Abstract-The rapid evolution of 6G wireless communication systems challenge intelligent network management that is capable of supporting heterogeneous services with the ultra-low latency, massive connectivity, high reliability, and dynamic quality-of-service (QoS) requirements. The conventional network slicing and resource orchestration techniques are relying on the static optimization or heuristic-based strategies that struggle in adapting to changing network environments, which is resulting in inefficient resource utilization, is increasing service latency, and reduced user satisfaction. To overcome these limitations, this paper proposes a Hierarchical Adaptive Deep Reinforcement Learning Resource Orchestration (HADRL-RO) framework for an intelligent network slicing in next-generation 6G networks. The proposed HADRL-RO is combining the hierarchical deep reinforcement learning with the adaptive policy optimization, multi-objective resource allocation, slice-aware traffic prediction, and continuous environment feedback to dynamically allocate computing, spectrum, and communication resources across the heterogeneous network slices. A reward-driven learning mechanism optimizes resource allocation while it is maintaining service-level agreements (SLAs) under the varying traffic conditions. Extensive simulations is showing that the proposed framework is significantly improving network performance when it is compared with the conventional deep Q-learning, heuristic slicing, and optimization-based resource management approaches. Extensive simulations is showing that the proposed framework significantly outperforms Conventional DQN-based Resource Orchestration, Multi-Agent Reinforcement Learning, and Graph Neural Network-Based Resource Optimization methods. The proposed HADRL-RO is achieving 98.43% resource utilization, 96.57% slice acceptance ratio, 97.18% QoS satisfaction, 108.62 Gbps throughput, and 3.34 ms service latency, is showing superior adaptability, efficient resource orchestration, and enhanced quality-of-service for heterogeneous next-generation 6G network slicing scenarios.