This paper investigates the challenge of optimizing communication performance in highway scenarios with discontinuous infrastructure coverage. First, we propose an Adaptive Cooperative Strategy (ACS) that dynamically integrates Vehicle-to-Infrastructure (V2I) and Vehicle-to-Vehicle (V2V) communications in both same-direction and opposite-direction traffic flows, adaptively selecting the optimal transmission path based on real-time vehicle positions and network conditions. Then, we develop a comprehensive analytical model, modeling the vehicular network as a stochastic process to capture the dynamic topology of the network and deriving a closed-form expression for the average achievable throughput. Finally, extensive simulations validate the accuracy of our analysis and evaluate the impact of key parameters on performance. The results demonstrate that the proposed ACS significantly outperforms traditional single-mode strategies in scenarios with discontinuous infrastructure coverage. Our approach aims to maximize achievable throughput in infrastructure-sparse vehicular networks and promote the development of intelligent transportation systems.