due to the study duration and low local concentrations of growth factor.

The functional and aesthetic rehabilitation of patients suffering from critical-size bone defects involving the facial skeleton remains a great challenge in the field of reconstructive surgery. Autologous bone grafts still represent the gold standard and widely preferred therapy, but the use of such grafts is associated with an increased operation time and additional donor site morbidity. In this context, tissue-engineered bone grafts represent a promising approach to this issue.

Among other techniques in bone tissue engineering, endocultivation aims to grow individually shaped vascularized bone grafts within customized scaffolds at extraskeletal sites. Due to the presence of its own arteriovenous vessel system, the latissimus dorsi muscle is the preferred site to grow bone replacements and can ultimately provide a free tissue flap transfer with an embedded engineered bone graft.

Although techniques in bone tissue engineering are constantly being improved, most studies have reported excessive bone growth outside the scaffold margins. Furthermore histological examinations have concluded that the characteristics of the newly formed bone are comparable to those of cancellous bone, leading to less stability of these bone grafts. Excessive heterotopic bone formation might theoretically be avoided by improving the biocompatibility of the scaffold materials, thereby directly influencing the function and behaviour of adhering osteogenic cells. Besides the scaffold material, the additional application of growth factors (e.g. bone morphogenetic proteins (BMPs)) plays a crucial role in bone induction processes.

In previous trials of endocultivation, BMPs have been applied simultaneously with scaffold implantation. Knowledge about the optimal time point at which to load the scaffold with growth factors is limited and is based mainly on empirical studies. Bone morphogenetic protein 2 (BMP-2) has been delivered via pre-loading of the scaffold simultaneously with scaffold implantation, or by delayed injection into the scaffold to allow for prior soft tissue ingrowth.

In addition to pharmacologically based drug delivery systems, an implantable pump represents another option for administering BMPs at a continuous flow rate. Mini-osmotic pumps are available and have been used widely in animal studies for the administration of various types of drug. These miniature infusion pumps (3 × 0.7 cm in size) are designed specifically for small animal studies. Miniature pumps allow for controlled and continuous drug delivery at the site of interest over a specified time period.

This study evaluated the use of mini-osmotic pumps for the continuous application of recombinant human BMP-2 (rhBMP-2) in endocultivation. It was hypothesized that the continuous application of rhBMP-2, in comparison to the simultaneous injection of rhBMP-2 at scaffold implantation, might have a different impact on hard tissue formation and, furthermore, might have measurable effects on the tissue morphology inside the scaffolds.

Materials and methods

Animals and surgical procedure

Ethical approval was obtained for this animal study. Forty-two female Lewis rats (Charles River Laboratories International, Wilmington, MA, USA), at an average age of 3 months and with an average weight of 220–240 g, were used in this study. Food and water were available ad libitum. The animals were kept in a room with a controlled environment and a circadian day–night rhythm of 12 h. All interventions were performed under general anaesthesia, through intraperitoneal injection of 0.2 μl/g body weight xylazine (Rompun, 2%; Bayer Healthcare, Leverkusen, Germany) and 1 μl/g body weight ketamine (Ketavet, 10%; Pfizer Pharma, Berlin, Germany).

After local disinfection of the shaved skin (Kodan; Schülke & Mayr, Norderstedt, Germany), an incision was made with a scalpel over the thoracic spine ( Fig. 1 A ), following which a subcutaneous blunt dissection was performed. A pouch was prepared in the latissimus dorsi muscle in which to implant a three-dimensionally printed scaffold ( Fig. 1 B–D).

Surgical procedure for scaffold implantation into the latissimus dorsi muscle, in chronological order: (A) skin incision over the thoracic spine; (B) preparation of a pouch to take the scaffold; (C) scaffold implantation; (D) closure of the pouch; (E) connection of a mini-osmotic pump to the tube leading into the centre of the scaffold; (F) positioning of the pump in a separate pouch.
Fig. 1
Surgical procedure for scaffold implantation into the latissimus dorsi muscle, in chronological order: (A) skin incision over the thoracic spine; (B) preparation of a pouch to take the scaffold; (C) scaffold implantation; (D) closure of the pouch; (E) connection of a mini-osmotic pump to the tube leading into the centre of the scaffold; (F) positioning of the pump in a separate pouch.

