Agreement between Histomorphometry and Microcomputed Tomography to Assess Bone Microarchitecture of Dental Implant Sites. (17th November 2013)

• Record Type: Journal Article
• Authors: Dias, Danilo Rocha; Leles, Cláudio Rodrigues; Batista, Aline Carvalho; Lindh, Christina; Ribeiro‐Rotta, Rejane Faria
• Is Part Of: Clinical implant dentistry and related research. Volume 17:Number 4(2015:Aug.)
• Page Start: 732
• Page End: 741
• Publication Date: 2013-11-17
• Subjects: Dental implants -- Periodicals; Dental Implantation -- Periodicals; Dental Implants -- Periodicals; 617.693
• DOI: 10.1111/cid.12176
• Languages: English
• ISSNs: 1523-0899

Abstract

<abstract abstract-type="main"> <title>Abstract</title> <sec id="cid12176-sec-0001" sec-type="section"> <title>Background:</title> <p>Histomorphometry and microcomputed tomography (microCT) have been used in implant studies but need better understanding before being used as equivalent methods.</p> </sec> <sec id="cid12176-sec-0002" sec-type="section"> <title>Purpose</title> <p>The purpose of this study was to investigate the agreement between 2D (histomorphometry) and 3D (microCT) reference methods for assessing jawbone microarchitecture in vivo<bold>.</bold></p> </sec> <sec id="cid12176-sec-0003" sec-type="section"> <title>Material and Methods</title> <p>Forty‐four bone specimens from 32 patients were obtained during implant placement and examined by microCT, followed by hematoxylin–eosin staining and histomorphometric analysis. The morphometric parameters included bone volume density (BV/TV), bone surface fraction (BS/TV), bone surface density, trabecular thickness, trabecular number, and trabecular separation (Tb.Sp). Bland–Altman plots were used for pairwise agreement analysis between the equivalent 3D and 2D parameters, and complemented with Mountain plots. The association between the two methods was tested using Pearson's correlation followed by Passing–Bablok regression.</p> </sec> <sec id="cid12176-sec-0004" sec-type="section"> <title>Results</title> <p>Systematic bias was observed in all Bland–Altman and Mountain plots, including constant bias for BV/TV and Tb.Sp, and proportional bias for all other parameters. Significant correlation was found for BV/TV (<italic>r</italic> = 0.80; <italic>p</italic> < .001) and BS/TV (<italic>r</italic> = 0.44; <italic>p</italic> = .003), and the Passing–Bablok regression showed constant bias for BV/TV and proportional bias for BS/TV.</p> </sec> <sec id="cid12176-sec-0005" sec-type="section"> <title>Conclusion</title> <p>Because of the poor agreement between measures obtained by histomorphometry and microCT, these methods should not be used interchangeably for jawbones.</p> </sec> </abstract>