The accuracy of radiographic methods in assessing the outcome of periodontal regenerative therapy

Background: The study of regenerative therapy in the periodontal intrabony defect has relied upon surgical re-entry as the gold standard of outcome assessment. The search for a non-invasive method has led to the application of various radiographic techniques in evaluating post-treatment bone fill. Methods: The purpose of this study was to determine the ability of 2 forms of radiographic analyses (linear measurement and computer assisted densitometric image analysis, CADIA) to assess postsurgical bone fill as measured at a re-entry procedure. A method that incorporates linear measurements and CADIA (linear-CADIA) was developed and tested as well. Forty-five intrabony defects in 15 patients were treated with open flap debridement, demineralized freeze-dried bone allograft (DFDBA), or a combination of DFDBA and tetracycline. Standardized radiographs were obtained at baseline and at 1-year postsurgery. Results: A 12-month surgical re-entry provided clinical measurements for post-treatment bone fill. All radiographs were digitally scanned and analyzed on a computer. Fifty-three percent of the defects were excluded from the study due to poor standardization or poor defect quality. Forty percent of all pairs of radiographs were judged to have poor standardization. in the first analysis, standardized images were subtracted and quantitatively analyzed utilizing CADIA. It was found that CADIA had the highest correlation with clinical bone fill when a region of interest (ROI) was examined in the middle portion of the defect. This quantitative evaluation provided very little clinically relevant information regarding actual bone fill. For the second analysis, pre- and post-treatment linear radiographic measurements were obtained. In only 43% of the sites, did linear radiographic measurements determine post-treatment bone fill within 1.0 mm of the clinical measurements. Overall, linear measurements underestimated bone fill by 0.96 mm (+/-1.2). These differences were statistically significant (paired Student t-test, P = 0.0023). A method, which incorporates the use of both CADIA and linear radiographic measurements (linear-CADIA), was tested. The linear-CADIA method underestimated bone fill by 0.26 mm (+/-1.4), but these differences were not statistically significant (paired Student t-test, P= 0.41). Conclusion: Linear radiographic measurements significantly underestimate post-treatment bone fill when compared to re-entry data. The linear-CADIA method provided the highest level of accuracy of the 3 methods tested. This study also emphasizes the importance of developing a consistent method of radiographic standardization.