The requested generalization of Jacobi's two-square theorem is a remarkably recent result: N. Bagis and M.L Glasser, <A HREF="http://arxiv.org/abs/1406.0466">On the Number of Representations of Integers by various Quadratic and Higher Forms</A> (2015):

<IMG SRC="http://ilorentz.org/beenakker/MO/sumofsquares_2.png"/>

where $r(n)$ is Jacobi's formula for the number of representations of $n=x^2+y^2$.

---

I understand from the comments that this formula is not what the OP was looking for. Specific cases are considered by Berkovich and Yesilyurt in <A HREF="http://arxiv.org/abs/math/0611300">Ramanujan's Identities and Representation of Integers by Certain Binary and Quaternary Quadratic Forms</A> (2006). One representative example is given below, there are several others ($n=x^2+ay^2$ with $a=5,6,15,27,\ldots$)

<IMG SRC="http://ilorentz.org/beenakker/MO/arXiv_0611300.png"/>