Detecting cancer through pathology imaging is crucial for early diagnosis and effective treatment. However, the performance of deep learning (DL) models is often undermined by domain shifts between training and test data—variations caused by differences in imaging acquisition device, staining protocols, and patient populations in real-world scenarios. To address this challenge, we propose PathTTT, a novel framework that combines Test-Time Training (TTT) with Model-Agnostic Meta-Learning (MAML) to enhance model robustness under domain shift conditions. Our experimental results demonstrate significant improvements over SOTA methods on benchmark datasets, highlighting PathTTT’s potential for robust cancer detection in real-world dataset applications.