Autogenous bone or autograft is bone transplanted within the same individual. In contrast to other bone grafting materials, autografts may contain viable cells as well as matrix proteins including collagen (mainly type 1), osteopontin, bone sialoprotein, osteonectin, osteocalcin, fibronectin, transforming growth factor-β-type 1 (TGFβ-1and bone morphogenetic proteins (Urist 1965, Gorski 2011). Cell viability in transplanted autografts may be preserved for up to two hours if cells are stored correctly in saline (Laursen et al. 2003). On the other hand, osteocytes degenerate and resorption of the graft increases when autografts are stored dry for more than 30 minutes before grafting (Rocha et al. 2013). The inorganic component of bone consists mainly of hydroxyapatite. However, based on systematic comparisons of synthetic and biologic apatites using X-ray diffraction, infrared spectroscopy, and chemical analysis, biologic apatites should be considered carbonate apatites with traces of magnesium, sodium, fluoride, and chloride (LeGeros 2008). Bone apatite crystals are approximately 5x40 nm in size and are arranged in a lattice parallel to the collagen fibrils (LeGeros 2008).

Donor site

Most autografts used in oral and maxillofacial osseous reconstruction are free grafts. Free autografts may be utilized as blocks or in particulate form and can be harvested intraorally or extraorally according to the clinical indication and the amount and type of graft needed (Nkenke & Neukam 2014). The character and quality of bone and thereby also of harvested autografts shows large inter- and intra-individual differences in terms of geometry, cortical-to-cancellous bone ratio, osteoconductivity, osteogenicity, and osteoinductivity depending on age, gender, systemic disease, medication, donor site, and harvesting method (Boskey & Coleman 2010, Seeman 2008). 

The most frequently used extraoral donor sites for harvesting bone blocks for oral and maxillofacial reconstruction include the iliac crest (anterior or posterior), the tibia, and the calvaria (Nkenke & Neukam 2014). The iliac crest is formed by endochondral ossification, whereas calvarial bone is of membranous origin. There has been discussion of whether the embryonic origin of a bone graft is of importance for its behavior as a bone graft, but today it is rather believed that the architecture (the cortico-cancellous ratio) of the bone graft determines the biological behavior (Ozaki et al. 1999). 

Bone from the iliac crest has a thin outer cortical layer and a large cancellous component, whereas calvarial and mandibular bone are mainly cortical. Correspondingly, grafts from the iliac crest have shown faster revascularization but also more initial resorption clinically than calvarial grafts. Extraoral bone harvesting generally requires general anesthesia, whereas intraoral harvesting procedures most often can be accomplished using local anesthesia with or without sedation. In addition, extraoral harvesting procedures are accompanied by more frequent and more severe complications and increased morbidity as compared to intraoral bone harvesting. 

The most frequent intraoral donor sites for bone block harvesting are the mandibular body/ramus and the chin. The architecture of bone blocks from the mandibular body/ramus is mainly cortical, whereas cortico-cancellous blocks may be harvested from the mandibular chin. This is probably the reason for faster revascularization of chin blocks than mandibular body blocks. The morbidity related to bone harvesting from the mandibular body/ramus is low compared to the mandibular chin area. Therefore, most authors consider the mandibular ramus as the donor site of choice for intraoral bone harvesting.

Harvesting methods

Due to their autogenous origin, autografts are biocompatible and safe, unless contaminated during the harvesting or application procedure. Autogenous bone is hydrophilic and thus easy to handle and mix with blood or saline. All autografts are resorbable. However, their resorbability depends on the donor, the donor site, and the harvesting method (Nkenke & Neukam 2014). Bone blocks have been shown to be more mechanically stable clinically and thereby better able to preserve the augmented volume than particulated autografts (Jensen & Terheyden 2009). 