Twenty-four animals (eight animals per group) were chosen for the implantation of mini-osmotic pumps (Alzet; Durect Corporation, Cupertino, CA, USA). After scaffold implantation, another pouch was prepared on the left side of the animal's back. A tube leading into the centre of the scaffold was connected to the mini-osmotic pump ( Fig. 1 E, F), which was then placed into the separate pouch. The mini-pumps were set to release their contents over the course of 1, 2, or 4 weeks after activation (for a total dose of 200 μg). The wound was closed with resorbable sutures (Vicryl; Ethicon, Norderstedt, Germany). Another group of animals (reference group, n = 9) did not undergo mini-pump implantation, but received 200 μg rhBMP-2 simultaneously with scaffold implantation. A control group of animals ( n = 9) underwent scaffold implantation, but did not undergo mini-pump implantation and did not receive any rhBMP-2. The group allocation and study protocol is further summarized in Fig. 2 .

Study protocol and group division.
Fig. 2
Study protocol and group division.

Weekly intraperitoneal fluorochrome injections were started at 2 weeks after implantation and were continued until the end of the trial. The fluorochromes applied were Xylenol Orange (6% in NaHCO 3 , 5 ml/kg; Sigma-Aldrich, Steinheim, Germany), Alizarin (3% in 2% NaHCO 3 , 0.8 ml/kg body weight; Sigma-Aldrich), doxycycline (1 mg/kg; Ratiopharm SF, Ulm, Germany) and Calcein Green (1% in 2% NaHCO 3 , ; Sigma-Aldrich). The specimens were assessed for the presence of each fluorochrome in all specimens, in a semi-quantitative manner: 0, not detected; 1, some detected; 2, widely detected, as described previously.

Conventional computed tomography examinations

Conventional computed tomography (CT) examinations were performed to evaluate bone density (Somatom; Siemens AG, Munich, Germany; settings of 120 kV, 210 mAs, and 46.82 mGy). A radiologist who was unfamiliar with the group allocation performed all examinations. The bone density value per animal was determined in the area of scaffold implantation. Scaffold areas were marked manually on each slice for every animal using the CT scanning software provided by the manufacturer. The bone density was measured in Hounsfield units (HU).

Sample preparation and histology

Animals were sacrificed after an 8-week observation period by CO 2 insufflation. The scaffold-bearing muscular pouch was then excised. The specimens were cut in half before embedding in wax or methyl methacrylate (MMA). Additional histopathological techniques for sample preservation and staining were routinely applied, as described previously.

In brief, after decalcification and dehydration in a graded series of alcohol solutions, the first half of each specimen was embedded in wax and then sliced to a width of approximately 5 μm. Histological staining was then performed (Masson–Goldner and haematoxylin–eosin). The other half of each specimen was embedded in MMA and sliced to a width of 100 μm for microradiography, 40 μm for fluorescence microscopy, and 10 μm for toluidine staining, respectively. Microphotographs were then obtained (Mikrophot-FXA; Nikon, Düsseldorf, Germany). Photoshop CS5.1 (Adobe Systems, Munich, Germany) was used for further image processing.

Histological assessment

To standardize the histological evaluations, the cross-sectional area of the scaffold was divided into four major zones ( Fig. 3 ).

Definition of the zones inside the scaffold (cross-sectional area) on a schematic drawing (left) in comparison with a control group microradiograph (right); scale bar = 2000 μm.
Fig. 3
Definition of the zones inside the scaffold (cross-sectional area) on a schematic drawing (left) in comparison with a control group microradiograph (right); scale bar = 2000 μm.

Blood vessels were counted in the haematoxylin–eosin-stained specimens in four fields of view, each peripheral and central, at 100× magnification. The number of blood vessel-containing zones was also determined in the haematoxylin–eosin specimens, as well as the bone tissue-containing scaffold zones in the microradiography and toluidine-stained specimens.

