A Pseudo-Likelihood Approach to Analysis of Nested Case-Control Studies