Different harvesting methods result in very different morphology of autograft particles (Saulacic et al. 2015). In general, the content of growth factors and viable cells in autografts decreases the more the bone is processed or manipulated during and/or after the harvesting procedure. More viable bone cells are found in unprocessed cancellous bone chips than in cortical chips. This is also reflected in the finding that more cells can be cultured from cancellous bone in the maxilla than from the more compact bone in the mandible. 

Bone blocks may be particulated using a bone mill. The character of the resulting autograft particles depends on the geometry of the original block and on the bone mill utilized (Erpenstein et al. 2001). Small autograft particles have a larger surface area, which potentially increases the presentation of growth factors and BMPs as compared to larger particles. As a result, the osteoconduction and osteoinduction are increased (Pallesen et al. 2002). However, osteoclasts too have increased access to the enlarged surface area, leading to a higher resorption rate (Pallesen et al. 2002). Preparation of a bone block in a bone mill reduces the amount of viable cells in the graft (Springer et al. 2004). However, the rate of revascularization of the augmented area is significantly higher with particles than with blocks.

Particulate autogenous bone may also be harvested by using bone scrapers, piezo or bone filters connected to the surgical suction device. Bone scrapers may be single- or multi-use devices that collect small bone particles from the bone surface. By scraping the surface, mainly cortical bone chips are harvested. The size of the particles is smaller than those prepared by bone milling and larger than the ones obtained with piezosurgery and bone filters (Miron et al. 2013). Despite their cortical origin, in vitro studies have documented viable osteogenic cells to be present in grafts harvested with bone scrapers (Miron et al. 2011). Experimentally, the ability of bone mill- and bone scraper-derived autografts to support osteoblastic differentiation and production of mineralized tissue was significantly increased compared to those harvested with a piezoelectric device and with a bone filter (Miron et al. 2011). In addition, the concentration of growth factors known to be involved in bone formation, like BMP-2 and VEGF, is higher in autografts prepared by bone mill and bone scraper than with piezosurgery and bone filters (Miron et al. 2013). However, the potential impact of this increased expression of bioactive molecules has not been documented in vivo (Saulacic et al. 2015).

Bone dust collected with bone filters is shown to be contaminated with bacteria from the oral micro flora (Lambrecht et al. 2006). Preoperative mouth rinsing with chlorhexidine gluconate, a stringent suction protocol where the device connected to the bone filter is exclusively used for suctioning at the site of implant bed preparation, and post-harvesting rinsing of the collected bone dust with an antibiotic solution has been shown to reduce the bacterial load of the graft without being able to eliminate it. The potential risk of infectious complications caused by this contamination has raised concern about the appropriateness of this graft for intraoral bone regeneration procedures. However, bacteria have also been cultured from grafts harvested by bone scrapers, trephines, rongeurs, and bone chisels as well, but in lower counts as compared to autografts from bone filters (Manzano-Moreno et al. 2015).

In general, autograft is the only available bone augmentation material that can stimulate new bone formation, whereas all bone substitute materials act as passive fillers, that may support new bone formation long-term, but does not accelerate it (Fig. 1).

Biocompatibility and safety

The risk of disease transmission with musculo-skeletal allografts is lower than for most other biological grafts like whole blood. Donor screening, organized recall systems implemented by medical authorities, and official guidelines for surgeons have increased the safety of allografts considerably (Mellonig 1995; Holtzclaw et al. 2008). According to local legislation, different regulations and guidelines may apply to the acquisition, processing, and use of allografts (Holtzclaw et al. 2008). 

Fresh allografts from living donors and minimally processed unmatched allografts bear a considerable risk of rejection (about 50%) and in general are not used. Remnants of bone marrow and periosteal cells in particular are highly antigenic. To reduce the risk of immunological reactions and disease transmission, allografts may be processed in different ways. After donor screening and pre-processing microbiological and serological testing, all allografts are manually cleaned for residual soft tissues. Thereafter, decontamination may be conducted using various solutions and antimicrobial agents (Holtzclaw et al. 2008). Transmission of HIV, hepatitis B, and hepatitis C has been documented in a few cases with the use of FFB (Salvucci 2011). However, a combination of established donor screening criteria, certification of tissue banks, and strict guidelines for donor bone processing has increased safety considerably, and no cases of disease transmission have been documented in recent years. Immunogenicity of FDBA and DFDBA are reduced in comparison to FFB, which is ascribed to the additional defatting, removal of bone marrow, denaturation of cellular elements, and elimination of moisture during the freeze-drying process (Holtzclaw et al. 2008). 