Soft tissues within the scaffolds were differentiated into fibrous or vacuole-containing types throughout the specimen surface and the proportions recorded as follows: 0%, none detected; 100%, one tissue type detected; presence and proportion of both tissue types: approximately 25%/75%, 50%/50%, 75%/25%, as established previously.

Statistical evaluation

Mean and 95% confidence interval (CI) values were calculated, and the Kruskal–Wallis test was applied to identify statistically significant differences between the groups. Subsequently, between-group comparisons were performed using Dunn's post hoc test with Bonferroni correction to provide the adjusted P -values. A P -value of <0.05 was considered statistically significant.

Results

Bone density evaluation

The highest mean bone density values in this trial were found in the group provided with 4 weeks of continuous rhBMP-2 application (mean 1342 HU, 95% CI 1274–1409 HU), whereas the lowest bone densities were measured when 200 μg rhBMP-2 was applied simultaneously with scaffold implantation (mean 1217 HU, 95% CI 1134–1300 HU). The Kruskal–Wallis test did not reveal any statistically significant relationship between the group allocation and bone density when comparing all groups ( P = 0.1502), and did not reveal any statistically significant differences for multiple comparisons between groups. A box-plot of the bone density values is displayed in Fig. 4 .

Box and whisker plot (min to max) of the bone density assessments, indicating no statistically significant differences between the groups.
Fig. 4
Box and whisker plot (min to max) of the bone density assessments, indicating no statistically significant differences between the groups.

Histological assessment

In general, blood vessels were detectable throughout the cross-sectional area, accompanied by bone tissue or solely by soft tissue, as observed in specimens of the control group.

The Kruskal–Wallis test revealed a highly significant relationship between the group allocation and number of blood vessels counted ( P < 0.0001). The highest mean blood vessel counts per field of view (BV/FOV) were found in the group provided with 1 week of continuous rhBMP-2 application (mean 17.17, 95% CI 14.94–19.39). The lowest mean blood vessel counts (BV/FOV) were found in the control group (mean 7.59, 95% CI 6.05–9.14). Figure 5 presents a box-plot summarizing the results of the blood vessel count.

Box and whisker plot (min to max) of the blood vessel counts, revealing statistically significant differences between the groups with continuous rhBMP-2 application (groups 2–4) and both the control group and the group receiving simultaneous rhBMP-2 injection at scaffold implantation (group 1). Adjusted P -values are listed separately in Table 1 . The bar inside the box displays the median value.
Fig. 5
Box and whisker plot (min to max) of the blood vessel counts, revealing statistically significant differences between the groups with continuous rhBMP-2 application (groups 2–4) and both the control group and the group receiving simultaneous rhBMP-2 injection at scaffold implantation (group 1). Adjusted P -values are listed separately in Table 1 . The bar inside the box displays the median value.

Multiple comparisons between the groups revealed highly significant differences between the control group and all of the continuous rhBMP-2 application groups (all P < 0.0001). Continuous rhBMP-2 treatment was also significantly superior to simultaneous injection of rhBMP-2 at scaffold implantation (all P < 0.01). P -values for the multiple comparisons (Dunn's post hoc test) are displayed in Table 1 .

Table 1
Comparisons of the numbers of blood vessels between the groups (Dunn's post hoc test with Bonferroni correction). a .
Multiple comparisons Adjusted P -value
Group 1 vs. group 2 0.0010
Group 1 vs. group 3 0.0096
Group 1 vs. group 4 0.0026
Group 1 vs. control group 0.5712
Group 2 vs. group 3 >0.9999
Group 2 vs. group 4 >0.9999
Group 2 vs. control group <0.0001
Group 3 vs. group 4 >0.9999
Group 3 vs. control group <0.0001
Group 4 vs. control group <0.0001

a Group 1: (reference group) received rhBMP-2 at the time of scaffold implantation only; group 2: received continuous application of rhBMP-2 over 1 week; group 3: received continuous application of rhBMP-2 over 2 weeks; group 4: received continuous application of rhBMP-2 over 4 weeks; control group: received only block implantation (no rhBMP-2).

