How can I perform a two-sample multivariate t-test where one group is a subset of the other?

Is it possible to compare two means (ideally multivariate) where one comes from a group that is the subset of the other group?

E.g. the first group consists of A, B, C, D, E subjects and the second one consists of A, B, C, D, E, F subjects. My goal is to test whether the F’th subject greatly impacts the result