Allogenic bone grafts or allografts are derived from a donor and used in another individual of the same species. Allografts are frequently used in orthopedic, oral, and maxillofacial surgery. The bone quality of allografts may be expected to vary to the same extent as that of autografts. However, donor screening and careful macroscopic evaluation of the harvested bone makes a limited degree of standardization of allografts possible.

Allografts are categorized according to the processing technique applied after harvesting. Fresh unprocessed allografts from living donors are rarely used. Fresh frozen bone (FFB) is exposed to temperatures below -70°C. To prepare a freeze-dried bone allograft (FDBA), in addition to freezing, the donor bone is defatted, the bone marrow is removed, and it is dehydrated using different solvents. Finally, for a demineralized freeze-dried bone allograft (DFDBA), the HA skeleton of the allograft is removed using hydrochloric acid, with the aim of exposing the osteoinductive molecules in the allograft additionally.

Handling and mechanical stability

Allografts come in unlimited quantities and in different forms: blocks (cortical or cortico-cancellous blocks of FFB, or FDBA), particles (cortical or cancellous particles of FFB, FDBA, or DFDBA), and putties (DFDBA). Blocks may be delivered individualized based on a preoperative 3D scan of the bone defect to be grafted (Schlee & Rothamel 2013). Duration of surgery and morbidity from donor sites can thus be reduced as compared to operations including harvesting of autografts.

The freezing of FFB to -70°C has a limited effect on mechanical stability. FDBA, on the other hand, is mechanically less stable than the original allogenic block. Micro cracks along collagen fibers caused by the dehydration process are believed to cause the structural changes leading to compromised mechanical properties of FDBA. These effects seem to be magnified additionally by combining freeze-drying and gamma irradiation for sterilization of allografts. In DFDBA, additional loss of mechanical resistance occurs with the removal of the inorganic structural components of the bone. Therefore, in load-bearing areas, mechanical stability must be ensured by other means than the use of DFDBA.

Degradation

The amount and pattern of degradation depends on the donor site and processing technique. Like autografts, a higher resorption rate is observed for cortico-cancellous than cortical allografts (Spin-Neto et al. 2015). However, FFB and FDBA may be expected to degrade even faster than corresponding autografts due to activation of the immune system by mismatch of the major histocompatibility complex (MHC). In addition, the mode of resorption differs between allografts and autografts. Whereas autografts are resorbed and remodeled through external osteoclastic activity on the surfaces and by bone metabolizing units (cutting cones), allografts seem to be remodeled by surface resorption on the periphery of the grafts and in already existing Haversian canals only (Enneking & Campanacci 2001).

Osteoconduction and osteoinduction

In contrast to autografts, viable osteogenic cells are not contained in any of the allografts, irrespective of the processing technique, and thus cannot contribute to bone formation.

FDBA and DFDBA have been shown to be biocompatible and to contain osteoinductive molecules such as BMPs. However, there are big differences in the concentration of osteoinductive molecules between commercially available batches of FDBA and DFDBA (intervariability and intravariability) (Boyan et al. 2006). The processing of allografts has been shown to influence revascularization of the grafts significantly. Histologically, FDBA and DFDBA allografts were observed embedded in newly formed bone as an indicator for osteoconductive properties (Buser et al. 1998). However, so far it has not been possible experimentally to demonstrate the clinical relevance of the osteoinductive molecules in FDBA and DFDBA (Boyan et al. 2006). 