Distinctions between cortical and cancellous bone structures could not be identified in the histological sections. The bone tissue in all specimens appeared to be of cancellous origin, which was on a comparable level whether adhering to the scaffold material or distributed in the interstitial spaces ( Fig. 6 ).

Microradiographs of scaffolds from the study groups receiving rhBMP-2, showing mineralized bone tissues adhering to the scaffold material and in the interstitial spaces inside the scaffold. Ectopic bone formation was also observed in all study groups.
Fig. 6
Microradiographs of scaffolds from the study groups receiving rhBMP-2, showing mineralized bone tissues adhering to the scaffold material and in the interstitial spaces inside the scaffold. Ectopic bone formation was also observed in all study groups.

Two different types of soft tissue could be identified. The fibrous soft tissue type was characterized mainly by extracellular fibres, the sporadic appearance of blood vessels, and rims of differentiated osteogenic cells delimiting the newly formed bone within the connective tissue. Osteoid was not always determinable within this soft tissue type. The histological appearance of the fibrous tissue type indicated that it was associated with active bone formation processes. The reticular or vacuole-containing soft tissue type was solely found when additional bone tissue was present. Blood vessels were observed in the septal spaces between vacuoles. The histological appearance of this tissue type was recognized as representing a state of completed differentiation ( Fig. 7 ).

Haematoxylin–eosin staining, revealing the two different soft tissue types present in the cross-sectional area (fibrous and vacuole-containing soft tissue). Blood vessels were detectable in all sections (red arrows). Bone tissue is indicated with the green arrows and was observed in the presence of both soft tissue types. Cortical bone structures were not observed within the bone tissues; bone formation was solely cancellous.
Fig. 7
Haematoxylin–eosin staining, revealing the two different soft tissue types present in the cross-sectional area (fibrous and vacuole-containing soft tissue). Blood vessels were detectable in all sections (red arrows). Bone tissue is indicated with the green arrows and was observed in the presence of both soft tissue types. Cortical bone structures were not observed within the bone tissues; bone formation was solely cancellous.

Regarding the proportional distribution of these two soft tissue types, the reticular soft tissue type was found to be predominant when rhBMP-2 was applied simultaneously (group 1: 66.7% reticular vs. 33.3% fibrous) or continuously for 1 week (group 2: 68.8% vs. 31.2%). The largest proportion of reticular tissue was seen in the group treated with 2 weeks of continuous rhBMP-2 (group 3: 75% vs. 25%). In contrast, an equal distribution of the two types was observed in the group treated with 4 weeks of continuous rhBMP-2.

With regard to the fluorochrome labelling index values, the highest cumulative labelling index was calculated for group 2 (5.8), whereas fluorochromes were absent in the control group (labelling index 0.0). The highest fluorochrome-specific index in all groups, except for the control group, was found for Calcein Green. In the groups given continuous rhBMP-2 application, fluorochromes were detected in newly formed bone 5 weeks after scaffold implantation. When rhBMP-2 was applied simultaneously at the operation, new bone formation was verified 4 weeks after scaffold implantation. Table 2 shows the fluorochrome-specific labelling indices.

Table 2
Appearance of fluorochromes after intraperitoneal injection (labelling index values); results are presented as the mean ± standard deviation. a
Fluorochrome (application week)
Xylenol Orange Calcein Green Alizarin Doxycycline
(Weeks 2 and 7) (Weeks 3 and 6) (Week 4) (Week 5)
Group 1 1.4 ± 0.7 2.0 ± 0.0 1.1 ± 0.9 0.0 ± 0.0
Xylenol Orange Alizarin Doxycycline Calcein Green
(Weeks 2 and 6) (Weeks 3 and 7) (Week 4) (Week 5)
Group 2 1.7 ± 0.3 1.5 ± 0.5 0.9 ± 0.6 1.7 ± 0.3
Group 3 1.4 ± 0.7 1.0 ± 0.8 0.7 ± 0.5 1.7 ± 0.4
Group 4 1.2 ± 1 1.1 ± 0.9 0.6 ± 0.5 1.3 ± 0.8
Control group 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0

a Group 1: (reference group) received rhBMP-2 at the time of scaffold implantation only; group 2: received continuous application of rhBMP-2 over 1 week; group 3: received continuous application of rhBMP-2 over 2 weeks; group 4: received continuous application of rhBMP-2 over 4 weeks; control group: received only block implantation (no rhBMP-2).

