Biotechnology Letters

, Volume 31, Issue 12, pp 1825–1835

Delivery of recombinant bone morphogenetic proteins for bone regeneration and repair. Part B: Delivery systems for BMPs in orthopaedic and craniofacial tissue engineering


  • Ziyad S. Haidar
    • Faculty of DentistryMcGill University
    • Department of Biomedical Engineering, Faculty of MedicineMcGill University
    • Center for Biorecognition and BiosensorsMcGill University
    • Shriners Hospital for Children and Division of OrthopaedicsMontreal Children’s Hospital, McGill University
  • Reggie C. Hamdy
    • Center for Biorecognition and BiosensorsMcGill University
    • Shriners Hospital for Children and Division of OrthopaedicsMontreal Children’s Hospital, McGill University
    • Faculty of DentistryMcGill University
    • Department of Biomedical Engineering, Faculty of MedicineMcGill University
    • Center for Biorecognition and BiosensorsMcGill University

DOI: 10.1007/s10529-009-0100-8

Cite this article as:
Haidar, Z.S., Hamdy, R.C. & Tabrizian, M. Biotechnol Lett (2009) 31: 1825. doi:10.1007/s10529-009-0100-8


Localized and release-controlled delivery systems for the sustained expression of the biologic potency of rhBMPs are essential. A substantial number of biomaterials have been investigated thus far. Most fail after implantation or administration mainly due to either being too soft, difficult to control and/or stabilize mechanically. In the second part of this review, we review a representative selection of rhBMP-2 and rhBMP-7 carrier materials and delivery systems ranging from simple nano/microparticles to complex 3-D scaffolds in sites of orthopaedic and craniofacial bone regeneration and repair.


BiomaterialsBone morphogenetic proteinsBone regenerationControlled-releaseNanoparticlesSynthetic polymers


Attempts to induce de novo bone formation for bridging defects using bone grafting procedures, segmental bone transport, distraction osteogenesis or biomaterials have been applied to a great extent (Kneser et al. 2006; Mussano et al. 2007). Several materials for the delivery of recombinant human (rh) forms of bone morphogenetic proteins (BMPs) have been developed in recent years with only specific collagen-based formulations for rhBMP-2 and rhBMP-7/rhOP-1 obtaining FDA approval for their restricted clinical use in humans, namely in orthopaedic and spinal fusion applications (Table 1). The ultimate goal would be to develop safe, proficient, user-friendly and cost-effective rhBMP delivery systems that would completely replace traditional grafting procedures in a number of diverse applications (Moioli et al. 2007; Schmidmaier et al. 2008). In this review, we classify and discuss the carrier biomaterials, particularly natural and synthetic polymers and their combinations in different formats for the delivery of rhBMP-2 and rhBMP-7 to preclinical and clinical sites of bone regeneration and repair.
Table 1

Overview of FDA-approved rhBMP-2 and rhBMP-7/rhOP-1 devices


FDA approval year


Delivery system

Type of FDA approval



Single-level anterior lumbar interbody fusion

Medtronic INFUSE® Bone Graft + LT-CAGE®


Acute, open tibial shall fractures

Medtronic INFUSE® Bone Graft + intramedullary nail



Posterolateral (Intertransverse) lumbar spinal fusion

Stryker® OP-1 putty

HDE program

Recalcitrant long bone non-unions

Stryker® OP-1 implant

ACS absorbable collagen sponge; LT-CAGE® lumbar tapered fusion device; HDE humanitarian device exemption; OP-1 putty; rhOP-1 + Type 1 bovine bone collagen matrix + carboxymethylcellulose as an additive; OP-1 implant rhOP-1 + purified bone-derived collagen particles as a scaffold