Xenogenic bone substitute materials or xenografts consist of either bone mineral derived from animals or alternative natural sources such as bone-like minerals derived from calcifying corals or algae. 

A multitude of animal species have been used as donors for animal-derived xenografts. Multiple systematic reviews of experimental, clinical, and histomorphometrical data have concluded that deproteinized bovine bone mineral (DBBM) is the best documented bone substitute material in periodontal, oral and maxillofacial reconstruction used alone or as a composite graft in combination with autogenous bone (Jensen & Terheyden 2009). The present article will therefore focus on DBBM among the animal-derived xenografts. The organic component of bone may be eliminated by chemical methods, heat treatment, and/or -radiation. Cortical, cancellous, or cortico-cancellous blocks or particles can thereafter be prepared according to clinical indication.

Bone substitute materials have been synthesized from different calcifying marine algae species (Kasperk et al. 1988). The CaCO3 skeleton of the algae is transformed into calcite (Ca5CO3) by pyrolysis (700°C) over 30 hours. Thereafter the calcite may be transformed into fluoro-HA, HA, or BCP with different HA/TCP ratios by hydrothermal exchange reactions (Schopper et al. 2005). Using these methods, the three-dimensional structure of the algae skeleton is maintained, including parallel-arranged pores with a mean diameter of 10 μm and interconnections of 1 μm (Schopper et al. 2005).

Certain reef-building corals have porous CaCO3 skeletons like calcifying algae. Pore size and configuration vary between species (Guillemin et al. 1987). However, the intra-species variability is low. The pores are arranged in parallel and in part mimic osteons in human cortical and cancellous bone, with mean pore sizes between 200 μm and 500 μm, and interconnections of 190 and 260 μm, respectively. Their pore sizes and interconnections are thus larger than the ones observed in calcifying algae. On the other hand, the total HA surface area of coral is smaller (Kasperk et al. 1988). In parallel with algae-derived bone substitute materials, most often, a hydrothermal exchange reaction is performed to convert the CaCO3 skeleton into HA (Jensen et al. 1996). In addition, a partial exchange reaction has been documented to convert the outer shell of the lattice into HA only, leaving the central portion as the original CaCO3 (Jensen et al. 1996).

Biocompatibility and safety

Xenografts of bovine origin have been tested as fresh, fresh frozen, and freeze-dried. Today, however, deproteinized bovine-derived products are almost exclusively used due to inconsistent results and a variety of inflammatory and immunologic complications with the products containing organic remnants. Organic components disappear from bone when exposed to temperatures of around 350°C (Haberko et al. 2006). Still, concern has been raised regarding possible residual protein content in unsintered DBBM introducing a risk of disease transmission (Schwartz et al. 2000). Importantly, there are no reports of disease transmission via unsintered DBBM. Concerns regarding safety of sintered DBBM, algae-, and coral-derived biomaterials have not been raised.

Handling and mechanical stability

DBBM and coral-derived bone substitute materials are available as porous blocks and particles, whereas algae-derived bone substitute materials come as particles only.

Cancellous DBBM blocks are considered too brittle to allow stable screw fixation, making their use as onlay grafts unpredictable (De Santis et al. 2015). 

