The performance of long-term power system studies are vital to evaluate the system reliability and efficiency under an increasing penetration of renewable energy sources (RES). This work develops a long-term unit commitment (UC) model considering combined-cycle (CC) units to enable the performance of annual studies in power systems. Specifically, a simplified configuration-based model of the CC units is used that decreases the number of configurations to reduce the computational complexity. The considered UC problem is formulated as a mixed-integer linear program (MILP), which is intractable when a long-term horizon is used. To make the problem tractable, a rolling horizon approach is proposed to solve the long-term MILP problem, generating high-quality approximate solutions. The proposed approach is applied to a real isolated power system and an annual study is carried out to examine the impact of an increasing RES penetration on the system operation.