F-value and Post-hoc tests conflict

What does it mean when you obtain a significant F-value for an ANOVA but when you perform the post-hoc test no pairwise comparisons are significant? How do I report this?

More information:

When there is heterogenity for the DV in question I use Game-Howell post-hoc test because it does not assume homogenity of variance. Most of the time when what i described above happens, it happens when using Games-Howell.

Thanks yall!