Degradation

Over the years, controversy has remained as to whether DBBM is truly resorbable. It has been shown in vitro and in vivo that osteoclast-like cells are able to proliferate on a bovine-derived xenograft and produce resorption pits (Jensen et al. 2015). However, in vitro, the osteoclast-like cells were reduced in number and size, and the resorption pits were less pronounced compared with native bovine bone. The presence of multinucleated giant cells on the surface of DBBM particles has been documented in several clinical as well as animal studies (Jensen et al. 2011, Mordenfeld et al. 2010). Positive TRAP staining has often been used to identify these cells as osteoclast-like, and in addition, scanning electron microscopy has provided ultrastructural indications that multinucleated giant cells on DBBM present ruffled borders and sealing zones that are unique for osteoclasts (Jensen et al. 2015). On the other hand, long-term animal studies have demonstrated very little if any volume reduction of DBBM particles over time (Thorwarth et al. 2006, Jensen et al. 2009). Human biopsies after sinus floor elevation and after contour augmentation in the anterior maxilla have confirmed that DBBM particles still can be observed without signs of significant resorption up to ten years postoperatively (Mordenfeld et al. 2010). Osteoclast-like cells on the surface of DBBM placed in an osseous environment are, therefore, suggested to take part in a very slow physiologic remodeling process and not to be related to an inflammatory reaction. On the other hand, pronounced resorption has been documented when DBBM particles are situated in a non-osseous environment during early healing (Jensen et al. 2015, Busenlechner et al. 2012).

Thermal treatment of animal-derived xenografts induces structural changes in the bone according to the temperature applied (Haberko et al. 2006, De Cavalho et al. 2019). At >500°C, the apatite crystals begin to increase in size, the nano-porosity in between the crystals reduces, and the surface area decreases. At temperatures exceeding 700°C, the CO3 groups of the inorganic lattice start to decompose, whereas CaO remains detectable. Therefore, sintered DBBM biomaterials should be expected to be less prone to undergo remodeling than DBBM biomaterials deproteinized using lower temperatures. In vitro, seeded osteoclasts produced fewer and less pronounced resorption lacunae on sintered versus unsintered DBBM particles (Taylor et al. 2002).

Algae-derived fluoro-HA, HA, or BCP xenografts have also been shown to take part in bone remodeling (Schopper et al. 2005). The degree of resorption may be tailored in algae-derived BCPs by changing the HA/TCP ratio. Multi-nucleated giant cells resembling osteoclasts have been observed on algae-derived biomaterial surfaces under degradation. However, no TRAP staining or ultrastructural analysis has further determined the nature of these cells.

Coral-derived HA materials have not shown signs of degradation over time, irrespective of the pore size (Jensen et al. 1996, Buser et al. 1998). However, partial conversion of only the original CaCO3 skeleton allows modification of the degradability of the biomaterial (Jensen et al. 1996). Whether the degradability of the partially converted CaCO3 is cellularly mediated or physico-chemical has not been reported.

Degradable and non-degradable xenografts may be used as blocks or particles. The rate of degradation should be expected to increase with an increase in the surface area for particles as compared to blocks of degradable biomaterials. On the other hand, non-degradable materials will not turn degradable by using particles instead of blocks or by reducing particle size (Jensen et al. 2015), unless the particle size comes below the threshold (≈50 μm) at which direct phagocytosis by macrophages or multinucleated giant cells becomes possible.

Osteoconduction and osteoinduction

All cells and proteins are deliberately removed during the processing of xenogenic biomaterials. Therefore, xenografts do not contain cells or bioactive molecules that can contribute to the attraction of mesenchymal stem cells and their differentiation into osteoblasts. Whenever xenografts are placed in an osseous environment, bone ingrowth has been documented almost exclusively to take place from the pre-existing bone walls, and not in the center of the grafts as could be expected for biomaterials with high osteoinductive potential. In addition, the rate of bone formation is always lower when xenografts are compared to autografts within the same experimental setting (Jensen et al. 2006, Jensen et al. 2009, Buser et al. 1998) (Fig. 2). When xenografts are compared to autografts as onlay grafts for horizontal or vertical ridge augmentation, corresponding to a one-wall bone defect, the lower osteopromotive potential is consistently documented, with bone formation only in the part closest to the recipient bone surface in animal experiments (Schwartz et al. 2010) as well as clinically (Pistilli et al. 2014). In contrast, autografts are usually completely penetrated by bone (De Santis et al. 2015, Araujo et al. 2002).