BMP delivery systems

Challenges in BMP delivery have been discussed in Part A of this review (preceding review). It has been for long suggested that to induce osteogenesis, a suitable delivery system is required for new bone to form due to the very short half-life (1–4 h) of these cytokines (Takaoka et al. 1991), thus requiring large single doses or multiple smaller applications (Talwar et al. 2001). Also, given that BMPs are not bony tissue-specific (Okubo et al. 2000), their localized (vs. systemic) and release-controlled (vs. un-controlled) delivery is necessary to prevent any un-desired and un-controlled ectopic bone formation in non-bony tissues of the body (Schmidmaier et al. 2008). The FDA of the USA approved bovine collagen-based delivery devices for rhBMP-2 and rhBMP-7 have limited indications in spinal interbody fusion and open tibial non-unions, respectively, mainly due to the large doses required to achieve the desired osteogenic effect where there is more exogenous BMP in a single dose than is present in 1000 humans, hence raising serious concerns regarding safety and cost (Kwon and Jenis 2005). Supra-physiological concentrations resulting from imperfect release kinetics of BMPs where 30% of the encapsulate is lost in the initial bust phase (Geiger et al. 2003) are additionally being related to severe clinical complications including generalized hematomas in soft tissue and para-implant bone resorption (Gautschi et al. 2003; Robinson et al. 2008).

BMP-specific carrier types and delivery materials

Researchers have commonly explored biomaterials with demonstrated osteoconductive properties. However, an ideal BMP carrier material needs to be osteoconductive, osteoinductive and osteogenic (Seeherman and Wozney 2005; De Long et al. 2007; Schmidmaier et al. 2008). Although intensively evaluated in animal studies as well as in clinical trials with satisfactory results, bovine collagen has also demonstrated some safety issues mainly owing to its xenogenic origin. Polymeric matrices have attracted attention over the last years to achieve the localized and controlled release of proteins over long periods of time and overcome limitations in enzymatic susceptibility, stability during storage and efficacy upon administration (Termaat et al. 2005; Moioli et al. 2007; Schmidmaier et al. 2008). Generally, the major categories of carrier materials (Table 2) include: (1) natural-origin polymers such as collagen, hyaluronans, soy- and alga-derived materials, and poly(hydroxyalkanoates); (2) inorganic materials and ceramics/cements such as hydroxyapatite, tri-calcium phosphates and -sulphates as well as bioglasses and metals; (3) synthetic biodegradable polymers such as poly(lactic acid) (PLA), polyglycolide (PLG) and their copolymers, poly l-lactic acid (PLLA), poly d,l-lactide-co-gycolic acid (PLGA) and poly ε-caprolactone (PCL); and (4) composites which are combinations that take advantage of each material class as well as other biomolecules (Fei et al. 2008).
Table 2

Major classes of carrier materials and selective experimental studies of BMP-incorporated delivery systems

Carrier material




BMP/material type/animal model site

Selected references

Natural polymers

Biocompatible, biodegradable, soluble in physiologic fluids with natural affinity for BMPs

Immunogenicity, pathogen transmission processing and sterilization difficulties

Collagen (gel, nano fibres, scaffolds and films), fibringlue, alginate and chilosan

BMP-2/getalin hydrogel/rabbit ulna

Yamamoto et al. (2006)

BMP-2/chilosan/rat mandible

Issa et al. (2008a, b)

Inorganic materials

Similar to bone degradable/non-degradable, osteoconductive, self-setting ability in vivo and have affinity for BMPs

Brittle, difficult mold and some calcium phosphate cements (CPC) are exothermic

CPC, bioactive glasses, Hydroxyapatite, hyaluronic acid, tri-calcium phosphates, metals, ceramics and calcium sulfate

BMP-2/Hyaluronic acid/in vitro/spine

Kim and Valentini (2002)

BMP-2/CPC in ACS/monkey/spine

Barnes et al. (2005)

BMP-2/CPC putty/rabbit cranium

Haddad et al. (2005)

Synthetic polymers

Excellent chemical and mechanical properties, easy to process and sterilize, flexible to tailor and reproducible

Inflammatory response, localized pH drop due to by-products acidity and limited biological function

Polymers of α-hydroxy esters such as PLA, PGA and copolymers of these two monomers (PLGA)

BMP-2/PLGA micro particle/rat cranium

Woo et al. (2001)


Saito and Takaoka (2003)

Injectable PLA-PEG/mouse/orthotropic

Saito et al. (2005)


According to combination from different material (and biomolecule) dasses

Complex manufacturing processes

Collagen-HA and titanium-PLLA

BMP-2/collagen PLG-alginate rat calvaria

Kenley et al. (1994)

BMP-7/PLGA-PLLA scaffold/rat s.c.

