Uncemented fixation has been a major method employed in arthroplasty for decades [1, 2]. To this end various rough and porous surfaces have been developed and applied in clinical use [3]. Aseptic loosening, however, is still a main cause of prosthesis failure [4]. In order to further improve bone-implant integration, highly porous or rough structures and surface coatings are continuously being investigated to enhance osteogenesis at the implant surface.

The recruitment and migration of osteogenic cells to the surface of implants to differentiate to osteoblasts forming new bone directly on the implant is referred to as contact osteogenesis [5, 6]. Porous or rough surfaces can greatly increase surface area so as to attach large amount of surface adsorbing fibrins, which in turn cause increased numbers of osteo-differentiating cells to migrate to the bone-implant interface [5, 6]. Plasma spraying is one of the most popular techniques used in the fabrication of porous surfaces for uncemented implantation [7, 8]. It has been recognised that plasma spraying produces highly porous surfaces with open and interconnected pores, which can vastly improve bone ingrowth characteristics [7, 9]. In addition, depending on porosity and the thickness of the porous coating, the compressive modulus of the porous substrate can be tailored to match that of cancellous bone, thus reducing the problems associated with stress shielding [7].

The osteoconductive nature of calcium phosphates can facilitate improved de novo bone formation at the bone-implant interface [10]. As such, they are often applied to implant substrates to improve bone-implant fixation [11, 12]. Conventional hydroxyapatite (HA) coatings are also typically applied by a plasma spraying technique [13]. A limitation of this particular method is that HA may interfere with the structure, openness and interconnectivity of pores. An alternative method, electrochemical cathodic deposition, is performed in a solution containing dissolved calcium and phosphorus ions resulting in the deposition of a thin and uniform layer of calcium phosphate compound on the 3D porous substrate, with grain size ranging from the sub-micron scale to several micrometers [14]. Dicalcium phosphate dihydrate (DCPD) is one such osteoconductive coating which can be applied to a porous substrate by this method, without compromising pore openness and interconnectivity [15]. However, DCPD exists in living bone in a metastable phase [16], meaning that the length of time present in vivo is limited.

We conducted this study to determine whether an electrochemically-deposited DCPD coating could improve the extent of ingrowth and ongrowth for a highly porous titanium surface and whether the coating could enhance bone-implant interfacial shear strength. Our null hypothesis was that the DCPD coating would have no effect on interfacial cortical shear strength and osseointegration in either cortical or cancellous sites.

Electrochemical cathodic deposition is a method employed to apply a thin and uniform layer of calcium phosphate coating on a porous implant surface. Metallic implants are submerged in an electrolyte bath containing dissolved calcium and phosphorus ions and connected to an external power supply [14]. A thin DCPD layer with grain size ranging from 1-3 μm is deposited on and within the porous implant surface, without compromising pore openness and interconnectivity [15]. DCPD dissolution is mainly affected by volume diffusion [19]. In this study the DCPD layer was found to be mostly dissolved at 1 week, with only trace amounts present at 2 weeks, which is consistent with other reports in the literature [20, 21].

DCPD is believed to act as a heterogeneous centre for HA growth in early bone formation [22]. For this reason it has been postulated that the thin calcium phosphate coating will improve bone ongrowth and ingrowth of porous implant surfaces to achieve rapid and early bone-implant interface integration and stability. Our results suggest that a DCPD coating has the potential to improve the extent of cancellous bone ingrowth in the early postoperative period (Figure 3). This finding is consistent with an in vitro study showing higher cell attachment ability on calcium phosphate compound samples in the early stages [23]. Simank et al [15] detected no significant difference in the mechanical fixation or bone formation throughout porous titanium implants coated with either an osteoinductive growth and differentiation factor-5 (GDF-5) or osteoconductive DCPD coating [15]. The mean bone ingrowth rate in the DCPD group was approximately 66% in cortical bone at 4 weeks, which compares well with values of 60% and 48% previously reported for a porous beaded coating with and without a 50 μm HA coating at 4 weeks in an ovine model [10].

In this study the cancellous implantation sites presented with higher mean bone ingrowth and ongrowth values than in the cortical bone sites at 4 weeks postoperatively for both DCPD and TiPS groups. Whilst this mean increase was not statistically distinguishable this finding is consistent with the knowledge that cancellous bone heals at a faster rate than cortical bone [24]. On the other hand, bone ingrowth and ongrowth in cortical bone sites showed generally higher percentage values than in cancellous bone at 12 weeks for both groups. Ostensibly, this result at 12 weeks is indicative of the compact nature of cortical bone. In joint arthroplasty the primary mode of fixation for uncemented tibial trays, femoral components and acetabular cups is indeed via cancellous bone ongrowth and ingrowth. Possible effects which the differential in ongrowth and ingrowth patterns observed in this study may have on uncemented fixation of joint components remains unknown.

