In any introductory book to GR where the Kerr metric is reviewed, they all apply transformations of the form:
$$cu' = cu + ia\cos\theta;\text{ } r' = r + ia\cos\theta$$
In the end, the result describes the gravitational field of an axisymmetrically rotating mass.
But why? I guess it has something to do with $z=r\cos\theta$, but I haven’t found any explanation on the Internet. I have read in the original paper from Boyer and Lindquist that you can presume $a$ to be the angular momentum per unit mass comparing the resulting equations to the Lense-Thirring effect ones, but I can’t see the connection between it and the transformation with $\cos\theta$.
Furthermore, I am curious to know what the transformations for $cdu$ and $r$ would be if the axis of rotation was not $z$ but rather $x$ or $y$; or even how a rotation along two axis would look like in terms of a Kerr transformation (for example, to describe a mass rotating like a pulsar, that could be described as rotating along the $z$ and $y$ axis at different speeds).