Wei et al. (2007)

BMP-7/liposomes coated with alginate and chitosan (layer-by-layer)/rabbit/libia

Haidar et al. (2008a,b)

Haidar et al. (2009)

Modified and updated from Li and Wozney (2001); Lee and Shin (2007)

Natural-origin polymeric carriers


Collagen is the most abundant protein in connective tissues of mammals and the major non-mineral component of bone with a well established role in cellular infiltration and wound healing (Geiger et al. 2003). It has been prepared in powders, membrane films and implantable absorbable sponges as well as in aqueous forms (Geiger et al. 2003; Lee and Yuk 2007). Despite being known for versatility and ease of manipulation, manufacturing collagen carriers is highly sensitive to several factors including mass, soaking time, protein concentrations, sterilization methods, buffer composition as well as pH and ionic strength (between collagen and the encapsulant) that directly affects the binding of rhBMPs (Gao and Uludag 2001; Lee and Yuk 2007). Absorbable collagen sponges (ACS) have been evaluated in numerous in vivo models and clinical trials. In the healing of a critical-sized radial defect stabilized by an external fixator, Sciadini and Johnson compared the efficacy of various dosages of rhBMP-2 delivered in an ACS to autogeneous bone grafts in 27 dogs. Defects treated with rhBMP-2 showed better healing than those treated with the ACS alone or autogeneous bone grafts (Sciadini and Johnson 2000). Paramount evidence, nonetheless, is derived from well-designed clinical trials. Table 3 lists selected examples of randomized clinical trials and clinical case series. The BESTT study investigated low (0.75 mg/ml) and high (1.5 mg/ml) doses of rhBMP-2 impregnated in an ACS. At 12 months, patients in the latter group had statistically significant accelerated healing, fewer invasive interventions and a lower rate of non-union than the control group (Govender et al. 2002). In patients requiring staged maxillary sinus floor augmentation, rhBMP-2/ACS safely induced adequate bone formation for the placement and functional loading of endosseous dental implants (Boyne et al. 2005). The use of rhBMP-2/ACS without concomitant bone grafting materials in critical-sized mandibular defects produced excellent regeneration in a very recent case review of 14 patients (Herford and Boyne 2008). On the other hand, no differences using rhBMP-7 incorporated in a type I collagen carrier over 24 months were detected in a prospective clinical trial (Friedlaender et al. 2001). It was concluded to be similar to autografts in the management of non-unions except for the pain factor associated with the donor site. Although eliminating the need to harvest autologous bone and alleviating the associated pain, animal-derived collagens are limited by their xenogenic nature (mostly bovine and porcine skin) where anti-type I collagen antibodies developed in almost 20% of patients treated with rhBMP-2/ACS (Sciadini and Johnson 2000, Geiger et al. 2003; Bessa et al. 2008). Also, sterilization is usually using ethylene oxide prior to soaking the sponge in the BMP solution; hence, with an effect on the release kinetics or the bioactivity of the protein within (Gittens and Uludag 2001). Furthermore, without delivery in situ, BMPs rapidly diffuse away from defect site (Sciadini and Johnson 2000) probably explaining some of the discrepancies noted in preclinical and clinical outcomes (Chen and Mooney 2003). Therefore, other sources of collagen (recombinant perhaps) and biomaterials are currently being evaluated for rhBMP delivery, though still at the preclinical stages.
Table 3

rhBMP-2 and rhBMP-7/rhOP-1 use in clinical studies




Total number of patients






Absorbable collagen sponge (ACS)

