ABSTRACT Cancer remains a major global health challenge, underscoring the urgent need for early diagnostic strategies. Altered cancer metabolism can be captured through serum analysis by laser desorption/ionization mass spectrometry (LDI‐MS). Here, we report the design of gold nanoparticle‐modified manganese–cobalt oxide heterojunctions (Au/MCOHs) as a high‐performance LDI‐MS nano‐matrix. The Au/MCOHs combine strong ultraviolet absorption, enhanced charge separation, and favorable surface potential to achieve efficient laser energy conversion and ionization. Leveraging this platform, we performed serum metabolomic analysis on 698 participants, comprising healthy controls and patients with six different cancer types. Machine learning models distinguished cancer patients from healthy controls with area under the curve (AUC) values up to 0.955. Critically, early‐stage cancers were also detected, with AUCs ranging from 0.776 to 0.938. Thirteen cancer‐associated metabolites were identified, and a combined diagnostic panel achieved an AUC of 0.865. This study demonstrates how rationally engineered plasmonic heterojunction nanomatrices can bridge materials innovation and clinical translation, offering a robust, non‐invasive route for pan‐cancer metabolomic screening.
Qi et al. (Sun,) studied this question.