CONCLUSIONS
Most of the models used to assess the adhesion of
probiotics in vitro represent simplifications of in vivo
conditions. However, some recommendations may be
made to bring the in vitro observations more in line with
the in vivo situation (Table II). Great care must be taken to
extrapolate from in vitro observations. In vitro models are
of great help in selecting probiotic candidates and are
indispensable instruments in the study of adhesion mechanisms.
Nonetheless, transient colonization and in vivo adhesion
of probiotics must be determined in vivo. Due to hostspecific
physiology, anatomy and intestinal microbiota
factors, this should be performed in the host for which
the probiotic is intended.
Considering all the factors influencing in vivo adhesion
that cannot be simulated in vitro, the validity of results from
in vitro adhesion assays remains extremely uncertain. The
assumed correlation between in vitro and in vivo adhesion
therefore requires urgent investigation, using adherent and
low-adherent isogenic strains with in vitro and in vivo