**Yes** we can factor an RSA modulus $n$ given $n$ and $k \cdot \phi(n)$, including where $k$ is a (reasonably) large prime. We just use $f\gets k \cdot \phi(n)+1$ instead of $f\gets e\,d-1$ in the algorithm of [this answer][1]. The algorithm is heuristic, and I do not claim a rigorous proof of the distribution of the runtime. Also, a larger $k$ tends to make it more costly. But in practice it works fine for $k$ of size comparable to $n$. [1]: https://crypto.stackexchange.com/a/62487/555