Billstrom, GH and Dawson, J and Hofmann, S and Muller, R and Stoddart, M and Alini, M and Redl, H and El Haj, AJ and Brown, R and Salih, V and Hilborn, J and Larsson, S and Oreffo, R (2016) Multicentre study reveals poor correlation between in vitro and in vivo assessments of biomaterials for bone-regeneration. In: UNSPECIFIED.

Full text not available from this repository.

Abstract

Introduction: Research on biomaterials for bone regeneration generates a plethora of new biomaterials requiring evaluation with reliable and comparable methods of biocompatibility and functionality for clinical translation. To reduce the burden of in vivo assessment, there is a need for refined in vitro assays that are predictive of in vivo outcomes. This retrospective study correlated in vitro results with in vivo outcomes observed for a range of biomaterials for bone regeneration.

Methods: Members from eight universities in the European consortium BioDesign kindly provided the data. Assessors from each lab scored both in vitro and in vivo outcomes of individual biomaterial experimental variables (1=poor to 5=very good) on the basis of assessor-defined criteria particular to each assay. The data included 36 in vivo studies 47 in vitro assays and 93 materials. Data were sorted into sub-groups of different in vitro assays. In addition, in vitro assays were combined in pairs to investigate if combinations of in vitro assays allowed better prediction of in vivo outcomes compared to single assays.

Results: There was no significant overall correlation between in vitro and in vivo outcome. The mean in vitro scores shared 58% of variance to the in vivo scores. The mean in vivo scores shared 51% of variance to the in vitro scores. All combinations of in vitro assays showed less than 10% covariance, except for one combination of alkaline phosphatase and biocompatibility assays, which shared 95% covariance though low statistical power limits the conclusions that can be drawn.

Conclusion: The conventional biomaterial-testing pipeline is susceptible to selection bias through a tendency to select only materials with positive in vitro outcomes for testing in subsequent in vivo studies. This makes the poor correlation surprising. Current findings encourage the development of novel approaches of biomaterial assessment able to reliably predict in vivo outcomes.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Subjects: Q Science > Q Science (General)
R Medicine > R Medicine (General)
Divisions: Faculty of Medicine and Health Sciences > Institute for Science and Technology in Medicine
Depositing User: Symplectic
Date Deposited: 22 Jun 2016 08:05
Last Modified: 19 May 2017 13:56
URI: http://eprints.keele.ac.uk/id/eprint/1917

Actions (login required)

View Item View Item