Development and internal validation of a titanium ion‐integrated diagnostic classification model for peri‐implantitis
Amar Ashok Thakare, Vaishali Mashalkar, Kavita Gupta, Rahul Tiwari, Harisha Dewan, Syed Shujaulla, Manish SharmaAbstract
Background
Titanium release from dental implants has been implicated in peri‐implantitis; however, its integration into multivariable models for disease classification remains limited. This study aimed to develop and internally validate a titanium ion‐integrated model for the classification of peri‐implantitis.
Methods
This cross‐sectional analysis with prospective recruitment included 120 participants (60 healthy implants and 60 peri‐implantitis). Titanium ion concentrations in peri‐implant crevicular fluid (PICF) and serum were quantified using inductively coupled plasma mass spectrometry. Candidate predictors comprised PICF and serum titanium levels, implant surface area, functional duration, occlusal load, and age. Multivariable logistic regression was used to derive the Titanium Release Score. Model discrimination was assessed using receiver operating characteristic analysis, while calibration was evaluated using the Hosmer–Lemeshow test. Internal validation was performed through bootstrap resampling (1000 iterations).
Results
Higher titanium concentrations in PICF (adjusted odds ratio [OR] = 1.52; 95% confidence interval [CI]: 1.22–1.90) and serum (adjusted OR = 1.79; 95% CI: 1.25–2.56) were associated with peri‐implantitis. The model demonstrated good discriminative performance (area under the curve [AUC] = 0.87; 95% CI: 0.78–0.95). Internal validation showed an optimism‐corrected AUC of 0.83. Calibration was acceptable (Hosmer–Lemeshow, p = 0.62).
Conclusion
The titanium‐integrated model demonstrated good discrimination between peri‐implantitis and healthy implants within the constraints of a cross‐sectional design. Titanium ion levels were associated with disease status alongside implant‐related variables. These findings support the potential role of titanium measurements as adjunctive biomarkers for disease classification rather than predictors of future risk; however, external validation and longitudinal studies are required before clinical application.
Key points
A titanium ion‐integrated classification model was developed to estimate peri‐implantitis risk using peri‐implant crevicular fluid, serum titanium levels, and implant variables. The Titanium Release Score demonstrated good discrimination (area under the curve 0.87) and acceptable calibration with stable performance after bootstrap validation. Titanium ion burden showed an independent associative value for peri‐implantitis, supporting its role as a biologically relevant risk indicator.
Plain Language Summary
Peri‐implantitis is a common complication of dental implants that leads to inflammation and progressive bone loss. Early detection of implants at risk remains a clinical challenge. This study developed a classification model that combines titanium ion levels measured in peri‐implant crevicular fluid and blood with implant‐related factors such as surface area, duration of function, and occlusal load. Titanium, although widely used in implants, may be released over time and contribute to inflammatory responses. The model demonstrated good accuracy in distinguishing healthy implants from those with peri‐implantitis, suggesting that titanium levels may provide additional information beyond traditional clinical parameters. These findings highlight the potential role of titanium measurements in improving risk assessment. However, further validation in larger and more diverse populations is required before this approach can be applied in routine clinical practice.