Resolving late failure of dental implant is difficult and costly; however, only few reviews have addressed the risk factors associated with late failure of dental implant. The aim of this literature review was to summarize the influences of different potential risk factors on the incidence of late dental implant failure. The protocol of this systematic review was prepared and implemented based on the PRISMA (Preferred reporting items for systematic reviews and meta-analyses) guideline. In December 2018, studies published within the previous 10 years on late dental implant failure were selected by fulfilling the eligibility criteria and the risk factors identified in qualified studies were extracted by using a predefined extraction template. Fourteen eligible studies were assessed. The common risk factors for late failure were divided into three groups according to whether they were related to (1) the patient history (radiation therapy, periodontitis, bruxism and early implant failure), (2) clinical parameters (posterior implant location and bone grade 4) or (3) decisions made by the clinician (low initial stability, more than one implant placed during surgery, inflammation at the surgical site during the first year or using an overdenture with conus-type connection). Clinicians should be cautions throughout the treatment process of dental implant—from the initial examination to the treatment planning, surgical operation and prosthesis selection—in order to minimize the risk of late failure of dental implant.

Dental implant surgery is a common treatment for missing teeth. Its survival rate is considerably affected by host bone quality and quantity, which is often assessed prior to surgery through dental cone-beam computed tomography (CBCT). Dental CBCT was used in this study to evaluate dental implant sites for (1) differences in and (2) correlations between cancellous bone density and cortical bone thickness among four regions of the jawbone. In total, 315 dental implant sites (39 in the anterior mandible, 42 in the anterior maxilla, 107 in the posterior mandible, and 127 in the posterior maxilla) were identified in dental CBCT images from 128 patients. All CBCT images were loaded into Mimics 15.0 to measure cancellous bone density (unit: grayscale value (GV) and cortical bone thickness (unit: mm)). Differences among the four regions of the jawbone were evaluated using one-way analysis of variance and Scheffe’s posttest. Pearson coefficients for correlations between cancellous bone density and cortical bone thickness were also calculated for the four jawbone regions. The results revealed that the mean cancellous bone density was highest in the anterior mandible (722 ± 227 GV), followed by the anterior maxilla (542 ± 208 GV), posterior mandible (535 ± 206 GV), and posterior maxilla (388 ± 206 GV). Cortical bone thickness was highest in the posterior mandible (1.15 ± 0.42 mm), followed by the anterior mandible (1.01 ± 0.32 mm), anterior maxilla (0.89 ± 0.26 mm), and posterior maxilla (0.72 ± 0.19 mm). In the whole jawbone, a weak correlation (r = 0.133, p = 0.041) was detected between cancellous bone density and cortical bone thickness. Furthermore, except for the anterior maxilla (r = 0.306, p = 0.048), no correlation between the two bone parameters was observed (all p > 0.05). Cancellous bone density and cortical bone thickness varies by implant site in the four regions of the jawbone. The cortical and cancellous bone of a jawbone dental implant site should be evaluated individually before surgery.

Small-diameter dental implants are associated with a higher risk of implant failure. This study used both three-dimensional finite-element (FE) simulations and in-vitro experimental tests to analyze the stresses and strains in both the implant and the surrounding bone when using one-piece (NobelDirect) and two-piece (NobelReplace) small-diameter implants, with the aim of understanding the underlying biomechanical mechanisms. Six experimental artificial jawbone models and two FE models were prepared for one-piece and two-piece 3.5-mm diameter implants. Rosette strain gauges were used for in-vitro tests, with peak values of the principal bone strain recorded with a data acquisition system. Implant stability as quantified by Periotest values (PTV) were also recorded for both types of implants. Experimental data were analyzed statistically using Wilcoxon's rank-sum test. In FE simulations, the peak value and distribution of von-Mises stresses in the implant and bone were selected for evaluation. In in-vitro tests, the peak bone strain was 42% lower for two-piece implants than for one-piece implants. The PTV was slightly lower for one-piece implants (PTV = -6) than for two-piece implants (PTV = -5). In FE simulations, the stresses in the bone and implant were about 23% higher and 12% lower, respectively, for one-piece implants than those for two-piece implants. Due to the higher peri-implant bone stresses and strains, one-piece implants (NobelDirect) might be not suitable for use as small-diameter implants.