Anterior lumbar interbody arthrodesis

46 and 279 (2 trials)

Burkus et al. (2002a, b)

Bovine ACS

Open tibial shaft intermedullary fixation


Govender et al. (2002)

Bovine ACS

Cervical anterior arthrodesis


Baskin et al. (2003)

Xenogenic bone (Bio-Oss)

Maxillary implant placement


Jung et al. (2003)


Anterior lumbar interbody arthrodesis


Burkus et al. (2005)


Maxillary sinus floor augmentation


Boyne et al. (2005)


Mandibular continuity defects


Herford and Boyne (2008)








Type I collagen

Fibular osteotomy (critical-sized defects)


Geesink et al. (1999)

Type I collagen

Maxillary sinus floor elevation


van den Bergh et al. (2000)

Type I collagen

Tibial intramedullary fixation (non-union)


Friedlaender et al. (2001)

Type I collagen

Posterolateral lumbar arthrodesis (non-instrumented)


Johnsson et al. (2002)


Fresh tibial fracture (external fixation)


Maniscalco et al. (2002)

OP-1 putty*

Posterolateral lumbar fusions


Vaccaro et al. (2005)

OP-1 putty

Posterolateral lumbar arthrodesis (instrumented)


Kanayama et al. (2006)

OP-1 putty

Posterolateral lumbar arthrodesis


Vaccaro et al. (2008)

*OP-1 putty consists of rhOP-1, Type I bovine bone collagen matrix and an additive; carboxymethylcellulose sodium

Alginate and chitosan

Alginate (AL) is a non-immunogenic polysaccharide found abundantly in the surface of seaweeds used in a wide range of tissue engineering applications due to its gel-forming properties (Tönnesen and Karlsen 2002). Injectable in situ-forming AL gels were prepared then loaded with an osteoinductive growth factor (IGF-I), for example. Significantly accelerated proliferation of osteoblast-like MG-63 cells favorable for the conformal filling of bone defects was demonstrated (Luginbuehl et al. 2004). Furthermore, encapsulant release from AL matrices can be modulated by different parameters such as particle size, viscosity and chemical composition. Liew et al. (2006) in a recent investigation found that particle size affected the extent of burst release and the higher the viscosity the slower the encapsulant release. Chitosan (CH) is a cationic copolymer of N-acetyl-d-glucosamine and glucosamine prepared by N-deacetylation of chitin well-known for its biological, chelating and adsorbing properties (Kumar et al. 2004; Lee and Yuk 2007). Practical use of CH has been mainly confined to un-modified forms as they have solely demonstrated favorable biodegradation kinetics slower than polymers such as collagen. However, the chemical-modification and graft co-polymerization onto CH further improved controlling the release profile of bioactive molecules and some have been described as osteoinductive materials (Prabaharan 2008). Novel chemically-modified CHs with controllable photo-curability showed enhanced biocompatibility and bone tissue repair in athymic rats (Qiu et al. 2008). Furthermore, rhBMP-2/CH accelerated osteogenesis in a rat critical-sized mandibular defect. CH adapted well to the defect and had favorable release kinetics as revealed by the amount of new bone tissue (Issa et al. 2008a, b).

Hyaluronic acid

Hyaluronic acid (HA) is another naturally-occurring biopolymer, which plays a significant role in wound healing. HA and its derivatives have been largely studied in biomedical and tissue engineering applications as gels, sponges and pads and as a viscous gel injected percutaneously in ophthalmic surgery (O’Regan et al. 1994; Bessa et al. 2008). HA also has an osteoinductive action itself where it has been shown to result in improved bone formation in mandibular defects in comparison to collagen sponges when both carriers were used to deliver BMP-2 in rats as well as in a human clinical trial (Arosarena and Collins 2005). That is probably due to that HA-based delivery vehicles might posses the capacity to retain more BMPs than collagen (Kim and Valentini 2002). In addition, hyaluronans seem to stimulate the proliferation of bone marrow stromal cells, expression of osteocalcin, enhance ALP activity and interact positively with BMPs to generate direct and specific cellular effects. This increased affinity is attributed to HA being anionic thus forming ionic bonds with the cationic BMPs significant for potential future clinical applications (Peng et al. 2008). Other less common natural polymers in BMP delivery include gelatin, dextran and fibrin as they are limited by their mechanical strength and fusion with other biomolecules or biomaterials seems necessary, nonetheless with promising applications in angiogenesis (Young et al. 2005).

