Dental implants have become a predictable, viable and functional alternative for many edentulous patients. However, between 3 to 10% of implants fail; and any factor that affects osseointegration (the process by which the implant becomes integrated with bone), will play a role in determining implant success or failure. Implant-related factors such as implant length and location have often been cited as risk factors for implant failure. Systemic factors have also been implicated in the failure of dental implants, but the role of systemic conditions and medications on implant outcomes remains unclear. At the present time, there is increasing evidence that osteonecrosis of the jaw (ONJ) is associated with intravenous and oral nitrogen-containing bisphosphonates. Bisphosphonates inhibit osteoclast-mediated bone resorption, thereby leading to decreased turnover of bone. Since the pathophysiological processes leading up to ONJ affect bone remodeling in the maxilla and mandible, there has been concern whether dental implants placed in women on oral bisphosphonate therapy would have worse implant outcomes. I therefore investigated the role of implant location and length in dental implant failure; and the association between oral bisphosphonate use at the time of implant placement and implant failure. The study involved 114 case and 233 control females, aged 40 years and older, who had 1181 implants placed in the Department of Periodontology and Implant Dentistry at New York University College of Dentistry between January 1997 and December 2004. Cases, defined as patients with one or more implant failures, were identified from the departmental database. Two controls, matched on year of implant placement were then randomly selected for each case. Clinical information pertaining to implant treatment was obtained through the database, while the medical history and medication use of cases and controls was obtained by a review of patient charts. Adjusted odds ratios were estimated using logistic regression models fitted through generalized estimating equations.

There was no association between implant location and the risk for implant failure (adjusted odds ratio [OR] for maxilla versus mandible = 1.05; 95% confidence interval [CI, 0.68 to 1.63). However, the odds of having a short (≤10.0 mm.) implant was 70% higher in patients with at least one implant failure compared to patients without any implant failure in this population (adjusted OR = 1.70; 95% CI 1.17 to 2.48), after adjusting for age, year of implant placement, smoking status, implant location, implant surface, implant diameter, and oral bisphosphonate use. This finding was observed regardless of location.

A history of oral bisphosphonate use at the time of implant placement was found to be significantly associated with dental implant failure, and the association remained nearly unchanged after adjustment of various covariates. The odds of oral bisphosphonate use was 2.69 (95% CI, 1.49 to 4.86) times higher in women for whom implants failed compared to those for whom implants did not fail, after adjusting for age, year of implant placement, smoking status, implant location, implant surface, implant diameter, and implant length. Moreover, although no significant interaction was observed (p = 0.4146), the stratified analyses suggested that the association between oral bisphosphonate use and dental implant failure may be stronger in the maxilla (OR = 2.60; 95% CI, 1.36 to 4.96) than in the mandible (OR = 1.38; 95% CI, 0.51 to 3.73).

Future research examining how other implant-related characteristics affect implant failure should be conducted using large databases and controlling for confounders. Findings from this study also suggest that dental practitioners should be aware of the increased risk of implant failure associated with oral bisphosphonate use in certain patient populations. More research is needed to validate these findings in other populations.