Another striking feature in the current study was the seeming preservation of trabecular bone structure to within the porous coating domain (Figure 7), despite the presence of intervening titanium. Because trabecular bone tends to adapt to direction of mechanical stress [25, 26] this phenomenon may indicate that mechanical loads were indeed transmitted through the thin titanium pore walls. This observation supports the potential of selective manufacturing to limit the effects of stress-shielding by tailoring the elastic modulus of mediums for hard tissue infiltration. Ryan and colleagues [7] have demonstrated that the compressive modulus of porous metals is better matched to cancellous bone as compared to solid metals. This phenomenon of the continuation of trabecular bone architecture to within the porous coating has not previously been observed for thick-walled porous surfaces, such as beaded constructs [18].

An implant exhibiting an osteoconductive coating can stimulate new bone growth directly on the implant surface [8, 9] and improve uncemented prosthesis fixation in the early postoperative period [27, 28]. In this study, the plasma sprayed titanium porous surface both with and without the electrochemically-deposited DCPD coating exhibited de novo bone formation on the implant surface as early as two weeks after implantation (Figure 5 and Figure 6). At this timepoint osteoblasts were seen lining new bone on both the implant surface and adjacent host bone (Figure 5). Contact osteogenesis in both DCPD and TiPS groups was in agreement with a report that porous concave coatings can stimulate osteogenic cells differentiating to osteoblasts [29].

The percentage ingrowth for both test materials in the current study averaged approximately 37% at 2 weeks, as compared to the 13% ingrowth obtained for a porous tantalum implant [30]. Tantalum has been recognized as having excellent bone and fibrous ingrowth properties, allowing for rapid and substantial bone and soft tissue attachment [31]. Direct comparison of these results is fraught with difficulty, though, due to differences between studies in terms of implant parameters (porosity and coating thickness), implantation site and species. Regardless, the results of the current study support the osteoconductive potential of a highly porous titanium surface with a DCPD coating.

Evidence of remodeling in the cancellous sites was observed in both DCPD and TiPS groups as early as 4 and 8 weeks, with Haversian canals identified at 12 weeks (Figure 5). In addition, considerable amounts of lamellar bone and evenly distributed osteocytes were clearly seen in surrounding bone on both DCPD and TiPS sections at 12 weeks. The rate of remodeling in the current study is in contrast to other previous uncemented implant fixation studies in sheep [32, 33] where woven bone and lamellar matrix persisted three months postoperatively. This remodeling rate may be attributed to the highly porous surface and the press-fit insertion manner adopted in current study.

Mechanical testing revealed no difference between DCPD and TiPS at either timepoint. When selecting a soluble material for coatings, the match of resorption rate and bone regeneration rate must be taken into account. If resorption rate is faster than regeneration, there may be a void left by the absorbed material, which can potentially compromise bone and implant contact [13]. The shear strength of DCPD group was not lower than the control group in the current study, although the DCPD coating appeared mostly absorbed at 1 week and almost completely at 2 weeks. The mechanical similarity between DCPD and TiPS group in the two early time points indicated the thin (20 μm) and highly soluble DCPD coating will not compromise bone-implant interface mechanical stability in early stage.

The failure mode for both implant types from 4 to 12 weeks postoperatively was primarily at the interface between de novo formed and host bone. The failure mode illustrated that shear strength depends on the amount and strength of surrounding new bone, which can also be correlated to a study showing that mechanical stability of rough titanium implants depends on the amount of bony tissue surrounding the implant [15]. This may be the reason why higher ingrowth did not result in higher shear strength in DCPD implants. The increase of mechanical strength with increasing time may be due to the increasing amounts of mature surrounding bone.

The study of plasma sprayed porous titanium surface coated with and without DCPD demonstrated electrochemically deposited thin layer of DCPD with fine grain size can improve bone ingrowth in vivo. Mechanical results indicate that the thin and soluble DCPD had neither a positive nor negative effect on interfacial shear strength and implant stability in cortical bone. Moreover, analysis of the failure mode suggests that the bone bonding strength of the porous surface depends on the amount and maturity of surrounding new bone for both groups. As expected, an improvement in interfacial shear strength for both implant types with time was observed, continuous with the mechanical advantage of bony remodeling.

Cancellous bone implantation was associated with higher bone ingrowth and ongrowth at the early stage, whilst cortical bone implantation had more bone ingrowth and ongrowth than cancellous bone at 12 weeks. The continuity of trabecular bone to within the porous coatings (Figure 7) also indicates the adaptation of the highly porous surface structure to cancellous bone. The implantation of the porous surface implants by press-fit insertion demonstrated excellent early new bone formation and remodelling.

Finally, electrochemical deposition has the potential to produce calcium phosphate compounds with sub-micron sized grains which may lead to higher cell adhesion and osteoblast activity [34]. The effect of such coatings may be examined in the future.