Given a rational curve $C:(f_1(t),f_2(t))$, where $f_i(t),i=1,2$ are rational functions with rational coefficients.

Question: Is there any criterion(proved or conjectural) for the existence of integral points on such a curve?