Dual-tracer PSMA/FDG PET/CT for lesion detection and risk stratification in biochemical recurrence with negative conventional imaging.

PURPOSE: To evaluate lesion detection patterns and the risk-stratification value of paired 68Ga-PSMA-11 and 18F-FDG PET/CT in patients with post-prostatectomy biochemical recurrence (BCR) and negative conventional imaging, and to explore the hypothesis-generating biological context of dual-tracer phenotypes using public single-cell transcriptomic data.

METHODS

This retrospective single-center study included patients with post-prostatectomy BCR who underwent paired 68Ga-PSMA-11 and 18F-FDG PET/CT within 14 days and had negative conventional imaging within 1 month before PET/CT. Overall dual-tracer PET positivity was defined as at least 1 positive lesion on either tracer. Among patients with PET-detectable disease on either tracer, phenotypes were classified as PSMA+/FDG-, PSMA+/FDG+, or PSMA-/FDG+. Public single-cell RNA sequencing data were analyzed to assess the relationship between FOLH1 expression and glycolytic activity.

RESULTS

Among 58 included patients, 21 (36.2%) had at least one positive lesion on either tracer. Overall dual-tracer PET positivity increased across PSA strata and with a greater number of prespecified risk features (both P for trend < 0.001). Older age and higher PSA at PET were independently associated with overall dual-tracer PET positivity. Among patients with PET-detectable disease on either tracer, FDG-avid disease was associated with pathologic stage ≥pT3, pathologic N1 disease, recurrence within 3 years after radical prostatectomy, and PSA doubling time ≤ 6 months. In the public single-cell cohort, FOLH1 expression and glycolytic activity showed only a weak association.

CONCLUSION

Paired 68Ga-PSMA-11 and 18F-FDG PET/CT revealed heterogeneity beyond lesion detection in patients with post-prostatectomy BCR and negative conventional imaging. Among patients with PET-detectable disease on either tracer, FDG-avid disease was enriched for adverse clinicopathologic features, and public single-cell transcriptomic analysis provided hypothesis-generating biological context for divergent dual-tracer phenotypes.