Discussion

In this study, the same overall cumulative dose of rhBMP-2 was used to investigate the morphological effects of a continuous rhBMP-2 application scheme compared to the conventional simultaneous application of this growth factor and a control group. The assessment of bone density did not reveal any significant difference in bone quantity among the study groups.

The efficacy of rhBMP-2 when applied exogenously might be dependent on the cellular environment at the application site itself. In a previous study, cellular colonization inside the scaffold was monitored with fibre-optic detection devices. The metabolic changes measured led to the conclusion that cellular ingrowth and tissue homeostasis was completed at 4 weeks after scaffold implantation. This indicated that a prolonged application of BMP-2 over the first 4 weeks might increase the effects of BMP-2 to induce chemotaxis in responsive osteogenic progenitor cells inside the scaffolds. In a previous study establishing the endocultivation model, lower bone density was determined after rhBMP-2 injection at 4 weeks after scaffold implantation, whereas an increase in bone density was detected when rhBMP-2 was applied at the time of the operation or at 1 or 2 weeks after the implantation procedure. Another systematic study with the same overall cumulative dose of rhBMP-2 evaluated the time-dependence of multiple doses on the hard and soft tissue morphology in comparison to simultaneous injection and no growth factor. In contrast to the present study, bone density was significantly higher in the groups with repetitive and delayed treatment in comparison to groups with simultaneous or no injection of growth factor, indicating a time-dependent effect.

A further factor influencing the efficacy of rhBMP-2 is that local proteinases within the extracellular fluid result in the rapid degradation and therefore inactivation of rhBMP-2, which consequently leads to a decrease in the local concentration of rhBMP-2 and reduced biological activity. The endogenous release of BMP inhibitors such as gremlin or noggin might have a further negative impact on the effects of exogenously administered rhBMP-2. A recently re-evaluated human fracture model showed that an imbalance between the presence of BMPs and their antagonists compromises callus formation.

Although the amount of mineralized tissue did not differ in the present study groups, it has to be mentioned that even the continuous release of rhBMP-2 over a 4-week period, as secured by the implanted mini-osmotic pump, was seemingly sufficient to exceed the necessary threshold to allow for osteogenic differentiation, even 4 weeks after the application of growth factor. As no bone formation was found in the control group receiving no rhBMP-2, the existence of new bone formation in every group with an implanted osmotic mini-pump may be interpreted as a proof of principle for this method, therefore supporting the already published principle of the mini-osmotic pumps.

As well as the histological signs of ongoing remodelling processes seen at the end of the study, soft tissue differentiation in favour of reticular soft tissue was found, in contrast to the results of the study that applied repetitive injections mentioned previously. A significantly higher blood vessel count was observed in animals that had received continuous rhBMP-2 application compared to the simultaneous injection of rhBMP-2 at scaffold implantation, which is in accordance with the results of the previously mentioned study applying repetitive injections.

Considering that capillary vascularization has been shown to be a potential indicator of new bone formation, that fluorochrome-labelled osteoid was evident at the half-point of the study, and that the fibrous tissue type representing a cellular base for osteogenic differentiation was present, it might be concluded that bone tissue remodelling and new bone formation processes were ongoing at the end of the study. Therefore the tissue remodelling processes were yet not completed. From a retrospective point of view, extending the length of the study might have shown a beneficial impact on the new bone formation processes.

In summary, the administration of rhBMP-2 via mini-osmotic pumps – as a proof of principle – was found to be capable of inducing mineralized bone tissue at extraskeletal sites, even at very low dosages. The different durations of rhBMP-2 application did not have an impact on the amount of new bone formation, probably due to insufficient local concentrations of the growth factor. On the other hand, significant differences in blood vessel counts were found, most probably indicating forthcoming new bone formation; in this regard, a longer study period would have been favourable. Future studies will have to further improve the effectiveness of the rhBMP-2 dosages applied.

You're Reading a Preview

Become a DentistryKey membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here