Physicochemical study of plasma-sprayed hydroxyapatite-coated implants in humans.

This study represents the first report of the physical and chemical changes occurring in coatings of failed hydroxyapatite (HA)-coated titanium implants obtained from a comprehensive, multicenter human dental implant study. A total of 53 retrieved samples were obtained and compared with unimplanted controls with the same manufacturer and similar manufacture dates. Forty-five retrieved implants were examined for surface characteristics and bulk composition. Implants were staged based on implantation history: stage 1 (implants retrieved between surgical placement and surgical uncovering), stage 2 (implants retrieved at surgical uncovering and evaluation), stage 3 (implants retrieved between surgical uncovering evaluation and occlusal loading), and stage 4 (implants retrieved after occlusal loading). Scanning electron microscopy showed progressive coating thinning with implantation time. At later stages, bare Ti metal was detected by energy-dispersive X-ray analysis and electron spectroscopy for chemical analysis. Increases in Ti and Al (2-7.5 atm % each) were detected at the apical ends of all stage 4 samples. In unimplanted coatings, X-ray diffraction analysis demonstrated the presence of amorphous calcium phosphate, beta-tricalcium phosphate, tetracalcium phosphate, and calcium oxide in addition to large hydroxyapatite crystals (c axis size, D002 = 429 +/- 13 A; a axis size, D300 = 402 +/- 11 A, a/c aspect ratio 0.92). The nonapatitic phases disappeared with increased implantation time, although there was a persistence of amorphous calcium phosphate. Bulk coating chemical analysis showed that Ca/P ratios for implant controls (1.81 +/- 0.01) were greater than stoichiometric HA (1.67) and decreased for implant stages 3 and 4 (1.69 +/- 0.09 and 1.67 +/- 0.09, respectively), explained by the dissolution of the non apatitic phases. Crystal sizes also changed with implantation times, being smaller than the control at all but stage 4. Fourier transform infrared analyses agreed with these results, and also indicated the accumulation of bone (protein and carbonate-apatite) in the retrieved coatings. The accumulation of bone was not stage dependent. These findings indicate that there was some biointegration with the surrounding bone, but the greatest changes occurred with the HA coating materials, their loss, and chemical change.