Inorganic materials


Hydroxyapatite (HAP) is well known for its osteoconductivity and has been widely used as a bone-substitute material clinically since the 1970s due of its ability to bond directly with bone (Li and Wozney 2001). Synthetic HAP comes in ceramic (porous and non-porous) or non-ceramic, cementable form (Moore et al. 2001) with only porous HAP being evaluated as a scaffold and a controlled release carrier (particles, powder, granules, disks or blocks) of BMPs for bone regeneration (Bessa et al. 2008). However, lack of bone induction due to the high affinity between the material and the BMPs in addition to the lack of resorption of the HAP and dependence on the geometry of the substratum was evident (Noshi et al. 2001). Therefore, HAP has often been combined with tri-calcium phosphates, collagen, other natural and synthetic polymers to form a more rigid, resorbable and porous BMP carrier. Generally, such composites have demonstrated better local BMP delivery and bone formation than HAP alone in various bone defects in vivo (Schopper et al. 2008; Kim et al. 2008).

Calcium phosphates and bioactive glasses

Inorganic materials such as calcium phosphate-based cements, ceramics and coatings (CPCs) have proven to be versatile carriers that can be formulated as implantable and injectable cements. They harden in vivo and can be used to deliver bioactive growth factors (in low temperature forms to prevent protein denaturation) with established prominent bone formation (Moore et al. 2001; Ginebra et al. 2006). Ceramics are known to mimic natural bony structure when implanted (Schmidmaier et al. 2008). Furthermore, overall, lower dosages of BMPs are required with the use of CPCs compared to other carriers (Ginebra et al. 2006). However, phase separation during injection, lack of intrinsic macroporosity to allow cell infiltration, intrinsic radiopaque nature and decreased mechanical tensile and shear properties compared to bone and other materials are among the main disadvantages of CPCs (Seeherman and Wozney 2005). Modifications to increase injectability and macroporosity in vivo have been recently reported (Bohner and Baroud 2005). Also, CPCs have been used as a bulking agent to improve the osteogenicity of ACS loaded with rhBMP-2 where it helped lowering the BMP dose (>3-fold) in a spinal fusion model in the non-human primate model (Barnes et al. 2005). Bioactive glasses (BG) are a group of hard and non-porous silica-based bioactive compounds which are known to bond directly to bone due to their good osteointegrative and osteoconductive properties. They have different resorption rates depending on their chemical compositions where solubility is proportional to the sodium oxide content (Välimäki and Aro 2006). BG, such as 45S5 Bioglass, are commonly used as filler material for fractured bone, augmentation of the alveolar ridge and vertebral implants. Recent studies have shown that BG induce a high local bone turnover in vitro and in vivo where they increase the BMP effect, support osteoblast growth and favor osteoblast differentiation by stimulating the synthesis of phenotypic markers like alkaline phosphatase, collagen Type I and osteocalcin (Moore et al. 2001; Välimäki and Aro 2006).

