High-sensitivity spin-exchange relaxation-free (SERF) magnetometry combined with magnetic-responsive iron oxide nanoparticles for real-time monitoring of malignant tumor therapy

<b>Rationale</b>: Early and accurate evaluation of chemotherapy efficacy remains essential, yet conventional imaging approaches rely on delayed morphological changes. Functional alterations such as apoptosis and reduced metabolic activity occur earlier but are difficult to detect noninvasively. Magnetic signal detection offers a promising alternative but is limited by signal instability and biological noise. <b>Methods</b>: We developed a magnetic signal-based monitoring platform by combining magnetically responsive ferromagnetic-superparamagnetic iron oxide nanoparticle (F-SPION) with a spin-exchange relaxation-free magnetometer, with signal amplification achieved through rubidium magnetization. <i>In vitro</i>, we assessed the linear correlation between magnetic signal intensity and tumor cell number, and further evaluated doxorubicin (DOX)-induced signal changes under constant cell conditions. Prussian blue staining was used to confirm changes in F-SPION uptake. <i>In vivo</i>, F-SPION was intravenously injected into tumor-bearing mice, and magnetic signals from tumor and normal tissues were measured at multiple time points after magnetization. The mice were randomly assigned to control or doxorubicin-treated groups, and tumor signals were monitored on Days 1, 7, 14, and 21. Biocompatibility was assessed through cytotoxicity, hemolysis, histology, and blood analysis. <b>Results</b>: <i>In vitro</i>, magnetic signal intensity strongly linearly correlated with tumor cell number (R² = 0.974). Doxorubicin treatment resulted in signal reduction despite the identical cell numbers (control: 267.88 ± 5.97 pT; 24 h: 206.02 ± 2.23 pT; 48 h: 122.74 ± 2.11 pT), with Prussian blue staining confirming reduced F-SPION uptake. <i>In vivo</i>, the signal peaked at 0.5 h post-injection (1528.54 ± 23.34 pT). The tumor signals were consistently greater than the signals of normal tissues at 5 min (802.7 ± 60.8 vs. 149.3 ± 16.2 pT) and 60 min (163.6 ± 3.2 vs. 42.8 ± 1.5 pT). On Day 1, the signal of the treatment group was 425.3 ± 24.4 pT and remained stable until Day 7 (425.4 ± 14.4 pT), whereas that of the control group increased from 481.4 ± 3.8 to 830.7 ± 5.9 pT. <b>Conclusions</b>: This magnetic signal-based platform enables noninvasive, real-time, and functional monitoring of tumor response, offering a sensitive and translational strategy for early-phase therapeutic evaluation.