Pancreatic cancer (PC) is one of the most malignant cancers, owing to extremely high aggressiveness and mortality. Yet, this condition currently incurs widely drug resistance and therapeutic deficiency. In this study, we proposed a novel functional metabolomics strategy as Spatial Temporal Operative Real Metabolomics (STORM) to identify the determinant functional metabolites in a dynamic and visualized pattern whose level changes are mechanistically associated with therapeutic efficiency of gemcitabine against PC. Integrating quantitative analysis and spatial-visualization characterization of functional metabolites in vivo, we identified that the AMP-cAMP axis was a novel therapeutic target of PC to intermediate therapeutic efficiency of gemcitabine. Gemcitabine could induce the dual accumulation of cyclic AMP (cAMP) and AMP in tumor tissues. Quantitative analysis of associated biosynthetic enzymes and genes revealed that two independent intracellular ATP derived biosynthetic pathways to promote the dual activation of AMP-cAMP axis in a lower-level energetic environment. Then, gemcitabine induced the dual accumulation of AMP and cAMP can separately activate signaling pathways of AMPK and PKA, leading to the inhibition of tumor growth by the upregulation of the downstream tumor suppressor GADD45A. Collectively, our new STORM strategy was the first time to identify novel target of PC from a metabolic perspective as the dual activation of AMP-cAMP axis induced by gemcitabine can efficiently suppress PC tumor growth. In addition, such discovery has the capability to lower drug resistance of gemcitabine by specifically interacting with novel target, contributing to the improvement of therapeutic efficiency.