Two new approaches to improve the analysis of BALB/c 3T3 cell transformation assay data

Validation activities of the BALB/c 3T3 cell transformation assay (CTA) - a test method used for the assessment of the carcinogenic potential of compounds – revealed the need for statistical analysis tailored to typical BALB/c 3T3 CTA data. Therefore, an international expert group reviewed commonly applied statistical approaches in view of a harmonised BALB/c 3T3 CTA test protocol and the specific data properties, such as variability. As it was concluded that none of the regularly applied approaches is entirely appropriate, two novel ways for data analysis were recommended. These were an approach based on the negative binomial generalised linear model (with William’s-type protected tests (GLM-NB) and an approach normalising the data by a specific transformation subsequently apply William’s-type protected tests (NT). These two approaches are presented and applied to exemplary data demonstrating their performance and the type of results generated. It is confirmed that both approaches are able to statistically handle the specific BALB/c 3T3 CTA data properties, such as non-monotone concentration-response curves. Furthermore, a procedure for data interpretation dichotomising results into negatives and positives with the option of re-testing for inconclusive cases is proposed. The statistical scripts to be used in a freely available statistical software including exemplary outputs are annexed in order promote the acceptance of the two recommended approaches and facilitate their application.

HOFFMANN Sebastian;  HOTHORN Ludwig;  EDLER Lutz;  KLEENSANG Andre;  SUZUKI Masaya;  PHRAKONKHAM Pascal;  GERHARD Daniel; 

2012-12-31

ELSEVIER SCIENCE BV

JRC66213

1383-5718,   

https://publications.jrc.ec.europa.eu/repository/handle/JRC66213,   

10.1016/j.mrgentox.2011.12.002,