Large Language Model-based Automated Program Repair (LLM-APR) has recently received significant attention as a debugging assistance. Our objective is to improve the performance of LLM-APR. In this study, we focus on semantic information contained in the source code. Semantic information refers to elements used by the programmer to understand the source code, which does not contribute to compilation or execution. We picked out specification, method names and variable names as semantic information. In the investigation, we prepared eight prompts, each consisting of all combinations of three types of semantic information. The experimental results showed that all semantic information improves the performance of LLM-APR, and variable names are particularly significant.