The eventual amount of newly formed bone in bone defects grafted with xenografts with a low substitution rate is most often reduced compared to defects grafted with more degradable bone substitute materials and autografts (Jensen et al. 2009, Buser et al. 1998). This is partly explained by the limited resorption of the biomaterials leaving less space for new bone formation, and partly by the fact that a bone marrow volume of 20%–30% is needed to preserve normal bone homeostasis when steady state is reached (Jensen et al. 2009, Buser et al. 1998, Jensen et al. 2014). 

The deproteinization technique of animal-derived bone substitute materials seems to have a significant impact on the osteoconductive properties of DBBM. In a comparative rabbit study, sintered DBBM has thus displayed less osteoconductive potential than unsintered DBBM (Jensen et al. 1996). Since no differences could be demonstrated with regard to biochemical composition and other material characteristics, this most likely reflects the production-related changes in surface characteristics in terms of crystal growth, reduced nano-porosity and surface area (Haberko et al. 2006). 

Alloplastic bone substitute materials are synthetically manufactured. They can be categorized in four groups: calcium phosphates (CaP), bioactive glasses, polymers, and metals. Of these, calcium phosphates, and especially hydroxyapatite (HA) and tricalcium phosphate (TCP), have been most intensively studied, due to their composition, which closely resembles the inorganic phase of bone (LeGeroz 2008; Bohner 2000). Therefore, the present article will focus on the CaP materials.

Preparation methods

CaP ceramics for medical applications can be prepared in different ways (LeGeroz 2008, Bohner 2000, LeGeroz et al. 2003):

• Precipitation from an aqueous solution around room temperature
• Solid-state reactions: calcium- and phosphate-containing compounds are mixed, compressed and sintered (>900°C)
• Hydrothermal reactions: the solid-state reactants are heated under pressure (<400°C)
• Sol-gel reactions: solutions or colloids are mixed, aged, dried, and sintered (>900°C).

Interconnecting macroporosity may be introduced by adding porogens (e.g., naphthalene, H2O2, polymeric porogens) or by using foaming methods (LeGeroz 2008, Tadic et al. 2004).

Biocompatibility and safety

Synthetic CaP biomaterials are in general considered biocompatible and bear no risk of disease transmission or immunologic reactions.

Handling and mechanical stability

Alloplastic CaP biomaterials may be prepared in any form desired. HA, TCP, and BCP biomaterials have been tested experimentally as particulate as well as blocks. However, alloplastic blocks have only been tested in protected or partly protected defects. Although technologic developments today allow the preparation of CaP materials with improved mechanical characteristics, and their combination with polymer technology to reduce brittleness, only a few experimental animal studies have tested allopastic CaP blocks for horizontal ridge augmentation (Yeo et al. 2012, Kirchhoff et al. 2011). So far, no alloplastic CaP block has been documented to perform predictably as an onlay or interpositional graft. 

Degradation

Degradation of alloplastic CaP bone substitute materials can be tailored to a larger extent than the other classes of bone substitute materials. Degradation is dependent on surface area (particle size, porosity, surface roughness), crystallinity, crystal size, Ca:P ratio, ionic substitution, and for BCP materials the HA:TCP ratio (Jensen et al. 2007, Jensen et al. 2009, Eggli et al. 1988, Doi et al. 1999, Detsch et al. 2010, Detsch et al. 2008, Chen et al. 2008).

Crystalline HA is generally considered non-resorbable, whereas β-TCP has a high resorption rate (LeGeroz 2008). However, more recently introduced nano-crystalline non-sintered HA materials have been shown to undergo resorption (Rumpel et al. 2006). Degradation rates of BCP biomaterials can be altered by changing the HA:TCP ratio. However, BCPs with an identical HA:TCP ratio of 60:40, surface area, and porosity produced by either a solid state reaction or by precipitation may have significantly different degradation rates (Gauthier et al. 1999). Therefore, documentation of the exact material characteristics needs to be available to predict in vivo behavior.

