A multivariate statistical analysis to follow the course of disease after infection of cats with different strains of the feline immunodeficiency virus (FIV)

A descriptive multivariate assay is described which is suitable to analyze results of a biological experiment with small sample size but high qualitative and quantitative complexity of variables. This type of assay allows evaluation of multiple variables observed in the course of an experimental virus infection (e.g. viremia, nucleic acid detection, antibody titers, clinical parameters, anti-microbial treatments or vaccination) in a single graph. In our study, a multiple correspondence analysis (MCA) was used to correlate a total of 145 measurements from each of a dozen of variables measured in five groups of three cats infected by five isolates of feline immunodeficiency virus (FIV). Three groups of virus isolates with distinct virulence were defined and correlation between dynamics of lymphocyte subset counts and viral virulence was established. Comparison between the primary stages of illness and follow-up examinations were of prognostic value and are thus helpful for development and monitoring of therapeutic strategies. (C) 2002 Elsevier Science B.V. All rights reserved.