Abstract Objective To evaluate the feasibility and accuracy of a sprayable γ-GGT fluorescent probe for visualizing epithelial ovarian cancer (EOC) across three hierarchical models (organoids, frozen sections, and ex vivo tissues), with a focus on optimizing its clinical applicability in intraoperative lesion detection. Methods This prospective study, leveraging a medical-engineering collaborative γ-GGT probe, was conducted in three phases: preliminary validation, ex vivo tissue optimization, and clinical validation. Initially, TCGA analysis and immunohistochemistry confirmed γ-GGT overexpression in epithelial ovarian cancer (EOC), with organoids and frozen sections employed to evaluate the probe’s specificity at both cellular and microstructural levels. Subsequently, fresh EOC tissues ( n = 42 lesions from 8 patients) were treated with varying concentrations of the probe (1 μM, 5 μM, or 10 μM) to ascertain the minimal effective concentration (10 μM) and the optimal observation window (1 min–1 h) using signal-to-background ratio (SBR) analysis. Finally, clinical validation involved imaging 27 lesions ( 0.05), with optimal imaging stability observed 1 min-1 h post-application (SBR range:1.26–1.29). Clinical validation in 27 sub-0.3 cm lesions demonstrated 96.3% sensitivity (26/27 true positives) and 81.5% macroscopic detection rate, with one false positive (3.7%) in normal peritoneal tissue. Immunohistochemistry confirmed γ-GGT membrane expression in 100% of cancerous lesions (42/42 experimental group, 26/27 validation group) versus absent expression in normal tissues. Univariate analysis identified preoperative serum γ-GGT levels (≥ 20 U/L: SBR 1.28 ± 0.13 vs < 20 U/L: 1.23 ± 0.18, P = 0.005) and treatment modality ( P = 0.001) as significant factors, though only treatment approach remained independently predictive in multivariate analysis ( P < 0.001). Conclusion The γ-GGT probe enables rapid, accurate EOC visualization at clinically relevant scales, with 10 μM concentration and 1 min–1 h window being optimal for intraoperative use. Hierarchical validation from cellular to tissue models underscores its translational potential.
Liu et al. (Sun,) studied this question.