The exact mechanism behind degradation of CaP biomaterials is not known. In vitro studies have documented that osteoclast-like cells may produce resorption pits on synthetic CaP biomaterials (Doi et al. 1999). CaP biomaterials with a high TCP content may, however, release Ca2+ ions in concentrations too high to allow cellular attachment (Detsch et al. 2008). These findings correspond well with in vivo indications that β-TCP degrades by physico-chemical dissolution of the surface, whereby smaller fragments are released from the biomaterial surface and phagocytosed by macrophages (Jensen et al. 2006). TRAP-positive multi-nucleated giant cells have frequently been identified on the surfaces of BCP with a lower degradation rate, indicative of an osteoclast-like nature. However, classic resorption lacunae (Howship’s lacunae) have not been observed (Jensen et al. 2007, Jensen et al. 2009), and an ultrastructural analysis of these cells failed to demonstrate osteoclast features like sealing zones and ruffled borders (Jensen et al. 2015). Instead, these multinucleated cells seem to dissolve the TCP compartment of the BCP biomaterial, undermining the surface, subsequently allowing phagocytosis of smaller BCP fragments directly from the surface (Jensen et al. 2015).

Osteoconduction and osteoinduction

Alloplastic bone substitute biomaterials do not contain osteogenic cells or osteoinductive molecules. When implanted in an osseous environment, bone formation exclusively progressed from pre-existing bone walls and not in the center of the grafted defects as might be expected from biomaterials with a high osteoinductive potential (Yuan et al. 2010). 

The amount of bone formation around various alloplastic CaP biomaterials has been studied in a vast number of animal studies. In self-contained bone defects, pure TCP-based biomaterials generally increase bone formation as compared to pure HA-based materials with otherwise identical material characteristics (Jensen et al. 2007, Eggli et al. 1988). Also, for BCP materials, the rate of bone formation increases with increasing TCP content in the composite material with otherwise identical material characteristics (decreasing HA/TCP ratio) (Jensen et al. 2009). On the other hand, bone formation has been observed mainly to take place at a distance from surfaces of biomaterials with a high TCP content in the early healing phases (Jensen et al. 2005, 2006, 2007, 2009, Yang et al. 2014). This may be explained by the high Ca2+ concentration close to the highly resorbable β-TCP surface rendering it less attractive for direct adhesion of osteoblasts (John et al. 2003, Rice et al. 2003, Wang et al. 2004, Curran et al. 2005). 

Most alloplastic CaP materials on the market today have interconnected porosities, with pore diameters and interconnections of >100 μm (Klein et al. 2009). Porosity does not by itself add to the osteoconductivity of a biomaterial. However, macro porosities facilitate bone ingrowth, and together with micro and nano porosities they increase the surface area available for osteoconduction. Comparative studies confirm that increasing the macro pore size and the size of interconnections of otherwise identical BCP materials has a greater positive effect on bone formation than increasing the total porosity (Gauthier et al. 1998, Tamai et al. 2002). Studies on different sized filler particles in injectable BCP cement have shown a reverse relationship between particle size and bone formation. However, it cannot be determined from the studies whether this increased bone formation is a result of a larger osteoconductive surface area of the smaller particles or if the additional resorption of the smaller particles merely provided space for more bone formation.

Surface characteristics of CaP biomaterials have been shown to have a marked influence on protein adsorption, cell adhesion, proliferation and osteoid secretion in vitro. Surface architecture of otherwise identical BCP blocks had a more pronounced effect on osteoinduction than micro porosity when implanted heterotopically in dogs (Zhang et al. 2015). How surface architecture drives osteogenesis is, however, not yet clear. 

Despite expanding knowledge of material characteristics influencing osteoinduction and osteoconduction, the ideal CaP biomaterial surface has still not been identified. In addition, classical preparation methods to produce macroporous CaP biomaterials suffer from limitations in controlling nanoporosity, pore interconnectivity, and surface topography. With the refinement of 3D printing techniques, it is expected to be possible to recreate ideal surface characteristics down to the nano-scale level in the future.