In this work, we propose a combined path planning algorithm and foothold adaptation for quadrupedal locomotion in challenging terrain. The proposed path planning algorithm, based on Rapidly-exploring Random Tree* (RRT*), efficiently determines kinematically feasible paths for quadruped considering the given terrain information. However, following a body reference alone is insufficient to overcome the entire of challenging terrain, as the path planning could not capture whole scenarios, the robot must adapt its plan based on its current situation. Therefore, we use reactive foothold adaptation and generate collision-free swing leg trajectories aligned with the adapted foothold positions. Through the experiments in a 3D simulation environment featuring a sparse stepping stone map, we demonstrate the efficacy of our combined framework, comprising path planning and foothold adaptation.