Discriminant analysis HW

I'm stuck on the last question of my HW. I've done a discriminant analysis and got 87,3% correct using cross-validation (leave-one-out). there are only 2 groups; spam & not spam. sample size is 4601. how would I solve the following?

Suppose you start using a spam-filter based on the discriminant model. In the sample almost 40% of the e-mails were tagged as spam. However, assume that the actual proportion in real life is 10%. Under these assumptions, what proportion of the incoming e-mails would be correctly classified by the spam-filter?