Synthetic biodegradable polymeric carriers

Unlike natural polymers and collagen, synthetic biodegradable polymers pose no danger of immunogenicity or possibility of disease transmission. A number of synthetic biodegradable polymeric delivery systems for BMP-2 were discussed in two recent reviews (Saito and Takaoka 2003, 2005). The most commonly used polymers herein are PLLA and PLGA. Their degradation is primarily via hydrolysis and different proportional combinations of PGA and PLA, for example, demonstrate various material properties which in turn affect biodegradability. Material crystallinity and scaffold morphology (pore size/number) also play an important role in biodegradability where a more porous scaffold degrades faster as will that comprised of low molecular weight polymers. Synthetic polymers can be processed into highly porous 3-D scaffolds, fibers, sheets, blocks or microspheres (Seeherman and Wozney 2005). BMP release is by means of diffusion, polymer swelling followed by fast diffusion-controlled release, and polymer erosion (Engstrand et al. 2008). Bioresorbable PLA/PGA beads were found to be superior to collagen when used to deliver rhBMP-2 to transosseous rat mandibular defects (Zellin and Linde 1997). PLGA was evaluated successfully in several canine defect models such as for BMP-2-induced periodontal regeneration, maxillary alveolar cleft repair, and segmental ulnar long-bone defects (Sigurdsson et al. 1996; Mayer et al. 1996; Itoh et al. 1998). Nonetheless, the main limitation is their acidic breakdown by-products and the associated risk of inflammatory response if not cleared quickly which is detrimental to the stability of the encapsulated BMPs and the overall therapeutic outcome (Saito et al. 2005). As a result, a continuous supply of osteoinductive factors would be crucial to compensate for polymer degradation (Schliephake et al. 2008).


Combinations of different material classes (Table 2) have been used to optimize the benefits and overcome the limitations of many of the above materials. Examples of recent composites include glycidyl methacrylated dextran (Dex-GMA)/gelatin scaffolds containing microspheres loaded with rhBMP-2 and implanted into the periodontal defects of dogs (Chen et al. 2007), PLGA-gelatin composites for the delivery of rhBMP-2 to vertical alveolar ridge augmentation in dogs (Kawakatsu et al. 2008), HAP-coated porous titanium/rhBMP-2/HA in the metaphysic of the distal femur of rabbits (Peng et al. 2008) and PEG hydrogel/rhBMP-2/HAP/TCP granules in the rabbit calvarial bones (Jung et al. 2008). A 3-D, highly porous PLA/HAP/collagen scaffold was prepared for use in healing of canine segmental bone defects. The HAP/collagen portion was to mimic the natural extracellular matrix of bone, with the collagen serving as a template for apatite formation. All defects healed satisfactorily (Hu et al. 2003). Fu et al. recently combined rabbit mesenchymal stem cells (MSCs) with AL/BMP-2 and implanted the composite in a posterolateral intertransverse fusion model in 24 rabbits. Results demonstrated that MSCs delivered with rhBMP-2 (2.5 mg) are more osteoinductive than MSCs without rhBMP-2 and that the composite material enhanced bone formation and spine fusion success (Fu et al. 2008).

Nano- and micro-particles from synthetic polymers and natural polymers (Fig. 1) are other dosage forms that have consummated much attention for the localized and release-controlled delivery of growth factors due to their attractive tendency to amass in sites of inflammation (Lee and Shin 2007). Enhanced in vivo tissue regeneration using PLGA and gelatin microparticles for growth factor release was reported (Park et al. 2005). Compared to microparticles, nanoparticle and nanofiber delivery systems have demonstrated superiority in terms of longer residencies in general circulation, consequently extending the bioactivity of the entrapped molecule (Nair and Laurencin 2008). In a recent example of a combined localized and release-controlled delivery system, PLGA nanospheres (NS) immobilized onto prefabricated nanofibrous PLLA scaffolds were used to load and deliver rhBMP-7 (Wei et al. 2007; Ma 2008). BMP-7 delivered from NS-scaffolds induced significant ectopic bone formation while passive adsorption of the protein into the scaffold resulted in failure of bone induction either due to the loss of protein bioactivity or its rapid release from the scaffolds upon implantation in vivo.
Fig. 1

Selective electron micrographic images of BMP carriers and delivery systems and carrier biomaterials evaluated for BMPs. a Porous collagen sponge, bar = 100 μm b PLGA-coated gelatin sponge c PLGA microspheres d Heparin-conjugated PLGA nanoparticles, bar = 100 μm e core-shell nanoparticles constituting liposomes, alginate and chitosan f PLGA nanofibres. Adopted from Moioli et al. (2007); Lee and Shin (2007); Kim et al. (2008); Haider et al. (2009)

