Securing edge networks against evolving cyber threats remains a critical challenge due to their distributed architecture and limited computational resources. This work introduces a hybrid framework that combines Quantum Neural Networks (QNNs) with Digital Twins (DTs) to enable proactive intrusion prevention. The DT continuously replicates the edge environment, allowing simulated attack scenarios and predictive threat analysis, while the QNN autoencoder captures complex traffic patterns in quantum feature space to detect subtle anomalies. Experimental results on the CICIDS2017 dataset indicate that the proposed system outperforms traditional machine learning models in accuracy, false-positive reduction, and real-time responsiveness. This study demonstrates the potential of integrating quantum computing with digital twin technology for adaptive, resilient edge cybersecurity.