Studies of identical twins, raised both apart and together, have indicated hereditary influence on personality traits. Traits. Not whole personalities. It's unscientific to confuse the two. Confusing things with what they ain't is a propaganda trick, not Scientific Method.
I really hate abuse of statistics. Like when people say, "Research shows that 65% of personality is inherited."
Not true. That's not what the research shows.
The percentages researchers came up with were a creature of their instruments (evaluations and questionnaires), not a measure of how alike the twins actually were. For example, if most of your questions are aimed at eliciting evidence of narcissism, you are going to find a high percentage of correlation due to just that one trait. But what percentage of the package of that person's total personality is that? Perhaps minuscule.
No ones claims to have an instrument that measures the whole spectrum of human personality, so that it can be used in such a study to determine "how alike" the personalities of identical twins are.
What's more, the trait of narcissism exists to some extent in everyone and is a far different thing than NPD.