We have successfully encapsulated a model protein, bovine serum albumin in monodisperse and non-toxic nanoparticles constituting a core of cationic liposomes and a shell constructed through the layer-by-layer electrostatic-based self-assembly of alternating layers of anionic AL and cationic CH. The release profile was characterized by an initial burst followed by sustained albumin release, highly desirable for growth factor delivery; particularly in large bony defects (Haidar et al. 2008a). In a subsequent work, the ability of the core-shell nanoparticulate delivery system to encapsulate a range of concentrations of BMP-7 was evaluated. The system exhibited high physical stability in simulated physiological media as well as an extended shelf-life allowing for immediate protein loading prior to administration, preventing degradation or loss of the encapsulant. A sustained triphasic linear release of BMP-7 was evident for an extended period of 45 days with the bioactivity of the protein maintained via enhancing pre-osteoblast differentiation (Haidar et al. 2008b). In a rabbit model of tibial distraction osteogenesis, accelerated osteogenesis and consolidation was evident following a single injection of the nanoparticles loaded with a dose of no more than 1 μg rhBMP-7 (Haidar et al. 2009) in comparison to earlier results from a single injection of rhBMP-7 (75 μg in saline), accentuating the role of the injectable localized and release-controlled nanoparticles (Mandu-Hrit et al. 2006). Other groups investigated injectable scaffolds and matrices for drug delivery in bone tissue engineering (Kretlow et al. 2007). A recent example is the work of Hosseinkhani et al. (2007) where an injectable hydrogel of self-assembled peptide-amphiphile and BMP-2 was fabricated. Significant homogenous ectopic bone formation in the back subcutis of rats was demonstrated. A 3-D scaffold for the sequential delivery of BMP-2 and BMP-7 was recently developed (Basmanav et al. 2008). The system consisted of microspheres of polyelectrolyte complexes of poly(4-vinyl pyridine) and AL loaded with both proteins and incorporated in PLGA scaffolds. Neither BMP-2 nor BMP-7 delivery had any direct effect on cellular proliferation; however, their co-administration enhanced osteogenic differentiation to a higher degree than with their single administration. This was suggested to be due to the physical properties (pore size and distribution) of the foams. In conclusion, no single clinically-effective delivery system for rhBMP-2, rhBMP-7 or any other bone growth factors has been developed to date.

Future prospects

The use of delivery systems is crucial for reliable bone formation and economic application of BMPs. Many carrier materials have been investigated. Collagen matrices have been clinically successful nevertheless with shortcomings including biodegradability, local retention and controlling release kinetics of BMPs. A limited number of other potential delivery systems have been developed and are still at the preclinical stage. Nonetheless, advances in the field will eventually lead to novel customized and optimized BMP-specific carrier materials. Better, safer and more integrated minimally-invasive drug delivery systems that utilize smaller amounts of BMPs effectively are needed. They will be insoluble under physiological conditions in order to retain the BMP yet absorbable by host tissue after implantation/administration so that be replaced by the regenerating bone, preferably following single application. Controlled prospective clinical trials should follow. The use of BMPs and other less understood morphogens will extend into much broader range of orthopaedic as well as craniofacial and oro-dental indications including bone, cartilage and tendon/ligament (and periodontal) tissue regeneration and repair.


The authors wish to thank grants from the Shriners of North America, Fonds de la Recherche en Santé du Québec, the National Science and Engineering Research Council, the Canadian Institutes of Health Research (CIHR)—Regenerative Medicine/Nanomedicine and the Center for Biorecognition and Biosensors (CBB), McGill University, Montréal, Québec, Canada. Dr. Haidar acknowledges scholarships from the Center for Bone and Periodontal Diseases Research and the Shriners Hospital, Montréal, Québec, Canada.

Copyright information

© Springer Science+Business Media B.V. 2009